Abstract
Power lithium-ion battery represents a new type of high-energy battery successfully developed in the 20th century. Compared with Ni-MH, lead-acid, and Ni-Cd batteries, lithium-ion batteries possess unparalleled advantages in terms of energy density, volume, cycle life, and environmental protection, and their application fields are expanding continuously. However, factors such as battery pack cost, safety performance, and cycle life still restrict the widespread application of power lithium-ion batteries. To address this situation, I designed an energy management system for power lithium-ion battery packs in this thesis, which ensures the safety and reliability of battery pack operation, displays real-time operating status data, and maximizes the service life of the battery pack.
This thesis first investigates the remaining capacity and State-of-Charge (SOC) of lithium-ion batteries, analyzes the factors affecting remaining battery capacity, compares several SOC estimation methods, and derives the SOC estimation approach adopted in this system. The method combines Ah counting with open-circuit voltage measurement as the primary strategy, while comprehensively incorporating compensation measures for charge-discharge efficiency, cycle life, and temperature. Based on practical application requirements, I determined the implementation scheme for each subsystem function and proposed the overall system architecture centered on the PIC18F452 microcontroller.
Regarding hardware, I designed a microcontroller-based management system, including battery voltage, current, and temperature sampling circuits, LED display, and function keys. For protection circuit design, I utilized the Seiko single-cell lithium-ion battery protection IC S-8261 to develop a low-cost, reliable, and well-expandable battery pack protection system. For equalization management, a simple energy-consuming equalization method was adopted to effectively achieve charge equalization of the battery pack.
Regarding software, I implemented modular program design, realized the functional algorithms of each module, and provided program flow charts.
Finally, I fabricated the circuit board according to the hardware design and performed system-level hardware and software debugging. Experimental results demonstrate that the management system ensures the safety and reliability of the battery pack in practical applications, realizes charge equalization functionality, and verifies the accuracy of the SOC estimation method.
Keywords: power lithium-ion battery pack, energy management system, protection circuit, state of charge, EV battery pack
Chapter 1 Introduction
1.1 Background and Significance of the Research
Power lithium-ion batteries are a new type of high-energy battery successfully developed in the 20th century, classified into liquid lithium-ion batteries and polymer lithium-ion batteries. Compared to traditional power batteries such as Ni-Cd and Ni-MH, lithium-ion power batteries exhibit advantages including high operating voltage, high energy density, light weight, small volume, low self-discharge rate, long cycle life, no memory effect, and environmental friendliness. They represent the development trend of the next generation of power batteries. Currently, they are widely used in portable mobile devices such as mobile phones, notebook computers, video cameras, and digital cameras. Large-capacity lithium-ion batteries have been trialed in electric vehicles and will become one of the primary power sources for electric vehicles in the 21st century, with applications in artificial satellites, aerospace, and energy storage.
With the expanding application fields of power lithium-ion batteries, the requirements for battery voltage and energy output capability continue to rise. Battery packs formed by series-parallel connections of multiple lithium-ion battery cells are increasingly being used. The use of such battery packs requires further consideration of safety and cost. Individual differences inevitably exist among cells in production, such as inconsistency in capacity and voltage, which accumulate over time and degrade the service life of the battery pack. Therefore, measures must be taken to effectively control the balance between cells. Additionally, safety is a significant factor hindering the widespread use of power lithium-ion battery packs. How to ensure safe and reliable operation during battery use is another critical issue that needs to be resolved.
Under such application demands, the battery management system has become increasingly important. By establishing an intelligent management system, the operating status of the battery pack can be effectively monitored, the SOC can be accurately measured, charge-discharge protection can be implemented, and various equalization methods can be adopted to achieve charge-discharge balance among cells, ensuring optimal battery operation and maximizing the service life of the battery pack.
1.2 Overview of Power Batteries
1.2.1 Development Status of Power Batteries
Power batteries generally refer to batteries with high rated capacity and output power that can serve as drive power sources for electric tools and electric vehicles. As the “heart” of electric vehicles, power batteries have been researched and developed worldwide for nearly two decades. The high-performance power batteries currently used in the power field mainly include valve-regulated lead-acid batteries, nickel-metal hydride batteries, lithium-ion batteries, and fuel cells.
Valve-regulated lead-acid batteries are widely used due to their reliable performance, mature production processes, and low prices, making them the most commonly used battery type in electric vehicles. However, they have critical drawbacks: low gravimetric energy density and volumetric energy density, resulting in heavy and bulky battery packs with limited energy. Additionally, lead-acid batteries have poor deep discharge capability and low-temperature discharge performance, cannot be rapidly charged, and cause secondary pollution when discarded. Therefore, in the long term, they may be replaced by other superior batteries.
Ni-MH batteries offer higher specific energy and specific power, stable discharge at high currents, good low-temperature discharge performance, long cycle life, safety, reliability, maintenance-free operation, no memory effect, and no environmental pollution with recyclability, aligning with the concept of sustainable development. They are currently the primary power batteries for HEVs. However, the price of Ni-MH batteries is significantly higher than lead-acid batteries, limiting their widespread adoption.
Lithium-ion batteries emerged almost simultaneously with Ni-MH batteries, using the lightest metal material—lithium—with an output voltage of 3.6V to 3.7V. They possess nearly all the advantages of batteries: high specific energy, high specific power, low self-discharge, no memory effect, good cycle characteristics, fast discharge capability, high efficiency, wide operating temperature range, and no environmental pollution. If safety issues can be resolved, lithium-ion batteries will become the preferred power source for the 21st century.
Fuel cells are electrochemical devices that convert the chemical energy of hydrogen and oxygen into electrical energy, producing no pollution or harmful emissions. Fuel cells operate quietly, with noise levels around 55dB, and achieve power generation efficiency exceeding 50%. However, because hydrogen is highly flammable and explosive, safely delivering hydrogen fuel remains a major obstacle to commercial fuel cell applications.
With the global energy shortage and increasing vehicle emission regulations, new power energy sources represent an inevitable development trend. Novel battery-powered vehicles are a major trend, but due to technological limitations, various power batteries currently cannot achieve perfection, and certain problems always exist. Therefore, future power batteries will be a key research direction for automobile manufacturers, with enormous market prospects in the global automotive market.
1.2.2 Characteristics of Power Lithium-ion Batteries
Lithium-ion batteries are improved products based on lithium batteries. Research on lithium batteries using metallic lithium as electrodes began as early as 1912. By the 1970s, non-rechargeable lithium batteries were first applied in commercial applications. However, lithium is a highly reactive metal that is unsafe, often experiencing combustion and rupture during charging. This led to the development of improved lithium-ion batteries, which incorporate components (such as cobalt and manganese) that suppress lithium reactivity, achieving genuine safety, efficiency, and convenience. In 1991, Sony Corporation achieved the first commercialization of lithium-ion batteries, and since then, they have developed rapidly with increasingly broad applications.
The rapid development and widespread application of lithium-ion batteries are closely related to their inherent characteristics. Compared with Ni-Cd, Ni-MH, and lead-acid batteries, lithium-ion batteries possess unique advantages:
| Characteristic | Lithium-ion | Ni-Cd | Ni-MH | Lead-Acid |
|---|---|---|---|---|
| Cell voltage (V) | 3.6 – 3.7 | 1.2 | 1.2 | 2.0 |
| Energy density (Wh/kg) | 180 | 50 | 70 – 80 | 30 – 40 |
| Cycle life (cycles) | >1000 | 500 | 500 | 300 – 500 |
| Self-discharge per month | 3% – 9% | 25% – 30% | 30% – 35% | 5% – 15% |
| Memory effect | None | Yes | Minimal | None |
| Environmental impact | Green | Toxic Cd | Minimal | Toxic Pb |
These characteristics make lithium-ion batteries particularly suitable for high-energy and high-power applications, especially in EV battery pack scenarios where power density and energy density are critical parameters affecting vehicle range and performance.
1.3 Overview of Power Lithium-ion Battery Energy Management Systems
1.3.1 Research Status of Power Battery Management Systems
Small and medium-sized power lithium-ion battery packs have been widely used in portable mobile devices such as mobile phones, notebook computers, video cameras, and digital cameras. The requirements for battery management technology in these mobile and handheld applications are continually increasing. Batteries must be small, lightweight, portable, safe, reliable, fast-charging, and equipped with accurate power indication functions so that users can monitor remaining battery power at any time and charge their devices promptly. Because these devices use relatively small-capacity lithium-ion batteries with fewer series-parallel cells (typically 3 to 14 cells), the battery management technology employed is relatively simple. Among these, dedicated chip-based lithium-ion battery energy management systems are widely used. These dedicated management ICs require minimal peripheral circuitry to achieve battery pack protection, charge-discharge management, and online remaining capacity estimation.
Currently, major international companies such as TI, Microchip, and Intersil have developed their own lithium-ion battery management ICs. Among them, TI’s BQ series management ICs have been developed earlier and have mature functionality, operating with microcontrollers to achieve management system functions, effectively realizing battery measurement, calculation, and protection. For instance, TI’s latest multi-cell battery management IC BQ20Z90 integrates advanced battery authentication technology and works collaboratively with TI’s latest analog front-end protection IC, suitable for 7.2V, 10.8V, or 14.4V battery packs. It can closely monitor battery capacity, impedance, charge balance, open-circuit voltage, and other critical parameters, and report information to the host system controller through serial communication buses, while driving 3, 4, or 5-segment LED displays to show remaining battery capacity. Japanese lithium-ion battery management chips, particularly protection ICs, feature comprehensive functionality and low power consumption, dominating the global market, such as Seiko’s S82XX series and RICOH’s R54XX series. Domestically, only a few companies in Taiwan and mainland China have developed relatively simple lithium battery protection chips. Research on management ICs integrating battery power measurement, protection, and equalization functions is still in its infancy, and no circuit with independent intellectual property rights has emerged domestically.
Large-scale power lithium-ion battery packs have become core components of various new energy electric vehicles. Many well-known automobile manufacturers are dedicated to developing power battery electric vehicles, including Ford, Chrysler, Toyota, Mitsubishi, Nissan, Hyundai, Courreges, and Ventury. Domestic manufacturers such as BYD, Geely, Chery, Lifan, and ZTE have also equipped their hybrid and pure electric vehicles with power lithium batteries. Although China’s new energy vehicle battery industry started later than Europe, America, and Japan, it has developed rapidly with substantial investment in lithium battery research and development. However, the current bottleneck hindering electric vehicle development lies in the safety performance of power lithium-ion batteries and the energy management system for vehicle power batteries. Electric vehicles use high-capacity power lithium-ion battery packs with demanding performance requirements. In such applications, dedicated battery management chips cannot meet the requirements. Instead, an energy management system centered on a microcontroller or DSP controller must be designed. This system collects battery data—including battery voltage, current, and temperature—through the controller, judges battery status based on these data, and implements protection functions. Through software programming based on specific algorithms, the remaining capacity is calculated, aiming to achieve the most effective use of battery capacity and extend battery service life. To this end, extensive and in-depth research on power battery energy management systems has been conducted both domestically and internationally.
1.3.2 Functions of Power Lithium-ion Battery Energy Management Systems
The battery management system primarily ensures the safety and reliability of power lithium-ion battery modules, displays operating status data in real time, and improves battery utilization efficiency, maximizing the battery pack’s service life. The main management functions encompass four aspects:
(1) Safety Protection Function: In idle, charging, and discharging states, the battery modules are monitored in real time. Once battery voltage, current, or temperature exceeds set limits, the management system immediately shuts down the battery module output and warns the user. In idle state, the main checks include excessive battery temperature and circuit shorts (voltage drops to 0V during short circuits). In operating states, protection functions include overcharge, overdischarge, charging over-temperature, discharging over-temperature, discharge overcurrent, and short circuit protection.
(2) Real-time Monitoring of Battery Pack Operating Status: The system continuously monitors and records performance parameters such as battery voltage, charge-discharge current, temperature, and internal resistance. Based on these parameters, it accurately estimates the SOC of the power battery pack, allowing users to understand the battery module’s operating status at any time.
(3) Battery Equalization Function: Differences between cells affect the battery pack’s service life and capacity utilization, requiring measures to achieve balance. The system typically employs charge equalization functionality. A bypass shunt circuit is designed to ensure every cell can be fully charged, slowing the aging process and extending battery service life.
(4) Battery Temperature Control Function: During charging and discharging, battery temperature generally rises and may reach safety limits. To ensure continuous battery module operation, installing and controlling cooling devices is necessary. Multi-level forced-air cooling can effectively reduce battery temperature, conserve energy, and benefit battery performance.
Current power lithium-ion battery pack management technology still has shortcomings in achieving management requirements, mainly in two aspects:
(1) The accuracy of data acquisition in battery monitoring is insufficient, primarily regarding the battery pack’s voltage and current measurement accuracy. These two parameters directly affect remaining battery capacity estimation. Therefore, it is necessary to improve voltage measurement accuracy and reduce current measurement errors as much as possible to enhance SOC estimation accuracy.
(2) Increasingly more applications demand higher battery voltage and output power. The use of series-parallel lithium-ion battery packs is growing. Due to differences between cells, voltage and capacity inconsistencies inevitably arise during use. Therefore, equalization devices must be added to the battery pack to ensure uniform charging and discharging among cells.
Because of these issues, the power performance of power lithium-ion battery packs cannot be fully realized, and battery life continues to decline. Therefore, a more comprehensive battery management system is essential to ensure safety during battery pack use, address equalization issues, maximize battery pack power performance, and extend battery service life.
1.4 Main Research Content
This thesis proposes a design method for an energy management system for power lithium-ion battery packs. The management system accurately estimates the SOC by detecting operating parameters such as voltage, charge-discharge current, and temperature in real time. An equalization protection system is designed to effectively resolve the overcharge and overdischarge issues of individual cells during charge-discharge processes, achieving charge equalization and thereby extending the service life of the power battery pack to achieve reasonable energy utilization. The main research content includes the following aspects:
(1) By analyzing the characteristics of several major power battery types, the advantages of power lithium-ion batteries are summarized. Knowledge related to power lithium-ion battery pack energy management systems is studied in depth, including specific functions and current implementation approaches.
(2) The definitions of remaining capacity and SOC for power lithium-ion batteries are analyzed, along with various factors affecting remaining battery capacity. Based on a comparison of major SOC estimation methods, the SOC estimation approach adopted in this system is further studied.
(3) A control circuit centered on the PIC18F452 microcontroller is designed to achieve real-time data acquisition of battery pack voltage, current, and temperature, along with SOC estimation. Protection and equalization circuits are designed for the lithium-ion battery pack to ensure system safety and reliability.
(4) Modular software design of the management system is conducted, completing functions such as data acquisition, SOC algorithm design, and display.
(5) Charge-discharge experiments are performed on the power lithium-ion battery pack to verify the correctness and feasibility of the SOC estimation method.
Chapter 2 SOC Estimation Methods for Power Lithium-ion Battery Remaining Capacity
2.1 Introduction
Power lithium-ion battery packs composed of multiple cells in series-parallel connections must be effectively managed during use to ensure battery safety and reliability, maximize power performance, and extend service life. The energy management system is centered on accurate SOC estimation. Many SOC estimation methods exist, among which open-circuit voltage method and Ah counting method are the two most widely used strategies. However, when used independently, they have certain limitations. To improve SOC estimation accuracy, these two methods are often combined, yet factors affecting remaining battery capacity remain insufficiently considered. Therefore, this thesis fully considers battery operating characteristics under various conditions, adopting an SOC estimation method primarily combining Ah counting with open-circuit voltage, while comprehensively incorporating compensation measures for charge-discharge efficiency, cycle life, and temperature.
2.2 Definitions of Remaining Capacity and SOC for Power Lithium-ion Batteries
SOC is a critical parameter describing battery status, reflecting the remaining capacity state of the battery. SOC is influenced by multiple factors and cannot be directly obtained from the battery but must be indirectly inferred through parameters such as voltage U, current I, internal resistance R, and temperature T. The relationship between these parameters and SOC changes with the battery aging process, containing many uncertainties.
Typically, SOC is defined from the perspective of electric quantity. The state of charge is the ratio of remaining capacity to battery capacity:
$$SOC = \frac{Q_c}{C_I} \tag{2-1}$$
Alternatively, based on the discharged capacity Q:
$$SOC = 1 – \frac{Q}{C_I} \tag{2-2}$$
In these formulations, SOC=1 indicates a fully charged battery, while SOC=0 indicates that the battery has been discharged to the cutoff voltage.
This SOC definition is commonly adopted internationally. However, it only describes the battery SOC under constant current discharge. Under variable current conditions, contradictions may arise. For example, discharging to the cutoff voltage at a large constant current would yield SOC=0 by definition, but continuing to discharge at a small current would release additional charge, implying SOC should be less than 0, which is unreasonable. This arises because the dischargeable capacity differs at different currents, and recoverable factors such as current and temperature are not considered. Therefore, this definition must be supplemented and corrected for variable current discharge applications.
2.3 Factors Affecting the Remaining Capacity of Lithium-ion Batteries
The discharge process of lithium-ion batteries is a complex electrochemical process. The remaining capacity is affected by multiple factors including battery temperature, discharge rate, self-discharge, and charge-discharge cycle count, making remaining capacity estimation challenging. The main factors are analyzed below:
(1) Discharge Current: The discharge capacity of a battery varies at different discharge currents. Higher discharge currents lead to lower discharge capacity, and vice versa. This is because the efficiency of converting chemical energy to electrical energy decreases at higher currents. Therefore, when discussing battery capacity, the discharge rate must be specified.
(2) Temperature: Temperature significantly affects capacity. At higher temperatures, the internal chemical active materials become more active, allowing reactions to proceed more fully with more chemical energy converted to electrical energy, resulting in increased total battery capacity. Conversely, at low temperatures or high discharge currents, the capacity provided by lithium batteries decreases significantly.
(3) Self-discharge Rate: This is the rate at which a battery loses capacity during storage, expressed as the percentage of capacity lost per unit time relative to the stored capacity. Higher ambient temperatures accelerate self-discharge. For different battery types, self-discharge rates vary. Lithium batteries exhibit self-discharge rates of 2% to 5% per month, significantly lower than Ni-Cd (25% to 30%) and Ni-MH (30% to 35%).
(4) Operating Life: Battery capacity decreases with increasing charge-discharge cycles, quantified as operating life—the number of cycles before capacity drops to 80% of the initial capacity. Typical lithium batteries have 300 to 500 cycles. Lithium battery life is also influenced by time; capacity begins to decline after leaving the factory regardless of usage.
Because battery capacity is affected by these uncertain factors during operation, accurately estimating SOC based on measurable battery parameters is a crucial challenge in energy management system design.
2.4 Common SOC Estimation Methods
(1) Discharge Test Method: This is the most direct and reliable SOC estimation method. A constant small current is used for continuous discharge, during which the battery can fully release stored energy. The product of discharge current and time equals the remaining capacity. The discharge test method is commonly used in laboratories and is applicable to almost all batteries. However, it has two significant drawbacks: it takes considerable time, and the battery must be disconnected from the system during the test. Therefore, it is not suitable for online battery capacity estimation and is typically used for battery maintenance.
(2) Ah Counting Method: The Ah counting method is the most commonly used SOC estimation method. It measures the charging and discharging current of the battery, integrates the product of current and time to calculate the accumulated charge, and estimates SOC accordingly:
$$SOC = SOC_0 – \frac{\eta}{C_N} \int_0^t i(t)dt \tag{2-3}$$
This method is simple to implement, minimally constrained by the battery itself, and does not require studying complex electrochemical reactions. The battery is viewed as a closed system with focus on its external characteristics. However, it requires high current measurement accuracy—measurement errors accumulate over time—and cannot estimate the initial SOC by itself, needing sufficient initial state data.
(3) Open-Circuit Voltage Method: This method obtains the SOC of a battery by measuring the open-circuit voltage and using a pre-established relationship curve between OCV and SOC. It is simple and works well in the early and late stages of charging. However, establishing the correspondence requires extensive charge-discharge experiments, and this method is affected by factors including battery state (charge vs. discharge), temperature, and aging. In addition, the open-circuit voltage is a static voltage without load and is not suitable for dynamic SOC measurement.
(4) Battery Internal Resistance Method: This method estimates SOC by measuring battery internal resistance based on the relationship between resistance and SOC. The resistance changes significantly during early discharge, making measurement difficult and reducing accuracy. However, this method demonstrates higher accuracy in the late discharge stage. It is rarely used alone and is typically combined with the Ah counting method.
(5) Kalman Filtering Method: This is a recently developed SOC estimation method. Based on the Ah counting method, it predicts initial battery capacity from open-circuit voltage, treats other factors affecting battery capacity as system noise, and uses the Kalman filtering minimum variance estimation optimization recursive algorithm for real-time filtering and estimation of SOC in the presence of measurement noise. It also provides the estimation error and is suitable for SOC estimation under highly variable current conditions. However, it requires high computational capability and does not comprehensively consider temperature and self-discharge factors.
Based on the above analysis, this thesis adopts a method that combines Ah counting and open-circuit voltage methods with modifications to address their limitations. The open-circuit voltage method is used to obtain the initial SOC when the battery is in a stable state, while the Ah counting method handles dynamic SOC calculation with compensation factors for temperature, cycle life, and charge-discharge efficiency.
2.5 SOC Estimation Method in this System
2.5.1 Determination of the SOC Estimation Method
From the preceding introduction, it is evident that various SOC estimation methods exist, each with its own characteristics suitable for different operating environments. Among them, the Ah counting method is the most widely used, capable of accurately estimating SOC when charge-discharge current is measured precisely and sufficient starting points exist for table lookup. The open-circuit voltage method provides the initial battery state point. Therefore, these two algorithms are often combined. However, open-circuit voltage measurement is affected by the self-recovery effect: after the load is disconnected, the working voltage gradually converts to open-circuit voltage, requiring time to stabilize. The time to reach equilibrium depends on the discharge current and battery characteristics.
The Ah counting method requires a fixed starting SOC state, but power batteries may begin operating at any SOC. Therefore, combining these two methods simply does not adapt to all operating environments. To accurately obtain SOC under all conditions, the self-recovery effect and temperature factors must be considered. Through the above analysis, this thesis adopts an SOC estimation approach based on the combination of the Ah counting method and open-circuit voltage method, with comprehensive compensation measures for charge-discharge efficiency, cycle life, and temperature.
The SOC estimation value is divided into two parts. The first part is the initial SOC value, determined by the battery’s starting state. When the battery is in a stable open-circuit state, the OCV-SOC curve provides the initial value. Otherwise, the SOC value at the end of the last charge-discharge cycle is adopted. The second part is the working SOC value, calculated by integrating the charge-discharge capacity using the Ah counting method. The final SOC is obtained by adding both parts.
2.5.2 Mathematical Model of SOC Calculation
Based on the preceding analysis, the SOC mathematical model is established:
$$Q(t) = \int_0^t \eta \cdot i(t) \cdot dt \tag{2-4}$$
$$SOC = K_0 \cdot SOC_0 + (1 – K_0) \cdot SOC_0′ + K_t \cdot K_c \cdot \frac{Q(t)}{Q_N} \tag{2-5}$$
where:
| Symbol | Description |
|---|---|
| \(Q_N\) | Rated capacity of the battery |
| \(Q(t)\) | Charge-discharge capacity at time t |
| \(SOC_0\) | Initial SOC corresponding to open-circuit voltage |
| \(SOC_0’\) | Terminal SOC of the last charge-discharge cycle |
| \(K_t\) | Temperature correction factor for SOC₀ (Kt=1 at room temperature) |
| \(K_c\) | Cycle life correction factor for SOC |
| \(K_0\) | Constant indicating whether the self-recovery effect has ended (K₀=1 if stable, otherwise 0) |
| \(\eta\) | Charge-discharge rate correction factor |
| \(i(t)\) | Working current at time t (positive for charging, negative for discharging) |
When the battery is in a stable state, the open-circuit voltage method determines the initial value \(SOC_0\). This avoids the adverse effects of the self-recovery effect on the open-circuit voltage method. Practically, when the battery terminal voltage change rate falls below a certain threshold, the battery voltage is considered stable. The SOC corresponding to this open-circuit voltage serves as \(SOC_0\). If the battery is still in self-recovery, the initial SOC value is the terminal SOC of the last charge-discharge period, and \(K_0=0\).
For the Ah counting method, the charge-discharge capacity is obtained by integrating the sampled current. For constant current discharge, the calculation directly multiplies current by time. For variable current discharge, the current-time curve is divided into small constant-current intervals of duration \(\Delta t\), equal to the current sampling period. For the k-th segment:
$$Q_k = Q_{k-1} + \Delta Q_k \tag{2-6}$$
$$\Delta Q_k = i_k \cdot \Delta t \tag{2-7}$$
where \(i_k\) represents the voltage value at the k-th point.
Chapter 3 Hardware Design of the Energy Management System
3.1 Introduction
System hardware design is the foundation for achieving energy management of power lithium-ion battery packs. Accurate SOC estimation requires precise data acquisition, stable system power supply, and robust system anti-interference capability. In actual implementation, the hardware structure design of each part is determined according to practical requirements. The energy management system designed in this thesis is directly powered by the lithium-ion battery pack without additional power supplies.
3.2 Overall Structure of System Hardware
The battery management system uses the Microchip PIC18F452 microcontroller as the control core, comprising four main parts: current, voltage, and temperature sampling circuits; battery protection and equalization circuits; battery SOC estimation; function keys; and display circuits. The system structure block diagram is shown below.
The system functions include battery information acquisition (cell voltage, battery pack total voltage, temperature, charge-discharge current), battery power status display, battery SOC estimation and display, and battery charge-discharge protection and charge equalization.
3.3 Microcontroller Selection
The microcontroller is the core unit of the energy management system. Its main tasks include collecting operating state parameters of the power battery pack, processing various data, and implementing functions through functional modules. Because the collected data are numerous with high demands on data acquisition accuracy and speed, the controller must possess adequate data processing capability and speed.
The PIC18F452 features a reduced instruction set computing (RISC) architecture, Harvard bus memory structure, two-stage pipeline instruction structure, and single-cycle instructions. It offers advantages in power consumption, drive capability, and peripheral module design. The PIC18F452 provides 32K bytes of FLASH program memory and 256 bytes of EEPROM, making external program memory expansion unnecessary. It includes a 10-bit A/D converter, 17 interrupt sources, built-in watchdog timer, and low-voltage detection circuitry.
The main characteristics of the PIC18F452 are summarized below:
| Feature | Specification |
|---|---|
| Architecture | RISC, Harvard bus, two-stage pipeline |
| Program memory | 32 Kbytes FLASH |
| Data memory | 256 bytes EEPROM |
| Instruction execution | 0.1 μs at 10 MHz internal clock |
| A/D converter | 10-bit, 8 channels |
| Interrupt sources | 17 independent sources |
| I/O drive capability | 25 mA sink/source |
| Operating voltage | 2.0 – 5.5 V |
The microcontroller’s peripheral circuit uses RA0, RA1, RA2 as the total voltage, current, and cell voltage sampling signal inputs. Port B’s RB4-RB7 form the independent keyboard circuit. Port C’s RC3-RC5 connect to the MAX7219 LED driver, while RC6 serves as the temperature signal input. An 8 MHz external crystal provides the system clock, yielding a 0.5 μs instruction cycle.
3.4 System Power Supply
The PIC18F452 operates at 2.0-5.5V. Since the A/D converter reference voltage is the microcontroller’s supply voltage, a stable power source is crucial. The system uses the LM7805 three-terminal voltage regulator to convert the lithium-ion battery pack voltage to a stable +5V. The current sampling circuit uses LM324 amplifiers requiring ±12V supplies. A +12V supply is achieved with an LM7812 regulator, and the -12V supply is generated using an ICL7662 DC-DC converter.
3.5 Voltage Acquisition Circuit
3.5.1 Total Voltage Measurement
The design uses seven lithium-ion batteries in series, with single-cell operating voltage of 3.7V, charge cutoff voltage of 4.2V, and discharge cutoff voltage of 3V. The total voltage measurement range is 29.4V to 21V. The total voltage must be converted to a 0-5V signal before entering the A/D converter. A simple resistor divider using two 0.1% precision resistors is employed for this purpose.
3.5.2 Individual Cell Voltage Measurement
Individual cell terminal voltage is an important parameter for battery state assessment and SOC estimation. This system uses an amplifier LM358 and a P-channel MOSFET F9630 to form the cell sampling circuit. The circuit operates by comparing the amplifier’s inverting and non-inverting inputs. At startup, the inverting input voltage exceeds the non-inverting input, causing the amplifier to output low, which turns on the MOSFET. The voltage divider action lowers the inverting input voltage until equilibrium is reached, at which point the voltages become equal with the cell terminal voltage. The resistor R3 voltage then represents the cell terminal voltage, which is sampled by the microcontroller A/D.
3.6 Current Acquisition Circuit
Current signals must be converted to voltage signals for measurement. The system uses a direct method employing a precision sampling resistor Rs connected in series with the battery main circuit. A 1Ω/5W resistor was selected to balance sampling accuracy and power loss. Since the microcontroller A/D can only accept positive voltages, a precision absolute value circuit is used to convert both positive and negative signals to positive signals. The current processing circuit amplifies the measured current by a factor of 5 before A/D conversion. This circuit works as follows:
When input voltage \(V_I < 0\), the output \(V_0\) is given by:
$$V_0 = -\frac{R_3}{R_7} V_I \tag{3-1}$$
When input voltage \(V_I > 0\):
$$V_0 = -\left(\frac{R_3}{R_2}V_I + \frac{R_3}{R_5}V_a\right) = -\frac{R_3}{R_5}V_I – \frac{R_3}{R_2}(-V_I) = 5|V_I| \tag{3-2}$$
Thus, regardless of polarity, the output voltage \(V_0\) is always positive and proportional to the absolute value of the input, with a gain of 5.
3.7 Temperature Detection Circuit
Battery pack temperature is essential for SOC estimation and determining safe operating conditions. The system employs the Dallas DS18B20 single-bus digital temperature sensor. The DS18B20 features a measurement range of -55°C to +125°C with an accuracy of ±0.5°C in the range of -10°C to +85°C. Its key characteristics include:
| Feature | Specification |
|---|---|
| Operating voltage | 3.0 – 5.5 V |
| Temperature range | -55°C to +125°C |
| Accuracy | ±0.5°C from -10°C to +85°C |
| Resolution | 9 to 12 bits (0.5°C to 0.0625°C) |
| Communication | Single-wire bus, supports multi-drop |
| Conversion time | 93.75 ms (9-bit) to 750 ms (12-bit) |
The external power supply mode is adopted for stable operation. A 4.7kΩ pull-up resistor connects the DQ line to +5V, and the DQ output connects to RC6 of the microcontroller.
3.8 Display Circuit
3.8.1 LED Digital Display Circuit
To provide user-friendly real-time monitoring, the system uses MAX7219 serially interfaced LED display drivers. The MAX7219 can directly drive eight 7-segment common-cathode digital tubes, containing an 8×8 static RAM, BCD decoder, multiplex scan circuitry, and segment/bit drivers. It supports digital brightness control, shutdown mode, and test mode. The display interface connects CLK, LOAD, and DIN pins to RC3, RC4, and RC5 of the microcontroller respectively. The MAX7219 communicates using a 16-bit serial frame: 8 bits of data and 4 bits of register address, transmitted MSB first with the data latched on the LOAD rising edge.
The MAX7219 control registers include decode mode, intensity, scan limit, shutdown, display test, and no-op registers. This system uses four 7-segment displays to show voltage, current, temperature, and SOC values.
3.8.2 LED Capacity Display Circuit
An auxiliary LED-based capacity display divides the battery pack capacity into 4 segments of 25% each, corresponding to voltage thresholds of 12.26V, 14.44V, 15.15V, and 15.92V. The circuit employs six LEDs: DS6 (yellow) indicates low capacity, DS5 (red) indicates circuit operation, and DS1-DS4 (green) indicate capacity levels. A TL431-based reference circuit provides voltage comparison. A push-button switch with a delay capacitor C1 maintains the display for approximately 2 seconds after release, reducing continuous power consumption.
3.9 Keyboard Circuit
The system uses independent key switches with each key connected to a separate I/O line. Four keys K1-K3 and a reset key are configured on RB4-RB7. External 10kΩ pull-up resistors ensure deterministic high levels when keys are released. The key functions are:
| Key | Function |
|---|---|
| K1 | Display battery voltage |
| K2 | Display battery charge-discharge current |
| K3 | Display battery temperature |
| Default | Display SOC value |
The system uses the software delay method to eliminate key bounce: upon detecting a key press, a 10ms delay is executed before confirming the key state.
3.10 Lithium Battery Protection Circuit Design
3.10.1 Study of Lithium Battery Protection
Lithium-ion batteries present high specific energy, which inherently carries high risk. Safety is the most critical issue in lithium-ion battery pack applications. Overcharging causes temperature rise, electrolyte decomposition, gas generation, and potentially fire or rupture. Overdischarging degrades battery characteristics and shortens cycle life. Because both overcharge and overdischarge cause irreversible damage, the protection circuit must provide overcharge protection, overdischarge protection, and overcurrent/short-circuit protection. This system’s protection circuit consists of protection ICs and two Power-MOSFETs, where the protection IC monitors cell voltages and the MOSFETs switch off the charge/discharge circuits when abnormal conditions are detected.
The protection IC selected for this design is the Seiko S-8261, a single-cell lithium-ion battery protection IC with built-in high-precision voltage detection and delay circuits. The S-8261 includes precision voltage detectors, logic control circuitry, overcurrent detection, delay circuits, 0V battery charging support, and MOSFET drive circuits.
3.10.2 Protection Circuit Design
The protection circuit uses four S-8261 ICs to protect four series-connected lithium-ion cells. The circuit employs PC817 optocouplers for output isolation because the protection ICs are connected in cascade without common ground. The first column of optocouplers is connected in series for discharge control, while the second column is connected in parallel for charge control.
The technical specifications of the protection board are listed below:
| Parameter | Value |
|---|---|
| Overcharge protection voltage | 4.28V ± 0.025V |
| Overcharge protection delay | <1.3s |
| Overdischarge protection voltage | 2.3V ± 0.050V |
| Overdischarge protection delay | <150ms |
| Charging voltage | <17V |
| Charging current | <2A |
| Discharging current | <2A |
| Overcurrent protection | 5A ± 10% |
| Overcurrent protection delay | <10ms |
| Short-circuit protection delay | <3ms |
| Static power consumption | <200μA |
| Under-voltage power consumption | <20μA |
The protection board functions include overcharge upper-limit protection, overdischarge lower-limit protection, lower-limit shutdown locking (disconnecting the load releases the lock), charge equalization management, overcurrent protection, and short-circuit protection.
3.10.3 Protection Circuit Operating Principle
(1) Overcharge Protection: In normal charging, the CO pins of all four S-8261 ICs output high voltage, keeping the second column of optocouplers off. The battery voltage is directly applied to the gate of the charging MOSFET Q5, keeping it on. When any protection IC detects that the cell voltage has reached the upper limit of 4.28V, its CO pin transitions from high to low, turning on the optocouplers and pulling the gate of Q5 to ground, which turns Q5 off and cuts off the charge circuit. Due to Q5’s body diode, the battery can still discharge to the external load.
(2) Overdischarge Protection: During normal discharge, the DO pins output high, keeping the first optocoupler column conducting. When any protection IC detects the cell voltage falling below 2.3V, its DO pin transitions from high to low, turning off the optocouplers and cutting off the discharge MOSFET Q6. The battery can no longer discharge to the load. The protection circuit enters a low-power state consuming less than 20μA.
(3) Overcurrent and Short-circuit Protection: The fourth protection chip S4 monitors the voltage across the sampling resistor R23. When the circuit current causes U > 1.2V, it is judged as overcurrent or short-circuit, and the DO pin transitions to low, turning off Q6. The short-circuit protection delay time is extremely short, typically less than 320μs.
(4) Lower-limit Self-locking Protection: Due to the self-recovery effect, the battery voltage rebounds after low-voltage shutoff. When the voltage recovers above the overdischarge release threshold, the discharge MOSFET would normally turn on again, causing repeated on-off oscillations with excessive switching losses. The design uses an MC14011 NAND gate to implement a self-locking circuit. When overdischarge occurs, the DO pin goes low, the first optocoupler column turns off, causing the NAND gate U9A output low, turning off Q6. The NAND gate U9C maintains the low gate voltage regardless of subsequent DO pin states as long as the load remains connected, effectively preventing the switch from oscillating.
This protection circuit uses single-cell protection ICs with improved application circuits, where each IC monitors one cell. The signals are processed and compared, and only two MOSFETs are ultimately used for four series cells, reducing cost and static power consumption. The circuit’s modular structure allows expansion to 10 or more cells by adding identical cell circuits.
3.11 Lithium Battery Equalization Circuit Design
Series-connected battery packs inevitably exhibit inconsistency in capacity, terminal voltage, and internal resistance among individual cells due to manufacturing and usage variations. This inconsistency accelerates the capacity degradation of weaker cells, and because the series pack capacity is determined by the weakest cell, the battery pack service life is shortened. Charge equalization is critically important because, during series charging, all cells receive the same current. If the charge terminates based on the largest-capacity cell, smaller cells will be overcharged; conversely, if based on the smallest-capacity cell, larger cells will be undercharged. With effective charge equalization, discharge equalization becomes less critical because each cell can be fully charged before use.
Equalization methods are categorized into energy-consuming (dissipative) and non-energy-consuming (non-dissipative) types.
3.11.1 Energy-Consuming Equalization Circuits
The dissipative equalization approach connects a bypass resistor in parallel with each cell. During the float charge stage, the bypass of fully charged cells is switched on to shunt current. The excess energy is dissipated as heat through the resistor. This method is the simplest and lowest cost, but energy efficiency is low. Three main types exist:
(1) Protection IC overcharge signal direct drive: This approach uses the protection IC’s CO pin to directly drive a switching transistor. When the cell voltage reaches the overcharge threshold, the CO pin outputs low, turning on the switch and the bypass resistor, dissipating energy until the voltage drops below the release threshold.
(2) Protection IC built-in equalization: Some protection ICs such as Seiko S-8233A and Intersil X3100 include built-in equalization functions with external resistors and power transistors.
(3) TL431-based equalization: The TL431 programmable reference is configured to a threshold voltage (e.g., 4.2V). When the cell voltage exceeds the threshold, the TL431 conducts, and energy is dissipated through a power resistor. This method has limited equalization capability due to TL431’s drive capability.
3.11.2 Non-Energy-Consuming Equalization Circuits
(1) Capacitor Equalization Method: This energy-transfer method uses capacitors as temporary energy storage to transfer charge between adjacent cells. Bidirectional switches connect a capacitor alternately between adjacent cells, transferring energy from higher-voltage cells to lower-voltage cells. N cells require N switches and N-1 capacitors. Energy loss is minimal, but energy transfer is slow for cells far apart, making it less suitable for larger battery packs.
(2) Parallel DC-DC Module Equalization: This energy-conversion method connects a DC-DC converter module to each cell. The module transfers energy from higher-voltage cells to other cells, enabling rapid, low-loss equalization. However, the system is complex, requires transformers, and has high cost.
3.11.3 Charge Equalization Circuit Design
Based on the analysis above, this system adopts the simplest dissipative approach: using the S-8241 protection IC overcharge signal to directly drive the bypass circuit. The S-8241 has an overcharge detection voltage of 4.275V and a release voltage of 4.1V. The equalization circuit connects the CO pin to the gate of a P-MOSFET (SI4435) with a bypass resistor of 68Ω/2W. When the cell voltage reaches the overcharge threshold, the CO pin outputs low, turning on the P-MOSFET, allowing the bypass resistor to dissipate energy until the cell voltage falls to the release voltage. The maximum equalization current is:
$$I_{eq} = \frac{V_{eq}}{R} = \frac{4.275V}{68\Omega} \approx 64mA \tag{3-3}$$
$$P_{diss} = I_{eq}^2 \times R = (0.064A)^2 \times 68\Omega \approx 0.28W \tag{3-4}$$
Each cell in the battery pack is equipped with this identical equalization module. This circuit is simple, low cost, and easily implemented. The main disadvantage is the energy dissipated as heat, which requires thermal management considerations. The equalization capability is limited (typically 10-20mA with the S-8241), making it suitable for applications where equalization demand is not excessive.
Chapter 4 Software Design of the Energy Management System
4.1 Introduction
Chapter 3 presented the hardware architecture. This chapter details the software design, which is modular and written in assembly language for the PIC microcontroller. The software uses the PICMATE2004 emulator and its integrated development environment for compilation, linking, and debugging.
4.2 Overall System Structure
The system software flow chart comprises the main program and subroutines. The main program completes initialization and determines key states to invoke corresponding subroutines. Two timers generate interrupts: TMR0 for data acquisition and TMR1 for data processing, SOC calculation, and display updates.
The software is divided into the following independent modules:
| Module | Function |
|---|---|
| System initialization | I/O port configuration, timer initialization, A/D initialization, MAX7219 initialization |
| Data acquisition | Voltage, current (A/D conversion), temperature (DS18B20 single-bus communication) |
| Power management | SOC estimation using modified Ah counting + OCV method with compensation |
| Key and display | Key scanning, debouncing, and LED display update |
4.3 System Initialization Module
The initialization program configures the I/O ports, timers, A/D converter, and MAX7219. The MAX7219 initialization sets the scan limit (number of digits), decode mode, shutdown mode, and display intensity. This process runs once after power-on reset.
4.4 Data Acquisition Module
4.4.1 Voltage and Current Acquisition
The PIC18F452’s internal 10-bit A/D converter acquires voltage and current signals. The A/D reference voltage is the +5V supply. In the program, channel 0 is first selected for total voltage sampling, then channel 2 for current sampling. Each result is stored in two registers (high and low bytes). The conversion uses the query method, and to improve accuracy and stability, a moving average filter is applied: eight consecutive samples are averaged to obtain the effective value, suppressing ripple noise and A/D reading deviations.
4.4.2 Temperature Acquisition Module
The DS18B20 temperature sensor communicates via the single-bus protocol. The temperature acquisition sequence involves three main steps: initialization (reset pulse and presence pulse), ROM commands, and DS18B20 function commands. The software implements the required timing signals with precise delays:
- Reset pulse: output low for 490μs, then release and sample the presence pulse after 60μs
- Write 0: output low for 65μs, then release for 2μs
- Write 1: output low for 2μs, then release for 65μs
- Read: output low for 2μs, then switch to input mode, read after 12μs, and wait 50μs
The temperature conversion command is sent, followed by a 1-second delay for conversion completion, then the scratchpad is read to obtain the temperature value.
4.5 Power Management Program
The SOC estimation flow chart operates as follows. After initialization, the battery’s initial SOC is read. This value depends on the battery’s starting state—if stable, the open-circuit voltage corresponding SOC is used; otherwise, the terminal SOC of the previous cycle is used. During operation, the system samples the charge-discharge current and integrates it using the Ah counting method, performed through the microcontroller’s timer interrupt. The SOC is then calculated based on the estimation model, and results are displayed.
The SOC estimation equations implemented in the software are:
For charging:
$$Q(t) = \int_0^t \eta \cdot i(t) \cdot K_c \cdot K_t \cdot dt \tag{4-1}$$
For discharging:
$$Q(t) = \int_0^t \eta \cdot i(t) \cdot K_c \cdot K_t \cdot dt \tag{4-2}$$
$$SOC = K_0 \cdot SOC_0 + (1 – K_0) \cdot SOC_0′ + \frac{Q(t)}{Q_N} \tag{4-3}$$
4.6 Key and Display Modules
4.6.1 Key Processing Subroutine
The keyboard uses a program-controlled scanning method. The software continuously scans the key states. When a key press is detected, a 10ms debounce delay is executed, and the key state is rechecked. If the key is still pressed, it is considered a valid key press; otherwise, it is treated as noise or bouncing. A new scan cycle begins only after the key is released, preventing unintended repeated triggers.
The key functions are: K1 displays battery voltage, K2 displays current, K3 displays temperature. The default display is SOC. Each key press updates the display temporarily, returning to SOC display after a delay.
4.6.2 Display Module Subroutine
The MAX7219 communicates with the PIC18F452 over a 16-bit serial frame. The display data and address are transmitted over the DIN line on the rising edge of CLK, and latched on the rising edge of LOAD. The control registers are programmed separately from the display registers. The display subroutine first loads the address byte, then the data byte, toggling the clock appropriately to shift each bit.
For the brightness control, the MAX7219 intensity register allows adjustment from 1/32 to 31/32 duty cycle. This design sets the scan limit to 4 digits (for the four-digit display) and uses BCD decode mode for direct segment display.
Chapter 5 Experimental Results and Analysis
5.1 Experimental Setup
The battery pack used in this experiment consists of seven series-connected lithium-ion batteries manufactured by Zhejiang Dongsheng Battery Co. Ltd. The battery specifications are:
| Parameter | Value |
|---|---|
| Battery model | TSE-086760 |
| Standard capacity | 2Ah |
| Rated voltage | 25.9V |
| Maximum charging voltage | 29.4V |
| Discharge cutoff voltage | 21V |
| Charging method | CC/CV (constant current/constant voltage) |
| Maximum continuous charging current | 2A |
| Maximum continuous discharge current | 2A |
| Battery pack weight | 350g |
| Operating temperature range | Charge: 0°C to +45°C; Discharge: -20°C to +45°C |
The experimental system hardware includes the battery pack, the management system circuit board, and a 20Ω power resistor for load testing.
5.2 Experimental Content
5.2.1 Relationship between Open-Circuit Voltage and SOC
When the battery is in a stable state, the open-circuit voltage method can estimate SOC accurately, making it suitable for obtaining the initial SOC. A test was performed at 21-26°C by discharging the battery at a constant small current of 0.05C (100mA). The battery was discharged to specific voltage points, allowed to rest until the open-circuit voltage stabilized, and then discharge resumed. Ten data points were collected. The experimental relationship between OCV and SOC is shown below:
| Voltage (V) | SOC |
|---|---|
| 3.35 | 0.08 |
| 3.50 | 0.14 |
| 3.60 | 0.20 |
| 3.68 | 0.30 |
| 3.75 | 0.40 |
| 3.82 | 0.50 |
| 3.88 | 0.60 |
| 3.95 | 0.72 |
| 4.03 | 0.85 |
| 4.10 | 0.95 |
The OCV-SOC curve exhibits a characteristic sigmoidal shape. The relationship between OCV and SOC is approximately linear in the mid-range (SOC 20% to 80%), deviating at the extremes. This non-linear behavior is important for accurately determining initial SOC in the EV battery pack application context.
5.2.2 Charging Curve
For lithium-ion batteries, the constant-current constant-voltage (CC/CV) charging method is widely adopted. In this experiment, the pack was charged at 0.3C (600mA) during the constant-current stage until the pack voltage reached 29.4V, followed by constant-voltage charging until the current dropped to 50mA. The initial battery condition was a minimum cell voltage of 3.35V, corresponding to an initial SOC of 8% based on the OCV-SOC relationship.
The charging voltage curve shows that the charging process lasted approximately 4 hours. During the constant-current phase, both voltage and SOC increased at a steady rate. When the pack voltage reached the charging limit of 29.4V, the charger transitioned to constant-voltage mode, and the voltage curve flattened while the charging current decreased gradually. This two-stage approach protects the lithium-ion battery by preventing overvoltage and maintaining a saturated charge state, which is crucial for extending the EV battery pack lifespan.
The charging current and SOC curves illustrate the relationship between current decay and charge accumulation during the constant-voltage stage. The SOC curve maintains a rapid initial rise, then decelerates as the battery approaches full charge, with the charging current tapering off exponentially.
5.2.3 Battery Constant Current Discharge Curve
Discharge tests were conducted at 0.2C (400mA), 0.4C (800mA), and 0.5C (1000mA) discharge rates at an ambient temperature of 25°C. The discharge curves reveal that the discharge current size affects the discharge characteristics significantly:
- At 0.2C discharge to 21V: about 3.6 hours
- At 0.4C discharge to 21V: about 2.2 hours
- At 0.5C discharge to 21V: about 1.3 hours
The experimental results confirm that higher discharge currents cause more rapid voltage decay. After discharge termination, the battery terminal voltage rises gradually due to the self-recovery effect. Larger discharge currents lead to higher stabilized open-circuit voltages, indicating that the battery’s usable capacity is affected by discharge rate. The relationship between discharge capacity and discharge rate is:
| Discharge rate | Discharge capacity (mAh) |
|---|---|
| 0.2C (400mA) | 1420 |
| 0.4C (800mA) | 1380 |
| 0.5C (1000mA) | 1229 |
The data show that the discharge capacity decreases with increasing discharge current, which aligns with the theoretical prediction and demonstrates the need for discharge rate compensation in SOC estimation. For EV battery pack applications, this highlights the importance of accurately tracking the current-dependent capacity behavior.
5.2.4 Battery Variable Current Discharge Curve
Discharge tests were also conducted with a variable load using a 40Ω resistance. The discharge voltage curve shows a relatively flat initial decay region—the main working region—followed by a steeper decline in the later discharge stage. The current flowing through the load decreased slightly throughout the discharge cycle, from approximately 0.65A to 0.55A, which is nearly equivalent to constant current discharge. The SOC curve decreases approximately linearly, confirming that the battery’s discharge characteristics under stable resistive loads are consistent with constant current discharge behavior.
These results validate the effectiveness of the SOC estimation method proposed in this thesis. For constant and stable-resistive loads, the Ah counting method combined with OCV-based initial SOC estimation provides accurate SOC tracking. The compensation for discharge rate, temperature, and cycle life further improves accuracy under varying operating conditions, making the system suitable for EV battery pack energy management.
Conclusion
This thesis proposed an overall design scheme for a power lithium-ion battery pack energy management system, addressing the safety, consistency, and SOC estimation accuracy problems encountered in practical applications. The system was designed and implemented from both hardware and software perspectives. Based on the joint hardware-software debugging results, the following conclusions are drawn:
(1) The designed control system centered around the PIC18F452 microcontroller, with its built-in 10-bit A/D converter, streamlined instruction set, and low-power modes, significantly simplified the hardware circuit design and promoted system integration and modularization.
(2) Considering the characteristics of power lithium-ion batteries, the traditional Ah counting method and open-circuit voltage method were improved by incorporating comprehensive compensation measures for charge-discharge efficiency, cycle life, and temperature. Experimental results demonstrate that this algorithm effectively overcomes the limitations of conventional methods, thereby enhancing SOC prediction accuracy for EV battery pack applications.
(3) For the equalization and safety issues in series-connected lithium-ion battery packs, a protection and equalization circuit based on the Seiko S-8261 was designed. Hardware debugging results confirm that this design effectively resolves the overcharge, overdischarge, and overcurrent problems in series-connected battery packs while realizing charge equalization.
(4) Based on the hardware design, the software was developed with a modular approach, fully exploiting the microcontroller’s programmable intelligence. The system achieves accurate SOC estimation and display by precisely acquiring battery parameters and applying the designed algorithm.
Nevertheless, due to design experience limitations and time constraints, the system requires further improvements. First, regarding the microcontroller system power supply, direct power extraction from the lithium-ion battery pack presents a potential imbalance risk; therefore, the power supply design requires optimization. Second, regarding SOC estimation, the calculation results still contain some errors due to constraints in experimental conditions.
Future work will continue to optimize both hardware and software designs to achieve more reliable and accurate energy management for EV battery pack systems.