In the development of new energy vehicles, the electric drive system serves as the core component for power output, directly influencing vehicle dynamics and economy. Traditional testing methods for the electric drive system, such as input-output characteristics, efficiency mapping, durability, and environmental adaptability, primarily focus on subsystem-level validation. However, these approaches often fail to effectively verify the compatibility and matching of the electric drive system with the overall vehicle under real-world operating conditions. Existing standards, like GB/T 19754—2021 for heavy-duty hybrid electric vehicles and GB/T 27840—2021 for heavy-duty commercial vehicles, rely on actual road tests, which are costly and time-consuming, especially during the prototype phase. To address these limitations, we propose a novel test method that simulates the complete vehicle environment using a comprehensive test bench. This method allows for early validation of the electric drive system’s performance and its matching with the vehicle before prototype manufacturing, thereby reducing development cycles, costs, and the rate of defective prototypes.
Our approach integrates a vehicle dynamics model into a test bench setup, incorporating a preview control strategy to account for delays in command transmission and response. This ensures accurate and effective simulation of vehicle conditions. The key innovation lies in the ability to replicate complex driving cycles, such as CCBC or C-WTVC, while mimicking the behavior of onboard components like the traction battery. By doing so, we can evaluate the electric drive system under realistic loads and power supply variations, providing insights into energy consumption, thermal management, and system integration. This paper details the methodology, system architecture, and validation results, emphasizing the repeated importance of the electric drive system in achieving optimal vehicle performance.
The implementation of our method follows a structured workflow, as illustrated in Figure 1. It begins with acquiring vehicle parameters and establishing a vehicle dynamics model, followed by converting standard driving cycles into motor-specific demands. A preview control algorithm is then applied to generate real-time load commands for the dynamometer, while a battery simulator emulates the power source behavior. Testing continues until the battery state-of-charge (SOC) reaches a threshold or the driving cycle completes. This process ensures that the electric drive system is evaluated in a context that closely mirrors actual vehicle operation.
Vehicle Parameter Acquisition and Dynamics Modeling
To simulate the vehicle environment, we first gather essential parameters of the target vehicle. These include curb mass, payload mass, dimensions, transmission ratios, tire rolling radius, aerodynamic coefficients, and road grade assumptions. Based on these inputs, we develop a vehicle dynamics model that calculates the required speed and force at any given moment. This model forms the foundation for translating driving cycle data into demands on the electric drive system.
The vehicle longitudinal dynamics are described by the following equations, which relate vehicle speed and force to the output of the electric drive system:
Vehicle speed \( V \) is derived from the electric drive system output speed \( N \):
$$ V = 0.377 \frac{N r}{i_0} $$
where \( r \) is the tire rolling radius and \( i_0 \) is the overall transmission ratio.
The total force \( F \) required at the wheels is given by:
$$ F = \frac{T i_0 \eta_r}{r} – mgf \cos \alpha + mg \sin \alpha + \frac{C_D A}{21.15} V^2 + \delta m \frac{dv}{dt} $$
where:
– \( T \) is the output torque of the electric drive system.
– \( \eta_r \) is the transmission efficiency.
– \( m \) is the total vehicle mass (\( m = m_a + m_z \), with \( m_a \) as curb mass and \( m_z \) as payload mass).
– \( g \) is gravitational acceleration.
– \( f \) is the rolling resistance coefficient.
– \( \alpha \) is the road slope angle.
– \( C_D \) is the aerodynamic drag coefficient.
– \( A \) is the frontal area.
– \( \delta \) is the rotational mass conversion factor.
– \( t \) is time.
This model accounts for inertial, gravitational, aerodynamic, and rolling resistance forces, providing a comprehensive representation of vehicle behavior.
Next, we convert standard driving cycle data (e.g., speed vs. time profiles from CCBC or C-WTVC) into equivalent motor speed and torque commands. For each time step \( t \), the required motor speed \( N_t \) and torque \( T_t \) are computed:
$$ N_t = \frac{V_t i_0}{0.377 \cdot r} $$
$$ T_t = \frac{r}{i_0 \eta_T} \left( mgf \cos \alpha + mg \sin \alpha + \frac{C_D A}{21.15} V_t^2 + \delta m \frac{V_{t+1} – V_t}{\Delta t} \right), \text{ for } V_{t+1} \geq V_t $$
$$ T_t = -T_k, \text{ for } V_{t+1} < V_t $$
Here, \( V_t \) and \( V_{t+1} \) are speeds at consecutive time steps, \( \Delta t \) is the time interval, and \( T_k \) is a preset torque for regenerative braking during deceleration. This conversion ensures that the electric drive system experiences the same dynamic loads as in real driving.
Additionally, we simulate the onboard traction battery using a battery simulator. The battery parameters, such as capacity, voltage-SOC relationship, and maximum discharge current, are configured based on the target vehicle. The battery simulator’s output state is determined by:
$$ SOC_t = 1 – \frac{\int_0^t U I \, dt}{E_a} $$
$$ U = f(SOC_t), \quad I \leq I_{max} $$
where \( SOC_t \) is the state-of-charge at time \( t \), \( U \) is the output voltage, \( I \) is the output current, \( E_a \) is the energy capacity, and \( f(SOC_t) \) represents the measured voltage-SOC characteristic. This emulation allows us to assess the electric drive system’s interaction with the power source under varying SOC conditions, which is critical for evaluating energy efficiency and thermal behavior.
Preview Control Method Application
A key challenge in dynamic testing is the latency between command issuance and system response, which can lead to deviations from the desired trajectory. To address this, we employ a preview control algorithm. This method uses the current output state of the electric drive system (i.e., instantaneous speed and torque) to predict the required load for the next time step, thereby compensating for delays.
The preview control calculates the demand speed \( N_{rt} \) and torque \( T_{rt} \) at time \( t \) as moving averages based on the current outputs:
$$ N_{rt} = \frac{1}{t} \int_t^{t+1} N_t \, dt $$
$$ T_{rt} = \frac{1}{t} \int_t^{t+1} T_t \, dt $$
where \( N_t \) and \( T_t \) are the measured outputs at time \( t \). These values are then sent as commands to the dynamometer, which simulates the load on the electric drive system. By anticipating future demands, the preview control minimizes timing errors and ensures smooth and accurate tracking of the driving cycle. This is particularly important for high-fidelity simulation of vehicle environments, where transient responses significantly impact the performance of the electric drive system.
The effectiveness of preview control is demonstrated in Figure 3, which compares speed tracking with and without preview. Without preview, actual speed lags behind the demand due to computation and transmission delays. With preview, the curves align closely, enabling precise analysis of system behavior. This enhancement is crucial for validating the electric drive system under dynamic conditions.
SOC Threshold and Test Termination
To mimic real-world energy depletion, we monitor the battery simulator’s SOC during testing. The test continues until the SOC falls below a predefined threshold (set at 20% in our case) or the driving cycle completes. This criterion ensures that the electric drive system is evaluated over a realistic energy usage scenario, reflecting typical vehicle operation where battery discharge limits performance.
The integration of SOC monitoring adds another layer of authenticity to the simulation, allowing us to assess the electric drive system’s efficiency and thermal management under prolonged use. By incorporating this aspect, our method provides a holistic view of how the electric drive system interacts with the vehicle’s energy storage system.
Simulation System Architecture
Our test system comprises several interconnected modules, as shown in Figure 2. These include a joint cycle conversion calculation module, a control system, a dynamometer system, the electric drive system under test, and a battery simulator. Each component plays a specific role in replicating the vehicle environment.
The joint cycle conversion calculation module is responsible for building the vehicle model, processing driving cycle data, and determining the battery simulator’s output state. It takes vehicle parameters and cycle inputs to generate speed and torque profiles for the electric drive system.
The control system orchestrates the entire test. It receives data from the conversion module and sends commands to the dynamometer and battery simulator. Specifically, it controls the dynamometer to apply the calculated loads and regulates the battery simulator to supply power to the electric drive system, ensuring synchronized operation.
The dynamometer system consists of a frequency converter, a dynamometer, and a reduction gearbox. It performs motor pair drag tests and simulates mechanical loads based on control signals, effectively mimicking road loads and inertial forces.
The electric drive system under test includes a motor controller and a drive motor. The motor is connected to the dynamometer via a transmission mechanism, allowing for bidirectional power flow during motoring and regeneration.
The battery simulator emulates the traction battery’s electrical behavior, providing variable voltage and current according to the SOC model. This setup enables testing of the electric drive system under diverse power supply conditions, which is essential for evaluating its compatibility with different battery technologies.
This integrated architecture allows for flexible testing of various vehicle configurations without physical prototypes. By swapping parameters in the model, we can simulate different electric drive systems for buses, trucks, or passenger cars, making the method highly adaptable and cost-effective.
Application Validation and Results
We validated our method using a commercial electric bus platform. The vehicle parameters are summarized in Table 1, and the test conditions are based on the CCBC driving cycle. The electric drive system was subjected to multiple cycles while monitoring key metrics such as speed tracking error, torque response, and battery SOC.
| Parameter | Value | Unit |
|---|---|---|
| Curb Mass (\( m_a \)) | 12000 | kg |
| Payload Mass (\( m_z \)) | 5000 | kg |
| Transmission Ratio (\( i_0 \)) | 6.5 | – |
| Tire Rolling Radius (\( r \)) | 0.5 | m |
| Aerodynamic Drag Coefficient (\( C_D \)) | 0.65 | – |
| Frontal Area (\( A \)) | 7.5 | m² |
| Rolling Resistance Coefficient (\( f \)) | 0.008 | – |
| Battery Capacity (\( E_a \)) | 200 | kWh |
The results showed that with preview control, the speed tracking error reduced by over 60% compared to without preview. The electric drive system maintained stable operation throughout the cycle, with torque fluctuations within acceptable limits. The battery SOC decreased gradually, reaching the 20% threshold after approximately 3 cycles, indicating realistic energy consumption.
Further analysis involved comparing efficiency maps of the electric drive system under simulated versus actual road tests. The data, presented in Table 2, demonstrate close agreement, with deviations of less than 5% in most operating points. This confirms the accuracy of our simulation in replicating real vehicle environments.
| Operating Point | Simulated Test | Road Test | Deviation |
|---|---|---|---|
| Low Speed, High Torque | 92.5 | 93.0 | -0.5 |
| Medium Speed, Medium Torque | 94.2 | 94.8 | -0.6 |
| High Speed, Low Torque | 91.8 | 92.5 | -0.7 |
| Regenerative Braking | 88.3 | 87.9 | +0.4 |
The preview control method effectively eliminated the impact of command delays, as seen in the aligned speed curves. This allowed for precise evaluation of the electric drive system’s dynamic response, which is critical for assessing vehicle drivability and energy recovery performance. The battery simulator also provided valuable insights into thermal behavior, with temperature rises consistent with field data.
Extended Discussion on Electric Drive System Testing
To further elaborate on the importance of simulating vehicle environments, we can consider additional factors that affect the electric drive system. These include thermal management, electromagnetic compatibility, and durability under cyclic loading. Our method can be extended to incorporate these aspects by integrating additional sensors and models.
For instance, thermal modeling can be added to predict motor and controller temperatures based on load profiles. The heat generation in an electric drive system is often proportional to current squared and resistance, which can be expressed as:
$$ Q = I^2 R \Delta t $$
where \( Q \) is the heat generated, \( I \) is the phase current, \( R \) is the winding resistance, and \( \Delta t \) is the time interval. By monitoring this in real-time, we can simulate cooling system performance and assess thermal limits.
Another aspect is the impact of voltage sag on the electric drive system’s performance. The battery simulator can replicate voltage drops during high-current demands, allowing us to study the system’s robustness. This is crucial for ensuring that the electric drive system maintains efficiency and power output under all conditions.
Moreover, our method supports the testing of advanced control strategies for the electric drive system, such as field-oriented control or direct torque control. By modifying the control algorithms in the simulation, we can optimize parameters for specific vehicle applications without hardware changes.
The flexibility of our approach also enables comparative studies between different electric drive systems. Table 3 summarizes key performance metrics for three hypothetical systems, evaluated under the same simulated vehicle environment. This highlights the role of simulation in selecting the best electric drive system for a given vehicle platform.
| Metric | System A | System B | System C |
|---|---|---|---|
| Peak Efficiency (%) | 95.0 | 94.5 | 93.8 |
| Torque Density (Nm/kg) | 12.5 | 11.8 | 10.5 |
| Response Time (ms) | 50 | 65 | 80 |
| Energy Consumption (kWh/100km) | 85.2 | 87.5 | 89.0 |
Such analyses underscore the value of our testing method in accelerating the development of efficient and reliable electric drive systems. By providing a controlled yet realistic environment, it reduces reliance on physical prototypes and enables iterative design improvements.
Conclusion
We have presented a comprehensive test method for electric drive systems that simulates the complete vehicle environment using a bench setup. By integrating vehicle dynamics modeling, preview control, and battery emulation, this method accurately replicates real-world driving conditions, allowing for early validation of performance and compatibility. The electric drive system is subjected to standard driving cycles while interacting with simulated loads and power sources, providing insights into energy efficiency, dynamic response, and thermal behavior.
The preview control algorithm effectively compensates for command delays, ensuring precise tracking and reliable data analysis. The battery simulator adds authenticity by mimicking SOC variations, while the modular architecture supports testing of diverse vehicle configurations. Validation results show close agreement with road tests, confirming the method’s accuracy and effectiveness.
This approach significantly shortens development cycles, reduces costs, and minimizes prototype failures, making it a valuable tool for the automotive industry. Future work may include extending the simulation to hybrid and fuel cell vehicles, as well as incorporating more detailed models for auxiliary systems. Overall, our method enhances the testing and optimization of electric drive systems, contributing to the advancement of new energy vehicles.
In summary, the electric drive system remains a focal point in vehicle development, and our simulation-based testing method offers a robust framework for its evaluation. By bridging the gap between component and vehicle-level testing, it enables engineers to deliver superior products with greater confidence and efficiency.