Direct current charging pile voltage stability optimization method and device and storage medium

Abstract

The invention relates to the technical field of charging piles, and provides a DC charging pile voltage stability optimization method and device and a storage medium, and the method comprises the steps: obtaining voltage sequence data of a DC output end of a charging pile, carrying out the segmentation statistics of the voltage sequence data, and obtaining an optimization adjustment matrix and a voltage compensation factor matrix; and performing dynamic adjustment on a preset direct current output control strategy of the charging pile based on the optimization adjustment matrix to obtain an adjusted control strategy parameter, performing compensation calculation on the adjusted control strategy parameter by using the voltage compensation factor matrix to obtain an optimization control instruction, and generating a dynamic voltage stabilization control signal according to the optimization control instruction. An output control strategy is dynamically adjusted by optimizing an adjustment matrix and a voltage compensation factor matrix, an efficient dynamic voltage stabilization control signal is generated, and the problems of response lag and insufficient prediction precision existing in the prior art when a complex voltage fluctuation mode is processed and a future voltage change trend is predicted are solved.

Description

Technical Field

[0001] This application relates to the technical field of charging piles, and in particular to a method, device, and storage medium for optimizing the voltage stability of DC charging piles. Background Art

[0002] With the development of new energy technologies and the increasing environmental protection requirements, electric vehicles have gradually become an important part of the transportation field. As a key device for electric vehicle charging facilities, the performance of DC charging piles has an important impact on the charging efficiency of electric vehicles and the user experience. Driven by the growing market demand for fast charging, improving the voltage stability of charging piles has become a hot and key research direction in the industry.

[0003] In related technical means, the voltage stability of DC charging piles is optimized by introducing a feedback control system. These systems use the voltage signals collected in real time and adjust the output voltage through a closed-loop control model to maintain it within a predetermined range. In addition, some technical solutions also combine filters to reduce voltage fluctuations and achieve the suppression of short-term fluctuations. These methods have generally improved the reliability of charging equipment and, to a certain extent, increased the charging efficiency of electric vehicles.

[0004] Regarding the above technical solutions, although the existing feedback control systems and filters and other technical means can slow down voltage fluctuations and improve the voltage stability of charging piles, there are problems of response lag and insufficient prediction accuracy in dealing with complex voltage fluctuation patterns and predicting future voltage change trends. This deficiency may lead to lagging voltage compensation or overcorrection, thus having a negative impact on the user charging experience. Summary of the Invention

[0005] In order to improve the problems of response lag and insufficient prediction accuracy existing in the prior art when dealing with complex voltage fluctuation patterns and predicting future voltage change trends, this application provides a method, device, and storage medium for optimizing the voltage stability of DC charging piles.

[0006] The present invention provides a method for optimizing the voltage stability of a DC charging pile, including: obtaining voltage sequence data at the DC output end of the charging pile, and performing segmented statistics on the voltage sequence data based on a set time sequence window to obtain a voltage time sequence feature matrix; performing pattern recognition on the voltage time sequence feature matrix, extracting key feature parameters of voltage fluctuations, constructing a voltage fluctuation map based on the key feature parameters of voltage fluctuations, and calculating the voltage fluctuation trend weight according to the voltage fluctuation map; performing voltage mapping relationship modeling based on the voltage fluctuation trend weight to obtain a voltage prediction model, using the voltage prediction model to perform prediction calculation on the real-time voltage data of the charging pile, generating a predicted voltage deviation value, adjusting the preset DC voltage stabilization control parameters of the charging pile according to the predicted voltage deviation value to obtain an optimized adjustment matrix and a voltage compensation factor matrix; dynamically adjusting the preset DC output control strategy of the charging pile based on the optimized adjustment matrix to obtain adjusted control strategy parameters, performing compensation calculation on the adjusted control strategy parameters using the voltage compensation factor matrix to obtain an optimized control instruction, and generating a dynamic voltage stabilization control signal according to the optimized control instruction.

[0007] As a preferred solution, the step of obtaining voltage sequence data at the DC output end of the charging pile and performing segmented statistics on the voltage sequence data based on a set time sequence window to obtain a voltage time sequence feature matrix includes: obtaining historical voltage data and real-time voltage data at the DC output end of the charging pile, and performing denoising processing on the historical voltage data to obtain denoised voltage data and noise feature data; constructing a noise distribution mapping matrix based on the noise feature data, and performing noise compensation on the real-time voltage data using the noise distribution mapping matrix to obtain compensated voltage data; performing matching analysis on the compensated voltage data and the denoised voltage data, extracting matching feature parameters, and arranging the compensated voltage data in sequence based on the matching feature parameters to obtain a voltage matching sequence; performing segmented statistics on the voltage matching sequence based on a set time sequence window to obtain a voltage time sequence adjustment matrix, and reconstructing the compensated voltage data based on the voltage time sequence adjustment matrix to obtain a voltage time sequence feature matrix.

[0008] As a preferred solution, the steps of performing pattern recognition on the voltage time-series feature matrix, extracting key voltage fluctuation characteristic parameters, constructing a voltage fluctuation map based on the key voltage fluctuation characteristic parameters, and calculating the voltage fluctuation trend weight according to the voltage fluctuation map include: performing pattern decomposition on the voltage time-series feature matrix to obtain voltage pattern components and pattern attribution factors, performing cluster analysis on the voltage pattern components through the pattern attribution factors to obtain voltage pattern clusters; performing a change trend analysis on the voltage pattern clusters to obtain mutation characteristic factors, and performing hierarchical induction on the voltage pattern clusters based on the mutation characteristic factors to obtain a hierarchical voltage pattern set; constructing a voltage fluctuation feature mapping relationship based on the hierarchical voltage pattern set to obtain a voltage fluctuation feature mapping matrix, and calculating a voltage fluctuation influence factor by using the voltage fluctuation feature mapping matrix; performing a fusion calculation on the voltage fluctuation influence factor and the pattern attribution factor to obtain key voltage fluctuation characteristic parameters, constructing a voltage fluctuation map based on the key voltage fluctuation characteristic parameters, and calculating the voltage fluctuation trend weight by using the voltage fluctuation map.

[0009] As a preferred solution, the steps of performing a change trend analysis on the voltage pattern clusters to obtain mutation characteristic factors, and performing hierarchical induction on the voltage pattern clusters based on the mutation characteristic factors to obtain a hierarchical voltage pattern set include: obtaining the historical change trends of the voltage pattern components in the voltage pattern clusters, performing time-series analysis on the historical change trends to obtain voltage pattern change rates and fluctuation frequency characteristics; constructing a voltage pattern change trend matrix based on the voltage pattern change rates and the fluctuation frequency characteristics, performing trend classification on the voltage pattern clusters by using the voltage pattern change trend matrix, and generating a fluctuation anomaly detection factor based on the classification results; using the fluctuation anomaly detection factor to perform anomaly screening on the voltage pattern clusters, extracting abnormal pattern components, calculating a voltage mutation threshold based on the abnormal pattern components, screening out mutation patterns according to the voltage mutation threshold, constructing a voltage mutation characteristic matrix based on the mutation patterns, and extracting mutation characteristic factors based on the voltage mutation characteristic matrix; using the mutation characteristic factors to classify the voltage pattern clusters, generating a voltage pattern hierarchical rule, and performing hierarchical induction on the voltage pattern clusters based on the voltage pattern hierarchical rule to obtain a hierarchical voltage pattern set.

[0010] As a preferred solution, the steps of modeling the voltage mapping relationship based on the voltage fluctuation trend weight to obtain a voltage prediction model, using the voltage prediction model to perform prediction calculations on the real-time voltage data of the charging pile to obtain a predicted voltage deviation value, and adjusting the preset DC voltage stabilization control parameters of the charging pile according to the predicted voltage deviation value to obtain an optimized adjustment matrix and a voltage compensation factor matrix include: using the voltage fluctuation trend weight to adaptively optimize the voltage fluctuation map to obtain an optimized voltage fluctuation map, extracting key fluctuation patterns based on the optimized voltage fluctuation map, and using the key fluctuation patterns to construct a voltage prediction correlation matrix; performing mapping calculations on the voltage fluctuation trend weight through the voltage prediction correlation matrix to obtain voltage trend prediction data and voltage disturbance prediction data; constructing a voltage prediction model based on the voltage trend prediction data and the voltage disturbance prediction data, and using the voltage prediction model to perform prediction calculations on the real-time voltage data of the charging pile to obtain a predicted voltage deviation value; using the predicted voltage deviation value to adjust the preset DC voltage stabilization control parameters of the charging pile to obtain a first optimized control parameter and a second optimized control parameter, generating an optimized adjustment matrix based on the first optimized control parameter, and constructing a voltage compensation factor matrix based on the second optimized control parameter.

[0011] As a preferred solution, the steps of using the predicted voltage deviation value to adjust the preset DC voltage stabilization control parameters of the charging pile to obtain a first optimized control parameter and a second optimized control parameter, generating an optimized adjustment matrix based on the first optimized control parameter, and constructing a voltage compensation factor matrix based on the second optimized control parameter include: calculating a voltage deviation compensation factor based on the predicted voltage deviation value, and using the voltage deviation compensation factor to adjust the preset DC voltage stabilization control parameters of the charging pile to obtain an initial optimized control parameter; extracting a control parameter correlation factor based on the initial optimized control parameter, and using the control parameter correlation factor to construct a control parameter optimization mapping matrix; performing non-linear transformation calculations on the initial optimized control parameter through the control parameter optimization mapping matrix to obtain a first optimized control parameter and a second optimized control parameter; constructing a voltage regulation coefficient matrix using the first optimized control parameter, and generating an optimized adjustment matrix based on the voltage regulation coefficient matrix; constructing a voltage error compensation matrix using the second optimized control parameter, calculating a dynamic compensation factor based on the voltage error compensation matrix, and generating a voltage compensation factor matrix based on the dynamic compensation factor.

[0012] As a preferred solution, the step of dynamically adjusting the preset DC output control strategy of the charging pile based on the optimization adjustment matrix to obtain the adjusted control strategy parameters, compensating and calculating the adjusted control strategy parameters by using the voltage compensation factor matrix to obtain an optimized control instruction, and generating a dynamic voltage stabilization control signal according to the optimized control instruction includes: performing a strategy adjustment calculation on the DC output control strategy of the charging pile by using the optimization adjustment matrix to obtain the adjusted control strategy parameters and a control strategy correction factor; performing a compensation calculation on the adjusted control strategy parameters by using the voltage compensation factor matrix to obtain the compensated control strategy parameters, performing a non-linear correction calculation on the compensated control strategy parameters by using the control strategy correction factor to obtain an optimized control instruction, and generating a dynamic voltage stabilization control signal based on the optimized control instruction.

[0013] The present application also provides a device for optimizing the voltage stability of a DC charging pile, including: an acquisition module, configured to acquire voltage sequence data at the DC output end of the charging pile, and perform segmented statistics on the voltage sequence data based on a set time sequence window to obtain a voltage time sequence feature matrix; an identification module, configured to perform pattern recognition on the voltage time sequence feature matrix, extract key voltage fluctuation characteristic parameters, construct a voltage fluctuation map based on the key voltage fluctuation characteristic parameters, and calculate a voltage fluctuation trend weight according to the voltage fluctuation map; a calculation module, configured to perform voltage mapping relationship modeling based on the voltage fluctuation trend weight to obtain a voltage prediction model, perform a prediction calculation on the real-time voltage data of the charging pile by using the voltage prediction model to generate a predicted voltage deviation value, and adjust the preset DC voltage stabilization control parameters of the charging pile according to the predicted voltage deviation value to obtain an optimization adjustment matrix and a voltage compensation factor matrix; an adjustment module, configured to dynamically adjust the preset DC output control strategy of the charging pile based on the optimization adjustment matrix to obtain the adjusted control strategy parameters, perform a compensation calculation on the adjusted control strategy parameters by using the voltage compensation factor matrix to obtain an optimized control instruction, and generate a dynamic voltage stabilization control signal according to the optimized control instruction.

[0014] The present application also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is run by a processor, the processor is caused to execute the method for optimizing the voltage stability of a DC charging pile as described in any one of the above.

[0015] Compared with the prior art, the present application has the following beneficial effects: rapid response and high prediction accuracy. By obtaining the voltage sequence data at the DC output end of the charging pile and performing segmented statistics on it, a voltage time-series feature matrix characterizing the voltage time-series characteristics is effectively constructed; using pattern recognition technology to analyze the voltage time-series feature matrix, key feature parameters of voltage fluctuations are extracted, a voltage fluctuation map is formed, and the voltage fluctuation trend weight is calculated; based on the voltage fluctuation trend weight, a voltage prediction model is established, which can predict voltage changes in real time and generate a predicted voltage deviation value; by optimizing the DC voltage stabilization control parameters of the charging pile through the predicted voltage deviation value, and dynamically adjusting the output control strategy through the optimization adjustment matrix and the voltage compensation factor matrix, an efficient dynamic voltage stabilization control signal is generated, improving the problems of response lag and insufficient prediction accuracy existing in the prior art when dealing with complex voltage fluctuation patterns and predicting future voltage change trends. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0017] The structures, proportions, sizes, etc. shown in the drawings of this specification are only used to cooperate with the content disclosed in the specification for those who are familiar with this technology to understand and read, and are not used to limit the limited conditions under which the present invention can be implemented. Therefore, they do not have technical substance significance. Any modification of the structure, change of the proportional relationship, or adjustment of the size, without affecting the effects that the present invention can produce and the purposes that can be achieved, should still fall within the scope that can be covered by the technical content disclosed in the present invention.

[0018] Figure 1 is a schematic flow chart of the method for optimizing the voltage stability of a DC charging pile provided by an embodiment of the present invention; Figure 2 is a schematic structural block diagram of the device for optimizing the voltage stability of a DC charging pile provided by an embodiment of the present invention.

[0019] Description of the reference numerals: 10. Device for optimizing the voltage stability of a DC charging pile; 11. Acquisition module; 12. Identification module; 13. Calculation module; 14. Adjustment module. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0020] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0021] The flowchart shown in the accompanying drawings is only an example illustration, and does not necessarily include all the content and operations / steps, nor does it necessarily need to be executed in the described order. For example, some operations / steps can also be decomposed, combined, or partially merged, so the actual execution order may change according to the actual situation.

[0022] It should also be understood that the terms used in the specification of this application are only for the purpose of describing specific embodiments and are not intended to limit this application. As used in the specification of this application and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an", and "the" are intended to include the plural forms.

[0023] It should be further understood that the term "and / or" used in the specification of this application and the appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0024] The technical solutions of the present invention will be further described below in conjunction with the accompanying drawings and through specific implementation manners.

[0025] Embodiment 1: As Figure 1 shown, this application provides a method for optimizing the voltage stability of a DC charging pile, including steps S100 to S400.

[0026] Step S100: Obtain the voltage sequence data at the DC output end of the charging pile, and perform segmented statistics on the voltage sequence data based on a set time sequence window to obtain a voltage time sequence feature matrix.

[0027] In this step, the voltage sequence data at the DC output end of the charging pile is continuously collected through a sensor to ensure the real-time and integrity of the data; specifically, according to the preset time sequence window, the collected voltage sequence data is segmented into multiple time periods, and the voltage data in each time period is statistically analyzed to extract characteristic values such as the mean and variance, and finally a voltage time sequence feature matrix with time sequence characteristics is generated.

[0028] For example, when the time sequence window is set to 1 second, the voltage sequence data at the DC output end of the charging pile is sampled 1000 times per second, and characteristic values such as the mean and variance are calculated based on these sampling data, so as to construct the characteristic matrix corresponding to this time period.

[0029] Step S200: Perform pattern recognition on the voltage time - series feature matrix, extract key feature parameters of voltage fluctuations, construct a voltage fluctuation map based on the key feature parameters of voltage fluctuations, and calculate the voltage fluctuation trend weight according to the voltage fluctuation map.

[0030] In this step, a pattern recognition algorithm is used to analyze the data in the voltage time - series feature matrix to detect the main patterns of voltage changes; specifically, clustering algorithms and classification algorithms are used to extract feature parameters, such as the amplitude, frequency, and duration of voltage fluctuations, and a voltage fluctuation map is constructed through these key feature parameters. This map reflects the voltage fluctuation law in different time periods, and at the same time, by calculating the voltage fluctuation trend weight, it provides data support for subsequent voltage prediction.

[0031] For example, the high - frequency components of voltage fluctuations are identified through the pattern recognition algorithm, and the weight values are calculated based on these high - frequency components to predict the future voltage fluctuation trend and its impact.

[0032] Step S300: Based on the voltage fluctuation trend weight, establish a voltage mapping relationship model to obtain a voltage prediction model. Use the voltage prediction model to perform prediction calculations on the real - time voltage data of the charging pile, generate a predicted voltage deviation value, and adjust the preset DC voltage stabilization control parameters of the charging pile according to the predicted voltage deviation value to obtain an optimized adjustment matrix and a voltage compensation factor matrix.

[0033] In this step, a multi - variable non - linear voltage prediction model is constructed using historical voltage sequence data and its fluctuation trend weight; specifically, this model uses a neural network algorithm, combines the input real - time voltage data, predicts the future voltage value, calculates the predicted voltage deviation value according to the prediction result, adjusts the preset DC voltage stabilization control parameters, and generates an optimized adjustment matrix and a voltage compensation factor matrix, so as to achieve more accurate dynamic compensation.

[0034] For example, a voltage prediction model is constructed using a multi - layer perceptron neural network, and the voltage fluctuation data in the past 10 seconds is input to predict the voltage change trend within the next 1 second, and the corresponding optimized adjustment matrix is generated.

[0035] Step S400: Dynamically adjust the preset DC output control strategy of the charging pile based on the optimized adjustment matrix to obtain adjusted control strategy parameters. Use the voltage compensation factor matrix to perform compensation calculations on the adjusted control strategy parameters to obtain an optimized control instruction, and generate a dynamic voltage stabilization control signal according to the optimized control instruction.

[0036] In this step, the control strategy parameters of the charging pile are adjusted in real time by analyzing and optimizing the adjustment matrix and the voltage compensation factor matrix. Specifically, according to the dynamically adjusted control strategy parameters, an optimized control instruction is generated, and with the help of a controller, the optimized control instruction is converted into a dynamically regulated voltage control signal with stable output, so as to realize the real-time dynamic regulation of the DC output voltage.

[0037] For example, for a scenario with large voltage fluctuation amplitude, the proportional-integral-differential (PID) control parameters of the output control strategy are adjusted in real time, and the adjusted parameters are further corrected in combination with the voltage compensation factor matrix to ensure that the output voltage is stable within a predetermined range.

[0038] In this embodiment, by obtaining the voltage sequence data at the DC output end of the charging pile and performing segmented statistics on the voltage sequence data based on a set time sequence window, a voltage time sequence feature matrix is generated; then, pattern recognition is performed on the voltage time sequence feature matrix, key feature parameters of voltage fluctuation are extracted, and a voltage fluctuation map is constructed based on these key feature parameters of voltage fluctuation, and the voltage fluctuation trend weight is calculated according to the voltage fluctuation map; then, based on the voltage fluctuation trend weight, a voltage mapping relationship model is established to obtain a voltage prediction model, the real-time voltage data of the charging pile is predicted and calculated using the voltage prediction model, a predicted voltage deviation value is generated, and the preset DC voltage regulation control parameters of the charging pile are adjusted according to the predicted voltage deviation value, and then an optimized adjustment matrix and a voltage compensation factor matrix are obtained; finally, based on the optimized adjustment matrix, the preset DC output control strategy of the charging pile is dynamically adjusted, and at the same time, the voltage compensation factor matrix is used to perform compensation calculation on the adjusted control strategy parameters to generate an optimized control instruction, and a dynamically regulated voltage control signal is generated according to the optimized control instruction. The fluctuation characteristics of the voltage sequence data of the charging pile are effectively extracted and analyzed, the change trend of the voltage is accurately predicted, the accuracy of voltage prediction is significantly improved, and by dynamically adjusting and compensating the DC output control strategy, not only the stability of the voltage output of the charging pile is enhanced, but also the problems of response lag and insufficient prediction accuracy existing in the prior art when dealing with complex voltage fluctuation patterns and predicting future voltage change trends are improved.

[0039] Embodiment 2: In step S100, historical voltage data and real-time voltage data at the DC output end of the charging pile are obtained, and the historical voltage data is denoised to obtain denoised voltage data and noise feature data.

[0040] By performing statistical analysis on the historical voltage data of the DC output end of the charging pile, the multi-scale wavelet transform method is used to separate the high-frequency noise signal and the low-frequency noise signal. At the same time, the Kalman filter is used to dynamically model the noise characteristics, and the voltage signal with higher stability is extracted to form denoised voltage data. Specifically, the historical voltage data must first be preprocessed during the processing process. After eliminating sampling errors and lost data, it is then fitted with a Gaussian distribution to separate the noise characteristic data.

[0041] For example, in the denoising process, for voltage data with high fluctuation frequency, wavelet decomposition is first used to extract the frequency band characteristics, and then a low-pass filter is used to further smooth the signal to minimize the impact of high-frequency fluctuations, and finally the denoised voltage data containing the true voltage change trend is obtained.

[0042] A noise distribution mapping matrix is ​​constructed based on the noise characteristic data, and the noise distribution mapping matrix is ​​used to perform noise compensation on the real-time voltage data to obtain compensated voltage data.

[0043] By analyzing the distribution law of noise characteristic data, a noise distribution mapping matrix is ​​established using a two-dimensional Gaussian kernel density estimation method, and the noise components in the real-time voltage data are dynamically compensated through a matrix interpolation method. Specifically, according to the differences in noise characteristics at different sampling time points, the compensation factor of the mapping matrix is ​​dynamically adjusted to correct the real-time voltage data point by point.

[0044] For example, if the noise in the real-time voltage data is mainly concentrated in the 50 Hz frequency range, the noise distribution mapping matrix is ​​used to perform special compensation for this frequency range, so that the final compensated voltage data can more accurately reflect the actual voltage value.

[0045] The compensation voltage data and the denoised voltage data are matched and analyzed to extract matching characteristic parameters, and the compensation voltage data are arranged in sequence based on the matching characteristic parameters to obtain a voltage matching sequence.

[0046] The cross-correlation algorithm is used to calculate the matching degree of the compensated voltage data and the denoised voltage data, and the matching characteristic parameters such as the voltage change rate and the fluctuation amplitude are extracted. Specifically, the dynamic time warping algorithm is used to align the time series of the two sets of data to ensure that the error in the matching analysis process is minimized, and the compensated voltage data is rearranged according to the matching characteristic parameters to generate a reasonably ordered voltage matching sequence.

[0047] For example, in the matching analysis, if the compensated voltage data has a time lag with the denoised voltage data, it is adjusted through interpolation and rearranged according to the rate of change so that the generated voltage matching sequence can accurately correspond to the actual voltage change.

[0048] Segment the voltage matching sequence based on the set time sequence window for statistical analysis to obtain a voltage time sequence adjustment matrix, and reconstruct the compensated voltage data based on the voltage time sequence adjustment matrix to obtain a voltage time sequence feature matrix.

[0049] The voltage matching sequence is segmented into multiple paragraphs at fixed time intervals, and statistical characteristics such as the average value, variance, and peak value of each paragraph are calculated to generate a voltage time sequence adjustment matrix. Specifically, in the statistical process, the weighted sliding window method is combined to dynamically adjust the window weight to ensure the continuity and accuracy of the statistical results. At the same time, the compensated voltage data is linearly interpolated and reconstructed using the statistical characteristics in the adjustment matrix, and finally a voltage time sequence feature matrix is generated.

[0050] For example, when the time sequence window size is set to 2 seconds, the voltage matching sequence data is segmented and statistically analyzed every 2 seconds, and the maximum and minimum values of each segment are calculated respectively. These statistical characteristics are used in the reconstruction process to generate an optimized voltage time sequence feature matrix.

[0051] In step S200, perform mode decomposition on the voltage time sequence feature matrix to obtain voltage mode components and mode attribution factors, and perform clustering analysis on the voltage mode components through the mode attribution factors to obtain voltage mode clusters.

[0052] By using the singular value decomposition (SVD) or principal component analysis (PCA) method on the voltage time sequence feature matrix, the main mode components in the voltage time sequence feature matrix are extracted, and the characteristics corresponding to each mode component are marked in combination with the mode attribution factors. Specifically, the voltage time sequence feature matrix is decomposed into a series of linearly independent mode components, and the clustering operation of the mode components is completed by calculating the attribution factor weights of each mode component, thereby forming multiple voltage mode clusters with independent characteristics.

[0053] For example, assume that the voltage time sequence feature matrix contains five main mode components. For each mode component, use the clustering algorithm for classification to generate voltage mode clusters such as "high-frequency short-term fluctuation cluster" and "low-frequency continuous fluctuation cluster".

[0054] Perform a change trend analysis on the voltage mode clusters to obtain mutation characteristic factors, and perform hierarchical induction on the voltage mode clusters based on the mutation characteristic factors to obtain a hierarchical voltage mode set.

[0055] By performing time series analysis on the historical fluctuation data of each mode component in the voltage mode clusters, the rate and frequency characteristics of the voltage fluctuations are extracted, and these characteristics are calculated to form mutation characteristic factors. Specifically, use the trend analysis model to fit the time series change trend of the voltage mode components, extract the mode components with stronger mutability from them, and use the hierarchical induction method to classify these modes into different levels of voltage mode sets.

[0056] For example, by analyzing the short-term large fluctuations in a certain voltage mode cluster, detecting that its mutation rate is higher than 1.5 times the threshold, and classifying these mutation patterns into the "abnormal mutation pattern set", a unique level of the hierarchical voltage mode set is formed.

[0057] Based on the hierarchical voltage mode set, construct the voltage fluctuation feature mapping relationship, obtain the voltage fluctuation feature mapping matrix, and calculate the voltage fluctuation influence factor using the voltage fluctuation feature mapping matrix.

[0058] By analyzing the interaction relationship between different levels in the hierarchical voltage mode set, use the graph theory model to construct the voltage fluctuation feature mapping relationship and generate the voltage fluctuation feature mapping matrix; specifically, define the voltage mode sets at different levels as nodes, and define the association strength between nodes as edge weights, and calculate the fluctuation influence factor of each mode with the help of the feature mapping matrix to characterize the contribution degree of each mode to the overall voltage fluctuation.

[0059] For example, assuming that the high-frequency fluctuation mode at a specific level contributes 45% to the overall fluctuation influence factor, the feature mapping matrix can be used to quantify this influence and guide subsequent control strategies.

[0060] Fuse the voltage fluctuation influence factor and the mode attribution factor for calculation to obtain the key voltage fluctuation characteristic parameters, construct the voltage fluctuation map based on the key voltage fluctuation characteristic parameters, and calculate the voltage fluctuation trend weight using the voltage fluctuation map.

[0061] By fusing the voltage fluctuation influence factor and the mode attribution factor, calculate the key voltage fluctuation characteristic parameters using the weighted combination formula. These key characteristic parameters include voltage fluctuation amplitude, mutation rate, and frequency distribution, etc.; specifically, use the key characteristic parameters as input variables to generate the voltage fluctuation map, on which the spatial distribution of voltage fluctuation characteristics is clearly marked, and calculate the voltage fluctuation trend weight according to the distribution density of the changing area in the map to predict the evolution law of voltage fluctuation.

[0062] For example, in the generated voltage fluctuation map, the fluctuation trend weight shown in a certain high-density area reaches 70%, indicating that this area has a dominant influence on future voltage fluctuations.

[0063] Among them, the steps of analyzing the change trend of the voltage mode cluster to obtain the mutation characteristic factor and hierarchically summarizing the voltage mode cluster based on the mutation characteristic factor to obtain the hierarchical voltage mode set include: obtaining the historical change trend of each voltage mode component in the voltage mode cluster, performing time series analysis on the historical change trend, and obtaining the voltage mode change rate and fluctuation frequency characteristics.

[0064] By jointly analyzing the historical change trend of the voltage mode component in the frequency domain and time domain, extracting the change rate and its fluctuation frequency, and using the sliding window method for dynamic tracking to ensure the continuity of the extracted features; specifically, segmenting the historical change trend into short-time sequences, calculating the change rate and fluctuation frequency of each segment, and then performing smoothing processing to obtain a core parameter set representing the mode characteristics.

[0065] For example, for a certain voltage mode component, its change rate is 0.5 volts per second, and the frequency characteristic shows a sudden increase near 50 Hz. These characteristics can be used as the main basis for mode classification.

[0066] Construct a voltage mode change trend matrix based on the voltage mode change rate and fluctuation frequency characteristics, use the voltage mode change trend matrix to classify the trends of the voltage mode cluster, and generate a fluctuation anomaly detection factor based on the classification results.

[0067] Through multi-dimensional clustering analysis of the voltage mode change rate and fluctuation frequency characteristics, use the classification results to generate a voltage mode change trend matrix, and construct a fluctuation anomaly detection factor based on the matrix; specifically, the generation process of the anomaly detection factor combines the statistical threshold method to mark the high-deviation mode as an abnormal mode to achieve a rapid screening of voltage mutations.

[0068] For example, for a certain mode cluster, if the classification results show that the frequencies of multiple mode points deviate from the average value by more than 30%, then these mode points will be marked as abnormal modes.

[0069] Use the fluctuation anomaly detection factor to screen the voltage mode cluster for anomalies, extract the abnormal mode components, calculate the voltage mutation threshold based on the abnormal mode components, screen out the mutation modes according to the voltage mutation threshold, construct a voltage mutation feature matrix based on the mutation modes, and extract the mutation feature factors based on the voltage mutation feature matrix.

[0070] Through quantitative analysis of the mode components screened for anomalies, calculate the voltage mutation threshold, mark the modes with fluctuation amplitudes exceeding the mutation threshold as mutation modes, and use these mutation modes to generate a voltage mutation feature matrix; specifically, extract the key variables in the mutation feature matrix as the mutation feature factors to provide data support for further optimizing the hierarchical rules.

[0071] For example, assume that the fluctuation amplitude of a certain mode is 1.2 volts and the mutation threshold is 1.0 volts, then this mode is marked as a mutation mode, and the feature matrix records parameters such as the mutation rate and the influence range.

[0072] Use the mutation feature factors to classify the voltage mode cluster, generate the voltage mode hierarchical rules, and hierarchically summarize the voltage mode cluster based on the voltage mode hierarchical rules to obtain a hierarchical voltage mode set.

[0073] By analyzing the correlation between mutation characteristic factors and pattern attribution factors, a hierarchical rule for voltage patterns is established, which includes the basis for hierarchical division and the corresponding classification criteria. Specifically, according to the hierarchical rule, the voltage pattern clusters are inductively summarized layer by layer to ensure that the hierarchical structure is clear and independent, thereby forming a hierarchical voltage pattern set.

[0074] For example, using the hierarchical rule, the high-frequency large-amplitude fluctuation pattern is classified into the first-level set, while the low-frequency small-amplitude fluctuation pattern is classified into the second-level set, obtaining a multi-level voltage pattern set.

[0075] In step S300, the voltage fluctuation map is adaptively optimized using the voltage fluctuation trend weight to obtain an optimized voltage fluctuation map. Based on the optimized voltage fluctuation map, key fluctuation patterns are extracted, and a voltage prediction correlation matrix is constructed using the key fluctuation patterns.

[0076] By adopting an adaptive iterative algorithm to dynamically adjust the voltage fluctuation map, the key fluctuation feature regions in the map are optimized using the voltage fluctuation trend weight to remove redundant information and enhance the visual representation of the fluctuation significance. Specifically, a weighting coefficient is introduced during the optimization process, and the characteristic parameters in the map are dynamically adjusted according to the voltage fluctuation trend weight to ensure that the key fluctuation patterns more accurately reflect the future voltage change law. Based on the optimized voltage fluctuation map, key fluctuation patterns representing the core behavior of voltage change are extracted, and these patterns are associated with the voltage fluctuation trend weight to construct a voltage prediction correlation matrix to describe the mapping relationship between the fluctuation pattern and the trend weight.

[0077] For example, the fluctuation trend weight in a certain area is optimized from the initial value of 30% to 50%. The finally optimized voltage fluctuation map can accurately reflect the high-frequency fluctuation characteristics of this area, and the constructed voltage prediction correlation matrix shows that the correlation coefficient between the high-frequency pattern and the trend weight is as high as 0.8.

[0078] Through the voltage prediction correlation matrix, mapping calculations are performed on the voltage fluctuation trend weight to obtain voltage trend prediction data and voltage disturbance prediction data.

[0079] By using the matrix mapping method to calculate the voltage prediction correlation matrix, the relationship between each fluctuation pattern and the trend weight is transformed into quantitative data using a mapping function, thereby generating voltage trend prediction data and voltage disturbance prediction data respectively. Specifically, the trend prediction data reflects the overall change trend of voltage fluctuations, while the disturbance prediction data depicts the change characteristics of local fluctuations, providing data support for voltage stability adjustment.

[0080] For example, through voltage prediction correlation matrix mapping calculation, the voltage trend prediction data shows that the voltage will show an upward trend within the next 5 seconds, while the voltage disturbance prediction data reflects that the instantaneous disturbance amplitude may fluctuate within the range of 0.2 volts.

[0081] Based on the voltage trend prediction data and the voltage disturbance prediction data, a voltage prediction model is constructed, and the real-time voltage data of the charging pile is predicted and calculated using the voltage prediction model to obtain the predicted voltage deviation value.

[0082] By combining linear regression and deep learning techniques, a multivariate voltage prediction model is constructed. This model uses the voltage trend prediction data and the voltage disturbance prediction data as input variables to comprehensively analyze the future change trend of the voltage. Specifically, the model is trained using historical voltage data and updated by real-time input data. Finally, the predicted value of the voltage is output, and the predicted voltage deviation value is calculated by comparing it with the current real-time voltage data.

[0083] For example, using the constructed voltage prediction model, the predicted voltage fluctuation range within the next 1 second is 50 volts to 52 volts, while the actual real-time voltage data shows 51 volts. After comparison, the predicted voltage deviation value is calculated to be 1 volt.

[0084] The preset DC voltage stabilization control parameters of the charging pile are adjusted using the predicted voltage deviation value to obtain the first optimized control parameter and the second optimized control parameter. An optimized adjustment matrix is generated based on the first optimized control parameter, and a voltage compensation factor matrix is constructed based on the second optimized control parameter.

[0085] Through dynamic feedback adjustment of the predicted voltage deviation value, the preset DC voltage stabilization control parameters of the charging pile are optimized using the proportional-integral-derivative (PID) algorithm to generate the first optimized control parameter and the second optimized control parameter. Specifically, the first optimized control parameter is used to adjust the main strategy of DC output control, and the second optimized control parameter is used to construct the auxiliary parameter matrix for compensating voltage, thus achieving refined control.

[0086] For example, based on the predicted voltage deviation value of 1 volt, the output sensitivity parameter of the control strategy is adjusted, and at the same time, the first optimized control parameter is generated as 0.8 times the dynamic gain, and the second optimized control parameter is generated as 0.2 times the compensation gain.

[0087] Among them, the steps of adjusting the preset DC voltage stabilization control parameters of the charging pile by using the predicted voltage deviation value to obtain the first optimized control parameter and the second optimized control parameter, generating an optimized adjustment matrix based on the first optimized control parameter, and constructing a voltage compensation factor matrix based on the second optimized control parameter include: calculating a voltage deviation compensation factor based on the predicted voltage deviation value, and using the voltage deviation compensation factor to adjust the preset DC voltage stabilization control parameters of the charging pile to obtain the initial optimized control parameter.

[0088] By performing a weighted process on the predicted voltage deviation value, calculate the voltage deviation compensation factor for correcting the original DC voltage stabilization control parameters; specifically, dynamically adjust the sensitivity and gain amplitude of the control parameters according to the size of the compensation factor, so as to generate the initial optimized control parameter.

[0089] For example, assuming that the predicted voltage deviation value is 2 volts and the calculated voltage deviation compensation factor is 1.5, then the gain amplitude of the original control parameter is adjusted by 150% accordingly to adapt to the current voltage change requirement.

[0090] Extract the control parameter correlation factor based on the initial optimized control parameter, and use the control parameter correlation factor to construct a control parameter optimization mapping matrix.

[0091] By analyzing the dependency relationship between the initial optimized control parameters, extract the control parameter correlation factor, and use these factors to construct an optimization mapping matrix, which describes the interaction between parameters and its influence degree on voltage regulation; specifically, perform dimensionality reduction processing on the optimization matrix by using principal component analysis to ensure a simple matrix structure and higher calculation efficiency.

[0092] For example, the control parameter correlation factor of an optimization mapping matrix shows that the correlation coefficient between the voltage output gain and the compensation delay is 0.85, indicating that there is a strong coupling relationship between the two in the adjustment strategy.

[0093] Perform a non-linear transformation calculation on the initial optimized control parameter through the control parameter optimization mapping matrix to obtain the first optimized control parameter and the second optimized control parameter.

[0094] By introducing a non-linear transformation function to adjust the initial optimized control parameter, use the optimization mapping matrix to quantitatively calculate the non-linear relationship of each control parameter, and finally generate the first optimized control parameter and the second optimized control parameter; specifically, determine the optimal solution of parameter adjustment through the curve fitting method.

[0095] For example, for the case where the gain value in the initial control parameter is too high, use non-linear transformation to adjust it to the optimal value range between 0.7 and 0.9, so as to generate a reasonable first optimized control parameter.

[0096] Construct a voltage regulation coefficient matrix using the first optimization control parameter, and generate an optimized adjustment matrix based on the voltage regulation coefficient matrix.

[0097] By taking the first optimization control parameter as the basic variable, construct a voltage regulation coefficient matrix using the interpolation algorithm, and dynamically update the matrix in combination with time series analysis, and finally generate an optimized adjustment matrix that reflects the real-time voltage change requirements; specifically, this matrix is directly used to guide the output adjustment strategy of the charging pile.

[0098] For example, based on the gain coefficient 0.85 and attenuation coefficient 0.15 shown in the voltage regulation coefficient matrix, the optimized adjustment matrix can dynamically adjust the output sensitivity to match the voltage fluctuation characteristics.

[0099] Construct a voltage error compensation matrix using the second optimization control parameter, calculate a dynamic compensation factor based on the voltage error compensation matrix, and generate a voltage compensation factor matrix based on the dynamic compensation factor.

[0100] By taking the second optimization control parameter as the input, construct a voltage error compensation matrix, and calculate real-time compensation factors in combination with the dynamic adjustment algorithm, and use these factors to generate a voltage compensation factor matrix; specifically, utilize the distribution characteristics of the compensation factors in the matrix to optimize the smoothness and stability of the voltage output.

[0101] For example, the range of the dynamic compensation factor shown in the voltage compensation factor matrix is 0.1 to 0.3, and the compensation parameters within this range can effectively offset the fluctuation error and ensure stable voltage output.

[0102] In step S400, perform a strategy adjustment calculation on the DC output control strategy of the charging pile using the optimized adjustment matrix to obtain the adjusted control strategy parameters and the control strategy correction factor.

[0103] By taking the optimized adjustment matrix as the input, dynamically adjust the DC output control strategy according to the weight values of the adjustment parameters in the matrix to optimize the output voltage control performance of the charging pile. Specifically, according to the adjustment coefficients in the optimized adjustment matrix, the key parameters of the control strategy are adjusted in real time, such as the gain value, integral time constant, and derivative time constant in the proportional-integral-derivative (PID) control parameters, to adapt to different voltage fluctuation conditions. In addition, further optimize the adjusted control strategy based on the control strategy correction factor, including correcting the real-time responsiveness and stability of the control strategy, and finally generate the adjusted control strategy parameters and the control strategy correction factor.

[0104] For example, when the output gain parameter in the optimization adjustment matrix is 0.85, by adjusting the proportional gain of the PID controller, the gain value is adjusted from the original 0.8 to 0.85, and at the same time, a control strategy correction factor is introduced to refine and correct the adjusted integral time constant to ensure that the control strategy remains stable under the condition of high current output.

[0105] The compensated control strategy parameters are obtained by compensating and calculating the adjusted control strategy parameters through the voltage compensation factor matrix. The optimized control instruction is obtained by performing a non-linear correction calculation on the compensated control strategy parameters through the control strategy correction factor, and a dynamic voltage stabilization control signal is generated based on the optimized control instruction.

[0106] By inputting the adjusted control strategy parameters into the voltage compensation factor matrix and utilizing the distribution characteristics of the compensation factors in the matrix, item-by-item compensation and correction are performed on the strategy parameters to generate the compensated control strategy parameters. Specifically, according to the dynamic compensation factors in each dimension of the voltage compensation factor matrix, the output sensitivity and response speed of the control strategy parameters are adjusted to ensure that the control strategy adapts to complex voltage fluctuation scenarios. Further, in combination with the control strategy correction factor, the compensated control strategy parameters are finely adjusted through a non-linear correction method to generate an optimized control instruction. Finally, by executing the optimized control instruction, a dynamic voltage stabilization control signal is generated in real time to ensure the stability and reliability of the DC output voltage of the charging pile.

[0107] For example, when the dynamic compensation factor range shown in the voltage compensation factor matrix is from 0.1 to 0.3, the output gain, integral time constant, and attenuation coefficient of the adjusted control strategy parameters are respectively compensated and corrected. The gain value is adjusted to 0.9, the integral time constant is adjusted to 2 seconds, and at the same time, the attenuation coefficient is non-linearly corrected by using the control strategy correction factor and optimized to 0.15. Finally, an optimized control instruction is generated for dynamically adjusting the output voltage. The generated dynamic voltage stabilization control signal can effectively mitigate the impact of 50Hz voltage fluctuations on charging stability.

[0108] In this embodiment, by obtaining the historical voltage data and real-time voltage data at the DC output terminal of the charging pile, denoising the historical voltage data and extracting the noise characteristic features, and performing noise compensation on the real-time voltage data based on the noise distribution mapping matrix, compensated voltage data is obtained. Further, the compensated voltage data and the denoised voltage data are used for matching analysis, the matching characteristic parameters are extracted and a voltage matching sequence is generated. Then, based on the set time series window, the voltage matching sequence is segmented and statistically analyzed to form a voltage time series adjustment matrix, and the compensated voltage data is reconstructed, and finally a voltage time series characteristic matrix is generated. Next, the voltage time series characteristic matrix is subjected to mode decomposition to obtain voltage mode components and mode attribution factors, a hierarchical voltage mode set is constructed through clustering and change trend analysis, and based on this, a voltage fluctuation characteristic mapping matrix is generated to quantify the voltage fluctuation influence factors. Through the fusion calculation of the voltage fluctuation trend weight and the key characteristic parameters, the voltage fluctuation map is further optimized, the key fluctuation modes are extracted to construct a voltage prediction correlation matrix, so as to predict the voltage trend and disturbance. The predicted voltage deviation value is used to optimize the DC voltage stabilization control parameters of the charging pile, and an optimized adjustment matrix and a voltage compensation factor matrix are generated to dynamically adjust the control strategy. Finally, an optimized control instruction is generated by combining the voltage compensation factor matrix and the control strategy correction factor, and a dynamic voltage stabilization control signal is output in real time. This embodiment effectively solves the stability problem under complex voltage fluctuations, and significantly improves the voltage prediction accuracy, voltage stabilization control efficiency and the overall operation performance of the charging pile.

[0109] Embodiment 3: As Figure 2 shown, the present application also provides a DC charging pile voltage stability optimization device 10, including an acquisition module 11, an identification module 12, a calculation module 13 and an adjustment module 14.

[0110] The acquisition module 11 is mainly used to acquire the voltage sequence data at the DC output terminal of the charging pile, and perform segmented statistics on the voltage sequence data based on the set time series window to obtain a voltage time series characteristic matrix.

[0111] The identification module 12 is mainly used to perform pattern recognition on the voltage time series characteristic matrix, extract the key voltage fluctuation characteristic parameters, construct a voltage fluctuation map based on the key voltage fluctuation characteristic parameters, and calculate the voltage fluctuation trend weight according to the voltage fluctuation map.

[0112] The calculation module 13 is mainly used to model the voltage mapping relationship based on the voltage fluctuation trend weight to obtain a voltage prediction model, use the voltage prediction model to perform prediction calculation on the real-time voltage data of the charging pile, generate a predicted voltage deviation value, and adjust the preset DC voltage stabilization control parameters of the charging pile according to the predicted voltage deviation value to obtain an optimized adjustment matrix and a voltage compensation factor matrix.

[0113] The adjustment module 14 is mainly used to dynamically adjust the preset DC output control strategy of the charging pile based on the optimization adjustment matrix to obtain the adjusted control strategy parameters, perform compensation calculation on the adjusted control strategy parameters using the voltage compensation factor matrix to obtain the optimized control instruction, and generate a dynamic voltage stabilization control signal according to the optimized control instruction.

[0114] In this embodiment, through the collaborative work of the acquisition module 11, the recognition module 12, the calculation module 13, and the adjustment module 14, the efficient and stable control of the voltage at the DC output end of the charging pile is achieved. The acquisition module 11 collects the voltage sequence data at the DC output end of the charging pile, and completes segmented statistics based on the set time series window to generate a voltage time series feature matrix, providing an accurate data basis for subsequent analysis. The recognition module 12 uses a pattern recognition algorithm to extract the key feature parameters of voltage fluctuations from the voltage time series feature matrix, combines the constructed voltage fluctuation map, and calculates the voltage fluctuation trend weight, so as to accurately reveal the regular characteristics of voltage fluctuations. The calculation module 13 generates a voltage prediction model by modeling the voltage mapping relationship based on the voltage fluctuation trend weight to predict the real-time voltage change of the charging pile and calculate the predicted voltage deviation value, dynamically adjusts the DC voltage stabilization control parameters according to the deviation value, and generates an optimization adjustment matrix and a voltage compensation factor matrix to support the strategy adjustment. The adjustment module 14 adjusts the control strategy of the charging pile through the optimization adjustment matrix, and further compensates and corrects the adjusted strategy using the voltage compensation factor matrix to generate an optimized control instruction and output a dynamic voltage stabilization control signal. Through this device, the voltage fluctuation can be accurately predicted and real-time regulated, significantly improving the voltage output stability and charging efficiency of the charging pile.

[0115] It should be noted that those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described device and each module can refer to the corresponding processes in the foregoing Embodiment 1 and will not be repeated here.

[0116] Embodiment 4: This application also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is run by a processor, the processor is caused to execute the DC charging pile voltage stability optimization method as in Embodiment 1.

[0117] In this embodiment, by storing a computer program on a computer-readable storage medium, when the program is run by a processor, the processor is enabled to execute the DC charging pile voltage stability optimization method in Embodiment 1, comprehensively realizing the intelligent optimization control of the charging pile voltage stability. Through multiple steps of operations such as acquiring voltage data, constructing a voltage fluctuation map, predicting the voltage change trend, and optimizing the control strategy, the computer program realizes the full-process automation from voltage data acquisition, feature extraction to predictive control by using an efficient algorithm. The logic in the program can not only accurately predict the trend and calculate the deviation of the voltage fluctuation of the charging pile, but also generate a dynamic voltage stabilization control signal by optimizing the control parameters to effectively cope with complex voltage fluctuation scenarios. Through the combination of this storage medium and the program, the optimization method of the present invention can be easily deployed on various hardware devices, with high versatility and scalability, further improving the voltage control efficiency and stability of the DC charging pile, and providing reliable technical support for the electric vehicle industry.

[0118] The structures, proportions, sizes, etc. shown in the drawings of this specification are only used to cooperate with the content disclosed in the specification for those skilled in this technology to understand and read, and are not used to limit the limited conditions under which the present invention can be implemented. Therefore, they do not have a substantial technical meaning. Any modification of the structure, change of the proportional relationship, or adjustment of the size, without affecting the effects that the present invention can produce and the purposes that can be achieved, should still fall within the scope that can be covered by the technical content disclosed in the present invention.

[0119] As mentioned above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A method for optimizing voltage stability of a DC charging pile, characterized in that: include: Obtaining voltage sequence data of the DC output terminal of the charging pile, and performing segmented statistics on the voltage sequence data based on a set timing window to obtain a voltage timing characteristic matrix; Performing pattern recognition on the voltage time series feature matrix, extracting key characteristic parameters of voltage fluctuation, constructing a voltage fluctuation map based on the key characteristic parameters of voltage fluctuation, and calculating a voltage fluctuation trend weight according to the voltage fluctuation map; Based on the voltage fluctuation trend weight, a voltage mapping relationship is modeled to obtain a voltage prediction model, and the voltage prediction model is used to predict and calculate the real-time voltage data of the charging pile to generate a predicted voltage deviation value, and the DC voltage regulation control parameters preset by the charging pile are adjusted according to the predicted voltage deviation value to obtain an optimization adjustment matrix and a voltage compensation factor matrix; Based on the optimization adjustment matrix, the DC output control strategy preset by the charging pile is dynamically adjusted to obtain the adjusted control strategy parameters, and the voltage compensation factor matrix is ​​used to perform compensation calculation on the adjusted control strategy parameters to obtain the optimization control instructions, and a dynamic voltage stabilization control signal is generated according to the optimization control instructions.

2. The DC charging pile voltage stability optimization method according to claim 1, characterized in that: The step of obtaining voltage sequence data of the DC output terminal of the charging pile and performing segmented statistics on the voltage sequence data based on a set timing window to obtain a voltage timing characteristic matrix includes: Acquire historical voltage data and real-time voltage data of the DC output terminal of the charging pile, and perform denoising on the historical voltage data to obtain denoised voltage data and noise characteristic data; Constructing a noise distribution mapping matrix based on the noise characteristic data, and performing noise compensation on the real-time voltage data using the noise distribution mapping matrix to obtain compensated voltage data; Performing matching analysis on the compensation voltage data and the denoised voltage data, extracting matching characteristic parameters, and arranging the compensation voltage data in sequence based on the matching characteristic parameters to obtain a voltage matching sequence; The voltage matching sequence is segmented and counted based on a set timing window to obtain a voltage timing adjustment matrix, and the compensation voltage data is reconstructed based on the voltage timing adjustment matrix to obtain a voltage timing characteristic matrix.

3. The DC charging pile voltage stability optimization method according to claim 1, characterized in that: The steps of performing pattern recognition on the voltage time series feature matrix, extracting key characteristic parameters of voltage fluctuation, constructing a voltage fluctuation map based on the key characteristic parameters of voltage fluctuation, and calculating the voltage fluctuation trend weight according to the voltage fluctuation map include: Performing pattern decomposition on the voltage time series feature matrix to obtain voltage pattern components and pattern attribution factors, and performing cluster analysis on the voltage pattern components through the pattern attribution factors to obtain voltage pattern clusters; Performing a change trend analysis on the voltage mode cluster to obtain a mutation characteristic factor, and hierarchically summarizing the voltage mode cluster based on the mutation characteristic factor to obtain a hierarchical voltage mode set; Constructing a voltage fluctuation characteristic mapping relationship based on the hierarchical voltage pattern set to obtain a voltage fluctuation characteristic mapping matrix, and calculating a voltage fluctuation influence factor using the voltage fluctuation characteristic mapping matrix; The voltage fluctuation influencing factor and the mode attribution factor are fused and calculated to obtain key characteristic parameters of voltage fluctuation, a voltage fluctuation map is constructed based on the key characteristic parameters of voltage fluctuation, and the voltage fluctuation trend weight is calculated using the voltage fluctuation map.

4. The DC charging pile voltage stability optimization method according to claim 3, characterized in that: The step of analyzing the change trend of the voltage mode cluster to obtain a mutation characteristic factor, and hierarchically summarizing the voltage mode cluster based on the mutation characteristic factor to obtain a hierarchical voltage mode set includes: Acquire the historical change trend of each voltage mode component in the voltage mode cluster, perform time series analysis on the historical change trend, and obtain the voltage mode change rate and fluctuation frequency characteristics; constructing a voltage mode change trend matrix based on the voltage mode change rate and the fluctuation frequency characteristics, performing trend classification on the voltage mode clusters using the voltage mode change trend matrix, and generating a fluctuation anomaly detection factor based on the classification result; The voltage pattern cluster is screened for abnormality by using the fluctuation abnormality detection factor, abnormal pattern components are extracted, voltage mutation thresholds are calculated based on the abnormal pattern components, mutation patterns are screened out according to the voltage mutation thresholds, a voltage mutation feature matrix is ​​constructed based on the mutation pattern, and mutation feature factors are extracted based on the voltage mutation feature matrix; The voltage mode clusters are classified by using the mutation characteristic factors to generate voltage mode stratification rules, and the voltage mode clusters are stratified and summarized based on the voltage mode stratification rules to obtain a stratified voltage mode set.

5. The DC charging pile voltage stability optimization method according to claim 1, characterized in that: The step of modeling the voltage mapping relationship based on the voltage fluctuation trend weight to obtain a voltage prediction model, using the voltage prediction model to predict and calculate the real-time voltage data of the charging pile to obtain a predicted voltage deviation value, adjusting the DC voltage regulation control parameters preset by the charging pile according to the predicted voltage deviation value, and obtaining an optimization adjustment matrix and a voltage compensation factor matrix includes: Adaptively optimizing the voltage fluctuation map by using the voltage fluctuation trend weight to obtain an optimized voltage fluctuation map, extracting key fluctuation patterns based on the optimized voltage fluctuation map, and constructing a voltage prediction association matrix by using the key fluctuation patterns; Mapping and calculating the voltage fluctuation trend weight through the voltage prediction association matrix to obtain voltage trend prediction data and voltage disturbance prediction data; Building a voltage prediction model based on the voltage trend prediction data and the voltage disturbance prediction data, and using the voltage prediction model to predict and calculate the real-time voltage data of the charging pile to obtain a predicted voltage deviation value; The DC voltage regulation control parameters preset for the charging pile are adjusted using the predicted voltage deviation value to obtain a first optimized control parameter and a second optimized control parameter, an optimized adjustment matrix is ​​generated based on the first optimized control parameter, and a voltage compensation factor matrix is ​​constructed based on the second optimized control parameter.

6. The DC charging pile voltage stability optimization method according to claim 5, characterized in that: The step of adjusting the DC voltage regulation control parameters preset by the charging pile by using the predicted voltage deviation value to obtain the first optimized control parameter and the second optimized control parameter, generating an optimized adjustment matrix based on the first optimized control parameter, and constructing a voltage compensation factor matrix based on the second optimized control parameter includes: Calculating a voltage deviation compensation factor based on the predicted voltage deviation value, and adjusting a DC voltage regulation control parameter preset by the charging pile using the voltage deviation compensation factor to obtain an initial optimized control parameter; Extracting control parameter correlation factors based on the initial optimized control parameters, and constructing a control parameter optimization mapping matrix using the control parameter correlation factors; Performing nonlinear transformation calculation on the initial optimized control parameters through the control parameter optimization mapping matrix to obtain first optimized control parameters and second optimized control parameters; constructing a voltage regulation coefficient matrix using the first optimization control parameter, and generating an optimization adjustment matrix based on the voltage regulation coefficient matrix; A voltage error compensation matrix is ​​constructed using the second optimized control parameters, a dynamic compensation factor is calculated based on the voltage error compensation matrix, and a voltage compensation factor matrix is ​​generated based on the dynamic compensation factor.

7. The DC charging pile voltage stability optimization method according to claim 1, characterized in that: The step of dynamically adjusting the preset DC output control strategy of the charging pile based on the optimization adjustment matrix to obtain the adjusted control strategy parameters, performing compensation calculation on the adjusted control strategy parameters using the voltage compensation factor matrix to obtain the optimized control instruction, and generating a dynamic voltage stabilization control signal according to the optimized control instruction includes: Using the optimization adjustment matrix to perform strategy adjustment calculation on the DC output control strategy of the charging pile, and obtain the adjusted control strategy parameters and control strategy correction factor; The adjusted control strategy parameters are compensated by the voltage compensation factor matrix to obtain compensated control strategy parameters, nonlinear correction calculation is performed on the compensated control strategy parameters by the control strategy correction factor to obtain optimized control instructions, and a dynamic voltage stabilization control signal is generated based on the optimized control instructions.

8. A DC charging pile voltage stability optimization device, characterized in that: include: An acquisition module is used to acquire voltage sequence data of a DC output terminal of a charging pile, and to perform segmented statistics on the voltage sequence data based on a set timing window to obtain a voltage timing characteristic matrix; an identification module, configured to perform pattern recognition on the voltage time series feature matrix, extract key characteristic parameters of voltage fluctuation, construct a voltage fluctuation map based on the key characteristic parameters of voltage fluctuation, and calculate a voltage fluctuation trend weight according to the voltage fluctuation map; A calculation module, used to model the voltage mapping relationship based on the voltage fluctuation trend weight to obtain a voltage prediction model, use the voltage prediction model to predict and calculate the real-time voltage data of the charging pile, generate a predicted voltage deviation value, adjust the DC voltage regulation control parameters preset by the charging pile according to the predicted voltage deviation value, and obtain an optimization adjustment matrix and a voltage compensation factor matrix; The adjustment module is used to dynamically adjust the DC output control strategy preset by the charging pile based on the optimization adjustment matrix, obtain the adjusted control strategy parameters, use the voltage compensation factor matrix to perform compensation calculation on the adjusted control strategy parameters, obtain the optimization control instruction, and generate a dynamic voltage stabilization control signal according to the optimization control instruction.

9. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed by a processor, the processor is enabled to execute the DC charging pile voltage stability optimization method as described in any one of claims 1 to 7.

Images