With the global energy transition and the pursuit of carbon neutrality goals, electric vehicles (EVs) have emerged as a core component of clean transportation, experiencing explosive growth. The proliferation of EV charging stations, as critical energy supply infrastructure, directly impacts user experience and industry development through their safety and intelligence levels. However, as the deployment density of EV charging stations increases, safety incidents related to these stations have become more frequent. Statistical analyses of publicly reported fire accidents involving EV charging stations indicate that most incidents originate from electrical faults, often associated with the plugging and unplugging of charging connectors. Traditional EV charging stations typically rely on single-sensor monitoring of electrical parameters, which struggles to address multidimensional risks during charging processes, particularly in monitoring physical state abnormalities. Therefore, constructing an efficient and accurate active safety protection system has become an urgent technological need for upgrading intelligent EV charging stations.
To address these challenges, this paper proposes an intelligent monitoring system for EV charging stations based on a single-chip microcomputer. The system innovatively integrates multi-sensor information fusion technology with machine vision to establish a dual-dimensional safety protection framework covering both electrical and physical aspects. First, at the hardware level, a high-precision sensor network is designed by incorporating temperature sensors, MQ-2 gas sensors, and Hall current sensors to enable real-time monitoring of electrical parameters and the external environment. Second, a visual monitoring and cloud data storage module is introduced to identify the position and mechanical damage of charging plugs, with image data uploaded to a cloud platform via Wi-Fi. This intelligent EV charging station monitoring system effectively addresses safety concerns during the operation of EV charging stations, offering significant application and promotion value.
The overall design of the intelligent EV charging station monitoring system is centered around the STM32F103C8T6 single-chip microcomputer as the main controller. Adopting a modular strategy, the system integrates various functional units based on specific requirements to achieve its objectives. The system architecture comprises the STM32F103C8T6 microcontroller, sensor modules, alarm modules, and cloud data storage modules. To ensure that the system can control all sensor modules while allowing each module to independently receive external information—and to prevent failure in one module from affecting others—a unified control protocol is essential. Commonly used protocols such as the Inter-Integrated Circuit (IIC) and serial communication are employed. Externally, cameras monitor the charging interface to observe usage conditions in real-time; internally, various sensors track parameters like current, ambient temperature, and smoke concentration, with information displayed on an OLED screen. If any parameter exceeds set thresholds, a buzzer activates to alert users. This modular design enhances the reliability, operational speed, and longevity of the EV charging station monitoring system.
The hardware design of the intelligent EV charging station monitoring system utilizes the STM32F103C8T6 single-chip microcomputer as the core controller, combined with multiple external sensors and execution modules. The STM32F103C8T6, based on the ARM Cortex-M3 core, operates at a maximum frequency of 72 MHz and supports single-cycle instruction execution. This microcontroller meets diverse design requirements with low cost and power consumption, making it suitable for small to medium-sized projects. The sensor module includes a temperature sensor, gas sensor, and Hall current sensor, each contributing to comprehensive monitoring of the EV charging station environment.
The DS18B20 digital temperature sensor employs a unique one-wire protocol and operates within a temperature range of -55°C to 125°C. In this system, the resolution is set to 10 bits to ensure measurement accuracy within ±0.5°C. A pull-up resistor is added to the bus for stable communication, and the STM32F103C8T6 reads temperature data via the one-wire protocol, identifying sensors through their ROM codes. The relationship between temperature and digital output can be expressed using the formula: $$T = \frac{D}{256}$$ where \(T\) is the temperature in degrees Celsius and \(D\) is the digital value read from the sensor. This temperature sensor is ideal for high-precision applications in EV charging stations.
The MQ-2 gas sensor uses tin dioxide as its sensing material, which has low conductivity in clean air. When smoke is present, the sensor’s conductivity increases with concentration. The analog output from the MQ-2 is converted to a digital signal via the ADC module of the STM32F103C8T6, enabling smoke concentration detection. Since the output voltage of the MQ-2 is non-linear with respect to smoke concentration, calibration is performed using its characteristic curve. The ratio \(R_s / R_0\) represents the sensor resistance in gas relative to that in clean air. A filtering capacitor is incorporated into the hardware circuit to ensure sampling accuracy and stability. The smoke concentration \(C\) can be approximated by: $$C = k \cdot \left( \frac{V_{out} – V_0}{V_{ref}} \right)^m$$ where \(k\) and \(m\) are constants derived from the characteristic curve, \(V_{out}\) is the output voltage, \(V_0\) is the baseline voltage in clean air, and \(V_{ref}\) is the reference voltage. This ensures reliable monitoring of environmental hazards in EV charging stations.
The ACS712 Hall-effect current sensor measures current non-invasively by detecting magnetic field-induced voltage offsets. With an output voltage range of 0.5 V to 4.5 V, and the STM32F103C8T6’s ADC reference voltage at 3.3 V, a voltage divider circuit scales the signal down to the appropriate range. The sensor includes a Hall element, amplifier, and linearization circuit, providing a direct analog output. The current \(I_p\) is calculated using: $$I_p = \frac{2V_{out} – V_{cc}}{2 \cdot \text{Sensitivity}}$$ where \(V_{out}\) is the ADC-read voltage, \(V_{cc}\) is the bias voltage, and Sensitivity is the sensor’s sensitivity (e.g., 100 mV/A for the ACS712). High sampling rates and averaging filters are employed to capture dynamic changes and ensure real-time monitoring. Decoupling capacitors are added between power and ground lines to enhance stability. This current detection module is crucial for preventing overloads in EV charging stations.
Visual monitoring and cloud data storage modules consist of an OV7670 camera and an ESP8266 Wi-Fi module. The STM32F103C8T6 controls the OV7670 to capture image data, which is then transmitted to a cloud platform via the ESP8266. This integration of image processing and IoT communication enables real-time image capture and upload for key monitoring points, such as charging interfaces. The OV7670 camera, configured via IIC interface, stores image data in a FIFO chip to handle high data rates. Preprocessed images are sent to the ESP8266 via UART, which connects to Wi-Fi and uses HTTPS to upload data to cloud storage. The cloud platform stores images in designated buckets, allowing API access for historical data retrieval. This supports advanced analytics and safety assurance for EV charging stations.
Software design for the intelligent EV charging station monitoring system is implemented in the Keil5 integrated development environment using C language. The program flow begins with system initialization, where sensor modules, visual monitoring modules, and communication interfaces are set up. Data acquisition involves reading analog signals from sensors and converting them to digital values via ADC. These values are displayed on an OLED screen and processed through filtering algorithms before being compared to predefined thresholds. If anomalies are detected, such as excessive temperature or smoke concentration, the system triggers alarms. The visual monitoring module continuously captures images, and the ESP8266 uploads them to the cloud for storage and analysis. This structured approach ensures high reliability and adaptability of the EV charging station monitoring system.
To validate system performance, module testing was conducted under normal and abnormal conditions. For instance, smoke volume fraction in normal environments is typically below 0.01%, but during fire incidents, it exceeds 0.5%. The STM32F103C8T6 operates at a current of approximately 20 mA; deviations beyond 10–30 mA trigger current sensor alerts. Thresholds for various parameters are summarized in Table 1, ensuring prompt response to hazards. Visual module testing revealed issues like data loss and image distortion due to high I/O rates, which were mitigated by incorporating FIFO chips or using microcontrollers with richer resources. Overall, testing confirmed that all modules meet design requirements, enabling effective safety monitoring for EV charging stations.
| Parameter | Normal Range | Alert Threshold |
|---|---|---|
| Temperature (°C) | -20 to 60 | >70 |
| Smoke Concentration (%) | <0.01 | >0.5 |
| Current (mA) | 10–30 | <10 or >30 |
The integration of multi-sensor networks and machine vision in this intelligent EV charging station monitoring system achieves comprehensive and smart safety oversight. By continuously monitoring electrical parameters and physical states during charging, the system ensures operational safety and enhances user trust in EV charging stations. Future work will focus on improving environmental factor acquisition capabilities and communication performance to handle complex real-world scenarios. For example, advanced algorithms for image recognition could be incorporated to detect plug misalignment or vandalism automatically. Additionally, expanding the sensor suite to include humidity and vibration sensors could provide a more holistic view of EV charging station conditions. The system’s scalability allows for integration with smart grid technologies, enabling dynamic load management and predictive maintenance. These advancements will further solidify the role of intelligent monitoring in the sustainable evolution of EV charging infrastructure.
In summary, the proposed system represents a significant step forward in EV charging station safety. By leveraging cost-effective hardware and robust software design, it addresses critical vulnerabilities while offering a platform for continuous improvement. As the adoption of electric vehicles accelerates, such innovations will be essential in building resilient and user-friendly charging networks. The ongoing development of IoT and AI technologies promises to unlock even greater potentials for automation and efficiency in EV charging station management.
The mathematical modeling of sensor interactions enhances system reliability. For instance, the overall risk score \(R\) for an EV charging station can be computed as a weighted sum of individual parameter deviations: $$R = w_t \cdot \Delta T + w_s \cdot \Delta S + w_c \cdot \Delta C$$ where \(w_t\), \(w_s\), and \(w_c\) are weights for temperature, smoke, and current deviations, respectively, and \(\Delta T\), \(\Delta S\), \(\Delta C\) are the differences between measured values and thresholds. This formula allows for prioritized responses based on real-time data.
| Sensor Type | Parameter Measured | Accuracy | Interface |
|---|---|---|---|
| DS18B20 | Temperature | ±0.5°C | One-Wire |
| MQ-2 | Smoke Concentration | ±5% | ADC |
| ACS712 | Current | ±1.5% | ADC |
| OV7670 | Image Data | VGA Resolution | IIC + FIFO |
Furthermore, the system’s energy efficiency is critical for sustainable EV charging station operation. The power consumption \(P\) of the monitoring system can be estimated as: $$P = V \cdot \sum I_i$$ where \(V\) is the supply voltage and \(I_i\) is the current drawn by each module. Optimization techniques, such as sleep modes and dynamic scaling, are implemented to minimize \(P\) without compromising safety.
In conclusion, the intelligent EV charging station monitoring system demonstrates how embedded technologies can transform infrastructure safety. Through continuous innovation and adaptation, it paves the way for smarter, more reliable EV charging networks globally.