Low-voltage charging pile identification method, system and device based on support vector machine model and storage medium
The support vector machine model is used to identify low-voltage charging pile users, which solves the problem of lagging charging infrastructure construction and improves the charging experience of new energy vehicles and the operating efficiency of power systems.
Patent Information
- Application Number
- CN202411019562.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-29
- Publication Date
- 2024-12-06
AI Technical Summary
It is difficult for the existing technology to accurately identify and manage low-voltage charging pile users, resulting in lagging charging infrastructure construction and affecting the development of new energy vehicles.
The support vector machine model is adopted to obtain data from the power consumption information collection system, perform preprocessing and feature extraction, build a support vector machine model, and use grid search to adjust hyperparameters to realize automatic identification of low-voltage charging pile users.
It significantly improves the accuracy and efficiency of user identification of low-voltage charging piles, reduces dependence on manual operations, improves the operating efficiency and energy management capabilities of the power system, and avoids energy waste.
Smart Images

Figure CN119089283A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of power systems, and in particular relates to a low-voltage charging pile identification method, system, device and storage medium based on a support vector machine model. Background Art
[0002] In recent years, with the increase of environmental awareness and the encouragement and support of relevant national policies, the new energy vehicle market has shown a rapid growth trend. The popularity of new energy vehicles has led to an increase in the demand for charging piles as supporting facilities. Although the new energy vehicle market is booming, the construction of charging infrastructure is relatively lagging, which has become one of the bottlenecks restricting the development of new energy vehicles. Factors such as the number of charging piles, the breadth of distribution, and the quality of service directly affect the charging experience of new energy vehicle users.
[0003] Different users have different requirements for charging piles, such as charging speed, charging cost, charging pile location, etc. Therefore, the power system needs to be able to accurately identify charging pile users in order to provide more personalized and efficient services. Summary of the invention
[0004] Purpose of the invention: The purpose of the present invention is to provide a low-voltage charging pile identification method, system, device and storage medium based on a support vector machine model, which realizes the automatic identification of low-voltage charging pile users.
[0005] Technical solution: A low-voltage charging pile identification method based on a support vector machine model of the present invention includes:
[0006] Measurement data acquisition: obtain the user's ledger data and active power measurement data at different time points from the electricity consumption information collection system;
[0007] Measurement data preprocessing: preprocess the acquired data, use the isnull function to identify missing values, use the box plot method to identify outliers, and use Newton interpolation to fill in the identified outliers and missing values;
[0008] Feature extraction: Extract statistical dimension and business dimension data features from the pre-processed active power measurement data, label the user as a charging pile based on the ledger data, form a sample set, and randomly divide the sample set into a training sample set and a test sample set;
[0009] Model construction: Select the support vector machine model in the Scikit-learn library, set the initial hyperparameter range of the support vector machine model according to the sample information, and build a basic support vector machine model;
[0010] Model training and testing: Import the feature data of the training sample set into the support vector machine model. In the process of training the basic support vector machine model, use the grid search method to adjust the hyperparameters within the range, select the model that achieves the best prediction accuracy on the training sample set, and finally obtain a low-voltage charging pile user recognition model based on the support vector machine model; use the test sample set to test the low-voltage charging pile user recognition model, and use the precision, recall rate and F1 score to evaluate the performance of the low-voltage charging pile user recognition model;
[0011] Low-voltage charging pile user identification: Obtain the full amount of non-charging pile user power data, use the low-voltage charging pile user identification model to identify, and finally obtain a list of users identified as charging piles.
[0012] Furthermore, the box plot method is used to identify outliers, including:
[0013] First, sort the acquired data, determine the first quartile Q1 and the third quartile Q3, and then calculate the interquartile range IQR, the formula is as follows:
[0014] IQR=Q3-Q1
[0015] According to the interquartile range, values less than Q1-1.5×IQR or greater than Q3+1.5×IQR were identified as outliers.
[0016] Furthermore, the expression of the Newton interpolation method is as follows:
[0017]
[0018] P(x) represents the polynomial function obtained by Newton interpolation method, which is used to estimate the interpolation result at the given position x value of the measured data point; p0 represents the initial power value point; x j Indicates the position of the jth power value point; Δ k p0 represents the k-order difference quotient of p0, and the difference quotient is defined as follows:
[0019] Δ k p i =Δ k-1 p i+1 -Δ k-1 p i
[0020] Δ 1 p i =p i+1 -p i
[0021] Among them, p i Represents the power value of the i-th position.
[0022] Furthermore, the statistical dimension data features include the kurtosis, skewness, maximum value, and median value of the user's time-divided active power measurement data.
[0023] Furthermore, the kurtosis of the user's time-divided active power measurement data is used to measure the steepness of the probability distribution of the random variable, and its expression is as follows:
[0024]
[0025] The skewness of the user's time-sharing active power measurement data is used to measure the asymmetry of the probability distribution of random variables, and its expression is as follows:
[0026]
[0027] Among them, S represents kurtosis; K represents skewness; X i It represents the specific value in the original data; μ represents the mean or average of the original data; σ represents the standard deviation of the original data.
[0028] Furthermore, the business dimension data features include the number of points where the user's time-sharing point active power is less than 1.2 times the minimum value, the number of points where the user's time-sharing point active power is greater than 0.8 times the maximum value, the number of points where the user's time-sharing point active power is greater than 1.5 times the median or less than 0.5 times the median, the number of points where the first-order difference of the user's time-sharing point active power is less than 0.1, the number of points where the first-order difference of the user's time-sharing point active power is greater than 0.8 times the maximum value, and the correlation coefficient between the user's time-sharing point active power on the current day and the time-sharing point active power on the previous day.
[0029] Furthermore, the construction process of the basic support vector machine model is as follows:
[0030] Select the support vector machine model in the Scikit-learn library, set the regularization coefficient C to a value range of 0.5 to 10, select the kernel function of the model from the linear kernel function, polynomial kernel function and Gaussian kernel function, use the grid search method during training, adjust the above hyperparameters, and build a basic support vector machine model.
[0031] Based on the same inventive concept, a low-voltage charging pile identification system based on a support vector machine model of the present invention includes:
[0032] The measurement data acquisition module is used to obtain the user's account data and time-sharing active power measurement data from the power consumption information collection system;
[0033] The measurement data preprocessing module is used to preprocess the acquired data, identify missing values using the isnull function, identify outliers using the box plot method, and fill in the identified outliers and missing values using the Newton interpolation method;
[0034] The feature extraction module is used to extract statistical dimension and business dimension data features from the pre-processed active power measurement data, label the user as a charging pile according to the ledger data, form a sample set, and randomly divide the sample set into a training sample set and a test sample set;
[0035] The model building module is used to select the support vector machine model in the Scikit-learn library, set the initial hyperparameter range of the support vector machine model according to the sample information, and build a basic support vector machine model;
[0036] The model training and testing module is used to import the feature data of the training sample set into the support vector machine model. In the process of training the basic support vector machine model, the grid search method is used to adjust the hyperparameters within a range, and the model is selected to achieve the best prediction accuracy on the training sample set, and finally a low-voltage charging pile user recognition model based on the support vector machine model is obtained; it is also used to test the low-voltage charging pile user recognition model using the test sample set, and to evaluate the performance of the low-voltage charging pile user recognition model using the precision, recall rate and F1 score;
[0037] The low-voltage charging pile user identification module is used to obtain the full amount of non-charging pile user power data, use the low-voltage charging pile user identification model to identify, and finally obtain a list of users identified as charging piles.
[0038] Based on the same inventive concept, a low-voltage charging pile identification device based on a support vector machine model of the present invention includes a processor and a memory, wherein the memory stores computer instructions, and the processor is used to execute the computer instructions stored in the memory. When the computer instructions are executed by the processor, the electronic device implements the steps of the low-voltage charging pile identification method based on the support vector machine model as described above.
[0039] Based on the same inventive concept, a computer-readable storage medium of the present invention stores a computer program thereon, and when the program is executed by a processor, the steps of the low-voltage charging pile identification method based on the support vector machine model as described above are implemented.
[0040] Beneficial effects: Compared with the prior art, the significant technical effects of the present invention are:
[0041] This solution obtains user records and active power data at different time points from the electricity consumption information collection system. It deeply mines data characteristics, obtains statistical dimension and business dimension features, and establishes a support vector machine model to identify charging pile users whose records are not maintained.
[0042] After experimental verification, the support vector machine model of the present invention has shown excellent performance and significant improvement in the task of low-voltage charging pile user identification. The model makes full use of the powerful ability of machine learning in processing sequence data, significantly improving the accuracy and efficiency of low-voltage charging pile user identification. By intelligently identifying and scheduling charging pile users, the power system can more effectively manage energy supply and avoid energy waste.
[0043] The identification method adopted by the present invention is highly automated, greatly reducing the reliance on manual operation, thereby effectively improving the operating efficiency and safety performance of the power system. BRIEF DESCRIPTION OF THE DRAWINGS
[0044] Figure 1 It is a flow chart of a low-voltage charging pile identification method based on a support vector machine model disclosed in an embodiment of the present invention;
[0045] Figure 2 It is a confusion matrix diagram of a binary classification model on a test set disclosed in an embodiment of the present invention;
[0046] Figure 3 It is a structural schematic diagram of a low-voltage charging pile identification system based on a support vector machine model disclosed in an embodiment of the present invention;
[0047] Figure 4 It is a structural schematic diagram of a low-voltage charging pile identification device based on a support vector machine model disclosed in an embodiment of the present invention. DETAILED DESCRIPTION
[0048] The technical solution of the present invention is described in detail below in conjunction with specific implementation methods and the accompanying drawings.
[0049] Example 1
[0050] like Figure 1 As shown, a low-voltage charging pile identification method based on a support vector machine model of the present invention comprises the following steps:
[0051] S1. Measurement data acquisition: Obtain the user's ledger data and time-based active power measurement data from the electricity consumption information collection system.
[0052] S2. Measurement data preprocessing: Preprocess the acquired data, use the isnull function to identify missing values, use the box plot method to identify outliers, and use the Newton interpolation method to fill in the identified outliers and missing values.
[0053] During the data collection process, due to the influence of various factors, the raw data often contains noise and outliers. In order to ensure the accuracy and reliability of the data, it is very important to preprocess the raw data. Preprocessing mainly includes data cleaning and eliminating outliers, thereby reducing the negative impact on model training. Step S2 specifically includes the following:
[0054] S2.1. Format and integrate the acquired measurement data into the Dataframe data structure in Python; use the isnull function in the Pandas library to identify missing values.
[0055] S2.2. Use the box plot method to identify outliers, including:
[0056] First, sort the acquired data, determine the first quartile Q1 and the third quartile Q3, and then calculate the interquartile range IQR, the formula is as follows:
[0057] IQR=Q3-Q1
[0058] According to the interquartile range, values less than Q1-1.5×IQR or greater than Q3+1.5×IQR were identified as outliers.
[0059] S2.3. Use Newton interpolation to fill in the identified outliers and missing values. The expression of Newton interpolation is as follows:
[0060]
[0061] P(x) represents the polynomial function obtained by Newton interpolation method, which is used to estimate the interpolation result at the given position x value of the measured data point; p0 represents the initial power value point; x j Indicates the position of the jth power value point; Δ k p0 represents the k-order difference quotient of p0, and the difference quotient is defined as follows:
[0062] Δ k p i =Δ k-1 p i+1 -Δ k-1 p i
[0063] Δ 1 p i =p i+1 -p i
[0064] Among them, p i Represents the power value of the i-th position.
[0065] S3. Feature extraction: Extract statistical dimension and business dimension data features from the pre-processed active power measurement data, label the user as a charging pile based on the ledger data, form a sample set, and randomly divide the sample set into a training sample set and a test sample set.
[0066] Extract 4-dimensional statistical features and 6-dimensional business features from the pre-processed 96-dimensional time-sharing active power data of the user on the same day, including: 4 statistical dimension features of the user's time-sharing active power measurement data, namely, kurtosis, skewness, maximum value, and median; 6 business dimension features of the number of points where the user's time-sharing active power is less than 1.2 times the minimum value, the number of points where the user's time-sharing active power is greater than 0.8 times the maximum value, the number of points where the user's time-sharing active power is greater than 1.5 times the median or less than 0.5 times the median, the number of points where the first-order difference of the user's time-sharing active power is less than 0.1, the number of points where the first-order difference of the user's time-sharing active power is greater than 0.8 times the maximum value, and the correlation coefficient between the user's time-sharing active power on the same day and the previous day's time-sharing active power. According to whether it is a charging pile user label, combined with the above characteristics, a sample set is formed; the sample set is divided into a training sample set and a test sample set. Step S3 specifically includes the following steps:
[0067] S3.1. Obtain statistical dimension features through user active power data, including the kurtosis, skewness, maximum value, and median of the user's active power data at different time points. The calculation process of skewness and kurtosis is as follows:
[0068] The kurtosis of the user's time-sharing active power measurement data is used to measure the steepness of the probability distribution of random variables. Its expression is as follows:
[0069]
[0070] The skewness of the user's time-sharing active power measurement data is used to measure the asymmetry of the probability distribution of random variables. Its expression is as follows:
[0071]
[0072] Among them, S represents kurtosis; K represents skewness; X i It represents the specific value in the original data; μ represents the mean or average of the original data; σ represents the standard deviation of the original data.
[0073] S3.2. Obtaining business dimension features through user active power data is conceived as follows:
[0074] S3.2.1. Considering that the active power of the charging pile user should be 0 kW or very small during the non-charging stage, the number of points at which the active power of the user's time-sharing point is less than 1.2 times the minimum value is calculated;
[0075] S3.2.2. Considering that the active power of charging pile users tends to be stable or fluctuates within a small range during the charging stage, and the charging time of a large number of low-voltage users is within a certain range, therefore, the number of points at which the active power of the user's time-sharing point is greater than 0.8 times the maximum value is calculated;
[0076] S3.2.3. Considering that the active power of the charging pile users in the charging and non-charging stages should hardly have points near the median, therefore, calculate the number of points where the active power of the user's time-sharing points is greater than 1.5 times the median or less than 0.5 times the median;
[0077] S3.2.4. Considering that the active power of charging pile users has almost no fluctuation or fluctuates within a very small range during the charging and non-charging stages, the number of points where the first-order difference of the active power of the user at each time point is less than 0.1 is calculated;
[0078] S3.2.5. Considering that the active power of the charging pile user changes suddenly during the charging and non-charging phases, the number of points where the first-order difference of the active power of the user at each time point is greater than 0.8 times the maximum value is calculated;
[0079] S3.2.6. Considering that street lamps and charging piles have very similar curves, and the street lamps are turned on and off at fixed times every day, calculate the correlation coefficient between the user's active power at the same time point on the day and the active power at the same time point on the previous day.
[0080] S3.3. Label the user as a charging pile based on the ledger data to form a sample set, and randomly divide the sample set into a 70% training sample set and a 30% test sample set.
[0081] S4. Model construction: Select the support vector machine model in the Scikit-learn library, set the initial hyperparameter range of the support vector machine model according to the sample information, and build a basic support vector machine model.
[0082] In this embodiment, the basic support vector machine model is constructed as follows:
[0083] Select the support vector machine model in the Scikit-learn library, set the regularization coefficient C to a value range of 0.5 to 10, select the kernel function of the model from the linear kernel function, polynomial kernel function and Gaussian kernel function, use the grid search method during training, adjust the above hyperparameters, and build a basic support vector machine model.
[0084] S5. Model training and testing: Import the feature data of the training sample set into the support vector machine model. In the process of training the basic support vector machine model, use the grid search method to adjust the hyperparameters within a range, and select the model that achieves the best prediction accuracy on the training sample set. Finally, obtain a low-voltage charging pile user recognition model based on the support vector machine model; use the test sample set to test the low-voltage charging pile user recognition model, and use the precision, recall rate and F1 score to evaluate the performance of the low-voltage charging pile user recognition model.
[0085] S6. Low-voltage charging pile user identification: obtain the full amount of non-charging pile user power data, use the low-voltage charging pile user identification model to identify, and finally obtain a list of users identified as charging piles.
[0086] The low-voltage charging pile identification method based on the support vector machine model of the present invention is further introduced below through specific embodiments.
[0087] like Figure 1 As shown, a low-voltage charging pile identification method based on a support vector machine model provided by an embodiment of the present invention specifically includes the following steps:
[0088] S1. Data acquisition: Obtain the records and active power data of all charging pile users and non-charging pile users. According to the electricity consumption characteristics of charging pile users and non-charging pile users, preliminarily eliminate some users with questionable labels in the records, and randomly obtain the records and active power data of 10,000 groups of charging pile users and 10,000 groups of non-charging pile users with correct labels.
[0089] S2. Data preprocessing: Preprocess the acquired data, use the isnull function to identify missing values, use the box plot method to identify outliers, and use Newton interpolation method to fill in the identified outliers and missing values.
[0090] S3. Feature extraction: Extract 4-dimensional statistical features and 6-dimensional business features from the pre-processed 96-dimensional time-sharing active power data of the user on the same day, including: four statistical dimension features of the user's time-sharing active power data, namely, the kurtosis, skewness, maximum value, and median value; six business dimension features of the user's time-sharing active power less than 1.2 times the minimum value, the number of points where the user's time-sharing active power is greater than 0.8 times the maximum value, the number of points where the user's time-sharing active power is greater than 1.5 times the median or less than 0.5 times the median, the number of points where the user's time-sharing active power first-order difference is less than 0.1, the number of points where the user's time-sharing active power first-order difference is greater than 0.8 times the maximum value, and the correlation coefficient between the user's time-sharing active power on the same day and the time-sharing active power of the previous day. Combined with whether the user is a charging pile label, a sample set is formed, and the 20,000 sample sets are randomly divided into 70% training sample sets and 30% test sample sets.
[0091] S4. Model construction: First, select the support vector machine model in the Scikit-learn library, set the value range of the regularization coefficient C to 0.5-10, select the kernel function of the model from the linear kernel function, polynomial kernel function and Gaussian kernel function, and use the grid search method during training to adjust the above hyperparameters to build a basic support vector machine model.
[0092] S5. Model training and testing: Import the feature data of the training sample set into the support vector machine model. In the process of training the basic support vector machine model, use the grid search method to adjust the hyperparameters within the range, select the model that achieves the best prediction accuracy on the training sample set, and finally select the support vector machine with a regularization parameter of 1.5 and a linear kernel function as the initial low-voltage charging pile user identification model. Use the test sample set to test the charging pile identification model, and use the precision, recall and F1 score to evaluate the model performance. Figure 2 As shown in the figure, the final model's recognition results on the test set of 3,000 groups of charging pile users and 3,000 groups of non-charging users are as follows: 2,967 of the 3,000 charging pile users were identified as charging piles, and 33 were identified as non-charging piles; 24 of the 3,000 non-charging pile users were identified as charging piles, and 2,976 were identified as non-charging piles; the model's precision on the test set reached 99.05%, the recall rate reached 99.05%, and the average F1 score was 99.05%. The model's final prediction results on the test set were excellent, and the final output was a low-voltage charging pile user recognition model based on the support vector machine model.
[0093] S6. Identification of low-voltage charging pile users: Obtain the untrained power data of the remaining 88,550 non-charging pile users in a certain city on a certain day, use the trained charging pile identification model to identify them, and finally obtain a list of 157 users identified as charging piles on that day.
[0094] Example 2
[0095] like Figure 3 As shown, a low-voltage charging pile identification system based on a support vector machine model of the present invention includes:
[0096] The measurement data acquisition module is used to obtain the user's account data and time-sharing active power measurement data from the power consumption information collection system;
[0097] The measurement data preprocessing module is used to preprocess the acquired data, identify missing values using the isnull function, identify outliers using the box plot method, and fill in the identified outliers and missing values using the Newton interpolation method;
[0098] The feature extraction module is used to extract statistical dimension and business dimension data features from the pre-processed active power measurement data, label the user as a charging pile according to the ledger data, form a sample set, and randomly divide the sample set into a training sample set and a test sample set;
[0099] The model building module is used to select the support vector machine model in the Scikit-learn library, set the initial hyperparameter range of the support vector machine model according to the sample information, and build a basic support vector machine model;
[0100] The model training and testing module is used to import the feature data of the training sample set into the support vector machine model. In the process of training the basic support vector machine model, the grid search method is used to adjust the hyperparameters within a range, and the model is selected to achieve the best prediction accuracy on the training sample set, and finally a low-voltage charging pile user recognition model based on the support vector machine model is obtained; it is also used to test the low-voltage charging pile user recognition model using the test sample set, and to evaluate the performance of the low-voltage charging pile user recognition model using the precision, recall rate and F1 score;
[0101] The low-voltage charging pile user identification module is used to obtain the full amount of non-charging pile user power data, use the low-voltage charging pile user identification model to identify, and finally obtain a list of users identified as charging piles.
[0102] In an optional embodiment, a low-voltage charging pile identification method based on a support vector machine model includes: a) obtaining a low-voltage user ledger and low-voltage user time-sharing power data; b) preprocessing the data to eliminate noise and outliers; c) extracting statistical dimension and business dimension features from the preprocessed data, marking whether the data set is a charging pile label, and dividing the data set into a training set and a test set; d) constructing a support vector machine model; e) training and testing the model; f) using a tested low-voltage charging pile identification model to identify users whose ledgers are not charging piles, and finally obtain a list of users who are actually charging piles.
[0103] Example 3
[0104] like Figure 4 As shown, the low-voltage charging pile identification device based on the support vector machine model of the present invention may include a processor and a memory, the memory stores computer instructions, the processor is used to execute the computer instructions stored in the memory, and when the computer instructions are executed by the processor, the electronic device implements the steps of the method described in the above embodiment and can achieve the technical effect consistent with the above method.
[0105] The memory may include computer system readable media in the form of volatile memory, such as random access memory (RAM) and / or cache memory. The device may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, the memory may be used to read and write non-removable, non-volatile magnetic media (commonly referred to as a "hard drive"). A program / utility having a set (at least one) of program modules may be stored in, for example, the memory, such program modules including but not limited to an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include the implementation of a network environment. The program modules typically perform the functions and / or methods in the embodiments described herein.
[0106] The processor executes various functional applications and data processing by running the program stored in the memory, such as implementing the method provided in the first embodiment of the present invention.
[0107] Example 4
[0108] Embodiment 4 of the present invention further provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the steps of the method described in the above embodiment are implemented and the technical effect consistent with the above method can be achieved.
[0109] The computer storage medium of the embodiment of the present invention can adopt any combination of one or more computer-readable media. The computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium can be, for example, but not limited to, a system, device or device of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination of the above. More specific examples (non-exhaustive list) of computer-readable storage media include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, a computer-readable storage medium can be any tangible medium containing or storing a program, which can be used by an instruction execution system, a device or a device or used in combination with it.
[0110] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, which carry computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.
[0111] The program code embodied on the computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.
[0112] Computer program code for performing the operation of the present invention may be written in one or more programming languages or combinations thereof, including object-oriented programming languages such as Java, Smalltalk, C++, and conventional procedural programming languages such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer through any type of network including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).
[0113] Of course, the storage medium containing computer executable instructions provided by an embodiment of the present invention, whose computer executable instructions are not limited to the above method operations, can also execute related operations in the method provided by any embodiment of the present invention.
[0114] The specific embodiments described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.
Claims
1. A low-voltage charging pile identification method based on a support vector machine model, characterized in that: include: Measurement data acquisition: obtain the user's ledger data and active power measurement data at different time points from the electricity consumption information collection system; Measurement data preprocessing: preprocess the acquired data, use the isnull function to identify missing values, use the box plot method to identify outliers, and use Newton interpolation to fill in the identified outliers and missing values; Feature extraction: Extract statistical dimension and business dimension data features from the pre-processed active power measurement data, label the user as a charging pile based on the ledger data, form a sample set, and randomly divide the sample set into a training sample set and a test sample set; Model construction: Select the support vector machine model in the Scikit-learn library, set the initial hyperparameter range of the support vector machine model according to the sample information, and build a basic support vector machine model; Model training and testing: Import the feature data of the training sample set into the support vector machine model. In the process of training the basic support vector machine model, use the grid search method to adjust the hyperparameters within the range, select the model that achieves the best prediction accuracy on the training sample set, and finally obtain a low-voltage charging pile user recognition model based on the support vector machine model; use the test sample set to test the low-voltage charging pile user recognition model, and use the precision, recall rate and F1 score to evaluate the performance of the low-voltage charging pile user recognition model; Low-voltage charging pile user identification: Obtain the full amount of non-charging pile user power data, use the low-voltage charging pile user identification model to identify, and finally obtain a list of users identified as charging piles.
2. The low-voltage charging pile identification method based on the support vector machine model according to claim 1 is characterized in that: Identify outliers using box plot methods, including: First, sort the acquired data, determine the first quartile Q1 and the third quartile Q3, and then calculate the interquartile range IQR, the formula is as follows: IQR=Q3-Q1 According to the interquartile range, values less than Q1-1.5×IQR or greater than Q3+1.5×IQR were identified as outliers.
3. The low-voltage charging pile identification method based on the support vector machine model according to claim 1 is characterized in that: The expression of the Newton interpolation method is as follows: P(x) represents the polynomial function obtained by Newton interpolation method, which is used to estimate the interpolation result at the given position x value of the measured data point; p0 represents the initial power value point; x j Indicates the position of the jth power value point; Δ k p0 represents the k-order difference quotient of p0, and the difference quotient is defined as follows: D k p i =D k-1 p i+1 -D k-1 p i Δ 1 p i =p i+1 -p i Among them, p i Represents the power value of the i-th position.
4. The low-voltage charging pile identification method based on the support vector machine model according to claim 1 is characterized in that: The statistical dimension data features include the kurtosis, skewness, maximum value and median value of the user's active power measurement data at different time points.
5. The low-voltage charging pile identification method based on the support vector machine model according to claim 4 is characterized in that: The kurtosis of the user's time-sharing active power measurement data is used to measure the steepness of the probability distribution of the random variable, and its expression is as follows: The skewness of the user's time-sharing active power measurement data is used to measure the asymmetry of the probability distribution of random variables, and its expression is as follows: Among them, S represents kurtosis; K represents skewness; X i It represents the specific value in the original data; μ represents the mean or average of the original data; σ represents the standard deviation of the original data.
6. The low-voltage charging pile identification method based on the support vector machine model according to claim 1 is characterized in that: The business dimension data features include the number of points where the user's time-sharing point active power is less than 1.2 times the minimum value, the number of points where the user's time-sharing point active power is greater than 0.8 times the maximum value, the number of points where the user's time-sharing point active power is greater than 1.5 times the median or less than 0.5 times the median, the number of points where the user's time-sharing point first-order difference of active power is less than 0.1, the number of points where the user's time-sharing point first-order difference of active power is greater than 0.8 times the maximum value, and the correlation coefficient between the user's time-sharing point active power on the day and the time-sharing point active power on the previous day.
7. The low-voltage charging pile identification method based on the support vector machine model according to claim 1 is characterized in that: The construction process of the basic support vector machine model is as follows: Select the support vector machine model in the Scikit-learn library, set the regularization coefficient C to a value range of 0.5 to 10, select the kernel function of the model from the linear kernel function, polynomial kernel function and Gaussian kernel function, use the grid search method during training, adjust the above hyperparameters, and build a basic support vector machine model.
8. A low-voltage charging pile identification system based on a support vector machine model, characterized in that: include: The measurement data acquisition module is used to obtain the user's ledger data and time-sharing active power measurement data from the power consumption information collection system; The measurement data preprocessing module is used to preprocess the acquired data, identify missing values using the isnull function, identify outliers using the box plot method, and fill in the identified outliers and missing values using the Newton interpolation method; The feature extraction module is used to extract statistical dimension and business dimension data features from the pre-processed active power measurement data, label the user as a charging pile according to the ledger data, form a sample set, and randomly divide the sample set into a training sample set and a test sample set; The model building module is used to select the support vector machine model in the Scikit-learn library, set the initial hyperparameter range of the support vector machine model according to the sample information, and build a basic support vector machine model; The model training and testing module is used to import the feature data of the training sample set into the support vector machine model. In the process of training the basic support vector machine model, the grid search method is used to adjust the hyperparameters within a range, and the model is selected to achieve the best prediction accuracy on the training sample set, and finally a low-voltage charging pile user recognition model based on the support vector machine model is obtained; it is also used to test the low-voltage charging pile user recognition model using the test sample set, and to evaluate the performance of the low-voltage charging pile user recognition model using the precision, recall rate and F1 score; The low-voltage charging pile user identification module is used to obtain the full amount of non-charging pile user power data, use the low-voltage charging pile user identification model to identify, and finally obtain a list of users identified as charging piles.
9. A low-voltage charging pile identification device based on a support vector machine model, characterized in that: The electronic device comprises a processor and a memory, wherein the memory stores computer instructions, and the processor is used to execute the computer instructions stored in the memory. When the computer instructions are executed by the processor, the electronic device implements the steps of the low-voltage charging pile identification method based on the support vector machine model as described in any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the low-voltage charging pile identification method based on a support vector machine model as described in any one of claims 1 to 7.
