A power load monitoring method and system based on data feedback

Abstract

The application relates to the technical field of power load monitoring, and particularly discloses a power load monitoring method and system based on data feedback, which comprises the following steps: S1, collecting relevant data of each working state of a charging gun, a charging gun environment and a corresponding charging vehicle in a charging station through a data collection layer; and S2, pre-processing the collected data through each edge computing layer; through comprehensive and targeted data collection and subsequent processing and analysis, the application achieves the effect of effectively monitoring the power load of the charging station; on one hand, comprehensive data collection can provide rich and accurate basic data for subsequent anomaly analysis, so that the anomaly analysis is more comprehensive and accurate; on the other hand, the layered processing mode can improve the data processing and transmission efficiency, abnormal conditions in the charging process can be found in time and a warning can be given, so that the safe and stable operation of the charging station is ensured.

Description

Technical Field

[0001] This invention relates to the field of power load monitoring technology, specifically to a power load monitoring method and system based on data feedback. Background Technology

[0002] With the booming development of the new energy vehicle industry, the number of charging stations is increasing daily. Their stable operation is crucial to ensuring the normal use of new energy vehicles and the stable power supply of the power grid. However, traditional power load monitoring methods have obvious technical bottlenecks.

[0003] Most existing technologies rely solely on threshold judgments of a single parameter to monitor the power load of charging stations. When faced with multi-dimensional data on charging guns, the charging gun environment, and the corresponding charging vehicles within the charging station, it is difficult to effectively integrate and analyze this complex data. This results in low accuracy in anomaly identification and an inability to detect potential safety hazards in a timely and precise manner. Furthermore, these traditional methods severely lack dynamic adaptability, failing to fully consider real-time changes in different environmental conditions and unable to dynamically adjust monitoring thresholds according to actual circumstances. Consequently, the monitoring effect is significantly reduced under complex operating conditions, making it difficult to ensure the stable and safe operation of charging stations. Summary of the Invention

[0004] The purpose of this invention is to provide a power load monitoring method and system based on data feedback, and to solve the following technical problems:

[0005] The question is how to improve the accuracy of anomaly identification during power load monitoring of charging stations.

[0006] The objective of this invention can be achieved through the following technical solutions:

[0007] A power load monitoring method based on data feedback, the method comprising:

[0008] S1. Collect relevant data on charging guns, charging gun environment, and corresponding charging vehicles in various working states within the charging station through the data acquisition layer.

[0009] S2. The collected data is preprocessed through each edge computing layer, and the preprocessed data is transmitted to the power load monitoring layer.

[0010] S3, the power load monitoring layer performs single-feature anomaly analysis and comprehensive anomaly analysis based on the received data, and generates corresponding early warning signals based on the analysis results.

[0011] Furthermore, the relevant data regarding the charging gun in operation, the charging gun environment, and the corresponding charging vehicle include:

[0012] Charging current, the charging current value of the charging pile, is obtained through the power sensor built into the charging pile;

[0013] Charging voltage, the charging voltage value of the charging pile, is obtained through the voltage sensor built into the charging pile;

[0014] The charging temperature, the charging temperature value of the charging pile, is obtained through the first temperature sensor built into the charging pile.

[0015] Ambient temperature: The ambient temperature value under the working state of the charging pile is obtained by a second temperature sensor near the charging pile.

[0016] Ambient humidity: The ambient humidity value during the operation of the charging pile is obtained by a humidity sensor near the charging pile.

[0017] Battery temperature, the battery temperature of the vehicle being charged while the charging pile is in operation, is obtained by interacting with the vehicle's BMS through the charging gun communication module.

[0018] Furthermore, the edge computing layer preprocesses the collected data, including:

[0019]

[0020] The change ΔA between adjacent data points of the a-th type of parameter is obtained by analyzing and calculating using formula (1). a ;

[0021] The number of parameters collected has N types, where a∈[1,N]. Let i be the value of the i-th data point of the a-th parameter. This represents the value of the (i-1)th data point of the a-th parameter.

[0022] The change ΔA between adjacent data points of the a-th parameter a Filtering threshold corresponding to preset parameters Perform a comparison;

[0023] when At that time, Mark as redundant data and remove;

[0024] when At that time, The data is marked as important and transmitted to the power load monitoring layer.

[0025] Furthermore, the process of generating the corresponding early warning signal includes:

[0026] S31. The power load monitoring layer interpolates and synchronizes the received data based on timestamps.

[0027] S32. The power load monitoring layer performs single-feature anomaly analysis based on the synchronized parameter data. If any type of feature data anomaly exists, a corresponding early warning signal is generated. If no type of feature data anomaly exists, step S33 is executed.

[0028] S33, the power load monitoring layer performs comprehensive anomaly analysis on various characteristics to determine whether to generate a corresponding early warning signal.

[0029] Furthermore, the process of single-feature anomaly analysis includes:

[0030] A sliding time window Q = [tk,t] is constructed through the power load monitoring layer;

[0031] The various parameter data within the sliding time window are processed at a preset frequency to obtain actual reference values ​​of M types of features, including average current, voltage change rate, and temperature rise slope.

[0032]

[0033] The deviation score ω of the b-th feature is obtained by analyzing and calculating using formula (2). b ;

[0034] Where t is the current time point, k is the length of the time window, b∈[1,M], W b W is the actual reference value for the b-th feature. b,std σ is the preset control value for feature type b. b The standard deviation of the b-th feature;

[0035] The deviation score ω of the b-th feature b The deviation score threshold range [ω] from the preset corresponding class features b,1 ,ω b,2 Compare;

[0036] If ω b <ω b,1 Then the anomaly score R of the b-th feature is b =0;

[0037] If ω b ∈[ω b,1 ,ω b,2 ], then the anomaly score R of the b-th feature is b =ω b ;

[0038] If ω b >ω b,2 If the corresponding abnormal characteristic is detected, a warning signal will be generated, and the charging circuit of the corresponding charging pile will be cut off.

[0039] Furthermore, the process of comprehensive anomaly analysis includes:

[0040]

[0041] The comprehensive anomaly score H is obtained by analysis and calculation using formula (3). risk ;

[0042] in, , where b is the weighting coefficient of the influence of the b-th type feature on the overall anomaly score;

[0043] The comprehensive abnormality score H risk Compare with the preset comprehensive anomaly score threshold range [H1,H2];

[0044] If H risk >H2, generates a warning signal for comprehensive analysis of abnormalities and cuts off the charging circuit of the corresponding charging pile;

[0045] If H1≤H risk ≤H2, the charging current of the corresponding charging pile will be limited to a preset proportional current;

[0046] If H risk

[0047] Furthermore, the method also includes:

[0048] S4. The power load monitoring layer adjusts the filtering threshold of the corresponding preset parameters at a preset frequency.

[0049]

[0050] The filter threshold corresponding to the preset parameters is obtained through formula (4).

[0051] Among them, T s The average ambient temperature within the sliding time window, T0 is the preset control temperature, and E... s The average ambient humidity within the sliding time window is given by E0, the preset control humidity is given by τ1, the temperature correction factor is given by τ2, and the humidity correction factor is given by τ2. To preset the basic threshold for the corresponding parameters, (T) s -T0)*τ1+(E s -E0)*τ2<1.

[0052] Furthermore, the method also includes:

[0053] S5. Analyze the historical data of all charging piles through the cloud optimization layer and adjust the preset frequency;

[0054] ​

[0055] The preset frequency P is obtained by analysis and calculation using formulas (5) and (6). s ;

[0056] Among them, P std The basic analysis frequency is θ, which is the frequency adjustment coefficient, and N is the frequency. s N represents the number of warnings triggered by all charging stations within a preset time period. std This is a preset threshold for the number of warnings. H is the average of all comprehensive anomaly scores for all charging stations within a preset time period. std The preset comprehensive abnormality score control value is γ1 and γ2, which are the weight coefficients of the factors affecting the frequency adjustment coefficient.

[0057] A power load monitoring system based on data feedback, the system being used in a power load monitoring method based on data feedback, the system comprising:

[0058] The data acquisition layer is used to collect relevant data on the charging guns, the charging gun environment, and the corresponding charging vehicles in various working states within the charging station.

[0059] The edge computing layer is used to preprocess the collected data and transmit the preprocessed data to the power load monitoring layer.

[0060] The power load monitoring layer is used to perform single-feature anomaly analysis and comprehensive anomaly analysis based on the received data, and generate corresponding early warning signals based on the analysis results.

[0061] The cloud-based optimization layer is used to analyze historical data from all charging stations and adjust preset frequencies.

[0062] The beneficial effects of this invention are:

[0063] (1) This invention collects relevant data on charging guns, charging gun environments, and corresponding charging vehicles in various working states within the charging station through a data acquisition layer. Then, the collected data is preprocessed through various edge computing layers, and the preprocessed data is transmitted to the power load monitoring layer. Finally, the power load monitoring layer performs single-feature anomaly analysis and comprehensive anomaly analysis based on the received data, and generates corresponding early warning signals based on the analysis results. Through such comprehensive and targeted data collection and subsequent processing and analysis, the power load of the charging station is effectively monitored. On the one hand, comprehensive data collection can provide rich and accurate basic data for subsequent anomaly analysis, making the anomaly analysis more comprehensive and accurate. On the other hand, the layered processing method can improve the efficiency of data processing and transmission, promptly detect abnormal situations in the charging process and issue early warnings, and ensure the safe and stable operation of the charging station. Attached Figure Description

[0064] The invention will now be further described with reference to the accompanying drawings.

[0065] Figure 1 This is a flowchart of the steps of a power load monitoring method based on data feedback proposed in this invention;

[0066] Figure 2 This is a schematic block diagram of a power load monitoring system based on data feedback proposed in this invention. Detailed Implementation

[0067] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0068] Please see Figure 1 As shown, in one embodiment, a power load monitoring method based on data feedback is provided, the method comprising:

[0069] S1. The data acquisition layer collects relevant data on the charging guns, charging gun environment, and corresponding charging vehicles at various working states within the charging station. This data includes: charging current (the charging current value of the charging pile, obtained through a built-in power sensor); charging voltage (the charging voltage value of the charging pile, obtained through a built-in voltage sensor); charging temperature (the charging temperature value of the charging pile, obtained through a built-in first temperature sensor); ambient temperature (the ambient temperature value of the charging pile during its working state, obtained through a second temperature sensor near the charging pile); ambient humidity (the ambient humidity value of the charging pile during its working state, obtained through a humidity sensor near the charging pile); and battery temperature (the battery temperature of the corresponding charging vehicle during its working state, obtained through interaction between the charging gun communication module and the vehicle's BMS).

[0070] S2. The collected data is preprocessed through each edge computing layer, and the preprocessed data is transmitted to the power load monitoring layer.

[0071] S3, the power load monitoring layer performs single-feature anomaly analysis and comprehensive anomaly analysis based on the received data, and generates corresponding early warning signals based on the analysis results.

[0072] Through the above technical solution, this embodiment provides a power load monitoring method based on data feedback. The method collects relevant data on charging guns, the charging gun environment, and corresponding charging vehicles at various operating states within the charging station through a data acquisition layer. This data covers multiple aspects such as charging current, charging voltage, charging temperature, ambient temperature, and battery temperature. Then, the collected data is preprocessed through various edge computing layers, and the preprocessed data is transmitted to the power load monitoring layer. Finally, the power load monitoring layer performs single-feature anomaly analysis and comprehensive anomaly analysis based on the received data, and generates corresponding early warning signals based on the analysis results. Through this comprehensive and targeted data collection and subsequent processing and analysis, the power load of the charging station is effectively monitored. On the one hand, comprehensive data collection provides rich and accurate basic data for subsequent anomaly analysis, making the analysis more comprehensive and precise. On the other hand, the layered processing method improves the efficiency of data processing and transmission, promptly detects anomalies during the charging process, and issues early warnings, ensuring the safe and stable operation of the charging station.

[0073] In one embodiment, the edge computing layer preprocesses the collected data by performing the following steps:

[0074]

[0075] The change ΔA between adjacent data points of the a-th type of parameter is obtained by analyzing and calculating using formula (1). a ;

[0076] The number of parameters collected can be categorized into N types, which is obtained by counting all the parameter categories that need to be collected. For example, if the parameters collected are charging current, charging voltage, charging temperature, ambient temperature, ambient humidity, and battery temperature, then N = 6, and a ∈ [1, N]. The value of the i-th data point for the a-th parameter is obtained by the corresponding sensor. For example, for the charging current parameter (a=1), the value of the i-th data point is the current value. The value of the (i-1)th data point of the a-th parameter is also collected by the sensor and is the value of the parameter corresponding to the previous data point.

[0077] The change ΔA between adjacent data points of the a-th parameter a Filtering threshold corresponding to preset parameters Perform a comparison;

[0078] when At that time, Mark as redundant data and remove;

[0079] when At that time, Mark the data as important and transmit it to the power load monitoring layer;

[0080] For example, suppose the charging current data collected is the current value collected at time i-1. The current value collected at time i The change ΔA between adjacent data points 1 =|22A-20A|=2A, while the preset charging current parameter filtering threshold at this time Then The data is marked as important and transmitted to the power load monitoring layer.

[0081] Through the above technical solution, this embodiment provides a method for edge computing layer to preprocess collected data. The change of adjacent data points of the a-type parameter is calculated by formula (1), and compared with the preset filtering threshold of the corresponding parameter to distinguish redundant data and important data. The important data is then transmitted to the power load monitoring layer, thereby reducing the amount of transmitted data, improving data transmission efficiency and processing targeting.

[0082] In one embodiment, the process of generating the corresponding warning signal includes:

[0083] S31. The power load monitoring layer performs interpolation synchronization on the received data based on the timestamp. The interpolation synchronization method includes, but is not limited to, spline interpolation, Lagrange interpolation, Kalman filter interpolation, etc., which are existing technologies and will not be described in detail here.

[0084] S32. The power load monitoring layer performs single-feature anomaly analysis based on the synchronized parameter data. If any type of feature data anomaly exists, a corresponding early warning signal is generated. If no type of feature data anomaly exists, step S33 is executed.

[0085] S33, the power load monitoring layer performs comprehensive anomaly analysis on various characteristics to determine whether to generate a corresponding early warning signal.

[0086] The process of single-feature anomaly analysis includes:

[0087] A sliding time window Q = [tk,t] is constructed through the power load monitoring layer;

[0088] The various parameter data within the sliding time window are processed at a preset frequency to obtain actual reference values ​​for M types of features. The M types of features include, but are not limited to, average current, voltage change rate, temperature rise slope, power change rate, etc., which can ensure that the features for monitoring electrical load are met. The specific features can be preset based on experience.

[0089]

[0090] The deviation score ω of the b-th feature is obtained by analyzing and calculating using formula (2). b ;

[0091] Where t is the current time point, obtained from the system clock, k is the length of the time window, set according to actual needs, generally set to 5 minutes or 10 minutes, b∈[1,M], W b The actual reference value for feature b is obtained by processing and calculating the data within the sliding time window. For example, for the feature of average current (b=1), W1 is the average current within that time window. b,std The preset reference value for feature b can be obtained through experience. Specifically, it can be preset by referring to the rated parameters of the charging pile, historical data statistics, or vehicle charging parameters read through the vehicle's BMS. For example, the preset reference value corresponding to the average current feature can be the minimum value between the charging pile's rated charging current and the vehicle's maximum allowable charging current. b The standard deviation of the b-th characteristic is obtained through statistical analysis of historical data and reflects the normal fluctuation range of this characteristic.

[0092] The deviation score ω of the b-th feature b The deviation score threshold range [ω] from the preset corresponding class features b,1 ,ω b,2 The comparison is performed, and the preset deviation scoring threshold range [ω] of the corresponding class features is determined. b,1 ,ω b,2 This can be preset based on experience;

[0093] If ω b <ω b,1 If the deviation of the b-th feature within the sliding time window is determined to be a normal deviation, then the abnormality score R of the b-th feature is... b =0;

[0094] If ω b ∈[ω b,1 ,ω b,2 If it is determined that the deviation of the b-th feature is slightly abnormal within the sliding time window, then the abnormality score R of the b-th feature is... b =ω b ;

[0095] If ω b >ω b,2 If it is determined that there is a severe abnormality in the deviation of the b-th type of feature within the sliding time window, an early warning signal for the corresponding type of feature abnormality is generated, and the charging circuit of the corresponding charging pile is cut off.

[0096] The process of comprehensive anomaly analysis includes:

[0097]

[0098] The comprehensive anomaly score H is obtained by analysis and calculation using formula (3). risk ;

[0099] in, The weighting coefficient for the influence of feature type b on the overall anomaly score is set according to the importance of each type of feature.

[0100] The comprehensive abnormality score H risk The score is compared with a preset comprehensive anomaly score threshold range [H1, H2], which can be obtained by preset based on experience.

[0101] If H risk >H2, if it is determined that there is a severe abnormality in the charging status of the charging pile within the sliding time window, a warning signal for comprehensive analysis of the abnormality is generated, and the charging circuit of the corresponding charging pile is cut off.

[0102] If H1≤H risk If ≤H2, it is determined that there is a slight abnormality in the charging status of the charging pile during the sliding time window, and the charging current of the corresponding charging pile is limited to a preset proportional current.

[0103] If H risk

[0104] ​Through the above technical solution, this embodiment provides a method for generating corresponding early warning signals during power load monitoring. Specifically, the power load monitoring layer first interpolates and synchronizes the received data based on the timestamp to ensure the consistency and accuracy of the data in the time dimension, providing a reliable data foundation for subsequent anomaly analysis. Then, single-feature anomaly analysis is performed. By constructing a sliding time window, various parameter data within the window are processed at a preset frequency to obtain actual reference values ​​of M types of features such as average current, voltage change rate, and temperature rise slope. Then, the deviation score of the b-th feature is calculated using formula (2). The deviation score of the b-th feature is compared with the preset deviation score threshold range of the corresponding type of feature. The anomaly score of the b-th feature is determined based on the comparison result. If it exceeds the threshold range, an early warning signal for the corresponding type of feature anomaly is generated, and the charging circuit of the corresponding charging pile is cut off. If no anomaly is found in the single-feature anomaly analysis, a comprehensive anomaly analysis is performed. The comprehensive anomaly score is calculated using formula (3). The comprehensive anomaly score is compared with the preset comprehensive anomaly score threshold range. If it exceeds the threshold range, a warning signal for comprehensive anomaly analysis is generated, and the charging circuit of the corresponding charging pile is cut off. If it is within a certain range, the charging current of the corresponding charging pile is limited to a preset proportional current. If it does not exceed the threshold range, no warning signal is generated. Through this dual anomaly analysis mechanism of first single-feature analysis and then comprehensive analysis, the effect of more comprehensive and accurate identification of anomalies and timely warning is achieved. Single-feature anomaly analysis can quickly discover the anomaly of a single feature and take timely measures to avoid the problem from expanding. Comprehensive anomaly analysis considers the mutual influence between various features from an overall perspective, further improving the accuracy and reliability of anomaly identification and ensuring the safe and stable operation of the charging station.

[0105] For example, in single-feature anomaly analysis, assuming the current time point t = 12:00, the time window length is 10 minutes, for the current mean feature (b = 1), the actual reference value W1 = 55A, and the preset control value W 1,std =50A, standard deviation σ1=2A, then the deviation score If the preset deviation score threshold range for the corresponding feature is 0-2, exceeding the threshold will generate an alarm signal indicating abnormal current mean characteristics and cut off the charging circuit of the corresponding charging pile.

[0106] In the comprehensive anomaly analysis, it is assumed that there are three types of features: mean current, rate of change of voltage, and slope of temperature rise (M=3). The anomaly score for the mean current feature is R1=0, and the weighting coefficient is... Voltage change rate anomaly score R² = 1.5, weighting coefficient Anomaly score for temperature rise slope characteristic R3 = 3, weighting coefficient Then the comprehensive abnormal score If the preset comprehensive anomaly score threshold range is 0-1, exceeding the threshold will generate a warning signal for comprehensive analysis anomalies and cut off the charging circuit of the corresponding charging pile.

[0107] In one embodiment, the method further includes:

[0108] S4. The power load monitoring layer adjusts the filtering threshold of the corresponding preset parameters at a preset frequency.

[0109]

[0110] The filter threshold corresponding to the preset parameters is obtained through formula (4).

[0111] Among them, T s The average ambient temperature within the sliding time window is calculated by averaging the data collected by the second temperature sensor within the time window. T0 is the preset control temperature, set according to the normal operating ambient temperature of the charging pile. E s The average ambient humidity within the sliding time window is calculated by averaging data collected from the corresponding humidity sensor. E0 is the preset control humidity, set empirically. τ1 is the temperature correction coefficient, representing the influence of temperature on the threshold. τ2 is the humidity correction coefficient, representing the influence of humidity on the threshold. τ1 and τ2 can be experimentally obtained based on the degree of influence of the environment on each parameter. The basic threshold for the preset corresponding parameters is the basic filtering threshold set based on historical experience.

[0112] Through the above technical solution, this embodiment provides a method for adjusting the filtering threshold of preset corresponding parameters. The filtering threshold of preset corresponding parameters is obtained by formula based on factors such as the average temperature and humidity of the environment within the sliding time window, thereby achieving the effect of enabling the filtering threshold to be adaptively adjusted according to environmental changes and improving the accuracy of data preprocessing.

[0113] For example, the average ambient temperature T within the sliding time window s =30 degrees, preset control temperature T0=25 degrees, average ambient humidity E within the sliding time window s =60%RH, preset control humidity E0 = 50%RH, temperature correction factor τ1 = 0.005, humidity correction factor τ2 = 0.003, preset current parameter base threshold is 1A, at this time the preset current parameter filtering threshold That is, when temperature and humidity rise, more monitoring of the data collected by the sensor is needed. If the change in current parameter between adjacent data points is ΔA... 1 =0.95A, then the corresponding current The data is marked as important and transmitted to the power load monitoring layer.

[0114] In one embodiment, the method further includes:

[0115] S5. Analyze the historical data of all charging piles through the cloud optimization layer and adjust the preset frequency;

[0116]

[0117] The preset frequency P is obtained by analysis and calculation using formulas (5) and (6). s ;

[0118] Among them, P std The basic analysis frequency is determined based on the initial system settings, typically taken as 1Hz, where θ is the frequency adjustment coefficient, and N... s N represents the number of warnings triggered by all charging stations within a preset time period, obtained through statistical analysis of historical data. std The preset warning frequency threshold is set based on the normal operating conditions of the charging pile. H is the average of all comprehensive anomaly scores for all charging piles within a preset time period. It is obtained by averaging the comprehensive anomaly scores of each charging pile within the preset time period. std The preset comprehensive anomaly score reference value is obtained based on experience. γ1 and γ2 are the weight coefficients of the factors affecting the frequency adjustment coefficient, which are determined according to the importance of the number of warnings and the comprehensive anomaly score to the frequency adjustment. The preset time period is set based on experience, and is generally 3 days or 5 days.

[0119] Through the above technical solution, this embodiment provides a method for adjusting the preset frequency through a cloud optimization layer. The preset frequency is obtained by analyzing and calculating factors such as the number of warnings and the average value of comprehensive anomaly scores within a preset time period using a formula. This achieves the effect of dynamically adjusting the monitoring frequency according to the actual operation of the charging pile, thereby improving monitoring efficiency and resource utilization.

[0120] For example, the fundamental analysis frequency P std =1Hz, the number of warnings N triggered by all charging piles within a preset time period. s =15 times, preset warning count reference value N std =10 times, the average of all comprehensive anomaly scores for all charging stations within a preset time period. Preset comprehensive abnormality score comparison value H std =1, weighting coefficients γ1=0.7, γ2=0.3, then the adjusted preset frequency At this time, the cloud optimization layer will control the power load monitoring layer to adjust the frequency of data analysis to 2.04Hz, thereby improving the monitoring efficiency. The cloud optimization layer adjusts the preset frequency once every 24 hours or 12 hours.

[0121] Please see Figure 2 As shown, in one embodiment, a data feedback-based power load monitoring system is provided. The system is used in a data feedback-based power load monitoring method, and includes:

[0122] The data acquisition layer is used to collect relevant data on the charging guns, the charging gun environment, and the corresponding charging vehicles in various working states within the charging station.

[0123] The edge computing layer is used to preprocess the collected data and transmit the preprocessed data to the power load monitoring layer.

[0124] The power load monitoring layer is used to perform single-feature anomaly analysis and comprehensive anomaly analysis based on the received data, and generate corresponding early warning signals based on the analysis results.

[0125] The cloud-based optimization layer is used to analyze historical data from all charging stations and adjust preset frequencies.

[0126] Through the above technical solution, this embodiment provides a power load monitoring system based on data feedback. The system includes a data acquisition layer, an edge computing layer, a power load monitoring layer, and a cloud optimization layer, which are used to implement the power load monitoring method based on data feedback. It achieves the effect of comprehensively and efficiently monitoring and optimizing the power load of charging stations through a layered architecture and collaborative work.

[0127] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A power load monitoring method based on data feedback, characterized in that, The method includes: S1. Collect relevant data on charging guns, charging gun environment, and corresponding charging vehicles in various working states within the charging station through the data acquisition layer. S2. The collected data is preprocessed through each edge computing layer, and the preprocessed data is transmitted to the power load monitoring layer. S3, the power load monitoring layer performs single-feature anomaly analysis and comprehensive anomaly analysis based on the received data, and generates corresponding early warning signals based on the analysis results; The process of generating the corresponding early warning signal includes: S31. The power load monitoring layer interpolates and synchronizes the received data based on timestamps. S32. The power load monitoring layer performs single-feature anomaly analysis based on the synchronized parameter data. If any type of feature data anomaly exists, a corresponding early warning signal is generated. If no type of feature data anomaly exists, step S33 is executed. S33. The power load monitoring layer performs comprehensive anomaly analysis on various characteristics to determine whether to generate a corresponding early warning signal. The process of single-feature anomaly analysis includes: Constructing a sliding time window through the power load monitoring layer ; The various parameter data within the sliding time window are processed at a preset frequency to obtain actual reference values ​​for M types of features, including average current, voltage change rate, and temperature rise slope. The deviation score of the b-th feature is obtained by analyzing and calculating using formula (2). ; in, At the current time point, The length of the time window, , This is the actual reference value for the b-th feature. This is the preset control value for feature type b. The standard deviation of the b-th feature; The deviation score of the b-th feature Deviation scoring threshold range from the preset corresponding class features Perform a comparison; like Then the anomaly score of feature b ; like Then the anomaly score of feature b ; like If the corresponding abnormal characteristic is detected, a warning signal will be generated and the charging circuit of the corresponding charging pile will be cut off. The process of comprehensive anomaly analysis includes: The comprehensive anomaly score is obtained by analysis and calculation using formula (3). ; in, , where b is the weighting coefficient of the influence of the b-th type feature on the overall anomaly score; Comprehensive abnormal score Compared with the preset comprehensive anomaly scoring threshold range Perform a comparison; like It generates a comprehensive analysis of abnormalities and cuts off the charging circuit of the corresponding charging pile. like The charging current of the corresponding charging pile will be limited to a preset ratio. like If not, no warning signal will be generated.

2. The power load monitoring method based on data feedback according to claim 1, characterized in that, The relevant data regarding the charging gun in operation, the charging gun environment, and the corresponding charging vehicle include: Charging current, the charging current value of the charging pile, is obtained through the power sensor built into the charging pile; Charging voltage, the charging voltage value of the charging pile, is obtained through the voltage sensor built into the charging pile; The charging temperature, the charging temperature value of the charging pile, is obtained through the first temperature sensor built into the charging pile. Ambient temperature: The ambient temperature value under the working state of the charging pile is obtained by a second temperature sensor near the charging pile. Ambient humidity: The ambient humidity value during the operation of the charging pile is obtained by a humidity sensor near the charging pile. Battery temperature, the battery temperature of the vehicle being charged while the charging pile is in operation, is obtained by interacting with the vehicle's BMS through the charging gun communication module.

3. The power load monitoring method based on data feedback according to claim 2, characterized in that, The edge computing layer preprocesses the collected data, including: The first result is obtained by analyzing and calculating using formula (1). Class parameter: change in adjacent data points ; The number of parameters collected falls into N categories. , For the first The value of the i-th data point in the class parameter. For the first The value of the (i-1)th data point in the class parameter; The first Variation of parameters between adjacent data points Filtering threshold corresponding to preset parameters Perform a comparison; when At that time, Mark as redundant data and remove; when At that time, The data is marked as important and transmitted to the power load monitoring layer.

4. The power load monitoring method based on data feedback according to claim 3, characterized in that, The method further includes: S4. The power load monitoring layer adjusts the filtering threshold of the corresponding preset parameters at a preset frequency. The filter threshold corresponding to the preset parameters is obtained through formula (4). ; in, The average ambient temperature within the sliding time window. To preset the control temperature, The average ambient humidity within the sliding time window. As a preset control humidity, This is a temperature correction factor. This is the humidity correction factor. To preset the basic threshold for the corresponding parameters, .

5. The power load monitoring method based on data feedback according to claim 4, characterized in that, The method further includes: S5. Analyze the historical data of all charging piles through the cloud optimization layer and adjust the preset frequency; The preset frequency is obtained by analysis and calculation using formulas (5) and (6). ; in, Based on the frequency of analysis, This is the frequency adjustment coefficient. This represents the number of warnings triggered by all charging stations within a preset time period. This is a preset threshold for the number of warnings. This is the average of all comprehensive anomaly scores for all charging stations within a preset time period. To preset the comprehensive anomaly score comparison value, , This refers to the weighting coefficients of the factors influencing the frequency adjustment coefficient.

6. A power load monitoring system based on data feedback, characterized in that, The system is used in the power load monitoring method based on data feedback as described in claim 5, and the system comprises: The data acquisition layer is used to collect relevant data on the charging guns, the charging gun environment, and the corresponding charging vehicles in various working states within the charging station. The edge computing layer is used to preprocess the collected data and transmit the preprocessed data to the power load monitoring layer. The power load monitoring layer is used to perform single-feature anomaly analysis and comprehensive anomaly analysis based on the received data, and generate corresponding early warning signals based on the analysis results. The cloud-based optimization layer is used to analyze historical data from all charging stations and adjust preset frequencies.

Images