Load control method based on charging pile ad hoc network

Abstract

The invention discloses a load control method based on a charging pile ad hoc network, and relates to the technical field of charging pile ad hoc network load control, and the method comprises the following steps: firstly, carrying out the real-time collection of various parameters in the operation process of a charging pile, guaranteeing the obtaining of high-quality original data, and providing a solid data basis for the subsequent analysis and model evaluation; according to the invention, through real-time data acquisition, intelligent feature extraction, deep learning evaluation and adaptive dynamic regulation and control, the problems of harmonic resonance, power distribution imbalance and power grid instability in the operation of the charging pile are solved. Charging pile parameters are collected in a high-precision mode, harmonic key features are extracted through a detection window mechanism, intelligent evaluation is conducted through a deep learning model, and the harmonic abnormal trend is predicted. When the resonance risk is detected, the power of the charging pile and the reactive compensation strategy are intelligently adjusted, the load distribution is optimized, the harmonic resonance risk is reduced, the power supply quality of the power grid is improved, and the intelligent charging network has higher stability, reliability and scheduling capability.

Description

Technical Field

[0001] The present invention relates to the technical field of load control for charging pile self-organizing networks, and particularly to a load control method based on charging pile self-organizing networks. Background Art

[0002] Load control for charging pile self-organizing networks refers to the process of dynamically adjusting the charging power or queuing strategy of each charging pile after multiple charging piles form a self-organizing network through wireless or wired communication technologies (such as based on Zigbee, LoRa, 5G, or PLC, etc.) to avoid grid overload, improve charging efficiency, and ensure charging fairness. This control method usually based on real-time grid load conditions, user demands, charging pile power limits, and new energy access and other factors, through distributed collaborative scheduling or centralized management strategies, to achieve optimal allocation of power resources. For example, when the grid load approaches the upper limit, the system can reduce the charging power of some charging piles or delay non-urgent charging tasks to prevent overload; while when the load is low, the charging power can be appropriately increased to improve charging efficiency and user experience.

[0003] The prior art has the following deficiencies: In the underground parking lot of a large commercial center, limited by the distribution capacity, if multiple charging piles operate simultaneously, it may cause grid overload. Therefore, the charging pile self-organizing network can balance the load by preferentially satisfying low-battery vehicles through intelligent power distribution to avoid peak power consumption shock to the power distribution system in a short period. However, during the dynamic power regulation process, the charging pile may resonate with reactive power compensation devices (such as capacitor banks) in the power grid, resulting in local voltage distortion and harmonic resonance. Such phenomena may cause the power supply module of the charging pile to overheat or even be damaged, causing a large area of charging piles to malfunction and affecting the overall charging capacity of the parking lot. At the same time, grid instability may affect other key power equipment in the mall, such as elevators, air conditioners, and lighting systems, and even trigger the power distribution protection mechanism of the entire commercial center, resulting in regional power supply interruption and seriously affecting normal operation.

[0004] The above information disclosed in the background art section is only used to enhance the understanding of the background of the present disclosure, and thus it may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0005] The object of the present invention is to provide a load control method based on the self - networking of charging piles. By means of real - time data acquisition, intelligent feature extraction, deep - learning evaluation and adaptive dynamic regulation, it solves the problems of harmonic resonance, unbalanced power distribution and grid instability during the operation of charging piles. The system accurately acquires the parameters of charging piles, constructs a standardized data set, extracts the key features of harmonics through a detection window mechanism, and uses a deep - learning model for intelligent evaluation to predict the abnormal trend of harmonics. When a resonance risk is detected, the system intelligently adjusts the power and reactive - power compensation strategies of the charging piles, optimizes the load distribution, reduces the risk of harmonic resonance, improves the power - supply quality of the grid, and makes the intelligent charging network more stable, reliable and schedulable, so as to solve the problems in the above - mentioned background technology.

[0006] To achieve the above object, the present invention provides the following technical solution: A load control method based on the self - networking of charging piles, comprising the following steps: First, various parameters during the operation of the charging piles are collected in real time to ensure obtaining high - quality original data, providing a solid data basis for subsequent analysis and model evaluation; After completing the real - time data acquisition, the operation parameters of the charging piles are uniformly stored in a data platform to construct a standardized data set, and at the same time, each parameter in the data set is pre - processed to ensure the integrity and accuracy of the data; Key features indicating that the harmonic frequency of the charging pile reaches the grid resonance point are mined and refined from the pre - processed data. Within the detection window, the extracted key features are analyzed to quantify the change trend of the harmonic frequency; The extracted key features are input into a pre - trained deep - learning model, and the current harmonic state is intelligently evaluated through the deep - learning model; When the evaluation result of the deep - learning model shows that the harmonic frequency of the charging pile reaches the grid resonance point, based on the dynamic change of the real - time harmonic features, the power - regulation strategy of the charging pile is intelligently reconstructed, its output power and the working frequency of the converter are accurately adjusted, and the harmonic frequency is actively guided away from the resonance interval. At the same time, the switching mode of the reactive - power compensation device is adaptively optimized to make it respond to the dynamic change of the load, flexibly adjust the resonance - frequency distribution, and construct an "anti - harmonic - resonance" intelligent barrier to ensure the stability of the grid and the efficient operation of the charging piles.

[0007] Preferably, first, high - precision sensors and intelligent electricity meters are deployed at key nodes of each charging pile and the distribution network to capture various key power parameters of the charging piles; Secondly, a high - speed data - acquisition module is used to collect and encapsulate data in real time to ensure a sampling accuracy of seconds, avoiding information lag and loss; Then, a time - synchronization mechanism is adopted to calibrate the timestamps of the data collected by different devices to ensure that all data points can be aligned under the same time reference; Finally, the real-time collected data is transmitted to the cloud or local server through a wireless or wired communication network and stored in the data platform to form a complete data set, so as to support subsequent harmonic feature extraction, intelligent analysis and dynamic control decision-making.

[0008] Preferably, key features indicating that the harmonic frequency of the charging pile reaches the grid resonance point are mined and refined from the preprocessed data. The refined features include the matching condition of the grid equivalent impedance at different frequency points and the proportion of the harmonic power fed back from the charging pile to the grid. The matching condition of the grid equivalent impedance at different frequency points and the proportion of the harmonic power fed back from the charging pile to the grid are analyzed under the detection window to generate a harmonic impedance matching reference value and a harmonic power return reference value respectively, and the change trend of the harmonic frequency is quantified through the harmonic impedance matching reference value and the harmonic power return reference value.

[0009] Preferably, the specific steps for analyzing the matching condition of the grid equivalent impedance at different frequency points under the detection window to generate a harmonic impedance matching reference value are as follows: In the power grid system, the equivalent impedance changes with frequency to form an impedance spectrum characteristic. When the charging pile generates harmonic frequencies, the output impedance of the charging pile at the harmonic frequency points is matched and analyzed with the grid equivalent impedance. The impedance matching error is defined as a measure of the matching degree between the grid and the charging pile at the harmonic frequency point, and the calculation expression is as follows: , where is the equivalent impedance of the grid at the harmonic frequency , is the equivalent output impedance of the charging pile at the harmonic frequency , is an extremely small positive number, is the harmonic impedance matching error; After calculating the impedance matching errors at different harmonic frequency points , a harmonic impedance matching reference value is further defined, which is used to quantify the harmonic resonance risk of the entire system within the detection window, and the calculation expression is as follows: , where is the harmonic impedance matching reference value, is the range of harmonic orders analyzed, is the power component of the harmonic voltage at , representing the influence of harmonic energy, is the impedance matching adjustment factor, is the harmonic order, is the weight factor.

[0010] Preferably, the specific steps for analyzing the proportion of harmonic power fed back by the charging pile to the power grid within the detection window to generate a harmonic power return reference value are as follows: First, calculate the harmonic power of each order output from the charging pile to the power grid, and compare it with the total harmonic power to quantify the proportion of harmonic feedback. The harmonic power of the charging pile is expanded by Fourier series, the voltage and current components of different harmonic orders are extracted, and then the power of each harmonic component is calculated. Define the proportion of harmonic power fed back from the charging pile to the power grid, and the calculation expression is as follows: , where is the proportion of harmonic power feedback, is the range of harmonic orders for analysis, is the harmonic order, from the 2nd order to the highest harmonic order , is the effective voltage value of the charging pile at the harmonic order , is the harmonic order , is the phase angle between the harmonic current and voltage at the power grid end, is the phase angle between the harmonic current output from the charging pile and the voltage; Based on the proportion of harmonic power feedback from the charging pile to the power grid, construct a harmonic power return reference value to comprehensively measure the feedback degree of different harmonic frequency components and their impact on the power grid resonance risk. The calculation expression is as follows: , where is the harmonic power return reference value, is the non-linear weight factor of the feedback power, is the equivalent impedance of the power grid at the harmonic order , is the reference impedance, is the weight parameter of the power grid impedance on the harmonic amplification effect.

[0011] Preferably, input the harmonic impedance matching reference value and the harmonic power return reference value after quantitative analysis into a pre-trained deep learning model. Generate a harmonic resonance risk coefficient through the deep learning model, and conduct an intelligent evaluation of the current harmonic state through the harmonic resonance risk coefficient.

[0012] Preferably, compare and analyze the harmonic resonance risk coefficient generated when conducting an intelligent evaluation of the current harmonic state through a pre-trained deep learning model with a pre-set harmonic resonance risk coefficient reference threshold, and classify the current harmonic state of the charging pile. The classification steps are as follows: If the harmonic resonance risk coefficient is greater than the harmonic resonance risk coefficient reference value threshold, the current charging pile harmonics are classified as harmonic frequencies that have reached the power grid resonance point; if the harmonic resonance risk coefficient is less than or equal to the harmonic resonance risk coefficient reference value threshold, the current charging pile harmonics are classified as harmonic frequencies that have not reached the power grid resonance point.

[0013] Preferably, when the evaluation result of the deep learning model indicates that the harmonic frequency of the charging pile reaches the power grid resonance point, based on the dynamic changes of the real-time harmonic characteristics, the power regulation strategy of the charging pile will be intelligently reconstructed. At the same time, the switching mode of the reactive power compensation device will be adaptively optimized. The specific steps for it to respond to the dynamic changes of the load are as follows: When the evaluation result indicates that the harmonic frequency of the charging pile reaches the power grid resonance point, the output power of the charging pile and the operating frequency of the converter are dynamically adjusted through an intelligent regulation strategy to guide the harmonic frequency away from the power grid resonance point. The specific regulation method is as follows: The dynamic adjustment calculation expression of the charging pile power is as follows: , , where, is the adjusted output power of the charging pile, is the original power of the charging pile before adjustment, is the power adjustment coefficient, is the harmonic resonance risk coefficient, is the harmonic resonance risk coefficient reference threshold, is the set maximum harmonic resonance risk coefficient for normalizing the adjustment amplitude, is the natural base, is the adjustment parameter, is the set minimum power limit to prevent the charging efficiency of the charging pile from decreasing too much due to excessive adjustment; The dynamic adjustment calculation expression of the converter switching frequency is as follows: , where, is the adjusted converter switching frequency, is the original converter switching frequency, is the frequency adjustment coefficient, is the sine adjustment factor; After adjusting the power and operating frequency of the charging pile, it is still necessary to optimize the switching strategy of the reactive power compensation device to dynamically adjust the resonance frequency of the power grid away from the current harmonic frequency. The optimization formula is as follows: The dynamic optimization calculation expression of the reactive power compensation capacity is as follows; , , where, Among them: is the adjusted reactive power compensation capacity, is the reactive power compensation capacity before adjustment, is the reactive power compensation adjustment coefficient, is the suppression parameter, is the minimum allowable value of reactive power compensation; The calculation expressions for dynamically adjusting the equivalent inductance and capacitance are as follows: , , where in the formula, is the adjusted system equivalent inductance, is the original inductance value, is the inductance adjustment coefficient, is the arctangent adjustment factor, is the adjusted equivalent capacitance, is the original capacitance value, is the capacitance adjustment coefficient, is the exponential decay factor, is the exponential decay parameter.

[0014] In the above technical solution, the technical effects and advantages provided by the present invention are: The present invention effectively solves problems such as harmonic resonance, unbalanced power distribution, and grid instability that may be caused during the operation of charging piles through real-time data acquisition, intelligent feature extraction, deep learning evaluation, and adaptive dynamic regulation. First, the system performs high-precision acquisition of key parameters during the operation of charging piles and constructs a standardized data set in the data platform to ensure the integrity and accuracy of the data. Then, through feature analysis and the detection window mechanism, key features representing harmonic frequencies reaching the grid resonance point are accurately identified, and an intelligent evaluation is carried out using a deep learning model to predict the abnormal trend of harmonics in advance. When the system detects that the harmonic frequency enters the dangerous range, the charging pile self-organized network will intelligently reconstruct the power regulation strategy, actively adjust the output power of the charging pile and the operating frequency of the converter, optimize the load distribution, and at the same time dynamically adjust the switching mode of reactive power compensation equipment, effectively reducing the risk of harmonic resonance and improving the power supply quality of the grid. The implementation of this solution enhances the adaptive ability of the charging pile group, not only improving the stability and safety of the charging system, but also reducing grid harmonic pollution, improving the energy utilization efficiency in complex power environments such as underground parking lots in shopping malls, and making the intelligent charging network have higher reliability, controllability, and intelligent scheduling capabilities. Description of the Drawings

[0015] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the accompanying drawings required in the embodiments. Obviously, the accompanying drawings in the following description are only some embodiments recorded in the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings.

[0016] Figure 1 This is the method flow chart of the load control method based on the self-organizing network of charging piles of the present invention. Detailed implementation manners

[0017] Now, the exemplary embodiments will be described more comprehensively with reference to the accompanying drawings. However, the exemplary embodiments can be implemented in various forms and should not be construed as limited to the examples set forth herein; on the contrary, these exemplary embodiments are provided so that the description of the present disclosure will be more comprehensive and complete, and the concept of the exemplary embodiments will be fully conveyed to those skilled in the art.

[0018] The present invention provides a load control method based on the self-organizing network of charging piles as Figure 1 shown, including the following steps: First, various parameters during the operation of the charging piles are collected in real time to ensure obtaining high-quality raw data, providing a solid data basis for subsequent analysis and model evaluation; Collect various parameters during the operation of the charging piles in real time to ensure obtaining high-quality raw data. The specific steps are as follows: First, deploy high-precision sensors (such as current transformers, voltage sensors, harmonic analyzers) and smart meters at key nodes of each charging pile and the power distribution network to capture key power parameters (voltage, current, power, harmonic components, power factor) of the charging piles; Secondly, use a high-speed data acquisition module (such as an edge computing gateway or an embedded data acquisition terminal) to collect and encapsulate data in real time to ensure sampling accuracy at the second level, avoiding information lag and loss; Then, adopt a time synchronization mechanism (such as GPS clock synchronization or IEEE 1588 PTP protocol) to calibrate the timestamps of the data collected by different devices to ensure that all data points can be aligned under the same time reference; Finally, through a wireless or wired communication network (such as 5G, LoRa, Wi-Fi, PLC or fiber optic Ethernet), transmit the real-time collected data to the cloud or local server, store it in the data platform, and form a complete data set to support subsequent harmonic feature extraction, intelligent analysis and dynamic control decision-making.

[0019] At this stage, it is necessary to continuously obtain dynamic data such as the voltage, current, power, temperature, and harmonic components of the charging pile through sensors, measuring devices, and the monitoring module of the charging pile itself. These data can reflect the harmonic characteristics generated by the charging pile under different loads and operating conditions. By constructing a real-time data acquisition system (such as installing high-precision electricity meters, harmonic analyzers, etc. on each charging pile), high-quality raw data can be obtained at millisecond or second-level time intervals, providing a solid data foundation for subsequent analysis and model evaluation.

[0020] After completing the real-time data acquisition, the operating parameters of the charging pile are uniformly stored in the data platform to construct a standardized data set. At the same time, preprocessing is performed on each parameter in the data set to ensure the integrity and accuracy of the data. This process includes aligning the timestamps and unifying the formatting of data from different devices and sources to ensure that various data information accurately corresponds to the same time axis and can be efficiently retrieved and processed in subsequent steps. In this way, a good data management foundation can be laid for subsequent deep learning models and feature analysis.

[0021] After the preliminary collation of the data set, further preprocessing is still required, including operations such as anomaly detection, missing value filling, and data denoising. First, abnormal data caused by sensor failures or communication interruptions need to be cleaned or removed. Second, occasional missing data can be filled through interpolation or other statistical methods. In addition, in order to better extract harmonic characteristics, it is often necessary to filter, normalize, or perform other denoising processes on the data to remove unnecessary random interference and make subsequent feature extraction and model training more accurate and stable.

[0022] Mine and refine the key features from the preprocessed data that characterize the harmonic frequency of the charging pile reaching the grid resonance point. Analyze the extracted key features within the detection window to quantify the change trend of the harmonic frequency. Mine and refine the key features from the preprocessed data that characterize the harmonic frequency of the charging pile reaching the grid resonance point. The extracted features include the matching situation of the grid equivalent impedance at different frequency points and the proportion of the harmonic power fed back from the charging pile to the grid. Analyze the matching situation of the grid equivalent impedance at different frequency points and the proportion of the harmonic power fed back from the charging pile to the grid under the detection window, and generate the harmonic impedance matching reference value and the harmonic power reflux reference value respectively. Quantify the change trend of the harmonic frequency through the harmonic impedance matching reference value and the harmonic power reflux reference value.

[0023] When the equivalent impedance of the power grid is successfully matched at a certain frequency point, it usually means that the harmonic components at that frequency point may be greatly amplified in the power grid, thus forming harmonic resonance. Therefore, if the harmonic frequency generated by the charging pile happens to be the same as the frequency point at which the equivalent impedance is matched, it indicates that the harmonic frequency of the charging pile has reached the power grid resonance point and may cause serious power quality problems. Specifically, the equivalent impedance of the power grid is jointly determined by inductive loads (such as transformers and power grid lines) and capacitive elements (such as reactive power compensation capacitor banks). These elements form an LC resonance circuit at certain specific frequencies. When the harmonic frequency generated by the charging pile falls into this resonance interval, the impedance of the power grid approaches the minimum or maximum at this frequency point, causing the harmonic voltage or current corresponding to this frequency to be significantly amplified, far exceeding the normal level. This phenomenon will cause the power electronic devices inside the charging pile to bear excessive harmonic current, overheat and even be damaged. At the same time, it may also affect the voltage stability of the entire power grid, causing surrounding equipment to malfunction or protection mechanisms to trigger. Therefore, when the harmonic frequency of the charging pile is the same as the matching frequency of the equivalent impedance of the power grid, it is usually considered that the harmonic frequency of the charging pile has reached the power grid resonance point. If dynamic regulation measures are not taken, it may lead to harmonic resonance and serious power quality problems.

[0024] The specific steps to analyze the matching situation of the equivalent impedance of the power grid at different frequency points under the detection window to generate the harmonic impedance matching reference value are as follows: In the power grid system, the equivalent impedance changes with frequency, forming an impedance spectrum characteristic. When the charging pile generates a harmonic frequency, the output impedance of the charging pile at the harmonic frequency point is matched with the equivalent impedance of the power grid. The impedance matching error is defined as a measure of the matching degree between the power grid and the charging pile at this harmonic frequency point. The calculation expression is as follows: , where is the equivalent impedance of the power grid at the harmonic frequency , is the equivalent output impedance of the charging pile at the harmonic frequency , is an extremely small positive number (to prevent the denominator from being zero), is the harmonic impedance matching error, which is a normalized matching index to measure whether the harmonic frequency of the charging pile is close to the power grid resonance point; The function of this step is to measure the impedance gap between the charging pile and the power grid at a specific harmonic frequency. The smaller the value, the higher the impedance matching degree between the two and the more likely to enter the resonance state; the larger the value, the lower the matching degree and the smaller the resonance risk.

[0025] After calculating the impedance matching error at different harmonic frequency points After that, the harmonic impedance matching reference value is further defined to quantify the harmonic resonance risk of the entire system within the detection window, and the calculation expression is as follows: , where is the harmonic impedance matching reference value, is the range of harmonic orders for analysis, is the power component of the harmonic voltage at , representing the influence of harmonic energy, is the impedance matching adjustment factor, a balancing factor used to prevent calculation anomalies or extreme situations, usually a small positive number, is the harmonic order, is the weighting factor, used to adjust the sensitivity of the harmonic impedance matching reference value to different frequency points.

[0026] The harmonic order (HarmonicOrder) refers to the integer multiple frequency components of the AC voltage or current waveform in the power grid or electrical equipment relative to the fundamental wave (usually the power frequency of 50 Hz or 60 Hz). The generation of harmonics mainly stems from the influence of non-linear loads (such as electric vehicle chargers, frequency converters, switching power supplies, etc.) on the power grid, causing the originally sinusoidal voltage or current waveform to become distorted and form multiple frequency components.

[0027] The function of this formula is to quantify the resonance matching degree of the entire system by means of weighted accumulation, combining the impedance matching error and harmonic voltage energy. When is larger, it indicates that the impedance matching degree at multiple harmonic frequency points is higher, the harmonic energy is more concentrated, and the risk of the power grid entering the resonance state increases significantly; when is smaller, it shows that the harmonic characteristics of the system are dispersed and resonance is not likely to occur. This reference value can be used as a real-time monitoring indicator to help the system dynamically adjust the charging pile power output or optimize the reactive power compensation strategy to effectively avoid harmonic resonance.

[0028] The larger the harmonic impedance matching reference value generated by analyzing the matching situation of the grid equivalent impedance at different frequency points under the detection window, the closer the harmonic frequency output by the charging pile is to the resonance frequency of the grid, and the system is at a higher harmonic resonance risk. When this reference value reaches or exceeds the set resonance critical value, it means that the harmonic frequency of the charging pile has successfully coupled to the resonance point of the grid, which may lead to problems such as local voltage distortion, harmonic current amplification, equipment overload, and even faults. On the contrary, if the harmonic impedance matching reference value is low, it indicates that the harmonic frequency of the charging pile is far from the grid resonance point, and there will be no significant harmonic amplification effect in the grid. Therefore, by real-time monitoring the change trend of the harmonic impedance matching reference value, it is possible to determine whether the current harmonic characteristics of the charging pile match the grid resonance point, and take timely measures such as power regulation, frequency adjustment, or reactive power compensation optimization when the harmonic impedance matching reference value increases, to avoid the occurrence of harmonic resonance and ensure the stability of the grid and the safe operation of the charging system.

[0029] A significant increase in the proportion of harmonic power fed back from the charging pile to the grid is usually one of the important indicators that the harmonic frequency of the charging pile reaches the grid resonance point. This is because when the charging pile is operating normally, its harmonic power is mainly absorbed by the load and will not be fed back to the grid in large amounts. However, when the proportion of harmonic power feedback increases abnormally, it may mean that the grid impedance matches the harmonic frequency output by the charging pile, forming a resonance circuit. In this case, the harmonic energy cannot be effectively consumed, but is continuously reflected and amplified between the grid and the charging pile, resulting in an increase in the local harmonic voltage distortion rate. At the same time, it may also cause the reactive power compensation equipment (such as capacitor banks) on the grid side to enter the resonance state, further enhancing the harmonic energy feedback effect. In addition, the power factor of the charging pile may fluctuate violently near the resonance point, the current waveform distortion intensifies, and even cause misoperation or overcurrent tripping of the grid protection device. Therefore, when the proportion of harmonic power feedback from the charging pile shows a significant increase, and is accompanied by a simultaneous increase in the grid harmonic distortion rate, the amplitude of specific sub-harmonics, or harmonic energy aggregation, it strongly indicates that the harmonic frequency of the charging pile has approached or reached the grid resonance point, and dynamic control measures need to be taken immediately to adjust the power output mode of the charging pile or optimize the reactive power compensation strategy to prevent the further deterioration of harmonic resonance.

[0030] The specific steps for generating the harmonic power return reference value by analyzing the proportion of harmonic power fed back from the charging pile to the grid under the detection window are as follows: First, calculate the harmonic power of each order output from the charging pile to the grid, and compare it with the total harmonic power to quantify the proportion of harmonic feedback. The harmonic power of the charging pile is expanded by Fourier series, the voltage and current components of different harmonic orders are extracted, and then the power of each harmonic component is calculated. Define the proportion of harmonic power fed back from the charging pile to the grid, and the calculation expression is as follows: , where is the proportion of harmonic power feedback, is the range of harmonic orders for analysis, is the harmonic order, from the 2nd order to the highest harmonic order , is the effective voltage value of the charging pile at the harmonic order ; is the harmonic order is the effective voltage value at this harmonic order, is the phase angle between the harmonic current and voltage at the grid side, is the phase angle between the harmonic current and voltage output by the charging pile; The above steps quantify the proportion of harmonic power feedback from the charging pile to the grid to evaluate the grid's absorption capacity for harmonic energy and identify potential harmonic resonance risks. By calculating the output and feedback relationship of each harmonic power, it can be determined whether the harmonic power generated by the charging pile is effectively consumed or refluxed and amplified within the grid. When the feedback power proportion is relatively high, it indicates that the grid has a weak absorption capacity for harmonics at this frequency and may be close to the resonance point, providing key input parameters for constructing a reference value for harmonic power reflux in the follow-up, so as to accurately identify harmonic resonance risks and take corresponding control measures.

[0031] Based on the proportion of harmonic power feedback from the charging pile to the grid, a reference value for harmonic power reflux is constructed to comprehensively measure the feedback degree of different harmonic frequency components and their impact on the grid resonance risk. The reference value for harmonic power reflux not only considers the proportion of feedback power but also introduces the grid harmonic impedance characteristics and the non-linear weight factor of the feedback power. The calculation expression is as follows: , where is the reference value for harmonic power reflux, is the non-linear weight factor of the feedback power, used to amplify or attenuate the influence of specific harmonic orders, is the equivalent impedance of the grid at the harmonic order ; is the reference impedance, is the weight parameter of the grid impedance for the harmonic amplification effect, usually taking .

[0032] After calculating the proportion of harmonic power fed back from the charging pile to the power grid in the above steps, by constructing a reference value for harmonic power reflux, the feedback degree of different harmonic orders is further comprehensively evaluated, and combined with the impedance characteristics of the power grid, the impact of harmonic feedback on power grid resonance is quantified. The reference value for harmonic power reflux not only considers the magnitude of harmonic feedback power, but also introduces the influence weight of the equivalent impedance of the power grid, enabling it to reflect the sensitivity of the power grid to different harmonic frequencies. By exponentially amplifying or attenuating the contributions of different order harmonics, the reference value for harmonic power reflux can more accurately identify whether the harmonics are approaching the power grid resonance point, providing a decision-making basis for intelligent regulation and ensuring the stable operation of the charging pile and power grid system.

[0033] The larger the reference value for harmonic power reflux generated after analyzing the proportion of harmonic power fed back from the charging pile to the power grid under the detection window, usually indicates that the harmonic frequency of the charging pile has reached the power grid resonance point; conversely, it means that the harmonic frequency of the charging pile has not reached the resonant state. Under normal operating conditions, most of the harmonic power of the charging pile is absorbed by the load or local filtering equipment, and the proportion fed back to the power grid is relatively low. However, when the harmonic frequency of the charging pile approaches the inherent resonance point of the power grid, the equivalent impedance of the power grid will decrease significantly, making the harmonic power output by the charging pile unable to be effectively consumed, and then forming a harmonic reflux channel between the charging pile and the power grid. At this time, the reference value for harmonic power reflux will increase significantly, reflecting the continuous accumulation and amplification of harmonic energy within the system, resulting in an increase in the harmonic distortion rate of the local power grid and even possibly triggering resonance phenomena in reactive power compensation equipment. Therefore, under the monitoring window, if the reference value for harmonic power reflux continues to rise and exceeds the set threshold, it can be used as an important basis for judging the occurrence of harmonic resonance; conversely, if the reference value for harmonic power reflux remains at a low level, it indicates that the harmonic frequency of the charging pile has not reached the power grid resonance point, and the power grid is still in a normal state of harmonic absorption and consumption.

[0034] Input the extracted key features into a pre-trained deep learning model, and use the deep learning model to intelligently evaluate the current harmonic state; Input the harmonic impedance matching reference value and the harmonic power reflux reference value after quantitative analysis into a pre-learned deep learning model, generate a harmonic resonance risk coefficient through the deep learning model, and use the harmonic resonance risk coefficient to intelligently evaluate the current harmonic state.

[0035] A pre-trained machine learning model refers to an intelligent model that has been trained using a large amount of historical data before formal deployment and has a certain generalization ability. In this scenario, such a model is mainly used to analyze the harmonic data of charging piles and evaluate whether the power grid enters the harmonic resonance risk range. Its training process usually includes multiple stages: First, a dataset covering the operating parameters of charging piles, the load status of the power grid, harmonic characteristics (such as harmonic impedance matching reference values, harmonic power reflux reference values), and historical resonance events needs to be constructed. Next, the data needs to go through feature engineering (such as principal component analysis PCA, time series decomposition, Fourier transform FFT, etc.) to extract key variables. Then, a suitable machine learning algorithm (such as deep neural network DNN, long short-term memory network LSTM, random forest RF, or gradient boosting decision tree GBDT) is selected, and supervised learning or semi-supervised learning is carried out so that the model can identify the characteristic patterns before harmonic resonance occurs. In addition, during the training process, the model is optimized using loss functions (such as mean square error MSE, cross-entropy loss CE, etc.) to ensure the accuracy of the prediction results, and cross-validation methods are used to prevent overfitting, thereby improving the adaptability of the model in different environments.

[0036] In the deployment stage, this pre-trained model is used for real-time inference, that is, based on the currently collected harmonic impedance matching reference values and harmonic power reflux reference values, a harmonic resonance risk coefficient is generated, and a dynamic risk assessment is carried out in combination with historical data. Since the model has learned the characteristics of harmonic resonance under different conditions during the training stage, it can make predictions within a millisecond response time and provide a quantified risk level (such as low risk, medium risk, high risk). When the risk coefficient exceeds the set threshold, the system can automatically trigger intelligent control strategies, such as dynamically adjusting the power output of the charging pile, optimizing the switching of reactive power compensation, changing the switching frequency of the charging device, etc. In addition, due to the dynamic change of the power grid load environment, the model can also perform online learning or transfer learning regularly, that is, adaptive optimization based on newly collected data to ensure that its prediction ability will not decline due to long-term operation. Through this intelligent evaluation mechanism, the charging pile system can take proactive prevention and control measures before harmonic resonance occurs, reduce the risk of power grid instability, and improve the reliability and safety of charging equipment.

[0037] The deep learning model is not limited here, and any deep learning model that can realize comprehensive analysis of the harmonic impedance matching reference value and the harmonic power reflux reference value to generate a harmonic resonance risk coefficient is acceptable. To implement the technical solution of the present invention, the present invention provides a specific implementation method: The harmonic resonance risk coefficient is generated according to the following formula: , where, , is the reference value of harmonic impedance matching and the reference value of harmonic power backflow for the preset proportionality coefficients, and , are all greater than 0.

[0038] Preset proportionality coefficients and represent the weighting factors for different variables (reference value of harmonic impedance matching and reference value of harmonic power backflow ), which are used to balance their relative contributions in the calculation of the harmonic resonance risk coefficient ( ). Since the reference value of harmonic impedance matching and the reference value of harmonic power backflow may have different numerical ranges, physical meanings, and degrees of influence, directly adding or calculating them may cause the influence of one item to be amplified or weakened. Therefore, the proportionality coefficients and are introduced to ensure that the degrees of influence of the two in the final risk assessment meet the requirements of the actual power grid. These coefficients are usually determined based on historical data analysis, experimental verification, or machine learning model training, and may involve statistical regression, optimization algorithms, or tuning based on expert experience to ensure that the calculation results can accurately reflect the harmonic resonance risk in different environments.

[0039] It can be seen from the harmonic resonance risk coefficient that the larger the reference value of harmonic impedance matching generated by analyzing the matching situation of the equivalent impedance of the power grid at different frequency points under the detection window, and the larger the reference value of harmonic power backflow generated by analyzing the proportion of harmonic power fed back from the charging pile to the power grid under the detection window, the larger the harmonic resonance risk coefficient generated when the current harmonic state is intelligently evaluated through a pre-trained deep learning model, indicating that the harmonic frequency of the charging pile has reached the power grid resonance point; conversely, it indicates that the harmonic frequency of the charging pile has not reached the power grid resonance point.

[0040] Compare and analyze the harmonic resonance risk coefficient generated when the current harmonic state is intelligently evaluated through a pre-trained deep learning model with the pre-set reference threshold of the harmonic resonance risk coefficient to classify the current harmonic state of the charging pile. The classification steps are as follows: If the harmonic resonance risk coefficient is greater than the reference value threshold of the harmonic resonance risk coefficient, then classify the current harmonic of the charging pile as the harmonic frequency has reached the power grid resonance point; if the harmonic resonance risk coefficient is less than or equal to the reference value threshold of the harmonic resonance risk coefficient, then classify the current harmonic of the charging pile as the harmonic frequency has not reached the power grid resonance point.

[0041] When the evaluation result of the deep learning model indicates that the harmonic frequency of the charging pile reaches the grid resonance point, based on the dynamic changes of real-time harmonic characteristics, the power regulation strategy of the charging pile will be intelligently reconstructed, and its output power and the operating frequency of the converter will be precisely adjusted to actively guide the harmonic frequency away from the resonance interval. At the same time, the switching mode of the reactive power compensation device will be adaptively optimized to make it respond to the dynamic changes of the load, flexibly adjust the resonance frequency distribution, and build an "anti-harmonic resonance" intelligent barrier to ensure the stability of the power grid and the efficient operation of the charging pile; When the evaluation result of the deep learning model indicates that the harmonic frequency of the charging pile reaches the grid resonance point, based on the dynamic changes of real-time harmonic characteristics, the power regulation strategy of the charging pile will be intelligently reconstructed. At the same time, the specific steps for adaptively optimizing the switching mode of the reactive power compensation device to make it respond to the dynamic changes of the load are as follows: When the evaluation result indicates that the harmonic frequency of the charging pile reaches the grid resonance point, the output power of the charging pile and the operating frequency of the converter will be dynamically adjusted through an intelligent regulation strategy to guide the harmonic frequency away from the grid resonance point. The specific regulation method is as follows: The calculation expression for dynamically adjusting the power of the charging pile is as follows: , , where is the adjusted output power of the charging pile, is the original power of the charging pile before adjustment, is the power adjustment coefficient, which affects the rate of power decline , is the harmonic resonance risk coefficient, is the reference threshold of the harmonic resonance risk coefficient, is the set maximum harmonic resonance risk coefficient, which is used to normalize the adjustment amplitude, is the natural base, is the adjustment parameter, which is used to control the smoothness of power adjustment to ensure that there will be no drastic power changes during the adjustment process, is the set minimum power limit, which prevents the charging efficiency of the charging pile from decreasing too much due to excessive adjustment; The calculation expression for dynamically adjusting the switching frequency of the converter is as follows: , where is the adjusted switching frequency of the converter, is the original switching frequency of the converter, is the frequency adjustment coefficient, which determines the adjustment amplitude of the operating frequency of the converter, is the sine adjustment factor, which makes the frequency adjustment show a periodic buffer to avoid excessive electromagnetic interference caused by sudden changes. The output range of is between, ensuring that the adjustment amplitude will not exceed the limit; After adjusting the power and operating frequency of the charging pile, it is still necessary to optimize the switching strategy of the reactive power compensation device to dynamically adjust the resonance frequency of the power grid, keep it away from the current harmonic frequency. The optimization formula is as follows: The dynamic optimization calculation expression of the reactive power compensation capacity is as follows; , , where, Among them: is the reactive power compensation capacity after adjustment, is the reactive power compensation capacity before adjustment, is the reactive power compensation adjustment coefficient , is the suppression parameter to prevent excessive switching of the reactive power compensation device and affect voltage stability, is the minimum allowable value of reactive power compensation to avoid excessive reduction of reactive power compensation leading to a decrease in power factor; The dynamic adjustment calculation expression of the equivalent inductance and capacitance is as follows: , , where, is the system equivalent inductance after adjustment, is the original inductance value, is the inductance adjustment coefficient, which determines the adjustment amplitude, is the arctangent adjustment factor to ensure that the inductance adjustment increases gradually and prevent sudden changes, is the equivalent capacitance after adjustment, is the original capacitance value, is the capacitance adjustment coefficient to control the reduction amount of the compensation capacitor, is the exponential decay factor to ensure smooth capacitance adjustment and prevent a drastic impact on the power grid compensation ability, is the exponential decay parameter used to control the smoothness and adjustment rate of the capacitance adjustment process.

[0042] Through intelligent dynamic regulation, harmonic resonance is suppressed in real time to ensure the stability of the power grid and optimize the operating efficiency of the charging pile. When the deep learning model detects that the harmonic frequency of the charging pile approaches or reaches the resonance point of the power grid, without intervention, the reactive power compensation devices (such as capacitor banks or SVG) in the power grid may resonate with specific harmonic frequencies, resulting in local voltage distortion, overload of the charging pile power module, and even triggering the power grid protection mechanism, causing a large-scale power outage. Therefore, this step aims to actively adjust the operating parameters of the charging pile through the coordinated optimization regulation of the charging pile and the reactive power compensation device, keep the harmonic frequency away from the resonance interval, and flexibly optimize the resonance characteristics of the power grid, thus avoiding the occurrence of resonance.

[0043] Specifically, this step first intelligently optimizes its power regulation strategy based on the dynamic changes in the harmonic characteristics of the charging pile. By adjusting the output power of the charging pile, the amplitude of specific harmonic components can be reduced, weakening their impact. At the same time, adjusting the operating frequency of the converter (such as adjusting the PWM switching frequency of the DC / DC or AC / DC converter) can shift the main harmonic frequency away from the resonance point of the power grid, thereby reducing the resonance risk. In addition, some charging piles can adopt power peak-shifting scheduling to avoid the resonance enhancement effect caused by multiple charging piles in the same time period, improving the overall harmonic distribution balance of the charging station.

[0044] Meanwhile, this step also dynamically optimizes the switching mode of reactive power compensation equipment to adapt to the changes in the power grid load. For example, when it is detected that a certain harmonic frequency matches the resonance point of the power grid, the switching strategy of the capacitor bank can be appropriately adjusted to change the equivalent impedance of the reactive power compensation equipment, so that the resonance frequency of the system shifts and avoids the main harmonic frequency of the charging pile. In addition, if the power grid is equipped with SVG (Static Var Generator) or TSC (Thyristor Switched Capacitor), the reactive power compensation amount can be optimized through intelligent control, and the equivalent resonance frequency of the power grid can be adjusted in real time and dynamically, so that it is always in a safe range far from the harmonic frequency of the charging pile.

[0045] Through the collaborative intelligent optimization of the charging pile and the reactive power compensation equipment, this step can not only effectively avoid the occurrence of harmonic resonance, reduce the risk of equipment damage caused by power grid fluctuations, but also improve the overall operating stability of the charging pile, reduce the power loss caused by harmonics, and make the power utilization rate higher. In addition, this strategy can also enhance the overall adaptability of the smart grid, enabling it to dynamically respond to the changes in harmonic characteristics under different load environments, form an "anti-harmonic resonance" intelligent barrier, and ensure the safety and efficient operation of the charging network in the underground parking lot.

[0046] The present invention effectively solves problems such as harmonic resonance, unbalanced power distribution, and grid instability that may occur during the operation of charging piles through real-time data collection, intelligent feature extraction, deep learning evaluation, and adaptive dynamic regulation. First, the system collects key parameters during the operation of charging piles with high precision and constructs a standardized data set in the data platform to ensure the integrity and accuracy of the data. Then, through feature analysis and the detection window mechanism, it accurately identifies the key features indicating that the harmonic frequency reaches the grid resonance point, and uses a deep learning model for intelligent evaluation to predict the harmonic anomaly trend in advance. When the system detects that the harmonic frequency enters the dangerous range, the charging pile self-organizing network will intelligently reconstruct the power regulation strategy, actively adjust the output power of the charging pile and the operating frequency of the converter, optimize the load distribution, and at the same time dynamically adjust the switching mode of the reactive power compensation device to effectively reduce the risk of harmonic resonance and improve the power supply quality of the grid. The implementation of this solution enhances the adaptive ability of the charging pile group, not only improves the stability and safety of the charging system, but also reduces the grid harmonic pollution, improves the energy utilization efficiency in complex power environments such as underground parking lots in shopping malls, and makes the intelligent charging network have higher reliability, controllability, and intelligent scheduling capabilities.

[0047] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data for software simulation to get a formula closest to the actual situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0048] Only some exemplary embodiments of the present invention have been described by way of illustration above. Undoubtedly, for those of ordinary skill in the art, without departing from the spirit and scope of the present invention, the described embodiments can be modified in various different ways. Therefore, the above drawings and description are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

[0049] It should be noted that in this article, if there are relational terms such as first and second, they are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "including one..." does not exclude the existence of another identical element in the process, method, article or device including the element.

[0050] It should be understood that in various embodiments of the present application, the sequence numbers of the above processes do not imply the order of execution, and the order of execution of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.

[0051] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professionals can use different methods for each specific application to implement the described functions, but such implementation should not be considered to exceed the scope of the present application.

[0052] Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments, and will not be elaborated herein.

[0053] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0054] In addition, in each embodiment of the present application, the functional units can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit.

[0055] As described above, the above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present application, and all should be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0056] Only some exemplary embodiments of the present invention have been described above by way of illustration. Undoubtedly, for those of ordinary skill in the art, without departing from the spirit and scope of the present invention, the described embodiments can be modified in various different ways. Therefore, the above drawings and description are illustrative in nature and should not be construed as limiting the protection scope of the claims of the present invention.

Claims

1. A load control method based on a charging pile self-organizing network, characterized in that: The following steps are involved: First, various parameters of the charging pile during operation are collected in real time to ensure the acquisition of high-quality raw data, providing a solid data foundation for subsequent analysis and model evaluation; After completing the real-time data collection, the operating parameters of the charging piles are uniformly stored in the data platform to build a standardized data set. At the same time, the various parameters in the data set are pre-processed to ensure the integrity and accuracy of the data; Mining and extracting key features that characterize the harmonic frequency of charging piles reaching the grid resonance point from the preprocessed data, analyzing the extracted key features within the detection window, and quantifying the change trend of the harmonic frequency; The extracted key features are input into the pre-trained deep learning model, and the current harmonic status is intelligently evaluated through the deep learning model; When the deep learning model evaluation results show that the harmonic frequency of the charging pile reaches the grid resonance point, the power control strategy of the charging pile will be intelligently reconstructed based on the dynamic changes of real-time harmonic characteristics, and its output power and converter operating frequency will be accurately adjusted to actively guide the harmonic frequency away from the resonance range. At the same time, the switching mode of the reactive compensation equipment will be adaptively optimized to respond to dynamic changes in the load, flexibly adjust the resonant frequency distribution, and build an intelligent barrier to "avoid harmonic resonance" to ensure the stability of the power grid and the efficient operation of the charging pile.

2. The load control method based on charging pile ad hoc network according to claim 1 is characterized in that: First, high-precision sensors and smart meters are deployed at key nodes of each charging pile and distribution network to capture key power parameters of the charging piles; Secondly, use high-speed data acquisition modules to collect and encapsulate data in real time, ensuring sampling accuracy in seconds and avoiding information lag and loss; Then, a time synchronization mechanism is used to calibrate the timestamps of the data collected by different devices to ensure that all data points can be aligned under the same time reference; Finally, the real-time collected data is transmitted to the cloud or local server through wireless or wired communication networks and stored in the data platform to form a complete data set to support subsequent harmonic feature extraction, intelligent analysis and dynamic control decisions.

3. The load control method based on charging pile ad hoc network according to claim 1 is characterized in that: The key features that characterize the harmonic frequency of the charging pile reaching the resonance point of the power grid are mined and extracted from the preprocessed data. The extracted features include the matching of the equivalent impedance of the power grid at different frequency points and the proportion of harmonic power fed back by the charging pile to the power grid. The matching of the equivalent impedance of the power grid at different frequency points and the proportion of harmonic power fed back by the charging pile to the power grid are analyzed under the detection window, and the harmonic impedance matching reference value and the harmonic power return reference value are generated respectively. The changing trend of the harmonic frequency is quantified by the harmonic impedance matching reference value and the harmonic power return reference value.

4. The load control method based on charging pile ad hoc network according to claim 3 is characterized in that: The specific steps for analyzing the matching of the grid equivalent impedance at different frequency points in the detection window to generate the harmonic impedance matching reference value are as follows: In the power grid system, the equivalent impedance changes with frequency, forming an impedance spectrum characteristic. When the charging pile generates a harmonic frequency, the output impedance of the charging pile at the harmonic frequency point is matched with the equivalent impedance of the power grid. The impedance matching error is defined as a measure of the degree of adaptation between the power grid and the charging pile at the harmonic frequency point. The calculation expression is as follows: , where The power grid is at harmonic frequency The equivalent impedance at The charging pile is at the harmonic frequency The equivalent output impedance at is a very small positive number. is the harmonic impedance matching error; When calculating different harmonic frequency points Impedance matching error at After that, the harmonic impedance matching reference value is further defined to quantify the harmonic resonance risk of the entire system within the detection window. The calculation expression is as follows: , where is the harmonic impedance matching reference value, is the harmonic order range to be analyzed, The harmonic voltage is The power component at , characterizing the influence of harmonic energy, is the impedance matching adjustment factor, is the harmonic order, is the weight factor.

5. The load control method based on charging pile ad hoc network according to claim 3 is characterized in that: The specific steps for analyzing the proportion of harmonic power fed back from the charging pile to the grid under the detection window to generate the harmonic power return reference value are as follows: First, calculate the harmonic power of each order output by the charging pile to the grid and compare it with the total harmonic power to quantify the proportion of harmonic feedback. The harmonic power of the charging pile is expanded by Fourier series to extract the voltage and current components of different harmonic orders. Then calculate the power of each harmonic component and define the proportion of harmonic power fed back to the grid by the charging pile. The calculation expression is as follows: , where is the harmonic power feedback ratio, is the harmonic order range to be analyzed, is the harmonic order, from 2nd to the highest harmonic order , The charging pile is in the harmonic order The effective value of the voltage at is the harmonic order The effective value of the voltage at is the phase angle between the harmonic current and voltage at the grid end, is the phase angle between the harmonic current and voltage output by the charging pile; The proportion of harmonic power feedback from charging piles to the grid ,The harmonic power backflow reference value is constructed to comprehensively measure the feedback degree of different harmonic frequency components and their impact on the grid resonance risk. The calculation expression is as follows: , where is the harmonic power return reference value, is the nonlinear weighting factor of the feedback power, The harmonic order of the power grid The equivalent impedance at is the reference impedance, It is the weight parameter of the grid impedance on the harmonic amplification effect.

6. The load control method based on charging pile ad hoc network according to claim 3 is characterized in that: The harmonic impedance matching reference value and the harmonic power return reference value after quantitative analysis are input into the pre-learned deep learning model, and the harmonic resonance risk coefficient is generated by the deep learning model. The current harmonic status is intelligently evaluated by the harmonic resonance risk coefficient.

7. The load control method based on charging pile ad hoc network according to claim 6 is characterized in that: The harmonic resonance risk coefficient generated by the pre-learned deep learning model when the current harmonic state is intelligently evaluated is compared with the pre-set reference threshold of the harmonic resonance risk coefficient to divide the current charging pile harmonic state. The division steps are as follows: If the harmonic resonance risk coefficient is greater than the harmonic resonance risk coefficient reference value threshold, the current charging pile harmonic is classified as the harmonic frequency has reached the grid resonance point; if the harmonic resonance risk coefficient is less than or equal to the harmonic resonance risk coefficient reference value threshold, the current charging pile harmonic is classified as the harmonic frequency has not reached the grid resonance point.

8. The load control method based on charging pile ad hoc network according to claim 7 is characterized in that: When the deep learning model evaluation results show that the harmonic frequency of the charging pile reaches the grid resonance point, the power control strategy of the charging pile will be intelligently reconstructed based on the dynamic changes of real-time harmonic characteristics. At the same time, the switching mode of the reactive compensation equipment will be adaptively optimized to respond to the dynamic changes of the load. The specific steps are as follows: When the evaluation results show that the harmonic frequency of the charging pile reaches the grid resonance point, the output power of the charging pile and the operating frequency of the converter are dynamically adjusted through the intelligent control strategy to guide the harmonic frequency away from the grid resonance point. The specific control method is as follows: The calculation expression for dynamic adjustment of charging pile power is as follows: ， , where is the adjusted charging pile output power, is the original power of the charging pile before adjustment, is the power regulation factor, is the harmonic resonance risk factor, is the reference threshold of harmonic resonance risk factor, is the maximum harmonic resonance risk factor set, used to normalize the adjustment amplitude, is the natural base, is the adjustment parameter, It is the minimum power limit set to prevent the charging efficiency of the charging pile from decreasing too much due to over-adjustment; The calculation expression for the dynamic adjustment of the converter switching frequency is as follows: , where is the adjusted converter switching frequency, is the original converter switching frequency, is the frequency adjustment factor, is the sinusoidal regulation factor; After adjusting the charging pile power and operating frequency, it is still necessary to optimize the switching strategy of the reactive compensation equipment so that the resonant frequency of the power grid can be dynamically adjusted to stay away from the current harmonic frequency. The optimization formula is as follows: The dynamic optimization calculation expression of reactive compensation capacity is as follows; ， , where in: is the adjusted reactive power compensation capacity, is the reactive power compensation capacity before adjustment, is the reactive power compensation adjustment coefficient, is the suppression parameter, It is the minimum allowable value of reactive power compensation; The dynamic adjustment of equivalent inductance and capacitance calculation expressions are as follows: ， , where is the adjusted system equivalent inductance, is the original inductance value, is the inductance adjustment factor, is the inverse tangent adjustment factor, is the adjusted equivalent capacitance, is the original capacitance value, is the capacitance adjustment factor, is the exponential decay factor, is the exponential decay parameter.

Images