With the rapid adoption of electric vehicles, EV charging stations have become critical infrastructure, and their operational noise can indicate potential faults or inefficiencies. However, noise data from EV charging stations often contain redundancies, leading to transmission delays and storage bottlenecks that compromise data integrity. This paper presents an intelligent acquisition and simulation method for noise data in 40kW three-phase AC EV charging stations, focusing on optimizing data compression and precision. By leveraging advanced algorithms and a structured acquisition framework, we aim to minimize data loss and enhance reliability. The approach integrates sensor-based data capture, analog-to-digital conversion, and adaptive compression techniques, validated through detailed simulations. Throughout this work, the term “EV charging station” is emphasized to underscore its relevance in modern energy systems.
The intelligent acquisition framework for EV charging station noise data comprises multiple interconnected units, as illustrated in the following diagram. This architecture ensures synchronized data flow from sensing to display, enabling efficient noise monitoring and analysis for EV charging station applications.
In this framework, sensors intelligently detect analog noise signals generated by the EV charging station, converting acoustic vibrations into electrical signals. These signals are then amplified to enhance their amplitude for subsequent processing. The absolute value circuit rectifies the amplified signals, ensuring uniformity by converting bipolar signals to unipolar form, which is crucial for consistent analysis in EV charging station environments. A timing control unit orchestrates the entire acquisition sequence, defining precise intervals for each operation to maintain synchronization and accuracy. Following this, an integrator processes the rectified signals to reduce randomness and smooth the noise data, making it more amenable to digital conversion. The A/D acquisition unit then transforms these analog signals into digital format using a successive approximation method, which involves comparing input signals with reference voltages generated by digital-to-analog converters (DACs). This process is encapsulated by the following equation for the digital output $D$ of an N-bit A/D converter: $$D = \frac{V_{in} – V_{ref}}{V_{ref}} \times (2^N – 1)$$ where $V_{in}$ is the input voltage from the EV charging station noise, $V_{ref}$ is the reference voltage, and N is the resolution in bits. This conversion is vital for capturing high-fidelity noise data from EV charging stations.
The core of data compression lies in the microcontroller, which employs a self-controlled precision rotating door algorithm to eliminate redundancies while preserving essential information. This algorithm dynamically adjusts compression parameters based on real-time error analysis, ensuring optimal performance for EV charging station noise data. Let $x = \{x_1, x_2, \dots, x_m\}$ represent the digital noise data points from the EV charging station, where m is the total number of samples. The compression process begins by initializing the maximum and minimum compression precision parameters $G_{max}$ and $G_{min}$, respectively. The current compression precision G is computed as: $$G = \frac{A(G_{max} + G_{min})}{2}$$ where A is the adjustment amplitude, given by: $$A = \frac{G_{max} – G_{min}}{10}$$ A random data point $x_i$ (with $0 < i \leq m$) is selected, and if the time interval between $x_i$ and the previous stored point exceeds the maximum allowable interval $T_{max}$, the previous point $x_{i-1}$ is stored directly. Otherwise, the algorithm proceeds to evaluate the slopes $\lambda_{i-1}$ and $\lambda_i$ of the rotating door segments. If $\lambda_{i-1} \geq \lambda_i$, $x_{i-1}$ is stored, and a new compression segment starts from $x_{i-1}$; otherwise, no point is stored, and the process continues. After compressing all m points, the reconstructed data $\hat{x}$ is obtained via linear interpolation, and the mean error e is calculated as: $$e = \frac{\sum_{i=1}^{m} (x_i^2 – \hat{x}_i^2)}{m}$$ The deviation $\epsilon$ from the expected error $\hat{e}$ is then: $$\epsilon = \hat{e} – e$$ If $|\epsilon| \geq \hat{e} \cdot u$, where u is the tolerance, compression is deemed complete. Otherwise, G is adjusted using: $$\hat{G} = \min \left( G + \frac{\epsilon}{\hat{e} \cdot u} A, G_{max} \right)$$ and the process iterates until convergence. This adaptive compression significantly reduces data volume for EV charging station noise, mitigating transmission delays and storage issues.
To validate the method, simulations were conducted using Actran software to model a 40kW three-phase AC EV charging station and its acoustic environment. The virtual EV charging station was configured with key operational parameters, as summarized in Table 1, which outlines electrical and physical characteristics critical for noise data acquisition in EV charging stations.
| Parameter Category | Parameter Name | Value |
|---|---|---|
| Electrical Parameters | Rated Voltage | AC 380V Three-Phase |
| Rated Current | 63A | |
| Input Frequency | 50Hz ± 1% | |
| Output Power | 0–40kW | |
| Physical Parameters | Device Dimensions | 620 × 270 × 1600 mm |
| Operating Temperature | -25°C to +50°C | |
| Operating Humidity | 5% to 95% non-condensing |
The simulation steps involved importing the 3D model of the EV charging station into Actran, defining material properties such as density and acoustic impedance, and setting boundary conditions to reflect real-world scenarios. The acoustic domain was discretized into a mesh, with refined density near noise sources like fans and transformers in the EV charging station. Noise sources were parameterized with specific frequencies and intensities, and the simulation computed the acoustic field over a defined frequency range. This setup enabled the generation of realistic noise data for the EV charging station, which was then processed using the intelligent acquisition method. The entire process emphasizes the importance of accurate modeling for EV charging station noise analysis.
Results demonstrate the effectiveness of the compression algorithm for EV charging station noise data. For instance, 10 random samples of digital noise signals were compressed, showing significant redundancy reduction while retaining critical features. The self-controlled precision rotating door algorithm achieved higher compression ratios compared to standard methods, as quantified by the compressibility prediction ratio, which approaches 1 under varying data loss scenarios. This ratio is defined as: $$C_{ratio} = \frac{\text{Compressed Data Size}}{\text{Original Data Size}}$$ where a value near 1 indicates optimal compression. As shown in Table 2, the proposed method maintains a high compressibility prediction ratio even with increasing data loss proportions, outperforming alternative approaches. This highlights the robustness of the method for EV charging station applications, where data integrity is paramount.
| Data Loss Proportion (%) | Proposed Method | Comparative Method A | Comparative Method B |
|---|---|---|---|
| 0 | 0.99 | 0.98 | 0.97 |
| 10 | 0.98 | 0.96 | 0.95 |
| 20 | 0.98 | 0.94 | 0.93 |
| 30 | 0.97 | 0.93 | 0.91 |
| 40 | 0.97 | 0.92 | 0.90 |
Furthermore, data loss rates were evaluated over a 10-minute acquisition period to assess precision. The proposed method achieved a 0% data loss rate throughout, whereas comparative methods exhibited increasing losses over time, as detailed in Table 3. This underscores the reliability of the intelligent acquisition system for EV charging station noise monitoring, where minimal data loss is essential for accurate fault diagnosis and performance optimization.
| Time (minutes) | Proposed Method (%) | Comparative Method A (%) | Comparative Method B (%) |
|---|---|---|---|
| 1 | 0 | 0 | 0 |
| 2 | 0 | 0 | 0 |
| 3 | 0 | 0.05 | 0 |
| 4 | 0 | 0.05 | 0 |
| 5 | 0 | 0.05 | 0.04 |
| 6 | 0 | 0.09 | 0.04 |
| 7 | 0 | 0.09 | 0.04 |
| 8 | 0 | 0.11 | 0.08 |
| 9 | 0 | 0.11 | 0.08 |
| 10 | 0 | 0.11 | 0.08 |
In conclusion, the intelligent acquisition and simulation method for 40kW three-phase AC EV charging station noise data effectively addresses challenges related to data redundancies and transmission delays. By integrating a robust acquisition framework with an adaptive compression algorithm, the method ensures high data precision and minimal loss, as evidenced by simulation results. This approach not only enhances the reliability of EV charging station monitoring but also supports proactive maintenance and optimization. Future work could explore real-time implementation in diverse EV charging station environments to further validate its applicability. The consistent focus on EV charging station noise data underscores its significance in advancing sustainable electric vehicle infrastructure.