Electric vehicle charging and discharging optimization scheduling method

Abstract

The invention discloses an electric vehicle charging and discharging optimization scheduling method, and relates to the technical field of electric vehicle charging and discharging, and the method comprises the following steps: firstly, based on historical data analysis, setting the monitoring frequency of a harmonic detection device, and deploying the monitoring frequency at a key node of electric vehicle charging and discharging equipment for data collection; according to the invention, based on a machine learning model of historical data, harmonic dynamic change is intelligently evaluated, a dynamic harmonic change coefficient is generated, and normal and dynamic change harmonics are accurately divided. The monitoring frequency of dynamic harmonic waves is improved, abnormity is rapidly captured, and risks are controlled; normal harmonic waves are monitored according to the preset frequency, and resource consumption is reduced. The method optimizes the monitoring resource configuration, guarantees the stability of the power grid, reduces the equipment loss, improves the efficiency, and reduces the cost.

Description

Technical Field

[0001] The present invention relates to the technical field of electric vehicle charging and discharging, and in particular to an electric vehicle charging and discharging optimization scheduling method. Background Art

[0002] Optimal dispatching of electric vehicle charging and discharging refers to coordinating the charging and discharging behaviors of multiple electric vehicles in a power system or microgrid through intelligent algorithms and optimization models to maximize resource utilization efficiency and achieve economic and stable system operation. This process comprehensively considers factors such as grid load, energy cost, electric vehicle charging demand, user time preference, battery life, and volatility of renewable energy, and formulates a dynamic dispatching strategy to balance supply and demand, reduce peak-to-valley differences, reduce electricity costs, and improve the utilization rate of new energy, while ensuring that the charging needs of electric vehicle users are met.

[0003] The prior art has the following deficiencies:

[0004] The power electronic characteristics of electric vehicle charging and discharging equipment are connected to the power grid through switching power supplies (such as inverters and rectifiers), which generate nonlinear loads during operation, thereby introducing harmonic pollution. Especially when large-scale electric vehicles are charged and discharged at the same time, harmonic loads may be superimposed and amplified, which will have a significant impact on the power grid, among which low-order harmonics such as the third and fifth harmonics are particularly prominent. Existing technologies usually deploy harmonic detection devices at key nodes of the power grid (such as substation outlets and charging station grid connection points) to evaluate their impact by regularly measuring harmonic data. However, the propagation of harmonic pollution in the power grid is dynamic, and different load switching may change the harmonic propagation path, and fixed frequency measurement is difficult to capture these real-time changes. If fixed frequency monitoring is continued, it may not be possible to detect the superimposed peak of harmonics during high-load switching or abnormal equipment operation in time, resulting in a sharp increase in harmonic amplitude, causing transformers, cables, filters and other equipment to withstand currents and voltages beyond the design range. This will cause problems such as equipment overheating and insulation layer breakdown, which will not only significantly shorten the life of the equipment, but may even cause the equipment to burn out, further causing local or full-grid power system failures, with serious consequences.

[0005] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0006] The purpose of the present invention is to provide an electric vehicle charging and discharging optimization scheduling method. Through a machine learning model based on historical data, the system can intelligently evaluate the dynamic changes of harmonic currents, generate dynamic harmonic change coefficients, and accurately divide normal changes and dynamically changing harmonics, breaking through the limitations of traditional fixed frequency monitoring. The monitoring frequency of dynamically changing harmonics is increased to quickly capture abnormal characteristics, identify risks in a timely manner, and take control measures; normal changing harmonics are monitored at a preset frequency to reduce resource consumption and optimize economic efficiency. The method realizes efficient monitoring of resource allocation, ensures safe and stable operation of the power grid, reduces equipment losses, improves management efficiency, and reduces operating costs to solve the problems in the above-mentioned background technology.

[0007] In order to achieve the above object, the present invention provides the following technical solution: a method for optimizing the charging and discharging of electric vehicles, comprising the following steps:

[0008] First, based on historical data analysis, the monitoring frequency of the harmonic detection device is set and deployed at key nodes of electric vehicle charging and discharging equipment for data collection;

[0009] Organizing the collected harmonic current data into a data set, and extracting key features reflecting the dynamic changes of harmonic current from the data set;

[0010] Analyze and process the extracted key features, and input the analyzed key features into a pre-trained machine learning model, so as to intelligently evaluate the abnormal dynamic changes of harmonic current through the machine learning model;

[0011] Based on the evaluation results of the machine learning model, the change state of harmonic current is divided into two categories: normal change harmonics and dynamic change harmonics;

[0012] For normal changing harmonics, regular measurements will continue to be carried out at the preset monitoring frequency, and their impact on the power grid will be evaluated;

[0013] For dynamically changing harmonics, the monitoring frequency is increased according to the evaluation results. Through higher frequency monitoring, abnormal changes in harmonic currents can be quickly captured, and corresponding control measures can be taken.

[0014] Preferably, key features reflecting the dynamic changes of harmonic currents are extracted from the data set, including the amplitude change ratio between different harmonic frequency components and the synchronization of harmonic currents in space and time. Within the detection window, the amplitude change ratio between different harmonic frequency components and the synchronization of harmonic currents in space and time are analyzed and processed to generate a harmonic amplitude-frequency ratio change index and a harmonic space-time synchronization index, respectively. The harmonic amplitude-frequency ratio change index quantifies the energy distribution and changes of the harmonic spectrum by analyzing the amplitude change ratio between different harmonic frequency components; the harmonic space-time synchronization index is used to quantify the synchronization of harmonic currents in space and time, and to reveal the changing relationship between harmonic currents in different monitoring nodes and time periods.

[0015] Preferably, after obtaining the harmonic amplitude-frequency ratio change index and the harmonic space-time synchronization index generated after analyzing and processing the extracted key features, the harmonic amplitude-frequency ratio change index and the harmonic space-time synchronization index are input into a pre-learned machine learning model, and a dynamic harmonic change coefficient is generated through the machine learning model. The dynamic harmonic change coefficient is used to perform an intelligent evaluation of the abnormal dynamic changes of harmonic current.

[0016] Preferably, the harmonic variation coefficient generated after analyzing the extracted key features is compared and analyzed with a preset reference threshold value of the harmonic variation coefficient, and the change state of the harmonic current is divided, and the division steps are as follows:

[0017] If the harmonic variation coefficient is greater than or equal to a preset harmonic variation coefficient reference threshold, the variation state of the harmonic current is classified as a dynamically varying harmonic;

[0018] If the harmonic variation coefficient is less than a preset harmonic variation coefficient reference threshold, the variation state of the harmonic current is classified as a normal variation harmonic.

[0019] Preferably, the monitoring frequency is increased according to the evaluation results, and the abnormal changes of harmonic currents can be quickly captured through higher frequency monitoring, and then the specific steps of taking corresponding control measures are as follows:

[0020] If the harmonic current is determined to be a dynamically changing harmonic, the monitoring frequency will be adjusted according to the value of the current harmonic change coefficient HVC, thereby improving the sensitivity of data acquisition. The calculation expression is as follows:

[0021] ,

[0022] In the formula, f new is the adjusted monitoring frequency, used to collect harmonic data at a higher frequency, f preset is the preset monitoring frequency, HVC ref is the preset reference threshold of harmonic variation coefficient, and α is the adjustment coefficient, which determines the sensitivity of monitoring frequency increase;

[0023] After adjusting the monitoring frequency, start to use the adjusted monitoring frequency f new Collect harmonic current data and perform real-time analysis to further determine the impact range and propagation path of dynamic harmonics. The calculation expression is as follows:

[0024] ,

[0025] Where D dynamic is the dynamic feature capture value, t 1 is the monitoring start time, t 2 is the monitoring end time, HAFRCI is the harmonic amplitude-frequency ratio change index, HSSI is the harmonic space-time synchronization index, It is the rate of change of the harmonic amplitude-frequency ratio change index over time, indicating the rapid fluctuation degree of the harmonic amplitude-frequency ratio change index HAFRCI over time within the detection window. It is the rate of change of the harmonic space-time synchronization index with respect to time, indicating the degree of synchronization destruction of the harmonic space-time synchronization index, HSSI, within the detection window;

[0026] According to the dynamic capture value D dynamic Based on the analysis results of the harmonic variation coefficient HVC, further control measures are taken to alleviate the impact of dynamic harmonics on the power grid. The calculation expression is as follows:

[0027] U control =β·max(HVC,D dynamic )

[0028] , where U control is the strength of the control signal, β is the control gain parameter, max(HVC,Dd ynamic ) is the maximum value operation.

[0029] Preferably, in the detection window, the specific steps of analyzing the amplitude change ratio between different harmonic frequency components and generating the harmonic amplitude-frequency ratio change index are as follows:

[0030] In the detection window, the fast Fourier transform is used to convert the collected time domain data into frequency domain data, and the amplitude information of different frequency components is extracted. The calculation expression is as follows:

[0031] ,

[0032] In the formula, x(t) is the input signal, which is the input signal at time point t, X(f) is the frequency domain signal, and e -j2πft is a complex exponential function, e is the natural base, j is the imaginary unit, and f is the frequency;

[0033] Calculate the amplitude change ratio between different harmonic frequency components. The calculation expression is as follows:

[0034] ,

[0035] In the formula, X(f i ,t) is the frequency f of the input signal x(t) after Fourier transformation at time point t i The complex spectrum at the frequency f i The amplitude and phase information of the frequency component at j ,t) is the frequency f of the input signal x(t) after Fourier transformation at time point t j The complex spectrum at the frequency f j The amplitude and phase information of the frequency component at R ij (t) is the frequency component f at time point t i and f j The amplitude ratio of

[0036] After calculating the amplitude ratio between different frequency components, the amplitude-harmonic-frequency ratio variation index is constructed. The calculation expression is as follows:

[0037] ,

[0038] Where HAFRCI is the harmonic amplitude-frequency ratio variation index, n is the total number of monitored frequencies, and R ij (t-1) is the frequency component f at the previous time t-1. i and f j The amplitude ratio of .

[0039] Preferably, within the detection window, the specific steps of analyzing the synchronization of harmonic currents in space and time and generating the harmonic spatiotemporal synchronization index are as follows:

[0040] Under the detection window, the harmonic amplitude and harmonic phase of each monitoring point are collected from multiple monitoring points to construct a harmonic current complex matrix, which is expressed as follows:

[0041] ,

[0042] In the formula, H k (t) is the complex harmonic current vector, which contains the complex current amplitude and phase of the kth harmonic frequency at time t. m,k (t) is the harmonic amplitude of monitoring point m, which means the amplitude of the kth harmonic frequency component at monitoring point m at time t, φ m,k (t) is the harmonic phase of monitoring point m, which means the phase angle of the kth harmonic frequency component located at monitoring point m at time t, M is the total number of monitoring points, e is the natural base, and j is the imaginary unit;

[0043] For each frequency component k, the synchronization covariance matrix between the monitoring points is calculated to quantify the correlation between the harmonic amplitude and phase of different monitoring points. The calculation expression is as follows:

[0044] ,

[0045] In the formula, C k (t) is the time-space synchronization matrix of the harmonic current, which represents the time-space synchronization matrix of the kth harmonic frequency component at time t. is the conjugate transpose of the complex vector of harmonic current;

[0046] In the time dimension, the synchronization offset vector of the kth harmonic frequency is calculated to quantify the synchronization difference of each monitoring point in time and space. The calculation expression is as follows:

[0047] ,

[0048] In the formula, Δ k (t) is the dynamic harmonic variation coefficient, which indicates the dynamic variation coefficient of the kth harmonic at time t, C p,pk (t) is the autocorrelation harmonic current matrix, which represents the autocorrelation matrix of the kth harmonic at the pth monitoring point, C p,qk (t) is the harmonic current correlation matrix across monitoring points, representing the correlation matrix of the kth harmonic between the pth and qth monitoring points;

[0049] The harmonic spatiotemporal synchronization index is generated by comprehensively quantifying the synchronization offset vectors of all frequency components in the detection window. The calculation expression is as follows:

[0050] ,

[0051] Where HSSI is the harmonic spatiotemporal synchronization index, K is the total number of harmonic frequency components, and w k is the weighting factor, which is the weighting factor of the kth harmonic and is used to adjust the dynamic change coefficient Δ of different harmonic frequency components. k (t)

[0052] Weighted.

[0053] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0054] The present invention uses a machine learning model based on historical data, and the system can intelligently evaluate the dynamic changes of harmonic currents, generate dynamic harmonic change coefficients, and compare them with reference thresholds to accurately divide normal change harmonics and dynamically change harmonics. This method breaks through the limitations of traditional fixed frequency measurement and can capture complex phenomena such as sudden changes in harmonic propagation paths or abnormal distribution of harmonic energy caused by load switching, power electronic equipment failures, etc. in real time. Especially for dynamically changing harmonics, by increasing the monitoring frequency, the system can quickly capture abnormal characteristics of harmonics, such as sudden increase in amplitude, spectrum expansion, and destruction of spatiotemporal synchronization, and timely identify potential risks. This highly sensitive monitoring and evaluation method provides important safety guarantees for key equipment in the power grid (such as transformers, cables, filters, etc.), avoids serious consequences such as equipment overheating, insulation layer breakdown, and even equipment burning caused by excessive harmonics, and effectively ensures the safety of the power grid and the long-term stable operation of equipment.

[0055] The present invention realizes efficient configuration of monitoring resources by dynamically adjusting the monitoring frequency. For normally changing harmonics, regular measurements are continued at the preset frequency, which avoids unnecessary high-frequency monitoring, reduces the overhead of computing and data storage, and ensures continuous tracking of the harmonic state of the power grid. For dynamically changing harmonics, through the accurate identification of machine learning models, the system can quickly increase the monitoring frequency, obtain more fine-grained data, and take targeted control measures based on real-time data analysis. This method of dynamically allocating monitoring resources according to the state of harmonic changes not only improves the monitoring efficiency, but also reduces the waste of resources and system overload problems that may be caused by traditional high-frequency monitoring. In addition, the classified management and precise processing of different harmonic states further optimizes the economy and management efficiency of power grid operation, and reduces operating costs while ensuring the stable operation of the power grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0057] Figure 1 The present invention is a method flow chart of an electric vehicle charging and discharging optimization scheduling method. DETAILED DESCRIPTION

[0058] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that the description of the present disclosure will be more comprehensive and complete, and the concept of the example embodiments will be fully conveyed to those skilled in the art.

[0059] The present invention provides Figure 1 An electric vehicle charging and discharging optimization scheduling method is shown, comprising the following steps:

[0060] First, based on historical data analysis, the monitoring frequency of the harmonic detection device is set and deployed at key nodes of electric vehicle charging and discharging equipment (such as charging station grid connection points and power electronic equipment output terminals) for data collection;

[0061] When setting the monitoring frequency of the harmonic detection device, it is first necessary to understand the harmonic change pattern during the charging and discharging process of electric vehicles by analyzing historical data. These historical data may include the harmonic current characteristics generated under different charging loads and the operating conditions of different electric vehicle charging equipment. Through the time series analysis of these data, the regular fluctuations of harmonic currents and possible abnormal fluctuations can be identified. When setting the monitoring frequency, the appropriate sampling frequency should be selected according to the fluctuation characteristics of the harmonics so as to accurately capture the harmonic changes in different operating environments. For example, for stable operation under low load, the monitoring frequency can be set lower; when the load is high or the charging equipment is frequently switched, the monitoring frequency needs to be increased to capture possible harmonic mutations.

[0062] Harmonic detection devices should be deployed at key nodes of electric vehicle charging and discharging equipment, such as charging station grid connection points and power electronic equipment output terminals, to ensure that all harmonic currents from the equipment to the grid can be monitored in real time. These nodes are the core locations where the grid is connected to charging and discharging equipment, and are also the main source of harmonic pollution that may be generated and propagated. At the charging station grid connection point, harmonic pollution may spread to other areas through the grid, so monitoring at this location is crucial; and the output terminal of the power electronic equipment is where the charging and discharging equipment directly generates harmonic currents. By deploying detection devices at these key nodes, harmonic data can be collected in a timely and accurate manner, ensuring that harmonic pollution in the grid can be analyzed and controlled in a targeted manner.

[0063] Organizing the collected harmonic current data into a data set, and extracting key features reflecting the dynamic changes of harmonic current from the data set;

[0064] Analyze and process the extracted key features, and input the analyzed key features into a pre-trained machine learning model, so as to intelligently evaluate the abnormal dynamic changes of harmonic current through the machine learning model;

[0065] Key features reflecting the dynamic changes of harmonic currents are extracted from the data set, including the amplitude change ratio between different harmonic frequency components and the synchronization of harmonic currents in space and time. Within the detection window, the amplitude change ratio between different harmonic frequency components and the synchronization of harmonic currents in space and time are analyzed and processed to generate the harmonic amplitude-frequency ratio change index and the harmonic spatiotemporal synchronization index respectively. The harmonic amplitude-frequency ratio change index quantifies the energy distribution and changes of the harmonic spectrum by analyzing the amplitude change ratio between different harmonic frequency components; the harmonic spatiotemporal synchronization index is used to quantify the synchronization of harmonic currents in space and time, and reveal the changing relationship between harmonic currents in different monitoring nodes and time periods.

[0066] After obtaining the harmonic amplitude-frequency ratio variation index and the harmonic space-time synchronization index generated by analyzing and processing the extracted key features, the harmonic amplitude-frequency ratio variation index and the harmonic space-time synchronization index are input into a pre-learned machine learning model, and a dynamic harmonic variation coefficient is generated through the machine learning model. The dynamic harmonic variation coefficient is used to perform an intelligent evaluation of the abnormal dynamic changes of harmonic currents;

[0067] A pre-learned machine learning model refers to a model that is trained and optimized through a large amount of historical data or simulation data, and the model can infer and predict new data. In this solution, the machine learning model learns the patterns and laws of the dynamic changes of harmonic currents and the relationship between them and the operating conditions of the power grid (such as load changes, switching of power electronic equipment, etc.) by analyzing the annotated harmonic data sets. The core task of this model is to capture and identify the characteristic changes of harmonic currents, including the changing trends of parameters such as amplitude changes, spectrum distribution, phase drift, and spatiotemporal synchronization. Through in-depth learning of these features, the machine learning model can accurately evaluate new harmonic current data, determine whether there are abnormal dynamic changes, and output the corresponding dynamic harmonic change coefficients according to preset standards. These change coefficients provide an intelligent assessment of the state of harmonic propagation, helping the system identify potential harmonic problems and power grid risks.

[0068] The training process of machine learning models usually involves two major steps: data preprocessing and model optimization. In the data preprocessing stage, the collected harmonic current data is first cleaned, denoised and standardized so that the data can be more effectively input into the model. Then, through feature extraction, the model extracts key features that reflect the dynamic changes of harmonics from the original data, such as changes in amplitude-frequency ratio, spatiotemporal synchronization, etc., and inputs these features into the model. In the model optimization stage, the machine learning algorithm is trained using annotated historical data sets so that the model can recognize different harmonic patterns and learn the relationship between different features and harmonic anomalies. Common machine learning methods include decision trees, support vector machines (SVMs), random forests, deep neural networks (DNNs), etc. These methods can be flexibly adjusted according to the complexity and nonlinear characteristics of the data to ensure that the model has high accuracy and robustness in practical applications.

[0069] By pre-learning the machine learning model, the system can realize the intelligent evaluation of harmonic current. In real-time operation, after the model receives the new harmonic amplitude-frequency ratio change index and the harmonic spatiotemporal synchronization index, it begins to infer these input data. Based on the knowledge it has learned, the machine learning model analyzes the similarity between the current harmonic current characteristics and historical data, and determines whether the current power grid operation status is abnormal and whether there are potential harmonic propagation problems. The dynamic harmonic change coefficient output by the model can characterize the degree of dynamic change of harmonic current and classify different types of changes according to preset thresholds. For example, if the dynamic harmonic change coefficient exceeds the set normal range, the model will mark it as a harmonic anomaly and further analyze the cause of the anomaly, such as sharp fluctuations in power grid load, failure of power electronic equipment, or harmonic fluctuations caused by equipment switching.

[0070] The advantage of this intelligent assessment is that the machine learning model can monitor harmonic currents and predict anomalies in real time, avoiding the limitations of manually setting fixed frequencies in traditional methods. Over time, the performance of the model will gradually improve through continuous interaction and feedback with actual data, and it will be able to adapt to the complex situations in different power grid operating environments. Especially in the scenario of large-scale electric vehicle charging and discharging connected to the grid, the harmonic pollution pattern of the power grid may change significantly. Traditional static measurement and fixed frequency monitoring methods may find it difficult to effectively identify these dynamic changes, while machine learning models can capture these subtle changes, issue warnings in a timely manner, and take corresponding adjustment measures. Through this real-time, intelligent assessment method, the power system can prevent harmonic problems before they develop into large-scale failures, ensuring the stability and safety of power grid operation.

[0071] The change in the amplitude change ratio between different harmonic frequency components can indeed indicate that the harmonic propagation path is changing dynamically. When harmonics propagate in the power grid, the propagation characteristics of different frequency components may be affected by many factors, such as load changes, start-up and shutdown of power electronic equipment, adjustment of power grid topology, and fluctuations of other nonlinear loads. These changes will cause the amplitudes of different harmonic frequency components to change in different proportions in time and space. For example, low-order harmonics may become more significant in some cases due to changes in grid impedance, while the propagation of high-order harmonics may be suppressed due to the filtering effect of the equipment. When the amplitude change ratio between these frequency components changes, it means that the harmonic characteristics of some parts of the power grid have changed significantly, which may be caused by changes in equipment status, adjustments in grid load distribution, or changes in harmonic propagation paths. These dynamic changes not only reflect the changes in the distribution characteristics of harmonics in the power grid, but also indicate that the propagation path, reflection, refraction and other phenomena of harmonics have been adjusted in the power grid, thereby affecting the overall power quality. Therefore, the change in the amplitude change ratio can be used as an important indicator of the dynamic changes in the harmonic propagation path, revealing the interaction between the propagation behavior of harmonics in the power grid and the environment.

[0072] In the detection window, the amplitude change ratio between different harmonic frequency components is analyzed, and the specific steps for generating the harmonic amplitude-frequency ratio change index are as follows:

[0073] In the detection window, the fast Fourier transform is used to convert the collected time domain data into frequency domain data, and the amplitude information of different frequency components is extracted. The calculation expression is as follows:

[0074] ,

[0075] In the formula, x(t) is the input signal, which is the input signal at time point t, and is a signal in the time domain. In other words, x(t) is a function that describes the change of the signal over time. X(f) is the frequency domain signal, which is the result obtained by Fourier transform and is the representation of the signal in the frequency domain. -j2πft is a complex exponential function, e is the natural base, j is the imaginary unit, and f is the frequency;

[0076] By using Fast Fourier Transform (FFT) to convert the collected time domain data into frequency domain data and extracting the amplitude information of different frequency components, the energy distribution of the signal at different frequencies can be revealed. This is crucial for analyzing the harmonic pollution generated during the charging and discharging process of electric vehicles. By converting the time domain signal into the frequency domain signal, the amplitude of each frequency component in the signal can be clearly identified, especially the intensity of the harmonic component. For example, the power electronic devices of electric vehicle charging piles will generate specific harmonic frequencies, such as third and fifth harmonics. If these frequency components exceed the normal range, they may have adverse effects on the power grid. By extracting the amplitude information of these frequency components, the intensity and change trend of the harmonics can be quantitatively analyzed, which can help optimize the charging and discharging scheduling strategy, reduce the harmonic pollution to the power grid, and ensure the stability and power quality of the power grid.

[0077] Calculate the amplitude change ratio between different harmonic frequency components. The calculation expression is as follows:

[0078] ,

[0079] In the formula, X(f i ,t) is the frequency f of the input signal x(t) after Fourier transformation at time point t i The complex spectrum at the frequency f i The amplitude and phase information of the frequency component at j ,t) is the frequency f of the input signal x(t) after Fourier transformation at time point t j The complex spectrum at the frequency f j The amplitude and phase information of the frequency component at R ij (t) is the frequency component f at time point t i and f j The amplitude ratio of

[0080] Frequency component f i and f j The amplitude ratio R ij (t) is used to measure the intensity relationship of different frequency components at a specific moment. Its role is to reveal the relative changes of each frequency component in the signal, especially in harmonic analysis, it can reflect the dynamic changes of harmonic propagation paths in the power system. When charging equipment or power electronic devices generate harmonics, some frequency components may become stronger or weaker, resulting in significant changes in the amplitude ratio between frequency components. By monitoring these changes, abnormal fluctuations in harmonic currents can be identified, and potential risks that may cause overload, overheating or failure of power equipment can be discovered in a timely manner, and then optimized scheduling and suppression measures can be taken to ensure the power quality and stability of the power grid. Therefore, the amplitude ratio R ij(t) is an important indicator used to monitor and evaluate the propagation and changing dynamics of harmonics in power systems in real time.

[0081] After calculating the amplitude ratio between different frequency components, the amplitude-harmonic-frequency ratio variation index is constructed. The calculation expression is as follows:

[0082] ,

[0083] Where HAFRCI is the harmonic amplitude-frequency ratio variation index, n is the total number of monitored frequencies, and R ij (t-1) is the frequency component f at the previous time t-1. i and f j The amplitude ratio of

[0084] Under the detection window, the larger the value of the harmonic amplitude-frequency ratio change index generated after analyzing the amplitude change ratio between different harmonic frequency components, the more it indicates that the harmonic propagation path is changing dynamically. The propagation path of harmonic current is not only affected by the grid load, but may also be dynamically adjusted by factors such as the equipment operating status, topology, and filter effect. When the harmonic amplitude-frequency ratio change index is large, it usually means that the amplitude ratio between different frequency components has changed significantly, which reflects that the propagation path of harmonics in the grid has changed. For example, some frequency components may be amplified or suppressed under certain conditions, indicating that the harmonic propagation mode in the grid has experienced dynamic fluctuations, which may be caused by load fluctuations, power equipment switching, or grid topology adjustments. On the contrary, if the amplitude-frequency ratio change index is small, it means that the spectrum distribution of the harmonic current is relatively stable, and the amplitude ratio between the frequency components changes little, which usually indicates that the harmonic propagation path has not changed significantly and is within the normal range of variation.

[0085] Abnormal changes in the synchronization of harmonic currents in space and time usually indicate that the harmonic propagation path is changing dynamically. The propagation of harmonic currents in the power grid is affected by many factors, including load changes, the working state of power electronic equipment, and the adjustment of the power grid topology. When the synchronization of harmonic currents at different locations is abnormal, it means that the changes in harmonic amplitude and phase at different monitoring points are no longer consistent, which is usually due to changes in the harmonic propagation path. For example, load fluctuations or switching of power equipment may cause the path of harmonic current propagation in the power grid to shift or refract, which in turn causes the harmonic signals received by each monitoring point to be out of sync in time and space. This synchronization anomaly reflects that the propagation of harmonics is not only affected by the power grid structure, but also may change the propagation mode of harmonics due to local faults or other nonlinear phenomena in the power grid. Therefore, by monitoring the synchronization of harmonic currents in space and time, the dynamic changes of the harmonic propagation path can be captured in real time, potential power grid problems can be identified in advance, and countermeasures can be taken in time to prevent harmonic pollution from harming the power grid.

[0086] Under the detection window, the synchronization of harmonic current in space and time is analyzed, and the specific steps of generating the harmonic space-time synchronization index are as follows:

[0087] Under the detection window, the harmonic amplitude and harmonic phase of each monitoring point are collected from multiple monitoring points to construct a harmonic current complex matrix, which is expressed as follows:

[0088] ,

[0089] In the formula, H k (t) is the complex harmonic current vector, which contains the complex representation of the current amplitude and phase of the kth harmonic frequency at time t. Each element of the vector represents the harmonic current data at a monitoring point m. m,k (t) is the harmonic amplitude of monitoring point m, which means the amplitude of the kth harmonic frequency component at monitoring point m at time t, φ m,k (t) is the harmonic phase of monitoring point m, which means the phase angle of the kth harmonic frequency component located at monitoring point m at time t, M is the total number of monitoring points, e is the natural base, and j is the imaginary unit;

[0090] The harmonic amplitude reflects the strength of each harmonic frequency component in the current or voltage, that is, the degree of their impact on the overall power system. The larger the amplitude of each harmonic component, the more significant the contribution of this frequency component to the power grid, which usually corresponds to more harmonic interference. By analyzing the harmonic amplitude, it can help identify which frequency band in the system has more serious harmonic pollution, and take corresponding control measures, such as adding filters or adjusting the operation mode of the equipment to reduce the amplitude of specific frequency components. The change in harmonic amplitude is also an important indicator for assessing the health of the power grid. If the harmonic amplitude increases significantly, it may indicate that the nonlinearity of the power grid load has increased, or that some power electronic equipment has abnormalities, resulting in a decrease in power quality.

[0091] Harmonic phase is mainly used to describe the relative displacement of current or voltage waveforms at different time points, and is a key parameter for analyzing the propagation and synchronization of harmonic signals. Phase information can reveal the time deviation and spatial distribution differences of harmonic currents between different monitoring points, which is of great significance for determining whether harmonic currents propagate synchronously at different locations. By monitoring phase changes, changes in harmonic propagation paths in the power grid can be analyzed, such as whether there is a phase imbalance when harmonics propagate between different areas, or whether the phase is out of sync due to load changes. This helps to identify possible harmonic propagation problems in advance, such as faults caused by harmonic interference between electrical equipment, or harmonic phase differences caused by uncoordinated operation of equipment.

[0092] For each frequency component k, the synchronization covariance matrix between the monitoring points is calculated to quantify the correlation between the harmonic amplitude and phase of different monitoring points. The calculation expression is as follows:

[0093] ,

[0094] In the formula, C k (t) is the time-space synchronization matrix of the harmonic current, which represents the time-space synchronization matrix of the kth harmonic frequency component at time t. is the conjugate transpose of the complex vector of harmonic current;

[0095] In the time dimension, the synchronization offset vector of the kth harmonic frequency is calculated to quantify the synchronization difference of each monitoring point in time and space. The calculation expression is as follows:

[0096] ,

[0097] In the formula, Δ k (t) is the dynamic harmonic variation coefficient, which indicates the dynamic variation coefficient of the kth harmonic at time t, C p,pk (t) is the autocorrelation harmonic current matrix, which represents the autocorrelation matrix of the kth harmonic at the pth monitoring point. It is usually obtained by performing an inner product operation on the harmonic current signal of the monitoring point and its own complex conjugate. Cp,qk (t) is the cross-monitoring point harmonic current correlation matrix, which represents the correlation matrix of the kth harmonic between the pth and qth monitoring points, and is usually obtained by performing a complex inner product operation on the harmonic current signal between the two monitoring points;

[0098] Autocorrelation harmonic current matrix C p,pk (t) quantifies the time synchronization of harmonic currents at the same monitoring point. It provides information on the stability and consistency of the harmonic currents at that point. When the autocorrelation is strong, it indicates that the harmonic currents at that monitoring point change more synchronously in time, otherwise it indicates that the harmonic current changes more randomly.

[0099] Cross-monitoring point harmonic current correlation matrix C p,qk (t) measures the synchronization of the harmonic currents between the two monitoring points. If the value is large, it means that the harmonic current phase and amplitude changes of the two monitoring points are relatively consistent, indicating the stability of the harmonic propagation path; if the value is small, it may indicate that the harmonic currents between the two points are not synchronized, indicating that the harmonic propagation path may change.

[0100] The dynamic harmonic variation coefficient is an indicator used to quantify the temporal and spatial synchronization changes of harmonic currents in the power grid. It reflects the degree of dynamic changes in the propagation path of harmonics between monitoring points. When the propagation path of harmonic currents changes or the synchronization of harmonic currents between different points is abnormal, the dynamic harmonic variation coefficient will fluctuate significantly. Its role is to timely identify potential harmonic problems in the power grid by detecting these changes, such as changes in harmonic propagation paths caused by equipment failures, load mutations, or grid structure adjustments, thereby providing intelligent and real-time decision support for the dispatch and maintenance of the power system, which helps to improve the stability and power quality of the power grid.

[0101] The harmonic spatiotemporal synchronization index is generated by comprehensively quantifying the synchronization offset vectors of all frequency components in the detection window. The calculation expression is as follows:

[0102] ,

[0103] Where HSSI is the harmonic spatiotemporal synchronization index, K is the total number of harmonic frequency components, and w k is the weighting factor, which is the weighting factor of the kth harmonic and is used to adjust the dynamic change coefficient Δ of different harmonic frequency components. k (t) is weighted.

[0104] Under the detection window, the larger the value of the harmonic spatiotemporal synchronization index generated after analyzing the synchronization of harmonic currents in space and time, the more it indicates that the harmonic propagation path is changing dynamically. This index quantifies the synchronization of harmonic currents in space and time, and reflects the similarity or consistency of harmonic signals between different monitoring points and times. When the synchronization of harmonic currents is abnormal, it means that the propagation path of the harmonics may be disturbed or changed, such as load fluctuations, startup or switching of power electronic equipment, and adjustment of grid topology. These changes will cause the harmonic currents to be asynchronous at different locations and times, thereby increasing the value of the harmonic spatiotemporal synchronization index, indicating that the harmonic propagation path has changed dynamically. On the contrary, if the synchronization of harmonic currents in space and time remains stable, the propagation path of the harmonics is relatively normal.

[0105] The machine learning model is not limited here, and any machine learning model that can generate a harmonic variation coefficient HVC after comprehensive analysis of the harmonic amplitude-frequency ratio variation index HAFRCI and the harmonic spatiotemporal synchronization index HSSI is acceptable. In order to implement the technical solution of the present invention, the present invention provides a specific implementation method;

[0106] The formula for generating the harmonic variation coefficient HVC is as follows:

[0107] ,

[0108] In the formula, f 1 、f 2 are the preset proportional coefficients of the harmonic amplitude-frequency ratio variation index HAFRCI and the harmonic space-time synchronization index HSSI, respectively, and f 1 、f 2 Both are greater than 0.

[0109] The preset proportionality factor f 1 and f 2 Refers to the artificially set weight coefficients when calculating the harmonic variation coefficient HVC in order to integrate the influence of different characteristics (i.e. the relative importance of the harmonic amplitude-frequency ratio variation index HAFRCI and the harmonic spatiotemporal synchronization index HSSI). Their role is to adjust the proportion of these two characteristics in the formula, so as to more reasonably reflect their comprehensive contribution to the dynamic change of harmonics.

[0110] For example, if f 1 >f 2 , which means that in the current scenario, the change of harmonic amplitude-frequency ratio HAFRCI has a greater impact on the dynamic change of harmonics; if f 2 >f 1 , it means that the change of harmonic spatiotemporal synchronization HSSI has a more significant impact on dynamic harmonics. 1 and f 2, the importance of different features can be highlighted according to specific application scenarios (such as power grid environment, load type, etc.), ensuring that the harmonic variation coefficient HVC generated by the model can more accurately quantify the abnormal dynamic changes of harmonic currents.

[0111] It can be seen from the calculation expression of the harmonic variation coefficient that, under the detection window, the larger the performance value of the harmonic amplitude-frequency ratio variation index generated after analyzing the amplitude variation ratio between different harmonic frequency components, the larger the performance value of the harmonic space-time synchronization index generated after analyzing the synchronization of harmonic currents in space and time, and the larger the performance value of the harmonic variation coefficient generated after analyzing the extracted key features under the detection window, indicating that the harmonic propagation path is changing dynamically, otherwise it means that the harmonic propagation path has not changed significantly and is within the normal range of variation.

[0112] Based on the evaluation results of the machine learning model, the change state of harmonic current is divided into two categories: normal change harmonics and dynamic change harmonics;

[0113] The harmonic variation coefficient generated after analyzing the extracted key features is compared with the preset reference threshold of the harmonic variation coefficient, and the change state of the harmonic current is divided. The division steps are as follows:

[0114] If the harmonic variation coefficient is greater than or equal to a preset harmonic variation coefficient reference threshold, the variation state of the harmonic current is classified as a dynamically varying harmonic;

[0115] If the harmonic variation coefficient is less than a preset harmonic variation coefficient reference threshold, the variation state of the harmonic current is classified as a normal variation harmonic;

[0116] Normally changing harmonics refer to the regular changes in frequency components and amplitude distribution caused by load operation or equipment switching during the propagation of harmonic currents, but these changes do not cause significant changes in the harmonic propagation path; dynamically changing harmonics refer to the phenomenon that the harmonic propagation path is significantly offset or frequently fluctuates due to factors such as load fluctuations, grid topology adjustments or equipment failures.

[0117] For normal changing harmonics, regular measurements will continue to be carried out at the preset monitoring frequency, and their impact on the power grid will be evaluated;

[0118] The purpose of regular measurement at the preset monitoring frequency for normal changing harmonics is to optimize the efficiency of resource use while ensuring the accuracy of harmonic monitoring. Since the propagation path and amplitude changes of normal changing harmonics are relatively stable and predictable, low-frequency monitoring is sufficient to capture their changing trends and impacts on the power grid without frequent sampling and calculation. This monitoring method can significantly reduce the system's computing burden and data storage pressure, avoid unnecessary waste of resources, and ensure continuous tracking of harmonic status, so that the power grid can maintain stable operation when the harmonic changes are small, saving monitoring costs and improving overall operating efficiency.

[0119] For dynamically changing harmonics, the monitoring frequency is increased according to the evaluation results. Through higher frequency monitoring, abnormal changes in harmonic currents can be quickly captured, and corresponding control measures can be taken.

[0120] The specific steps to increase the monitoring frequency according to the evaluation results, quickly capture abnormal changes in harmonic currents through higher-frequency monitoring, and then take corresponding control measures are as follows:

[0121] If the harmonic current is determined to be a dynamically changing harmonic, the monitoring frequency will be adjusted according to the value of the current harmonic change coefficient HVC, thereby improving the sensitivity of data acquisition. The calculation expression is as follows:

[0122] ,

[0123] In the formula, f new is the adjusted monitoring frequency, used to collect harmonic data at a higher frequency, f preset It is the preset monitoring frequency, usually used for regular sampling of normal changing harmonics. It is the initial monitoring frequency of the system. ref It is the preset reference threshold of harmonic variation coefficient, and α is the adjustment coefficient, which determines the sensitivity of monitoring frequency increase and is generally set according to the real-time response capability of the system;

[0124] The adjustment coefficient α is a key parameter used to control the adjustment amplitude of the harmonic monitoring frequency. It determines the sensitivity of the monitoring frequency to the harmonic variation coefficient HVC. Its role is to balance the response speed of the system and the consumption of monitoring resources. When the harmonic changes are large, a higher adjustment coefficient can quickly increase the monitoring frequency to capture abnormal changes; when the changes are small, a lower adjustment coefficient can avoid excessive frequency adjustment and waste of resources. The setting of the adjustment coefficient needs to comprehensively consider the real-time response capability of the system, the limitation of computing resources, and the dynamic characteristics of harmonic changes. It is usually determined based on historical data experience or through experimental optimization. For example, in high-load fluctuation scenarios, a larger adjustment coefficient can be set to improve the response sensitivity; in stable operation scenarios, the adjustment coefficient can be appropriately reduced to save resources and reduce system disturbances.

[0125] After adjusting the monitoring frequency, start to use the adjusted monitoring frequency f new Collect harmonic current data and perform real-time analysis to further determine the impact range and propagation path of dynamic harmonics. The calculation expression is as follows:

[0126] ,

[0127] Where D dynamic is the dynamic feature capture value, t 1 is the monitoring start time, t 2 is the monitoring end time, HAFRCI is the harmonic amplitude-frequency ratio change index, HSSI is the harmonic space-time synchronization index, It is the rate of change of the harmonic amplitude-frequency ratio change index over time, indicating the rapid fluctuation degree of the harmonic amplitude-frequency ratio change index HAFRCI over time within the detection window. It is the rate of change of the harmonic space-time synchronization index with respect to time, indicating the degree of synchronization destruction of the harmonic space-time synchronization index, HSSI, within the detection window;

[0128] Dynamic feature capture value D dynamic It is a comprehensive indicator used to quantify the intensity of dynamic changes in harmonic current within a specific time window. It captures the dynamic fluctuations of harmonic propagation path, amplitude distribution and synchronization by analyzing the rate of change of key harmonic characteristics (such as amplitude-frequency ratio change index HAFRCI and spatiotemporal synchronization index HSSI) over time. Its role is to provide an accurate description of the dynamic behavior of harmonics, especially in the case of load fluctuations, grid topology changes or equipment abnormalities, to provide data support for identifying harmonic anomalies, adjusting monitoring frequency and taking control measures. dynamic The combined impact of the harmonic amplitude-frequency ratio change rate and the synchronization change rate is quantified, which is an important reference for judging whether the abnormal harmonic change is significant.

[0129] According to the dynamic capture value D dynamic Based on the analysis results of the harmonic variation coefficient HVC, further control measures are taken to alleviate the impact of dynamic harmonics on the power grid. The calculation expression is as follows:

[0130] U control =β·max(HVC,D dynamic ),

[0131] Where U control is the strength of the control signal, β is the control gain parameter, max(HVC,D dynamic ) is the maximum value operation, which means the value D captured from the harmonic variation coefficient HVC and the dynamic characteristic dynamic Select the maximum value as the main basis for the current system response;

[0132] Control signal strength Ucontrol It is a quantitative indicator used to indicate the strength or priority of the system's response measures to the current harmonic abnormality. It is composed of the overall characteristics of the harmonic change (such as the harmonic change coefficient HVC and the dynamic feature capture value D dynamic ) is calculated comprehensively to guide specific control actions, such as adjusting the charging and discharging loads of electric vehicles, activating harmonic filtering devices, and optimizing the grid operation topology. The greater the strength of the control signal, the more serious the harmonic changes detected by the system, requiring faster and more powerful intervention measures to ensure the stability of grid operation and the safety of equipment. By generating control signals of appropriate strength, the system can dynamically respond to abnormal harmonic changes, quickly alleviate their impact on the grid, and avoid more serious fault spread.

[0133] The control gain parameter β is an adjustment coefficient used to amplify or suppress the control signal strength U control Response to input harmonic characteristic parameters. Its role is to balance the system's response sensitivity and the risk of over-control. When β is large, the system's response to harmonic anomalies will be significantly enhanced, which is suitable for rapid intervention in high-risk scenarios; when β is small, the system's response is relatively mild, which is suitable for scenarios with stable operating environments to avoid resource waste or system overload. By adjusting β, the system's control strategy can be flexibly set according to the actual needs of the grid's operating conditions. For example, in a grid environment with frequent load fluctuations or high sensitivity of power equipment, setting a larger β value can ensure a quick response of the system, while in a stable operating scenario, reducing the β value can save control costs and reduce system disturbances.

[0134] For dynamically changing harmonics, the role of increasing the monitoring frequency is to quickly capture abnormal changes in harmonic currents through more precise sampling and analysis, and to obtain dynamic changes in their propagation paths and amplitudes in a timely manner, providing accurate basis for subsequent grid control and protection measures. Dynamically changing harmonics often exhibit irregular and sudden characteristics, which may be caused by severe load fluctuations, grid topology adjustments, equipment failures, and other reasons. If the monitoring frequency is insufficient, the system may not be able to identify these anomalies in a timely manner, causing harmonic currents to exceed the safety threshold, thereby causing overload damage to equipment such as transformers and cables, and even causing regional power failures. By increasing the monitoring frequency, the system can obtain more detailed data at the early stage of abnormal changes, including amplitude mutation points, spectrum expansion range, and the degree of destruction of spatiotemporal synchronization, so as to identify the source of abnormalities and propagation directions as early as possible. In addition, this real-time, high-frequency monitoring method provides necessary data support for subsequent control measures, such as dynamically adjusting the load of charging and discharging equipment, activating harmonic filtering devices, or adjusting the grid topology to disperse harmonic pollution pressure. The core function of this step is to achieve early warning and precise control of dynamic harmonic changes, ensure the stability of power grid operation and the safety of equipment, and avoid greater economic losses and operational risks caused by harmonic problems turning into systemic failures.

[0135] The present invention uses a machine learning model based on historical data, and the system can intelligently evaluate the dynamic changes of harmonic currents, generate dynamic harmonic change coefficients, and compare them with reference thresholds to accurately divide normal change harmonics and dynamically change harmonics. This method breaks through the limitations of traditional fixed frequency measurement and can capture complex phenomena such as sudden changes in harmonic propagation paths or abnormal distribution of harmonic energy caused by load switching, power electronic equipment failures, etc. in real time. Especially for dynamically changing harmonics, by increasing the monitoring frequency, the system can quickly capture abnormal characteristics of harmonics, such as sudden increase in amplitude, spectrum expansion, and destruction of spatiotemporal synchronization, and timely identify potential risks. This highly sensitive monitoring and evaluation method provides important safety guarantees for key equipment in the power grid (such as transformers, cables, filters, etc.), avoids serious consequences such as equipment overheating, insulation layer breakdown, and even equipment burning caused by excessive harmonics, and effectively ensures the safety of the power grid and the long-term stable operation of equipment.

[0136] The present invention realizes efficient configuration of monitoring resources by dynamically adjusting the monitoring frequency. For normally changing harmonics, regular measurements are continued at the preset frequency, which avoids unnecessary high-frequency monitoring, reduces the overhead of computing and data storage, and ensures continuous tracking of the harmonic state of the power grid. For dynamically changing harmonics, through the accurate identification of machine learning models, the system can quickly increase the monitoring frequency, obtain more fine-grained data, and take targeted control measures based on real-time data analysis. This method of dynamically allocating monitoring resources according to the state of harmonic changes not only improves the monitoring efficiency, but also reduces the waste of resources and system overload problems that may be caused by traditional high-frequency monitoring. In addition, the classified management and precise processing of different harmonic states further optimizes the economy and management efficiency of power grid operation, and reduces operating costs while ensuring the stable operation of the power grid.

[0137] The above description is only by way of illustration of certain exemplary embodiments of the present invention. It is undoubted that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A method for optimizing the charging and discharging scheduling of electric vehicles, characterized in that: The following steps are involved: First, based on historical data analysis, the monitoring frequency of the harmonic detection device is set and deployed at key nodes of electric vehicle charging and discharging equipment for data collection; Organizing the collected harmonic current data into a data set, and extracting key features reflecting the dynamic changes of harmonic current from the data set; Analyze and process the extracted key features, and input the analyzed key features into a pre-trained machine learning model, so as to intelligently evaluate the abnormal dynamic changes of harmonic current through the machine learning model; Based on the evaluation results of the machine learning model, the change state of harmonic current is divided into two categories: normal change harmonics and dynamic change harmonics; For normal changing harmonics, regular measurements will continue to be carried out at the preset monitoring frequency, and their impact on the power grid will be evaluated; For dynamically changing harmonics, the monitoring frequency is increased according to the evaluation results. Through higher frequency monitoring, abnormal changes in harmonic currents can be quickly captured, and corresponding control measures can be taken.

2. The method for optimizing charging and discharging of electric vehicles according to claim 1, characterized in that: Key features reflecting the dynamic changes of harmonic currents are extracted from the data set, including the amplitude change ratio between different harmonic frequency components and the synchronization of harmonic currents in space and time. Within the detection window, the amplitude change ratio between different harmonic frequency components and the synchronization of harmonic currents in space and time are analyzed and processed to generate the harmonic amplitude-frequency ratio change index and the harmonic spatiotemporal synchronization index respectively. The harmonic amplitude-frequency ratio change index quantifies the energy distribution and changes of the harmonic spectrum by analyzing the amplitude change ratio between different harmonic frequency components; the harmonic spatiotemporal synchronization index is used to quantify the synchronization of harmonic currents in space and time, and reveal the changing relationship between harmonic currents in different monitoring nodes and time periods.

3. The method for optimizing charging and discharging of electric vehicles according to claim 2, characterized in that: After obtaining the harmonic amplitude-frequency ratio change index and the harmonic space-time synchronization index generated by analyzing and processing the extracted key features, the harmonic amplitude-frequency ratio change index and the harmonic space-time synchronization index are input into a pre-learned machine learning model, and a dynamic harmonic change coefficient is generated through the machine learning model. The dynamic harmonic change coefficient is used to perform an intelligent evaluation of the abnormal dynamic changes of harmonic current.

4. The method for optimizing charging and discharging of electric vehicles according to claim 3, characterized in that: The harmonic variation coefficient generated after analyzing the extracted key features is compared with the preset reference threshold of the harmonic variation coefficient, and the change state of the harmonic current is divided. The division steps are as follows: If the harmonic variation coefficient is greater than or equal to a preset harmonic variation coefficient reference threshold, the variation state of the harmonic current is classified as a dynamically varying harmonic; If the harmonic variation coefficient is less than a preset harmonic variation coefficient reference threshold, the variation state of the harmonic current is classified as a normal variation harmonic.

5. The method for optimizing charging and discharging of electric vehicles according to claim 4, characterized in that: The specific steps to increase the monitoring frequency according to the evaluation results, quickly capture abnormal changes in harmonic currents through higher-frequency monitoring, and then take corresponding control measures are as follows: If the harmonic current is determined to be a dynamically changing harmonic, the monitoring frequency will be adjusted according to the value of the current harmonic change coefficient HVC, thereby improving the sensitivity of data acquisition. The calculation expression is as follows: ， In the formula, f new is the adjusted monitoring frequency, used to collect harmonic data at a higher frequency, f presset is the preset monitoring frequency, HVC ref is the preset reference threshold of harmonic variation coefficient, and α is the adjustment coefficient, which determines the sensitivity of monitoring frequency increase; After adjusting the monitoring frequency, start to use the adjusted monitoring frequency f new Collect harmonic current data and perform real-time analysis to further determine the impact range and propagation path of dynamic harmonics. The calculation expression is as follows: ， Where D dynamic is the dynamic feature capture value, t1 is the monitoring start time, t2 is the monitoring end time, HAFRCI is the harmonic amplitude-frequency ratio change index, HSSI is the harmonic space-time synchronization index, It is the rate of change of the harmonic amplitude-frequency ratio change index over time, indicating the rapid fluctuation degree of the harmonic amplitude-frequency ratio change index HAFRCI over time within the detection window. It is the rate of change of the harmonic space-time synchronization index with respect to time, indicating the degree of synchronization destruction of the harmonic space-time synchronization index, HSSI, within the detection window; According to the dynamic capture value D dynamic Based on the analysis results of the harmonic variation coefficient HVC, further control measures are taken to alleviate the impact of dynamic harmonics on the power grid. The calculation expression is as follows: IN control =β·max(HVC,D dynamic ), Where U control is the strength of the control signal, β is the control gain parameter, max(HVC, D dynamic ) is the maximum value operation.

6. The method for optimizing charging and discharging of electric vehicles according to claim 2, characterized in that: In the detection window, the amplitude change ratio between different harmonic frequency components is analyzed, and the specific steps for generating the harmonic amplitude-frequency ratio change index are as follows: In the detection window, the fast Fourier transform is used to convert the collected time domain data into frequency domain data, and the amplitude information of different frequency components is extracted. The calculation expression is as follows: ， In the formula, x(t) is the input signal, which is the input signal at time point t, X(f) is the frequency domain signal, and e -j2πft is a complex exponential function, e is the natural base, j is the imaginary unit, and f is the frequency; Calculate the amplitude change ratio between different harmonic frequency components. The calculation expression is as follows: ， In the formula, X(f i , t) is the frequency f of the input signal x(t) after Fourier transformation at time point t i The complex spectrum at the frequency f i The amplitude and phase information of the frequency component at j , t) is the frequency f of the input signal x(t) after Fourier transformation at time point t j The complex spectrum at the frequency f j The amplitude and phase information of the frequency component at R ij (t) is the frequency component f at time point t i and f j The amplitude ratio of After calculating the amplitude ratio between different frequency components, the amplitude-harmonic-frequency ratio variation index is constructed. The calculation expression is as follows: ， Where HAFRCI is the harmonic amplitude-frequency ratio variation index, n is the total number of monitored frequencies, and R ij (t-1) is the frequency component f at the previous time t-1. i and f j The amplitude ratio of .

7. The method for optimizing charging and discharging of electric vehicles according to claim 2, characterized in that: Under the detection window, the synchronization of harmonic current in space and time is analyzed, and the specific steps of generating the harmonic space-time synchronization index are as follows: Under the detection window, the harmonic amplitude and harmonic phase of each monitoring point are collected from multiple monitoring points to construct a harmonic current complex matrix, which is expressed as follows: ， In the formula, H k (t) is the complex harmonic current vector, which contains the complex current amplitude and phase of the kth harmonic frequency at time t. m，k (t) is the harmonic amplitude of monitoring point m, which means the amplitude of the kth harmonic frequency component at monitoring point m at time t, φ m，k (t) is the harmonic phase of monitoring point m, which means the phase angle of the kth harmonic frequency component located at monitoring point m at time t, M is the total number of monitoring points, e is the natural base, and j is the imaginary unit; For each frequency component k, the synchronization covariance matrix between the monitoring points is calculated to quantify the correlation between the harmonic amplitude and phase of different monitoring points. The calculation expression is as follows: ， In the formula, C k (t) is the time-space synchronization matrix of the harmonic current, which represents the time-space synchronization matrix of the kth harmonic frequency component at time t. is the conjugate transpose of the complex vector of harmonic current; In the time dimension, the synchronization offset vector of the kth harmonic frequency is calculated to quantify the synchronization difference of each monitoring point in time and space. The calculation expression is as follows: ， In the formula, Δ k (t) is the dynamic harmonic variation coefficient, which indicates the dynamic variation coefficient of the kth harmonic at time t, C p，pk (t) is the autocorrelation harmonic current matrix, which represents the autocorrelation matrix of the kth harmonic at the pth monitoring point, C p，qk (t) is the harmonic current correlation matrix across monitoring points, representing the correlation matrix of the kth harmonic between the pth and qth monitoring points; The harmonic spatiotemporal synchronization index is generated by comprehensively quantifying the synchronization offset vectors of all frequency components in the detection window. The calculation expression is as follows: ， Where HSSI is the harmonic spatiotemporal synchronization index, K is the total number of harmonic frequency components, and w k is the weighting factor, which is the weighting factor of the kth harmonic and is used to adjust the dynamic change coefficient Δ of different harmonic frequency components. k (t) is weighted.

Images