Electric vehicle charging demand characteristic probability quantification method based on multivariate Gaussian mixture distribution

Abstract

The invention discloses an electric vehicle charging demand characteristic probability quantification method based on multivariate Gaussian mixture distribution, and relates to the field of vehicles, and the method comprises the steps: obtaining platform original data; preprocessing the original data of the platform to obtain charging demand characteristics of the electric vehicle; based on the initial model parameters and the charging demand characteristics of the electric vehicle, optimizing the initial model parameters by using an expectation maximization algorithm to obtain optimized model parameters; the initial model parameters are obtained through a random generation method; the model parameters comprise a weight, a mean value, a variance and a skewness coefficient; constructing a Gaussian mixture model based on the optimized model parameters; and quantifying the charging demand characteristics of the electric vehicle through a Gaussian mixture model. According to the invention, the accuracy and effectiveness of electric vehicle charging demand feature quantification are improved.

Description

Technical Field

[0001] The present application relates to the field of vehicles, and in particular to a method for probabilistic quantification of electric vehicle charging demand characteristics based on multivariate Gaussian mixture distribution. Background Art

[0002] In recent years, the number of electric vehicles in China has increased dramatically. This trend is accompanied by a significant increase in charging demand, which in turn has led to a substantial increase in charging load. This poses a challenge to the safe and stable operation of the power system and increases the risk of safe and stable operation of the distribution network. However, as a mobile energy storage carrier, electric vehicles have a high degree of flexibility in their charging and discharging capabilities. They can be used as both energy consumption and energy storage units, showing great potential as flexible loads in the power system. In addition, the disordered charging behavior of electric vehicles is highly random and uncertain, which is completely different from the behavior pattern of traditional internal combustion engine vehicles. In view of the above situation, in-depth analysis and quantification of the characteristics of electric vehicle charging demand are crucial for the reasonable site selection and capacity planning of charging infrastructure, the optimization of operation strategies, and the overall optimization of power system operation.

[0003] At present, traditional electric vehicle charging demand analysis methods mostly rely on underlying modeling, which is to make predictions by building underlying models based on the characteristics of different users. However, due to the significant differences in behavioral patterns and the variety of characteristics among users, this modeling process becomes extremely complicated. The current data-driven methods are mainly based on offline questionnaire survey data of car owners, actual operation data of fuel vehicles, and small-scale electric vehicle operation data. In the early stages of electric vehicle deployment, these data did provide valuable parameter inputs for charging demand feature modeling. However, with the explosive growth in the number of electric vehicles and the increasing diversification of user needs, existing methods have been unable to ensure the accuracy and effectiveness of the quantification of charging demand characteristics. Summary of the invention

[0004] The purpose of this application is to provide a probability quantification method for electric vehicle charging demand characteristics based on multivariate Gaussian mixture distribution, which can improve the accuracy and effectiveness of quantification of electric vehicle charging demand characteristics.

[0005] To achieve the above objectives, this application provides the following solutions:

[0006] In a first aspect, the present application provides a method for probabilistically quantifying characteristics of electric vehicle charging demand based on multivariate Gaussian mixture distribution, comprising:

[0007] Get the original data of the platform;

[0008] Preprocess the original data of the platform to obtain the characteristics of electric vehicle charging demand;

[0009] Based on the initial model parameters and the electric vehicle charging demand characteristics, the initial model parameters are iteratively optimized using the Expectation-Maximization algorithm to obtain the optimized model parameters; the initial model parameters are obtained by the random generation method; the model parameters include: weights, means, variances, and skewness coefficients;

[0010] Based on the optimized model parameters, a Gaussian mixture model is constructed;

[0011] The electric vehicle charging demand characteristics are quantified through the Gaussian mixture model.

[0012] Optionally, the electric vehicle charging demand characteristics include: vehicle charging start time, vehicle charging start SOC, vehicle charging end time, vehicle charging end SOC, vehicle battery capacity, and vehicle charging power.

[0013] Optionally, preprocessing the platform raw data specifically includes:

[0014] Cleaning and reconstructing the platform raw data.

[0015] Optionally, based on the initial model parameters and the electric vehicle charging demand characteristics, using the Expectation-Maximization algorithm to iteratively optimize the initial model parameters to obtain the optimized model parameters, specifically including:

[0016] According to the initial model parameters, calculate the expected value of the conditional probability distribution of the electric vehicle charging demand characteristics with respect to the latent variable;

[0017] According to the expected value, estimate the initial model parameters to obtain the estimated model parameters;

[0018] Calculate the maximum likelihood estimate value of the estimated model parameters;

[0019] Judge whether the maximum likelihood estimate value is less than the preset value or reaches the maximum number of iterations;

[0020] If so, determine the estimated model parameters as the optimized model parameters;

[0021] If not, according to the estimated model parameters, calculate the expected value of the conditional probability distribution of the electric vehicle charging demand characteristics with respect to the latent variable, and recalculate the maximum likelihood estimate value.

[0022] Optionally, the calculation formula for the expected value of the conditional probability distribution of the electric vehicle charging demand characteristics with respect to the latent variable is:

[0023] Q i (z i )=p(z i |x i ,θ k));

[0024] Among them, Q i (z i ) is the expected value of the conditional probability distribution of the electric vehicle charging demand characteristics with respect to the latent variable; p(g|g) is the posterior probability of the sub-distribution function; x i is the electric vehicle charging demand characteristic at the i-th moment; z i is the latent variable corresponding to x i ; θ k is the model parameter of the (k - 1)-th iteration.

[0025] Optionally, the calculation formula of the latent variable is:

[0026]

[0027] Among them, m is the number of electric vehicle charging demand characteristics.

[0028] Optionally, the calculation formula for estimating the initial model parameter is:

[0029]

[0030] Among them, l(g) is the weighted complete data log-likelihood function; p(g; g) is the sub-distribution function.

[0031] Optionally, the calculation formula of the estimated model parameter is:

[0032] θ k+1 = argmax l(z i , θ k ), k > 0, k ∈ Z;

[0033] Among them, θ k+1 is the model parameter of the k-th iteration.

[0034] Optionally, the expression of the Gaussian mixture model is:

[0035]

[0036] Among them, f(x) is the probability corresponding to the electric vehicle charging demand characteristic; x is the electric vehicle charging demand characteristic; n is the number of Gaussian distributions; A j is the weight occupied by the j-th Gaussian distribution; μ j is the mean of the j-th Gaussian distribution; σ j is the variance of the j-th Gaussian distribution; b j is the skewness coefficient of the j-th Gaussian distribution.

[0037] According to the specific embodiments provided by the present application, the present application has the following technical effects:

[0038] The present application provides a method for probabilistically quantifying the charging demand characteristics of electric vehicles based on a multivariate Gaussian mixture distribution. Based on the initial model parameters and the charging demand characteristics of the electric vehicles, the maximum expectation algorithm is used to optimize the initial model parameters to obtain optimized model parameters, thereby constructing a Gaussian mixture model. Since the model parameters include skewness coefficients, the constructed Gaussian mixture model can more accurately fit the asymmetric peak characteristics, can more accurately fit the charging demand characteristics of different types of electric vehicles, does not need to make assumptions based on existing rules, can effectively reduce the deviation of the fitting results, and improve the accuracy and effectiveness of the quantification of the charging demand characteristics of electric vehicles. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following will briefly introduce the drawings required to be used in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0040] Figure 1 It is an application environment diagram of a method for probabilistically quantifying the charging demand characteristics of electric vehicles based on a multivariate Gaussian mixture distribution in an embodiment of the present application;

[0041] Figure 2 It is a schematic flowchart of a method for probabilistically quantifying the charging demand characteristics of electric vehicles based on a multivariate Gaussian mixture distribution provided in an embodiment of the present application;

[0042] Figure 3 For Figure 1 It is a detailed flowchart of step 3 in

[0043] Figure 4 It is a histogram of the vehicle charging start time provided in an embodiment of the present application;

[0044] Figure 5 It is a schematic flowchart of the maximum expectation algorithm provided in an embodiment of the present application;

[0045] Figure 6 It is a schematic diagram of the fitting curve of the Gaussian mixture model of the present application for the vehicle charging start time and the fitting curve of the conventional Gaussian mixture model for the vehicle charging start time;

[0046] Figure 7 It is a schematic structural diagram of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0047] The following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, 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 application without creative efforts shall fall within the scope of protection of the present application.

[0048] At present, due to the large differences and many characteristics in the charging demand laws of electric vehicles, the following drawbacks will occur in traditional methods: (1) Traditional methods make assumptions based on existing laws and have poor adaptability. As electric vehicles are continuously promoted, their behavior laws are also constantly changing and becoming more diverse, resulting in the inability of traditional methods to accurately describe the charging demand laws of electric vehicles. Taking the most typical time distribution of electric vehicles connecting to the power grid as an example, previous studies usually assumed that it follows a single Gaussian distribution. This assumption is only applicable to fitting and modeling parameters with a single-peak curve distribution of characteristics, while the actual time when electric vehicles connect to the power grid often shows a multi-peak distribution, which leads to large deviations in the fitting by traditional methods; (2) Traditional methods often analyze and model based on small-scale sample data, which will result in certain limitations in the obtained results and may have a large deviation from the overall characteristics, thus making it difficult to ensure the accuracy and effectiveness of the quantification of charging demand characteristics.

[0049] To solve the above problems, the present application proposes a method for probabilistic quantification of electric vehicle charging demand characteristics based on a multivariate Gaussian mixture distribution, which improves the adaptability when fitting and extracting variable characteristics, avoids the deviation caused by relying on assumptions about existing laws, and makes the obtained quantification results of charging demand characteristics more representative by analyzing a larger range of sample data, thereby improving the accuracy and effectiveness of characteristic quantification.

[0050] To make the above objects, features, and advantages of the present application more obvious and understandable, the following further details the present application with reference to the accompanying drawings and specific embodiments.

[0051] A method for probabilistic quantification of electric vehicle charging demand characteristics provided by an embodiment of the present application can be applied to, for example Figure 1In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be set separately, integrated on the server 104, placed on the cloud or other servers. The terminal 102 can send the platform original data to the server 104. After the server 104 receives the platform original data, for the platform original data, the server 104 preprocesses the platform original data to obtain the electric vehicle charging demand characteristics; then, based on the initial model parameters and the electric vehicle charging demand characteristics, using the expectation-maximization algorithm, the initial model parameters are optimized to obtain the optimized model parameters; finally, based on the optimized model parameters, a Gaussian mixture model is constructed, so as to quantify the electric vehicle charging demand characteristics through the Gaussian mixture model. The server 104 can feedback the quantization result of the electric vehicle charging demand characteristics to the terminal 102. In addition, in some embodiments, the method for quantifying the probability of electric vehicle charging demand characteristics based on the multivariate Gaussian mixture distribution can also be implemented by the server 104 or the terminal 102 alone. For example, the terminal 102 can directly perform the quantization process of the electric vehicle charging demand characteristics on the platform original data, or the server 104 can obtain the platform original data from the data storage system and perform the quantization process of the electric vehicle charging demand characteristics on the platform original data.

[0052] Among them, the terminal 102 can be, but is not limited to, various desktop computers, laptop computers, smart phones, tablet computers, Internet of Things devices and portable wearable devices. The Internet of Things devices can be smart speakers, smart TVs, smart air conditioners, smart in-vehicle devices, etc. The portable wearable devices can be smart watches, smart bracelets, head-mounted devices, etc. The server 104 can be implemented by an independent server or a server cluster composed of multiple servers, and can also be a cloud server.

[0053] In an exemplary embodiment, as Figure 2 shown, a method for quantifying the probability of electric vehicle charging demand characteristics based on the multivariate Gaussian mixture distribution is provided. The method includes the following steps:

[0054] Step 1: Obtain the platform original data.

[0055] Step 2: Preprocess the platform original data to obtain the electric vehicle charging demand characteristics.

[0056] Step 3: Based on the initial model parameters and the electric vehicle charging demand characteristics, use the expectation-maximization algorithm to iteratively optimize the initial model parameters to obtain the optimized model parameters; the initial model parameters are obtained by the random generation method; the model parameters include: weight, mean, variance and skewness coefficient.

[0057] Step 4: Construct a Gaussian mixture model based on the optimized model parameters.

[0058] Step 5: Quantify the electric vehicle charging demand characteristics through the Gaussian mixture model.

[0059] In a specific embodiment, the platform original data obtained in Step 1 is shown in Table 1.

[0060] Table 1 Platform Original Data Items and Their Meanings

[0061]

[0062]

[0063] In a specific embodiment, Step 2 specifically includes:

[0064] After preprocessing the platform original data through cleaning, reconstruction, etc., the electric vehicle charging demand characteristics obtained include: vehicle charging start time, vehicle charging start SOC (State of Charge), vehicle charging end time, vehicle charging end SOC, vehicle battery capacity, and vehicle charging power.

[0065] In a specific embodiment, as Figure 3 shown, Step 3 specifically includes:

[0066] Step 31: Calculate the expected value of the conditional probability distribution of the electric vehicle charging demand characteristics with respect to the latent variable according to the initial model parameters.

[0067] Step 32: Estimate the initial model parameters according to the expected value to obtain the estimated model parameters.

[0068] Step 33: Calculate the maximum likelihood estimate value of the estimated model parameters.

[0069] Step 34: Determine whether the maximum likelihood estimate value is less than the preset value or reaches the maximum number of iterations.

[0070] Step 35: If so, determine the estimated model parameters as the optimized model parameters.

[0071] Step 36: If not, calculate the expected value of the conditional probability distribution of the electric vehicle charging demand characteristics with respect to the latent variable according to the estimated model parameters, and recalculate the maximum likelihood estimate value.

[0072] Specifically, the calculation formula for the expected value of the conditional probability distribution of the electric vehicle charging demand characteristics with respect to the latent variable is:

[0073] Q i (z i )=p(zi |x i , θ k )(1)

[0074] Among them, Q i (z i ) is the expected value of the conditional probability distribution of the electric vehicle charging demand characteristics with respect to the latent variable; p(g|g) is the posterior probability of the sub-distribution function; x i is the electric vehicle charging demand characteristic at the i-th moment; z i is the latent variable corresponding to x i ; θ k is the model parameter of the (k - 1)-th iteration.

[0075] Specifically, the calculation formula of the latent variable is:

[0076]

[0077] Among them, m is the number of electric vehicle charging demand characteristics.

[0078] Specifically, the calculation formula for estimating the initial model parameter is:

[0079]

[0080] Among them, l(g) is the weighted complete data log-likelihood function; p(g; g) is the sub-distribution function.

[0081] Specifically, the calculation formula of the estimated model parameter is:

[0082] θ k+1 = argmax l(z i , θ k ), k > 0, k ∈ Z(4)

[0083] Among them, θ k+1 is the model parameter of the k-th iteration.

[0084] Next, taking the vehicle charging start time as an example, using the Expectation-Maximization algorithm, the initial model parameter is optimized to obtain the optimized model parameter. Set the vehicle charging start time y = (y1, y2, y3... y t ), where y t is the vehicle charging start time at the t-th moment, and take the vehicle charging start time y as the total sample data to be fitted, Figure 4 is the histogram of the vehicle charging start time, Figure 4 in which the abscissa y t is the vehicle charging start time at the t-th moment, and the ordinate is the probability density corresponding to y t .

[0085] As Figure 5 shown, Figure 5 in Figure 5 , Y represents yes and N represents no. First, the initial model parameters θ0 = (A0, μ0, σ0, b0) are obtained by the random generation method, where A0 is the initial weight, μ0 is the initial mean, σ0 is the initial variance, and b0 is the initial skewness coefficient. Then, the following E-step is performed on the initial model parameters: According to the current model parameters, calculate the expected value of the conditional probability distribution of the vehicle charging start time at each moment with respect to the latent variable, and use it as the current estimated value of the latent variable z. The specific formula is:

[0086] Q t (z t ) = p(z t |y t , θ k )(5)

[0087]

[0088] where Q t (z t ) is the expected value of the conditional probability distribution of the vehicle charging start time with respect to the latent variable; p(g|g) is the posterior probability of the sub-distribution function; y t is the vehicle charging start time at the t-th moment; z t is the latent variable corresponding to y t ; T is the number of vehicle charging start times; θ k is the model parameter at the (k - 1)-th iteration.

[0089] It can be understood that Q t (z t ) is the probability that the vehicle charging start time yt follows the sub-distribution zt. In equation (5), the probability expectations of the sampled y t with respect to each sub-distribution z are calculated respectively, and the sum of the probability expectations of y t with respect to each sub-distribution z is obtained. The sum is used as the current estimated value of the latent variable z. The intention of this step is: Under the current model parameters, map the sample data to different sub-distributions z to make the fitting degree between the model and the samples the best.

[0090] Then, the following M-step is performed on the expected value: According to the expected value calculated in the E-step, perform a maximum likelihood estimation on the initial model parameters to obtain the estimated model parameters θ1 = (A1, μ1, σ1, b1). The calculation formula is:

[0091]

[0092] θ k+1 = argmaxl(z t , θ k), k > 0, k ∈ Z(8)

[0093] Among them, l(g) is the weighted complete data log-likelihood function; p(g; g) is the sub-distribution function, and each sampled sample corresponds to each sub-distribution, that is, the sampled sample will follow a certain sub-distribution in the Gaussian mixture model, representing y t follows the sub-distribution z t , z t The various model parameters corresponding to the sub-distribution are θ; θ k+1 is the model parameter of the k-th iteration.

[0094] Then, calculate the maximum likelihood estimate value of the estimated model parameter, and the calculation formula is:

[0095]

[0096] Among them, is the maximum likelihood estimate value of the estimated model parameter; T is the number of vehicle charging start times; Q t (z t ) is the posterior probability; y t is the vehicle charging start time at the t-th moment; z t is the latent variable corresponding to y t ; θ k+1 is the model parameter corresponding to z t ; p(g; g) is the sub-distribution function. The intention of this step is: under the probability expectations of the current various sub-distributions z, obtain the best model parameter θ of each sub-distribution to make the fitting degree of the model and the sample the best.

[0097] After that, according to the maximum likelihood estimate value of the estimated model parameter, obtain the optimized model parameter θ k+1 , and it is necessary to maximize the likelihood function of the optimized model parameter θ k+1 , and at this time, the θ k+1 corresponding to the maximum value of the likelihood function is the optimized model parameter. If it is judged has converged (reached the maximum number of iterations or the maximum likelihood estimate value is less than the preset value), then determine the estimated model parameter as the optimized model parameter; otherwise, return to execute step 31 and use θ k+1 as the initial model parameter to recalculate the model parameter.

[0098] Finally, through two steps of continuous iteration of the z parameter and the θ k+1 parameter, gradually approach the optimal solution so that the fitting curve of the Gaussian mixture model for the vehicle charging start time and Figure 4has the best goodness of fit (generally speaking: first fix θ to find the current optimal z, then inversely find the current optimal θ based on the obtained current z. After completion, return for iteration. Again, optimize z based on the new θ, and inversely optimize θ based on the optimized z, continuously iterating to approach the optimal solution).

[0099] In an exemplary embodiment, the expression of the Gaussian mixture model in step 4 is as follows:

[0100]

[0101] where f(x) is the probability corresponding to the electric vehicle charging demand characteristics; x is the electric vehicle charging demand characteristics; n is the number of Gaussian distributions; A j is the weight occupied by the j-th Gaussian distribution; μ j is the mean of the j-th Gaussian distribution; σ j is the variance of the j-th Gaussian distribution; b j is the skewness coefficient of the j-th Gaussian distribution.

[0102] Figure 6 In (a) of Figure 6 is the fitting curve of the Gaussian mixture model of the present application for the vehicle charging start time, Figure 6 In (b) of Figure 6 is the fitting curve of the conventional Gaussian mixture model for the vehicle charging start time. It can be seen from this that the present application can more accurately fit the asymmetric peak characteristics. This is because the present application adds a fourth term (skewness coefficient, which determines the skewness degree and direction of the i-th Gaussian distribution) on the basis of the conventional Gaussian mixture model. Since the Gaussian distribution is symmetric, the conventional Gaussian mixture model often has a large deviation when fitting the asymmetric peak characteristics, while the improved Gaussian mixture model of the present application can more accurately fit the asymmetric peak characteristics after adding the skewness coefficient.

[0103] A method for quantifying electric vehicle charging demand characteristics based on a multivariate Gaussian mixture distribution proposed by the present application uses the expectation-maximization algorithm, takes the initial model parameters and electric vehicle charging demand characteristics as inputs, optimizes the initial model parameters, obtains the optimized model parameters, and constructs a Gaussian mixture model through the optimized weights, means, variances, and skewness coefficients, so that the constructed Gaussian mixture model can fit the fitting curves conforming to different types of characteristics without making assumptions according to existing rules, enabling different types of characteristics to be quantified. At the same time, the introduced skewness coefficient can effectively reduce the deviation of the fitting results of the asymmetric peak characteristics, thereby improving the accuracy and effectiveness of characteristic quantification.

[0104] Based on the same inventive concept, an embodiment of the present application further provides a system for implementing the above-mentioned probability quantization system of electric vehicle charging demand characteristics based on multivariate Gaussian mixture distribution. The implementation solutions provided by this system for solving problems are similar to those described in the above method. Therefore, the specific limitations in one or more embodiments of the probability quantization system of electric vehicle charging demand characteristics based on multivariate Gaussian mixture distribution provided below can refer to the limitations on the probability quantization method of electric vehicle charging demand characteristics based on multivariate Gaussian mixture distribution in the above text, and will not be repeated here.

[0105] In an exemplary embodiment, a system for quantifying the probability of electric vehicle charging demand characteristics based on multivariate Gaussian mixture distribution is provided, including:

[0106] An acquisition module, configured to acquire platform original data.

[0107] A preprocessing module, configured to preprocess the platform original data to obtain electric vehicle charging demand characteristics.

[0108] An iteration module, configured to optimize the initial model parameters based on the initial model parameters and the electric vehicle charging demand characteristics by using the expectation-maximization algorithm to obtain optimized model parameters; the initial model parameters are obtained by the random generation method; the model parameters include: weights, means, variances, and skewness coefficients.

[0109] A model construction module, configured to construct a Gaussian mixture model based on the optimized model parameters.

[0110] A quantization module, configured to quantify the electric vehicle charging demand characteristics through the Gaussian mixture model.

[0111] In an exemplary embodiment, a computer device is provided. The computer device can be a server or a terminal, and its internal structure diagram can be as Figure 7As shown in the figure. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O), and a communication interface. Among them, the processor, the memory, and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store platform raw data. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with external terminals through a network connection. When the computer program is executed by the processor, it implements a method for quantifying the probability of electric vehicle charging demand characteristics based on a multivariate Gaussian mixture distribution.

[0112] Those skilled in the art can understand that Figure 7 the structure shown in the figure is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine some components, or have different component arrangements. In an exemplary embodiment, a computer device is provided, including a memory and a processor. A computer program is stored in the memory, and when the processor executes the computer program, the steps in the above method embodiments are implemented.

[0113] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by the processor, the steps in the above method embodiments are implemented.

[0114] In an exemplary embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by the processor, the steps in the above method embodiments are implemented.

[0115] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use, and processing of relevant data need to comply with relevant regulations.

[0116] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, database, or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memories. Non-volatile memories can include read-only memory (ROM), magnetic tapes, floppy disks, flash memories, optical memories, high-density embedded non-volatile memories, resistive random access memories (ReRAMs), magnetoresistive random access memories (MRAMs), ferroelectric random access memories (FRAMs), phase change memories (PCMs), graphene memories, etc. Volatile memories can include random access memories (RAMs) or external caches, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0117] The databases involved in the embodiments provided in the present application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., without limitation. The processors involved in the embodiments provided in the present application can be general-purpose processors, central processors, graphics processors, digital signal processors, programmable logics, data processing logics based on quantum computing, etc., without limitation.

[0118] The technical features of the above embodiments can be combined arbitrarily. For the sake of concise description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope described in this specification.

[0119] Specific examples are used in this article to elaborate on the principles and implementation manners of the present application. The descriptions of the above embodiments are only used to help understand the method and its core idea of the present application; at the same time, for those of ordinary skill in the art, according to the idea of the present application, there will be changes in the specific implementation manners and application scopes. In summary, the content of this specification should not be construed as a limitation to the present application.

Claims

1. A method for quantifying the probability of electric vehicle charging demand characteristics based on multivariate Gaussian mixture distribution, characterized in that: The electric vehicle charging demand characteristic probability quantification method based on multivariate Gaussian mixture distribution includes: Get the original data of the platform; Preprocess the original data of the platform to obtain the characteristics of electric vehicle charging demand; Based on the initial model parameters and the characteristics of the electric vehicle charging demand, the initial model parameters are iteratively optimized using the maximum expectation algorithm to obtain optimized model parameters; the initial model parameters are obtained by a random generation method; the model parameters include: weight, mean, variance and skewness coefficient; Based on the optimized model parameters, a Gaussian mixture model is constructed; The Gaussian mixture model is used to quantify the charging demand characteristics of electric vehicles.

2. The method for quantifying the probability of electric vehicle charging demand characteristics based on multivariate Gaussian mixture distribution according to claim 1 is characterized in that: The electric vehicle charging demand characteristics include: vehicle charging start time, vehicle charging start SOC, vehicle charging end time, vehicle charging end SOC, vehicle battery capacity and vehicle charging power.

3. The method for quantifying the probability of electric vehicle charging demand characteristics based on multivariate Gaussian mixture distribution according to claim 1 is characterized in that: Preprocess the platform's raw data, including: Clean and reconstruct the platform's original data.

4. The method for quantifying the probability of electric vehicle charging demand characteristics based on multivariate Gaussian mixture distribution according to claim 1 is characterized in that: Based on the initial model parameters and the electric vehicle charging demand characteristics, the initial model parameters are iteratively optimized using the maximum expectation algorithm to obtain optimized model parameters, which specifically include: According to the initial model parameters, the expected value of the conditional probability distribution of the electric vehicle charging demand characteristics with respect to the hidden variables is calculated; According to the expected value, the initial model parameters are estimated to obtain the estimated model parameters; Calculate the maximum likelihood estimates of the estimated model parameters; Determine whether the maximum likelihood estimate is less than the preset value or reaches the maximum number of iterations; If so, the estimated model parameters are determined to be optimized model parameters; If not, then the expected value of the conditional probability distribution of the electric vehicle charging demand characteristics with respect to the latent variables is calculated based on the estimated model parameters, and the maximum likelihood estimate is recalculated.

5. The method for quantifying the probability of electric vehicle charging demand characteristics based on multivariate Gaussian mixture distribution according to claim 4 is characterized in that: The calculation formula for the expected value of the conditional probability distribution of the electric vehicle charging demand characteristics with respect to the hidden variables is: Q i (z i )=p(z i |x i ,θ k ): Among them, Q i (z i ) is the expected value of the conditional probability distribution of the electric vehicle charging demand characteristics with respect to the hidden variables; p(g|g) is the posterior probability of the sub-distribution function; x i is the charging demand characteristics of electric vehicles at the i-th moment; z i For x i The corresponding hidden variable; θ k are the model parameters for the k-1th iteration.

6. The method for quantifying the probability of electric vehicle charging demand characteristics based on multivariate Gaussian mixture distribution according to claim 5 is characterized in that: The calculation formula of the hidden variable is: Where m is the number of electric vehicle charging demand characteristics.

7. The method for quantifying the probability of electric vehicle charging demand characteristics based on multivariate Gaussian mixture distribution according to claim 6 is characterized in that: The calculation formula for estimating the initial model parameters is: Among them, l(g) is the weighted complete data log-likelihood function; p(g; g) is the subdistribution function.

8. The method for quantifying the probability of electric vehicle charging demand characteristics based on multivariate Gaussian mixture distribution according to claim 7 is characterized in that: The calculation formula of the estimated model parameters is: θ k+1 =argmaxl(from i ,θ k ),k>0,k∈Z; Among them, θ k+1 are the model parameters for the kth iteration.

9. The method for quantifying the probability of electric vehicle charging demand characteristics based on multivariate Gaussian mixture distribution according to claim 8 is characterized in that: The calculation formula of the maximum likelihood estimate is: in, is the maximum likelihood estimate.

10. The method for probabilistic quantification of electric vehicle charging demand characteristics based on multivariate Gaussian mixture distribution according to claim 1, characterized in that: The expression of Gaussian mixture model is: Where f(x) is the probability corresponding to the electric vehicle charging demand characteristics; x is the electric vehicle charging demand characteristics; n is the number of Gaussian distributions; A j is the weight of the jth Gaussian distribution; μ j is the mean of the jth Gaussian distribution; σ j is the variance of the jth Gaussian distribution; b j is the skewness coefficient of the j-th Gaussian distribution.

Images