Power grid stability evaluation method and system for large-scale electric vehicle access

Abstract

The invention discloses a power grid stability evaluation method and system for large-scale electric vehicle access, and the method comprises the steps: firstly collecting a power grid data set related to an electric vehicle in a power grid, and carrying out the classification marking, and obtaining an original power grid data training set; then training the data expansion model based on a generative adversarial network algorithm of Riemannian manifold feature diffusion to obtain an expanded power grid data training set; inputting the expanded power grid data training set into a power grid stability evaluation model for training based on an extreme learning machine algorithm of nonlinear system coupling learning; and finally, inputting a new power grid data set sample related to the electric vehicle in the power grid into the trained power grid stability evaluation model for evaluation to obtain an evaluation grade of the power grid stability. According to the method, stability evaluation modeling is carried out for the high nonlinearity and dynamic change characteristics of the electric vehicle data in the power grid, the multi-dimensional influence of the access of the electric vehicle to the power grid is comprehensively reflected, and the stability evaluation of the access of the large-scale electric vehicle to the power grid is realized.

Description

Technical Field

[0001] The present invention belongs to the field of power grid stability assessment, and in particular relates to a power grid stability assessment method and system for large-scale electric vehicle access. Background Art

[0002] With the development and popularization of electric vehicles, the number of electric vehicles connected to the power grid has increased rapidly, and more and more vehicles need to be charged frequently, resulting in drastic fluctuations in the power load in the power grid and highly dynamic power demand. The power grid not only needs to meet the growing power demand, but also needs to adapt to the randomness and concentration of electric vehicle charging behavior. Especially during peak charging hours, the power grid is prone to problems such as a sharp increase in power load, frequency fluctuations, and reduced power factor. This has brought great uncertainty and pressure to the stability of traditional power grids, directly affecting the safe operation of the power grid. Summary of the invention

[0003] The purpose of the present invention is to provide a method and system for evaluating the stability of a power grid with large-scale electric vehicles connected in order to solve the above problems in the prior art.

[0004] To achieve the above objectives, the technical solution of the present invention is as follows:

[0005] In a first aspect, the present invention proposes a method for evaluating the stability of a power grid with large-scale electric vehicle access, comprising:

[0006] S1. Collect the grid data sets related to electric vehicles in the grid, classify and label them, and obtain the original grid data training set;

[0007] S2. Based on the generative adversarial network algorithm of Riemannian manifold feature diffusion, the original power grid data training set is expanded, and the data expansion model is trained to obtain the expanded power grid data training set;

[0008] S3, based on the extreme learning machine algorithm of nonlinear system coupling learning, the expanded power grid data training set is input into the power grid stability assessment model for training, and a trained power grid stability assessment model is obtained;

[0009] S4. Input new grid data set samples related to electric vehicles in the grid into the trained grid stability assessment model for assessment, and then obtain the assessment level of grid stability.

[0010] The S2 includes:

[0011] S21, initialize the maximum number of iterations of the preset generative adversarial network, and use the original power grid data training set as the quantum ground state set, which contains Ke quantum ground states |φ k>, using quantum computing strategies in the generator input layer of the generative adversarial network to generate a series of initial quantum state power grid data points;

[0012] The quantum computing strategy is:

[0013]

[0014] In the above formula, ψ c is the generated initial quantum state grid data point, Ke is the total number of quantum ground states, α c,k is the complex coefficient of the quantum state, |φ k > is the quantum ground state;

[0015] S22, generator G c Receiving power grid data points in an initial quantum state, and by performing data diffusion on a Riemann manifold, enabling the power grid data points in the initial quantum state to obtain new power grid data features, thereby generating new power grid data points;

[0016] The data diffusion method on the Riemann manifold is:

[0017]

[0018] In the above formula, is the grid data point of the t-th iteration. The grid data point when t is 0 is the grid data point of the initial quantum state. c is the diffusion step length, For gradient-based The manifold diffusion operator is is the loss function on the Riemann manifold, According to the grid data point The Riemannian metric matrix, μ, is dynamically adjusted for position on the Riemannian manifold. c is the mean of the current iteration batch power grid data, σ c is the scale parameter of the current iteration batch power grid data, and I is the unit matrix;

[0019] S23, based on the generated new power grid data points, using the loss function to optimize the generator and discriminator of the generative adversarial network, and performing feature correction on the new power grid data points;

[0020] The constraints of the generator and discriminator of the loss function to optimize the generative adversarial network are:

[0021]

[0022] In the above formula, G is the loss function of the generative adversarial network. c is the generator, D c is the discriminator, is the loss of the discriminator, is the loss of the generator, D c (x) is the score of the discriminator for the data point x in the original power grid data training set, G c (z) is the power grid data generated by the generator according to the random noise z, and the power grid data generated by the generator according to the random noise z is p data (x) is the distribution of data points in the original power grid data training set, p z (z) is the probability distribution of input noise;

[0023] The method of performing feature correction on the new power grid data point is:

[0024]

[0025] In the above formula, is the corrected power grid data, Proj S ( ) is the operation of projecting the power grid data onto the Riemann manifold, η c is the learning rate, is the gradient of the t+1th iteration, RS( ) is the regularization function of the correction, λ c，i To calibrate the intensity parameters, is the i-th grid data point in the t+1-th iteration, is the average value of all grid data points in the t+1th iteration;

[0026] S24, calculate the quality score of the corrected new power grid data point, and use the discriminator D c Evaluate the similarity between the new power grid data and the real power grid data. When the quality score of the new power grid data point is greater than the preset similarity threshold, retain the generated new power grid data point; otherwise, discard the new power grid data point.

[0027] The quality score of the corrected new grid data point is calculated using the following formula:

[0028]

[0029] In the above formula, is the quality score of the new power grid data point, d( ) is the distance measurement function between the means of two power grid data point sets, X real is the real power grid data, || || is the Euclidean distance;

[0030] S25, using the adaptive meaning space mapping strategy to the new grid data point G of the generator retained by S24 c (z) Perform meaning adjustment to obtain new power grid data points after meaning adjustment:

[0031] Sc (G c (z)) = G c (z)+β c ·tanh(Ω c BERT (G c (z)));

[0032] In the above formula, S c ( ) is the meaning adjustment function to adapt to the meaning space mapping module, β c is the weight parameter, tanh() is the hyperbolic tangent activation function, Ω c is a learnable transformation matrix, BERT (G c (z)) is the deep meaning feature extracted using the pre-trained BERT model;

[0033] S26, adopting a dynamic meaning perception adjustment mechanism, performing secondary adjustment on the new power grid data points after the meaning adjustment, and dynamically adjusting the generation process of the new power grid data points:

[0034]

[0035] In the above formula, X c,adjusted is the new sample after dynamic meaning perception adjustment, that is, the final sample generated, γ c To adjust the weight, softmax( ) is the Softmax function, Θ c is the learnable transformation matrix, Attn( ) is the self-attention function, is the before and after data vector of the power grid data;

[0036] S27, add 1 to the number of iterations, return to step S22, input the second adjusted power grid data points into the generator of the generative adversarial network, repeat the above steps until the preset maximum number of iterations is reached, the data expansion model training is completed, and the generated new power grid data points are superimposed on the original power grid data training set to obtain the expanded power grid data training set.

[0037] The S3 includes:

[0038] S31. Randomly initialize the hidden layer neurons of the extreme learning machine:

[0039] W u =rande(N u , D u );

[0040] b u =rande(n u , 1);

[0041] In the above formula, W uis the weight matrix of the neurons in the hidden layer of the extreme learning machine, N u is the number of neurons in the hidden layer of the extreme learning machine, D u is the dimension of the feature input to the extreme learning machine, b u is the bias vector of the hidden layer neuron of the extreme learning machine, rande(N u , D u ) function represents the generation of an N u ×D u A matrix whose elements are random numbers, rande(N u , 1) The function represents the generation of an N u ×1 vector, where the elements are random numbers;

[0042] S32, input the expanded power grid data training set into the randomly initialized extreme learning machine hidden layer neurons, and calculate the output of the extreme learning machine hidden layer neurons:

[0043] y u，j =f u (W u，j ·x u，j +b u，j );

[0044] In the above formula, y u，j is the output of the jth hidden neuron of the extreme learning machine, f u () is the activation function of the extreme learning machine, W u，j is the weight of the jth hidden neuron of the extreme learning machine, x u，j is the power grid data input into the jth hidden neuron of the extreme learning machine, b u，j is the bias vector of the jth hidden neuron of the extreme learning machine;

[0045] S33, calculating an adaptive threshold of the error according to the error between the output of the hidden layer neuron of the extreme learning machine and the target output, and dynamically adjusting the learning rate of the extreme learning machine through the adaptive threshold, thereby dynamically adjusting the weight and bias of the hidden layer neuron of the extreme learning machine;

[0046] The adaptive threshold of the error is calculated using the following formula:

[0047]

[0048] The learning rate of the extreme learning machine is calculated using the following formula:

[0049]

[0050] The weights and biases of the neurons in the hidden layer of the extreme learning machine are calculated using the following formula:

[0051]

[0052] In the above formula, is the adaptive threshold of the t-th iteration error, β is the smoothing coefficient, std() is the function for calculating the standard deviation, E (t-1) is the error set of all samples in the previous iteration, η(t) is the learning rate of the extreme learning machine in the tth iteration, η0 is the initial learning rate, is the average absolute value of the error at the tth iteration, is the weight of the extreme learning machine at the tth iteration, is the gradient of the loss function L() with respect to the weight, L() is the loss function of the extreme learning machine, X u , Y u are the input and output of the extreme learning machine respectively, μ is the influence factor of the change of the extreme learning machine, is the weight change of the extreme learning machine between this iteration and the previous iteration, The bias of the extreme learning machine at the tth iteration, is the gradient of the loss function L() with respect to the bias, is the change in bias of the extreme learning machine between this iteration and the previous iteration, M u is the number of samples input to the current iteration batch of hidden layer neurons in the extreme learning machine, is the cross entropy loss of a single sample, λ is the regularization coefficient of the extreme learning machine, is the regularization term of the extreme learning machine;

[0053] S34. Based on the dynamically adjusted weights and biases of the neurons in the hidden layer of the extreme learning machine, the gradient of each neuron in the hidden layer of the extreme learning machine is calculated through the reversible neuron, and the weights and biases of the neurons in the hidden layer of the extreme learning machine are further updated:

[0054]

[0055]

[0056] In the above formula, ← is the parameter update operation, a is the learning rate of the reversible neuron, and δ u，j is the gradient of the jth hidden neuron of the extreme learning machine, is the transpose of the input grid data of the jth hidden neuron of the extreme learning machine, is the predicted label of the extreme learning machine for the jth hidden neuron sample, f′ u ( ) is the derivative of the extreme learning machine activation function, z u，j is the activation function feature vector input to the jth hidden neuron of the extreme learning machine;

[0057] S35, using the updated weights and biases of the hidden layer neurons of the extreme learning machine, return to S32 to recalculate the outputs of the hidden layer neurons of the extreme learning machine, and define the convergence function of the output power grid data, and determine whether the convergence function is less than the set threshold. If so, it is considered that the power grid stability assessment model has stabilized and the training is terminated. If not, return to S33 to readjust and update the weights and biases of the hidden layer neurons of the extreme learning machine until the convergence function of the output power grid data is less than the set threshold, and a trained power grid stability assessment model is obtained;

[0058] The convergence function of the output power grid data is:

[0059]

[0060] In the above formula, C u is the convergence function of the output power grid data.

[0061] In S1, the grid data set related to electric vehicles in the grid includes: data on the use frequency and driving mode of electric vehicles obtained from the transportation department, data on the distribution of electric vehicles obtained from the electric vehicle registration database, and data on the geographical location and operation of charging stations obtained from the charging infrastructure database;

[0062] The grid is classified and labeled, and the labeled category is the stability assessment level, which is specifically divided into 5 levels, including unstable, less stable, relatively stable, stable, and high stability. The higher the level, the more stable the grid.

[0063] In a second aspect, the present invention proposes a grid stability assessment system for large-scale electric vehicle access, including a training set data acquisition module, a grid data set expansion module, a grid stability assessment model training module, and a grid stability level assessment module;

[0064] The training set data acquisition module is used to collect power grid data sets related to electric vehicles in the power grid, classify and annotate them, and obtain the original power grid data training set;

[0065] The power grid data set expansion module is used to perform data expansion on the original power grid data training set based on the generative adversarial network algorithm of Riemann manifold feature diffusion, and train the data expansion model to obtain the expanded power grid data training set;

[0066] The power grid stability assessment model training module is used for an extreme learning machine algorithm based on nonlinear system coupling learning, and the expanded power grid data training set is input into the power grid stability assessment model for training to obtain a trained power grid stability assessment model;

[0067] The power grid stability level assessment module is used to input new power grid data set samples related to electric vehicles in the power grid into the trained power grid stability assessment model for assessment, thereby obtaining an assessment level of power grid stability.

[0068] The power grid data set expansion module includes a quantum state power grid data point generation unit, a new power grid data point generation unit, an optimization and feature correction unit, a quality scoring unit, a meaning adjustment unit, a secondary adjustment unit, and an iteration unit;

[0069] The quantum state power grid data point generation unit is used to initialize the maximum number of iterations of the preset generative adversarial network, and use the original power grid data training set as the quantum ground state set, which contains Ke quantum ground states |φ k >, using quantum computing strategies in the generator input layer of the generative adversarial network to generate a series of initial quantum state power grid data points;

[0070] The quantum computing strategy is:

[0071]

[0072] In the above formula, ψ c is the generated initial quantum state grid data point, Ke is the total number of quantum ground states, α c,k is the complex coefficient of the quantum state, |φ k > is the quantum ground state;

[0073] The new power grid data point generation unit is used to generate G c Receiving power grid data points in an initial quantum state, and by performing data diffusion on a Riemann manifold, enabling the power grid data points in the initial quantum state to obtain new power grid data features, thereby generating new power grid data points;

[0074] The data diffusion method on the Riemann manifold is:

[0075]

[0076] In the above formula, is the grid data point of the t-th iteration. The grid data point when t is 0 is the grid data point of the initial quantum state. c is the diffusion step length, For gradient-based The manifold diffusion operator is is the loss function on the Riemann manifold, According to the grid data point The Riemannian metric matrix, μ, is dynamically adjusted for position on the Riemannian manifold. c is the mean of the current iteration batch power grid data, σ cis the scale parameter of the current iteration batch power grid data, and I is the unit matrix;

[0077] The optimization and feature correction unit is used to optimize the generator and discriminator of the generative adversarial network based on the generated new power grid data points by using the loss function, and perform feature correction on the new power grid data points;

[0078] The constraints of the generator and discriminator of the loss function to optimize the generative adversarial network are:

[0079]

[0080] In the above formula, G is the loss function of the generative adversarial network. c is the generator, D c is the discriminator, is the loss of the discriminator, is the loss of the generator, D c (x) is the score of the discriminator for the data point x in the original power grid data training set, G c (z) is the power grid data generated by the generator according to the random noise z, and the power grid data generated by the generator according to the random noise z is p data (x) is the distribution of data points in the original power grid data training set, p z (z) is the probability distribution of input noise;

[0081] The method of performing feature correction on the new power grid data point is:

[0082]

[0083] In the above formula, is the corrected power grid data, Proj S () is the operation of projecting the power grid data onto the Riemann manifold, η c is the learning rate, is the gradient of the t+1th iteration, RS() is the regularization function of the correction, and λ c，i To calibrate the intensity parameters, is the i-th grid data point in the t+1-th iteration, is the average value of all grid data points in the t+1th iteration;

[0084] The quality scoring unit is used to calculate the quality score of the corrected new power grid data point, using the discriminator D c Evaluate the similarity between the new power grid data and the real power grid data. When the quality score of the new power grid data point is greater than the preset similarity threshold, retain the generated new power grid data point; otherwise, discard the new power grid data point.

[0085] The quality score of the corrected new grid data point is calculated using the following formula:

[0086]

[0087] In the above formula, is the quality score of the new power grid data point, d() is the distance measurement function between the means of two power grid data point sets, X real is the real power grid data, || || is the Euclidean distance;

[0088] The meaning adjustment unit is used to adopt an adaptive meaning space mapping strategy to adjust the new power grid data point G of the generator retained by the quality scoring unit. c (z) Perform meaning adjustment to obtain new power grid data points after meaning adjustment:

[0089] S c (G c (z)) = G c (z)+β c ·tanh(Ω c BERT (G c (z)));

[0090] In the above formula, S c () is the meaning adjustment function to adapt to the meaning space mapping module, β c is the weight parameter, tanh( ) is the hyperbolic tangent activation function, Ω c is a learnable transformation matrix, BERT (G c (z)) is the deep meaning feature extracted using the pre-trained BERT model;

[0091] The secondary adjustment unit is used to adopt a dynamic meaning perception adjustment mechanism to perform secondary adjustment on the new power grid data point after the meaning adjustment, and dynamically adjust the generation process of the new power grid data point:

[0092]

[0093] In the above formula, X c，adjusted is the new sample after dynamic meaning perception adjustment, that is, the final sample generated, γ c To adjust the weight, softmax( ) is the Softmax function, Θ c is the learnable transformation matrix, Attn( ) is the self-attention function, is the before and after data vector of the power grid data;

[0094] The iteration unit is used to increase the number of iterations by 1, return the new power grid data point generation unit, input the second adjusted power grid data point into the generator of the generative adversarial network, repeat the above operation steps until the preset maximum number of iterations is reached, the data expansion model training is completed, and the generated new power grid data point is superimposed on the original power grid data training set to obtain the expanded power grid data training set.

[0095] The power grid stability assessment model training module includes a random initialization unit, an output calculation unit, a weight and bias dynamic adjustment unit, a weight and bias update unit, and a circulation unit;

[0096] The random initialization unit is used to randomly initialize the hidden layer neurons of the extreme learning machine:

[0097] W u =rande(N u , D u );

[0098] b u =rande(N u , 1);

[0099] In the above formula, W u is the weight matrix of the neurons in the hidden layer of the extreme learning machine, N u is the number of neurons in the hidden layer of the extreme learning machine, D u is the dimension of the feature input to the extreme learning machine, b u is the bias vector of the hidden layer neuron of the extreme learning machine, rande(N u , D u ) function represents the generation of an N u ×D u A matrix whose elements are random numbers, rande(N u , 1) The function represents the generation of an N u ×1 vector, where the elements are random numbers;

[0100] The output calculation unit is used to input the expanded power grid data training set into the randomly initialized extreme learning machine hidden layer neurons, and calculate the output of the extreme learning machine hidden layer neurons:

[0101] y u，j =f u (W u，j ·x u，j +b u，j );

[0102] In the above formula, y u，j is the output of the jth hidden neuron of the extreme learning machine, f u () is the activation function of the extreme learning machine, Wu，j is the weight of the jth hidden neuron of the extreme learning machine, x u，j is the power grid data input into the jth hidden neuron of the extreme learning machine, b u，j is the bias vector of the jth hidden neuron of the extreme learning machine;

[0103] The weight and bias dynamic adjustment unit is used to calculate the error adaptive threshold according to the error between the output of the extreme learning machine hidden layer neuron and the target output, and dynamically adjust the learning rate of the extreme learning machine through the adaptive threshold, thereby dynamically adjusting the weight and bias of the extreme learning machine hidden layer neuron;

[0104] The adaptive threshold of the error is calculated using the following formula:

[0105]

[0106] The learning rate of the extreme learning machine is calculated using the following formula:

[0107]

[0108] The weights and biases of the neurons in the hidden layer of the extreme learning machine are calculated using the following formula:

[0109]

[0110] In the above formula, is the adaptive threshold of the t-th iteration error, β is the smoothing coefficient, std() is the function for calculating the standard deviation, E (t-1) is the error set of all samples in the previous iteration, η(t) is the learning rate of the extreme learning machine in the tth iteration, η0 is the initial learning rate, is the average absolute value of the error at the tth iteration, is the weight of the extreme learning machine at the tth iteration, is the gradient of the loss function L( ) with respect to the weight, L( ) is the loss function of the extreme learning machine, X u , Y u are the input and output of the extreme learning machine respectively, μ is the influence factor of the change of the extreme learning machine, is the weight change of the extreme learning machine between this iteration and the previous iteration, The bias of the extreme learning machine at the tth iteration, is the gradient of the loss function L() with respect to the bias, is the change in bias of the extreme learning machine between this iteration and the previous iteration, M u is the number of samples input to the current iteration batch of hidden layer neurons in the extreme learning machine, is the cross entropy loss of a single sample, λ is the regularization coefficient of the extreme learning machine, is the regularization term of the extreme learning machine;

[0111] The weight and bias updating unit is used to calculate the gradient of each neuron in the hidden layer of the extreme learning machine through the reversible neuron based on the weight and bias of the neurons in the hidden layer of the extreme learning machine after the dynamic adjustment, and further update the weight and bias of the neurons in the hidden layer of the extreme learning machine:

[0112]

[0113] In the above formula, ← is the parameter update operation, α is the learning rate of the reversible neuron, and δ u，j is the gradient of the jth hidden neuron of the extreme learning machine, is the transpose of the input grid data of the jth hidden neuron of the extreme learning machine, is the predicted label of the extreme learning machine for the jth hidden neuron sample, f′ u ( ) is the derivative of the extreme learning machine activation function, z u，j is the activation function feature vector input to the jth hidden neuron of the extreme learning machine;

[0114] The circulation unit is used to use the updated weights and biases of the hidden layer neurons of the extreme learning machine to return to the output calculation unit to recalculate the outputs of the hidden layer neurons of the extreme learning machine, and define a convergence function of the output power grid data, and determine whether the convergence function is less than a set threshold. If so, it is considered that the power grid stability assessment model has stabilized and the training is terminated. If not, it returns to the weight and bias dynamic adjustment unit to readjust and update the weights and biases of the hidden layer neurons of the extreme learning machine until the convergence function of the output power grid data is less than the set threshold, thereby obtaining a trained power grid stability assessment model;

[0115] The convergence function of the output power grid data is:

[0116]

[0117] In the above formula, C u is the convergence function of the output power grid data.

[0118] In the training set data acquisition module, the grid data set related to electric vehicles in the grid includes: data on the use frequency and driving mode of electric vehicles obtained from the transportation department, data on the distribution of electric vehicles obtained from the electric vehicle registration database, and data on the geographical location and operation of charging stations obtained from the charging infrastructure database;

[0119] The grid is classified and labeled, and the labeled category is the stability assessment level, which is specifically divided into 5 levels, including unstable, less stable, relatively stable, stable, and high stability. The higher the level, the more stable the grid.

[0120] In a third aspect, the present invention provides a large-scale electric vehicle access grid stability assessment device, including a processor and a memory;

[0121] The memory is used to store computer program code and transmit the computer program code to the processor;

[0122] The processor is used to execute the aforementioned method for evaluating the stability of a power grid with large-scale electric vehicle access according to the instructions in the computer program code.

[0123] In a fourth aspect, the present invention provides a computer storage medium having a computer program stored thereon;

[0124] When the computer program is executed by a processor, the steps of the aforementioned method for evaluating the stability of a power grid with large-scale electric vehicle access are implemented.

[0125] Compared with the prior art, the present invention has the following beneficial effects:

[0126] 1. The present invention proposes a method and system for evaluating the stability of a power grid with large-scale electric vehicle access. The method first collects a power grid data set related to electric vehicles in the power grid, classifies and annotates it, and obtains an original power grid data training set; then, based on the generative adversarial network algorithm of Riemann manifold feature diffusion, the original power grid data training set is expanded, and the data expansion model is trained to obtain an expanded power grid data training set; and based on the extreme learning machine algorithm of nonlinear system coupling learning, the expanded power grid data training set is input into the power grid stability evaluation model for training to obtain a trained power grid stability evaluation model; finally, the new power grid data set samples related to electric vehicles in the power grid are input into the trained power grid stability evaluation model for evaluation, and then the evaluation level of power grid stability is obtained. The method collects power grid data sets related to electric vehicles in the power grid from multiple aspects, comprehensively reflects the multi-dimensional impact of electric vehicles accessing the power grid, and annotates and preprocesses the training set power grid data to reduce data distortion and improve the training performance of the model.

[0127] 2. The present invention proposes a method and system for evaluating the stability of a power grid with large-scale electric vehicles connected. Based on a generative adversarial network algorithm of Riemann manifold feature diffusion, data expansion is performed on the original power grid data training set, and a data expansion model is trained to obtain an expanded power grid data training set. On the one hand, the method generates initial power grid data points based on the randomness of quantum states at the input layer of the generative adversarial network generator, using a quantum computing strategy, so that the initial power grid data points have a high degree of randomness and coverage in distribution, which helps to explore the potential space of power grid data, thereby generating richer and more diverse power grid data; on the other hand, the method performs data diffusion on a Riemann manifold, maintains the essential attributes of data features, generates power grid data that is highly similar to real data, enhances the diversity and authenticity of the data, and in the training process of the data expansion model, through feature correction and meaning adjustment mechanisms, ensures the accuracy and stability of the generated data, and further improves the quality of data expansion.

[0128] 3. The present invention proposes a method and system for evaluating the stability of a power grid with large-scale electric vehicle access. Based on the extreme learning machine algorithm of nonlinear system coupling learning, the expanded power grid data training set is input into the power grid stability evaluation model for training to obtain a trained power grid stability evaluation model. On the one hand, the method adopts an adaptive threshold adjustment mechanism of error to dynamically adjust the learning rate of the extreme learning machine in real time, thereby enhancing the flexibility and convergence speed of the extreme learning machine when facing different types of data and reducing the training time of the model. On the other hand, the method targets the highly nonlinear and dynamic changes of power grid data. On the basis of the traditional extreme learning machine algorithm, the method adopts reversible neurons to adaptively adjust the gradients of the hidden layer neurons of the extreme learning machine during the training process, allowing the model to dynamically update the weights and biases of the hidden layer neurons of the extreme learning machine according to error feedback during the training process, so as to better match and identify the influencing factors of power grid stability, thereby improving the model's adaptability to complex nonlinear data and its ability to evaluate the stability of complex power grid operating environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0129] Figure 1 The figure is an overall flow chart of the method of the present invention.

[0130] Figure 2 The structure diagram of the system of the present invention is shown in FIG.

[0131] Figure 3 This is a structural diagram of the device described in Example 3. DETAILED DESCRIPTION

[0132] The present invention is further described in detail below in conjunction with specific implementations and drawings.

[0133] The present invention proposes a grid stability assessment method and system for large-scale electric vehicle access. The method utilizes the collection, preprocessing and annotation of multi-source training set grid data, combines quantum computing with a generative adversarial network of Riemann manifold feature diffusion to expand the grid data, and generates high-quality grid data samples. Subsequently, an extreme learning machine algorithm based on nonlinear system coupling learning is used to model the complex nonlinear characteristics of the grid for grid stability assessment, thereby realizing grid stability assessment for large-scale electric vehicle access.

[0134] Embodiment 1:

[0135] like Figure 1 As shown, a method for evaluating the stability of a power grid with large-scale electric vehicle access is performed in the following steps:

[0136] 1. Collect the grid data sets related to electric vehicles in the grid, classify and label them, and obtain the original grid data training set;

[0137] The grid data sets related to electric vehicles in the power grid include: data on the frequency of use and driving patterns of electric vehicles obtained from the transportation department, data on the distribution of electric vehicles obtained from the electric vehicle registration database, and data on the geographical location and operation of charging stations obtained from the charging infrastructure database; the collection method of grid data sets related to electric vehicles relies on the data transmission of the API interface and the on-site data collection equipment, and the collected data are stored in JSON format;

[0138] The attributes of the power grid dataset are shown in Table 1, including: Ra represents the regional power grid code, Da represents the increase or decrease of power demand in the time period, Ca represents the power load of the charging station, Pa represents the peak charging period, Ta represents the impact of temperature on charging efficiency, Ma represents the power grid dispatch response time, Na represents the proportion of renewable energy input, Fa represents the frequency regulation sensitivity, Ga represents the power factor, and Ha represents the voltage fluctuation range; In practical applications, the attributes of the data are usually more than 10 attributes, and the number of attributes of the data may reach dozens or even hundreds;

[0139] Table 1 Power grid dataset attribute table

[0140] Ra Da Ca Pa Ta Ma Na Fa Ga Ha 101 10% increase 60MW 18:00-20:00 -5℃ Impact 2min 30％ high 0.95 ±5％ 102 5% reduction 45MW 12:00-14:00 No impact 5min 50％ middle 0.98 ±2％ 103 20% increase 80MW 06:00-08:00 -10℃ Impact 1min 20％ Low 0.90 ±10％ 104 Stablize 50MW 21:00-23:00 5℃ Impact 3min 40％ high 0.93 ±3％ 105 15% reduction 35MW 09:00-11:00 15℃ Impact 6min 60％ middle 0.97 ±1％ ;

[0141] The grid data sets related to electric vehicles in the grid are manually classified and labeled. The labeled categories are the stability assessment levels, which are divided into 5 levels, including unstable, unstable, relatively stable, stable, and high stability. The higher the level, the more stable the grid.

[0142] The methods of preprocessing the labeled power grid data set include missing value processing, outlier processing, standardization, normalization and other operations. The labeled power grid data set is preprocessed to ensure that the power grid data set related to electric vehicles in the power grid is in a unified format when entering the subsequent algorithm processing;

[0143] If an attribute a of the power grid dataset i If the data is continuous and there are missing values, the mean interpolation method is used to process the missing values:

[0144]

[0145] In the above formula, a i is the attribute of the power grid data set, N is the number of samples in the power grid data set, a ij is the attribute value of the jth sample;

[0146] If an attribute a of the power grid dataset i If the data is discrete power grid data and there are missing values, the majority interpolation method is used to handle the missing values;

[0147] For attribute a i The outlier data of the power grid is processed, and the outliers are screened and deleted. The following formula is used to determine the attribute a i The outlier range of the power grid data is:

[0148] Q ycz <Q1(a i )-1.5×IQR(a i );

[0149] Q ycz <Q3(a i )+1.5×IQR(a i );

[0150] Q1(a i )=percentile(a i , 25);

[0151] Q3(a i )=percentile(a i , 75);

[0152] IQR(a i )=Q3(a i )-Q1(a i );

[0153] In the above formula, Q ycz is an outlier, percentile is the percentage function, Q1 and Q3 are the first quartile and the third quartile respectively, and IQR is the interquartile range;

[0154] In order to make the attributes of the power grid data on the same scale, the attribute a i Standardize the power grid data:

[0155]

[0156] In the above formula, b i is the standardized attribute a i The grid data, μ(a i ) is attribute a i The mean value of the power grid data, σ(a i ) is attribute a i The standard deviation of the power grid data;

[0157] Set the attribute to a i The power grid data is converted to the range of [0, 1] and normalized:

[0158]

[0159] In the above formula, c i is the normalized attribute a i of power grid data.

[0160] 2. Based on the generative adversarial network algorithm of Riemannian manifold feature diffusion, the original power grid data training set is expanded, and the data expansion model is trained to obtain the expanded power grid data training set;

[0161] The generative adversarial network algorithm based on Riemannian manifold feature diffusion is used to generate samples of the original power grid data training set, thereby realizing the expansion of power grid data. The algorithm consists of two parts: a generator and a discriminator. The generator is used to generate approximately real power grid data vectors, and the discriminator is used to distinguish between the generated vectors and the real power grid data vectors.

[0162] The specific process of generating power grid data set samples based on the generative adversarial network algorithm based on Riemann manifold feature diffusion is as follows:

[0163] The maximum number of iterations of the initial preset generative adversarial network is 500, and the original power grid data training set is used as the quantum ground state set, which contains Ke quantum ground states |φ k >, in the generator input layer of the generative adversarial network, the randomness of the quantum state is utilized, and the quantum computing strategy is used to generate a series of initial quantum state power grid data points from the superposition state of the original power grid data training set of random qubits. The generated initial power grid data points are distributed as vectors with high randomness and coverage, which helps to explore the potential space of power grid data, thereby generating richer and more diverse power grid data;

[0164] The quantum computing strategy is:

[0165]

[0166] In the above formula, ψ c is the generated initial quantum state grid data point, Ke is the total number of quantum ground states, set to 50, α c，k is the complex coefficient of the quantum state, |φ k > is the quantum ground state;

[0167] Generator G c Receive the power grid data points in the initial quantum state, and by performing data diffusion on the Riemann manifold, obtain a new power grid data vector feature representation while maintaining the essential vector attributes of the power grid data points in the initial quantum state, thereby generating new power grid data points;

[0168] The data diffusion method on the Riemann manifold is:

[0169]

[0170] In the above formula, is the grid data point of the t-th iteration. The grid data point when t is 0 is the grid data point of the initial quantum state. c is the diffusion step size, set to 0.01, For gradient-based The manifold diffusion operator is is the loss function on the Riemann manifold, According to the grid data point The Riemannian metric matrix, μ, is dynamically adjusted for position on the Riemannian manifold. c is the mean of the current iteration batch power grid data, σ c is the scale parameter of the current iteration batch power grid data, and I is the unit matrix;

[0171] Based on the generated new power grid data points, the loss function is used to optimize the generator and discriminator of the generative adversarial network, and the features of the new power grid data points are corrected to ensure the correctness of the power grid data features on the Riemann manifold;

[0172] The constraints of the generator and discriminator of the loss function to optimize the generative adversarial network are:

[0173]

[0174] In the above formula, G is the loss function of the generative adversarial network. c is the generator, D c is the discriminator, is the loss of the discriminator, is the loss of the generator, D c (x) is the score of the discriminator for the data point x in the original power grid data training set, G c (z) is the power grid data generated by the generator according to the random noise z, and the power grid data generated by the generator according to the random noise z is p data (x) is the distribution of data points in the original power grid data training set, p z (z) is the probability distribution of input noise;

[0175] The method of performing feature correction on the new power grid data point is:

[0176]

[0177] In the above formula, is the corrected power grid data, Proj S ( ) is the operation of projecting the power grid data onto the Riemann manifold, η c is the learning rate, set to 0.01, is the gradient of the t+1th iteration, RS( ) is the regularization function of the correction, λ c，i To calibrate the intensity parameter, set it to 0.05. is the i-th grid data point in the t+1-th iteration, is the average value of all grid data points in the t+1th iteration;

[0178] After the generator generates new grid data point vectors, the discriminator evaluates the similarity between these data point vectors and the real grid data point vectors;

[0179] Calculate the quality score of the corrected new grid data point and use the discriminator D c Evaluate the similarity between the new power grid data and the real power grid data. When the quality score of the new power grid data point is greater than the preset similarity threshold, retain the generated new power grid data point; otherwise, discard the new power grid data point.

[0180] The quality score of the corrected new grid data point is calculated using the following formula:

[0181]

[0182] In the above formula, is the quality score of the new power grid data point, d( ) is the distance measurement function between the means of two power grid data point sets, X real is the real power grid data, || || is the Euclidean distance, that is, the L2 norm;

[0183] Adopting adaptive meaning space mapping strategy to retain the new grid data point G of the generatorc (z) Perform meaning adjustment to obtain new power grid data points after meaning adjustment:

[0184] S c (G c (z)) = G c (z)+β c ·tanh(Ω c BERT (G c (z)));

[0185] In the above formula, S c ( ) is the meaning adjustment function adapted to the meaning space mapping module. This function extracts the deep meaning features of the power grid data based on the existing meaning analysis model and adjusts these features to optimize them through an adaptive learning mechanism. The meaning analysis model is a pre-trained BERT model. β c is the weight parameter, set to 0.1, tanh() is the hyperbolic tangent activation function, Ω c is a learnable transformation matrix, BERT (G c (z)) is the deep meaning feature extracted using the pre-trained BERT model;

[0186] A dynamic meaning perception adjustment mechanism is used to make secondary adjustments to the new power grid data points after the meaning adjustment. A self-attention mechanism is used to capture the contextual relationship in the power grid data. By analyzing the contextual factors that generate the power grid data, the generation process of the new power grid data points is dynamically adjusted:

[0187]

[0188] In the above formula, X c，adjusted is the new sample after dynamic meaning perception adjustment, that is, the final sample generated, γ c To adjust the weight, set it to 0.1, softmax() is the Softmax function, Θ c is a learnable transformation matrix, Attn() is the self-attention function, is the before and after data vector of the power grid data;

[0189] Let the number of iterations increase by 1, input the power grid data points after the second adjustment into the generator of the generative adversarial network, repeat the above steps until the preset maximum number of iterations is reached, the data expansion model training is completed, and the generated new power grid data points are superimposed on the original power grid data training set to obtain the expanded power grid data training set. Assuming that the original acquired power grid data training set samples are 800 and the generated new power grid data are 200, the expanded power grid data set contains 1000 samples.

[0190] 3. Based on the extreme learning machine algorithm of nonlinear system coupling learning, the expanded power grid data training set is input into the power grid stability assessment model for training to obtain a trained power grid stability assessment model;

[0191] The extreme learning machine algorithm based on nonlinear system coupling learning is based on the traditional extreme learning machine algorithm. It uses hidden layer reversible neurons to adaptively adjust the activation function during the training process, allowing the power grid stability assessment model to dynamically adjust the parameters of neurons according to error feedback during the training process to better match and identify nonlinear, complex, and stability-influencing factors in the power grid.

[0192] The specific training process of the extreme learning machine algorithm based on nonlinear system coupling learning includes:

[0193] Randomly initialize the hidden layer neurons of the extreme learning machine:

[0194] W u =rande(N u , D u );

[0195] b u =rande(N u ,1);

[0196] In the above formula, W u is the weight matrix of the neurons in the hidden layer of the extreme learning machine, N u is the number of neurons in the hidden layer of the extreme learning machine, D u is the dimension of the feature input to the extreme learning machine, b u is the bias vector of the hidden layer neuron of the extreme learning machine, rande(N u , D u ) function represents the generation of an N u ×D u A matrix whose elements are random numbers, rande(N u , 1) The function represents the generation of an N u ×1 vector, where the elements are random numbers;

[0197] Input the expanded power grid data training set into the randomly initialized extreme learning machine hidden layer neurons, and calculate the output of the extreme learning machine hidden layer neurons;

[0198] y u，j =f u (W u，j ·X u，j +b u，j );

[0199] In the above formula, y u，jis the output of the jth hidden neuron of the extreme learning machine, f u () is the activation function of the extreme learning machine, W u，j is the weight of the jth hidden neuron of the extreme learning machine, x u，j is the power grid data input into the jth hidden neuron of the extreme learning machine, b u，j is the bias vector of the jth hidden neuron of the extreme learning machine;

[0200] According to the error between the output of the neurons in the hidden layer of the extreme learning machine and the target output, the adaptive threshold of the error is calculated, and the learning rate of the extreme learning machine is dynamically adjusted through the adaptive threshold, thereby dynamically adjusting the weights and biases of the neurons in the hidden layer of the extreme learning machine;

[0201] The adaptive threshold of the error is calculated using the following formula:

[0202]

[0203] The learning rate of the extreme learning machine is calculated using the following formula:

[0204]

[0205] The weights and biases of the neurons in the hidden layer of the extreme learning machine are calculated using the following formula:

[0206]

[0207] In the above formula, is the adaptive threshold of the t-th iteration error, β is the smoothing coefficient, which is used to adjust the influence of historical errors, std() is the function for calculating the standard deviation, E (t-1) is the error set of all samples in the previous iteration, η(t) is the learning rate of the extreme learning machine in the tth iteration, η0 is the initial learning rate, is the average absolute value of the error at the tth iteration, is the weight of the extreme learning machine at the tth iteration, is the gradient of the loss function L with respect to the weight. L is the loss function of the extreme learning machine, which is a composite function integrating classification accuracy and model complexity. u , Y u are the input and output of the extreme learning machine respectively, μ is the influence factor of the change of the extreme learning machine, is the weight change of the extreme learning machine between this iteration and the previous iteration, The bias of the extreme learning machine at the tth iteration, is the gradient of the loss function L with respect to the bias, is the change in bias of the extreme learning machine between this iteration and the previous iteration, Mu is the number of samples input to the current iteration batch of hidden layer neurons in the extreme learning machine, is the cross entropy loss of a single sample, λ is the regularization coefficient of the extreme learning machine, is the regularization term of the extreme learning machine;

[0208] Based on the characteristics of real-time data streams, a dynamic error correction mechanism with adaptive thresholds is used to dynamically adjust the current error and correct the threshold to achieve a more accurate learning process and faster model adaptability. The adaptive threshold is dynamically adjusted based on the average error calculated in each iteration. The purpose of the adaptive threshold is to adjust the learning sensitivity according to the actual error situation at each step of model training, so that the model can show higher flexibility and efficiency when facing different types of data distributions.

[0209] Based on the dynamically adjusted weights and biases of the neurons in the hidden layer of the extreme learning machine, the gradient of each neuron in the hidden layer of the extreme learning machine is calculated through the reversible neurons, and the weights and biases of the neurons in the hidden layer of the extreme learning machine are further updated:

[0210]

[0211] After each forward propagation, the gradient of each parameter is calculated using the back propagation algorithm through the reversible neurons. Specifically, for each neuron in the hidden layer of the extreme learning machine, the gradient calculation method is expressed as:

[0212]

[0213] In the above formula, ← is the parameter update operation, α is the learning rate of the reversible neuron, and δ u，j is the gradient of the jth hidden neuron of the extreme learning machine, is the transpose of the input grid data of the jth hidden neuron of the extreme learning machine, is the predicted label of the extreme learning machine for the jth hidden neuron sample, f′ u ( ) is the derivative of the extreme learning machine activation function, z u，j is the activation function feature vector input to the jth hidden neuron of the extreme learning machine;

[0214] Using the updated weights and biases of the neurons in the hidden layer of the extreme learning machine, the outputs of the neurons in the hidden layer of the extreme learning machine are recalculated, and a convergence function of the output power grid data is defined to determine whether the convergence function is less than a set threshold. If so, it is considered that the power grid stability assessment model has stabilized and the training is terminated. If not, the weights and biases of the neurons in the hidden layer of the extreme learning machine are readjusted and updated until the convergence function of the output power grid data is less than the set threshold, thereby obtaining a trained power grid stability assessment model.

[0215] The convergence function of the output power grid data is:

[0216]

[0217] In the above formula, C u is the convergence function of the output power grid data.

[0218] 4. Input new grid data set samples related to electric vehicles in the grid into the trained grid stability assessment model for assessment, and then obtain the grid stability assessment level;

[0219] After the training is completed, new electric vehicle data set samples in the power grid are collected and input into the trained power grid stability assessment model to evaluate the impact of large-scale electric vehicle access on power grid stability. The stability assessment levels include unstable, understable, relatively stable, stable, and high stability. The higher the level, the more stable the power grid.

[0220] Embodiment 2:

[0221] like Figure 2 As shown, a grid stability assessment system for large-scale electric vehicle access includes a training set data acquisition module, a grid data set expansion module, a grid stability assessment model training module, and a grid stability level assessment module;

[0222] The training set data acquisition module is used to collect power grid data sets related to electric vehicles in the power grid, classify and annotate them, and obtain the original power grid data training set;

[0223] The power grid data set expansion module is used to perform data expansion on the original power grid data training set based on the generative adversarial network algorithm of Riemann manifold feature diffusion, and train the data expansion model to obtain the expanded power grid data training set;

[0224] The power grid stability assessment model training module is used for an extreme learning machine algorithm based on nonlinear system coupling learning, and the expanded power grid data training set is input into the power grid stability assessment model for training to obtain a trained power grid stability assessment model;

[0225] The power grid stability level assessment module is used to input new power grid data set samples related to electric vehicles in the power grid into the trained power grid stability assessment model for assessment, thereby obtaining an assessment level of power grid stability.

[0226] In the training set data acquisition module, the grid data set related to electric vehicles in the grid includes: data on the use frequency and driving mode of electric vehicles obtained from the transportation department, data on the distribution of electric vehicles obtained from the electric vehicle registration database, and data on the geographical location and operation of charging stations obtained from the charging infrastructure database;

[0227] The grid is classified and labeled, and the labeled category is the stability assessment level, which is specifically divided into 5 levels, including unstable, less stable, relatively stable, stable, and high stability. The higher the level, the more stable the grid.

[0228] The power grid data set expansion module includes a quantum state power grid data point generation unit, a new power grid data point generation unit, an optimization and feature correction unit, a quality scoring unit, a meaning adjustment unit, a secondary adjustment unit, and an iteration unit;

[0229] The quantum state power grid data point generation unit is used to initialize the maximum number of iterations of the preset generative adversarial network, and use the original power grid data training set as the quantum ground state set, which contains Ke quantum ground states |φ k >, using quantum computing strategies in the generator input layer of the generative adversarial network to generate a series of initial quantum state power grid data points;

[0230] The quantum computing strategy is:

[0231]

[0232] In the above formula, ψ c is the generated initial quantum state grid data point, Ke is the total number of quantum ground states, α c,k is the complex coefficient of the quantum state, |φ k > is the quantum ground state;

[0233] The new power grid data point generation unit is used to generate G c Receiving power grid data points in an initial quantum state, and by performing data diffusion on a Riemann manifold, enabling the power grid data points in the initial quantum state to obtain new power grid data features, thereby generating new power grid data points;

[0234] The data diffusion method on the Riemann manifold is:

[0235]

[0236] In the above formula, is the grid data point of the t-th iteration. The grid data point when t is 0 is the grid data point of the initial quantum state. c is the diffusion step length, For gradient-based The manifold diffusion operator is is the loss function on the Riemann manifold, According to the grid data point The Riemannian metric matrix, μ, is dynamically adjusted for position on the Riemannian manifold. c is the mean of the current iteration batch power grid data, σ c is the scale parameter of the current iteration batch power grid data, and I is the unit matrix;

[0237] The optimization and feature correction unit is used to optimize the generator and discriminator of the generative adversarial network based on the generated new power grid data points by using the loss function, and perform feature correction on the new power grid data points;

[0238] The constraints of the generator and discriminator of the loss function to optimize the generative adversarial network are:

[0239]

[0240] In the above formula, G is the loss function of the generative adversarial network. c is the generator, D c is the discriminator, is the loss of the discriminator, is the loss of the generator, D c (x) is the score of the discriminator for the data point x in the original power grid data training set, G c (z) is the power grid data generated by the generator according to the random noise z, and the power grid data generated by the generator according to the random noise z is p data (x) is the distribution of data points in the original power grid data training set, p z (z) is the probability distribution of input noise;

[0241] The method of performing feature correction on the new power grid data point is:

[0242]

[0243] In the above formula, is the corrected power grid data, Proj S ( ) is the operation of projecting the power grid data onto the Riemann manifold, η c is the learning rate, is the gradient of the t+1th iteration, RS() is the regularization function of the correction, and λ c,i To calibrate the intensity parameters, is the i-th grid data point in the t+1-th iteration, is the average value of all grid data points in the t+1th iteration;

[0244] The quality scoring unit is used to calculate the quality score of the corrected new power grid data point, using the discriminator D c Evaluate the similarity between the new power grid data and the real power grid data. When the quality score of the new power grid data point is greater than the preset similarity threshold, retain the generated new power grid data point; otherwise, discard the new power grid data point.

[0245] The quality score of the corrected new grid data point is calculated using the following formula:

[0246]

[0247] In the above formula, is the quality score of the new power grid data point, d() is the distance measurement function between the means of two power grid data point sets, X real is the real power grid data, |||| is the Euclidean distance;

[0248] The meaning adjustment unit is used to adopt an adaptive meaning space mapping strategy to adjust the new power grid data point G of the generator retained by the quality scoring unit. c (z) Perform meaning adjustment to obtain new power grid data points after meaning adjustment:

[0249] S c (G c (z)) = G c (z)+β c ·tanh(Ω c BERT (G c (z)));

[0250] In the above formula, S c () is the meaning adjustment function to adapt to the meaning space mapping module, β c is the weight parameter, tanh() is the hyperbolic tangent activation function, Ω c is a learnable transformation matrix, BERT (G c (z)) is the deep meaning feature extracted using the pre-trained BERT model;

[0251] The secondary adjustment unit is used to adopt a dynamic meaning perception adjustment mechanism to perform secondary adjustment on the new power grid data point after the meaning adjustment, and dynamically adjust the generation process of the new power grid data point:

[0252]

[0253] In the above formula, X c，adjusted is the new sample after dynamic meaning perception adjustment, that is, the final sample generated, γ c To adjust the weight, softmax() is the Softmax function, Θ cis a learnable transformation matrix, Attn() is the self-attention function, is the before and after data vector of the power grid data;

[0254] The iteration unit is used to increase the number of iterations by 1, return the new power grid data point generation unit, input the second adjusted power grid data point into the generator of the generative adversarial network, repeat the above operation steps until the preset maximum number of iterations is reached, the data expansion model training is completed, and the generated new power grid data point is superimposed on the original power grid data training set to obtain the expanded power grid data training set.

[0255] The power grid stability assessment model training module includes a random initialization unit, an output calculation unit, a weight and bias dynamic adjustment unit, a weight and bias update unit, and a circulation unit;

[0256] The random initialization unit is used to randomly initialize the hidden layer neurons of the extreme learning machine:

[0257] W u =rande(N u , D u );

[0258] b u =rande(N u , 1);

[0259] In the above formula, W u is the weight matrix of the neurons in the hidden layer of the extreme learning machine, N u is the number of neurons in the hidden layer of the extreme learning machine, D u is the dimension of the feature input to the extreme learning machine, b u is the bias vector of the hidden layer neuron of the extreme learning machine, rande(N u , D u ) function represents the generation of an N u ×D u A matrix whose elements are random numbers, rande(N u , 1) The function represents the generation of an N u ×1 vector, where the elements are random numbers;

[0260] The output calculation unit is used to input the expanded power grid data training set into the randomly initialized extreme learning machine hidden layer neurons, and calculate the output of the extreme learning machine hidden layer neurons:

[0261] y u，j =f u (W u，j ·x u，j +b u，j );

[0262] In the above formula, y u，j is the output of the jth hidden neuron of the extreme learning machine, f u () is the activation function of the extreme learning machine, W u，j is the weight of the jth hidden neuron of the extreme learning machine, x u，j is the power grid data input into the jth hidden neuron of the extreme learning machine, b u，j is the bias vector of the jth hidden neuron of the extreme learning machine;

[0263] The weight and bias dynamic adjustment unit is used to calculate the error adaptive threshold according to the error between the output of the extreme learning machine hidden layer neuron and the target output, and dynamically adjust the learning rate of the extreme learning machine through the adaptive threshold, thereby dynamically adjusting the weight and bias of the extreme learning machine hidden layer neuron;

[0264] The adaptive threshold of the error is calculated using the following formula:

[0265]

[0266] The learning rate of the extreme learning machine is calculated using the following formula:

[0267]

[0268] The weights and biases of the neurons in the hidden layer of the extreme learning machine are calculated using the following formula:

[0269]

[0270] In the above formula, is the adaptive threshold of the t-th iteration error, β is the smoothing coefficient, std() is the function for calculating the standard deviation, E (t-1) is the error set of all samples in the previous iteration, η(t) is the learning rate of the extreme learning machine in the tth iteration, η0 is the initial learning rate, is the average absolute value of the error at the tth iteration, is the weight of the extreme learning machine at the tth iteration, is the gradient of the loss function L( ) with respect to the weight, L( ) is the loss function of the extreme learning machine, X u , Y u are the input and output of the extreme learning machine respectively, μ is the influence factor of the change of the extreme learning machine, is the weight change of the extreme learning machine between this iteration and the previous iteration, The bias of the extreme learning machine at the tth iteration, is the gradient of the loss function L() with respect to the bias, is the change in bias of the extreme learning machine between this iteration and the previous iteration, Mu is the number of samples input to the current iteration batch of hidden layer neurons in the extreme learning machine, is the cross entropy loss of a single sample, λ is the regularization coefficient of the extreme learning machine, is the regularization term of the extreme learning machine;

[0271] The weight and bias updating unit is used to calculate the gradient of each neuron in the hidden layer of the extreme learning machine through the reversible neuron based on the weight and bias of the neurons in the hidden layer of the extreme learning machine after the dynamic adjustment, and further update the weight and bias of the neurons in the hidden layer of the extreme learning machine:

[0272]

[0273] In the above formula, ← is the parameter update operation, α is the learning rate of the reversible neuron, and δ u，j is the gradient of the jth hidden neuron of the extreme learning machine, is the transpose of the input grid data of the jth hidden neuron of the extreme learning machine, is the predicted label of the extreme learning machine for the jth hidden neuron sample, f′ u () is the derivative of the extreme learning machine activation function, z u，j is the activation function feature vector input to the jth hidden neuron of the extreme learning machine;

[0274] The circulation unit is used to use the updated weights and biases of the hidden layer neurons of the extreme learning machine to return to the output calculation unit to recalculate the outputs of the hidden layer neurons of the extreme learning machine, and define a convergence function of the output power grid data, and determine whether the convergence function is less than a set threshold. If so, it is considered that the power grid stability assessment model has stabilized and the training is terminated. If not, it returns to the weight and bias dynamic adjustment unit to readjust and update the weights and biases of the hidden layer neurons of the extreme learning machine until the convergence function of the output power grid data is less than the set threshold, thereby obtaining a trained power grid stability assessment model;

[0275] The convergence function of the output power grid data is:

[0276]

[0277] In the above formula, C u is the convergence function of the output power grid data.

[0278] Embodiment 3:

[0279] like Figure 3 As shown, a power grid stability assessment device for large-scale electric vehicle access includes a processor and a memory;

[0280] The memory is used to store computer program code and transmit the computer program code to the processor;

[0281] The processor is used to execute the method for evaluating the stability of a power grid with large-scale electric vehicle access as described in Example 1 according to the instructions in the computer program code.

[0282] Embodiment 4:

[0283] A computer storage medium having a computer program stored thereon;

[0284] When the computer program is executed by the processor, the steps of a method for evaluating the stability of a power grid with large-scale electric vehicle access described in this solution are implemented.

Claims

1. A method for evaluating the stability of a power grid with large-scale electric vehicles connected, characterized in that: The method comprises: S1. Collect the grid data set related to electric vehicles in the grid, classify and label it, and obtain the original grid data training set; S2. Based on the generative adversarial network algorithm of Riemannian manifold feature diffusion, the original power grid data training set is expanded, and the data expansion model is trained to obtain the expanded power grid data training set; S3, based on the extreme learning machine algorithm of nonlinear system coupling learning, the expanded power grid data training set is input into the power grid stability assessment model for training, and a trained power grid stability assessment model is obtained; S4. Input new grid data set samples related to electric vehicles in the grid into the trained grid stability assessment model for assessment, and then obtain the assessment level of grid stability.

2. A method for evaluating the stability of a power grid with large-scale electric vehicle access according to claim 1, characterized in that: The S2 includes: S21, initialize the maximum number of iterations of the preset generative adversarial network, and use the original power grid data training set as the quantum ground state set, which contains Ke quantum ground states |φ k >, using quantum computing strategies in the generator input layer of the generative adversarial network to generate a series of initial quantum state power grid data points; The quantum computing strategy is: In the above formula, ψ c is the generated initial quantum state grid data point, Ke is the total number of quantum ground states, α c，k is the complex coefficient of the quantum state, |φ k > is the quantum ground state; S22, generator G c Receiving power grid data points in an initial quantum state, and by performing data diffusion on a Riemann manifold, enabling the power grid data points in the initial quantum state to obtain new power grid data features, thereby generating new power grid data points; The data diffusion method on the Riemann manifold is: In the above formula, is the grid data point of the t-th iteration. The grid data point when t is 0 is the grid data point of the initial quantum state. c is the diffusion step length, For gradient-based The manifold diffusion operator is is the loss function on the Riemann manifold, According to the grid data point The Riemannian metric matrix, μ, is dynamically adjusted for position on the Riemannian manifold. c is the mean of the current iteration batch power grid data, σ c is the scale parameter of the current iteration batch power grid data, and I is the unit matrix; S23, based on the generated new power grid data points, using the loss function to optimize the generator and discriminator of the generative adversarial network, and performing feature correction on the new power grid data points; The constraints of the generator and discriminator of the loss function to optimize the generative adversarial network are: In the above formula, G is the loss function of the generative adversarial network. c is the generator, D c is the discriminator, is the loss of the discriminator, is the loss of the generator, D c (x) is the score of the discriminator on the data points in the original power grid data training set, G c (z) is the power grid data generated by the generator according to the random noise z, and the power grid data generated by the generator according to the random noise z is p data (x) is the distribution of data points in the original power grid data training set, p z (z) is the probability distribution of input noise; The method of performing feature correction on the new power grid data point is: In the above formula, is the corrected power grid data, Proj S () is the operation of projecting the power grid data onto the Riemann manifold, η c is the learning rate, is the gradient of the t+1th iteration, RS() is the regularization function of the correction, and λ c，i To calibrate the intensity parameters, is the i-th grid data point in the t+1-th iteration, is the average value of all grid data points in the t+1th iteration; S24, calculate the quality score of the corrected new power grid data point, and use the discriminator D c Evaluate the similarity between the new power grid data and the real power grid data. When the quality score of the new power grid data point is greater than the preset similarity threshold, the new power grid data point of the generator is retained, otherwise the new power grid data point is discarded; The quality score of the corrected new grid data point is calculated using the following formula: In the above formula, is the quality score of the new power grid data point, d() is the distance measurement function between the means of two power grid data point sets, X real is the real power grid data, |||| is the Euclidean distance; S25, using the adaptive meaning space mapping strategy to the new grid data point G of the generator retained by S24 c (z) Perform meaning adjustment to obtain new power grid data points after meaning adjustment: S c (G c (z))=G c (z)+β c ·tanh(Ω c ·BERT(G c (z))); In the above formula, S c () is the meaning adjustment function to adapt to the meaning space mapping module, β c is the weight parameter, tanh() is the hyperbolic tangent activation function, Ω c is a learnable transformation matrix, BERT (G c (z)) is the deep meaning feature extracted using the pre-trained BERT model; S26, adopting a dynamic meaning perception adjustment mechanism, making a secondary adjustment to the new power grid data point after the meaning adjustment, and dynamically adjusting the generation process of the new power grid data point: In the above formula, X c，adjusted is the new sample after dynamic meaning perception adjustment, that is, the final sample generated, γ c To adjust the weight, softmax() is the Softmax function, Θ c is a learnable transformation matrix, Attn() is the self-attention function, is the before and after data vector of the power grid data; S27, add 1 to the number of iterations, return to step S22, input the second adjusted power grid data points into the generator of the generative adversarial network, repeat the above steps until the preset maximum number of iterations is reached, the data expansion model training is completed, and the generated new power grid data points are superimposed on the original power grid data training set to obtain the expanded power grid data training set.

3. The method for evaluating the stability of a power grid with large-scale electric vehicle access according to claim 1, characterized in that: The S3 includes: S31. Randomly initialize the hidden layer neurons of the extreme learning machine: W u large(N u ,D u ); b u =rande(N u ,1); In the above formula, W u is the weight matrix of the neurons in the hidden layer of the extreme learning machine, N u is the number of neurons in the hidden layer of the extreme learning machine, D u is the dimension of the feature input to the extreme learning machine, b u is the bias vector of the hidden layer neuron of the extreme learning machine, rande(N u , D u ) function represents the generation of an N u ×D u A matrix whose elements are random numbers, rande(N u , 1) The function represents the generation of an N u ×1 vector, where the elements are random numbers; S32, input the expanded power grid data training set into the randomly initialized extreme learning machine hidden layer neurons, and calculate the output of the extreme learning machine hidden layer neurons: y u，j =f u (W u，j ·x u，j +b u，j ); In the above formula, y u，j is the output of the jth hidden neuron of the extreme learning machine, f u () is the activation function of the extreme learning machine, W u，j is the weight of the jth hidden neuron of the extreme learning machine, x u，j is the power grid data input into the jth hidden neuron of the extreme learning machine, b u，j is the bias vector of the jth hidden neuron of the extreme learning machine; S33, calculating an adaptive threshold of the error according to the error between the output of the hidden layer neuron of the extreme learning machine and the target output, and dynamically adjusting the learning rate of the extreme learning machine through the adaptive threshold, thereby dynamically adjusting the weight and bias of the hidden layer neuron of the extreme learning machine; The adaptive threshold of the error is calculated using the following formula: The learning rate of the extreme learning machine is calculated using the following formula: The weights and biases of the neurons in the hidden layer of the extreme learning machine are calculated using the following formula: In the above formula, is the adaptive threshold of the t-th iteration error, β is the smoothing coefficient, std() is the function for calculating the standard deviation, E (t-1) is the error set of all samples in the previous iteration, η(t) is the learning rate of the extreme learning machine in the tth iteration, η0 is the initial learning rate, is the average absolute value of the error at the tth iteration, is the weight of the extreme learning machine at the tth iteration, is the gradient of the loss function L() with respect to the weight, L() is the loss function of the extreme learning machine, X u , Y u are the input and output of the extreme learning machine respectively, μ is the influence factor of the change of the extreme learning machine, is the weight change of the extreme learning machine between this iteration and the previous iteration, The bias of the extreme learning machine at the tth iteration, is the gradient of the loss function L() with respect to the bias, is the change in bias of the extreme learning machine between this iteration and the previous iteration, M u is the number of samples input to the current iteration batch of hidden layer neurons in the extreme learning machine, is the cross entropy loss of a single sample, λ is the regularization coefficient of the extreme learning machine, is the regularization term of the extreme learning machine; S34. Based on the dynamically adjusted weights and biases of the neurons in the hidden layer of the extreme learning machine, the gradient of each neuron in the hidden layer of the extreme learning machine is calculated through the reversible neuron, and the weights and biases of the neurons in the hidden layer of the extreme learning machine are further updated: In the above formula, ← is the parameter update operation, α is the learning rate of the reversible neuron, and δ u，j is the gradient of the jth hidden neuron of the extreme learning machine, is the transpose of the input grid data of the jth hidden neuron of the extreme learning machine, is the predicted label of the extreme learning machine for the jth hidden neuron sample, f′ u () is the derivative of the extreme learning machine activation function, z u，j is the activation function feature vector input to the jth hidden neuron of the extreme learning machine; S35, using the updated weights and biases of the hidden layer neurons of the extreme learning machine, return to S32 to recalculate the outputs of the hidden layer neurons of the extreme learning machine, and define the convergence function of the output power grid data, and determine whether the convergence function is less than the set threshold. If so, it is considered that the power grid stability assessment model has stabilized and the training is terminated. If not, return to S33 to readjust and update the weights and biases of the hidden layer neurons of the extreme learning machine until the convergence function of the output power grid data is less than the set threshold, and a trained power grid stability assessment model is obtained; The convergence function of the output power grid data is: In the above formula, C u is the convergence function of the output power grid data.

4. The method for evaluating the stability of a power grid with large-scale electric vehicle access according to claim 1, characterized in that: In S1, the grid data set related to electric vehicles in the grid includes: data on the use frequency and driving mode of electric vehicles obtained from the transportation department, data on the distribution of electric vehicles obtained from the electric vehicle registration database, and data on the geographical location and operation of charging stations obtained from the charging infrastructure database; The grid is classified and labeled, and the labeled category is the stability assessment level, which is specifically divided into 5 levels, including unstable, less stable, relatively stable, stable, and high stability. The higher the level, the more stable the grid.

5. A grid stability assessment system for large-scale electric vehicle access, characterized in that: The system includes a training set data acquisition module, a power grid data set expansion module, a power grid stability assessment model training module, and a power grid stability level assessment module; The training set data acquisition module is used to collect power grid data sets related to electric vehicles in the power grid, classify and annotate them, and obtain the original power grid data training set; The power grid data set expansion module is used to perform data expansion on the original power grid data training set based on the generative adversarial network algorithm of Riemann manifold feature diffusion, and train the data expansion model to obtain the expanded power grid data training set; The power grid stability assessment model training module is used for an extreme learning machine algorithm based on nonlinear system coupling learning, and the expanded power grid data training set is input into the power grid stability assessment model for training to obtain a trained power grid stability assessment model; The power grid stability level assessment module is used to input new power grid data set samples related to electric vehicles in the power grid into the trained power grid stability assessment model for assessment, thereby obtaining an assessment level of power grid stability.

6. A grid stability assessment system for large-scale electric vehicle access according to claim 5, characterized in that: The power grid data set expansion module includes a quantum state power grid data point generation unit, a new power grid data point generation unit, an optimization and feature correction unit, a quality scoring unit, a meaning adjustment unit, a secondary adjustment unit, and an iteration unit; The quantum state power grid data point generation unit is used to initialize the maximum number of iterations of the preset generative adversarial network, use the original power grid data training set as a quantum ground state set, the ground state set includes Ke quantum ground states |φk>, and use a quantum computing strategy in the generator input layer of the generative adversarial network to generate a series of initial quantum state power grid data points; The quantum computing strategy is: In the above formula, ψ c is the generated initial quantum state grid data point, Ke is the total number of quantum ground states, α c，k is the complex coefficient of the quantum state, |φ k > is the quantum ground state; The new power grid data point generation unit is used to generate G c Receiving power grid data points in an initial quantum state, and by performing data diffusion on a Riemann manifold, enabling the power grid data points in the initial quantum state to obtain new power grid data features, thereby generating new power grid data points; The data diffusion method on the Riemann manifold is: In the above formula, is the grid data point of the t-th iteration. The grid data point when t is 0 is the grid data point of the initial quantum state. c is the diffusion step length, For gradient-based The manifold diffusion operator is is the loss function on the Riemann manifold, According to the grid data point The Riemannian metric matrix, μ, is dynamically adjusted for position on the Riemannian manifold. c is the mean of the current iteration batch power grid data, σ c is the scale parameter of the current iteration batch power grid data, and I is the unit matrix; The optimization and feature correction unit is used to optimize the generator and discriminator of the generative adversarial network based on the generated new power grid data points by using the loss function, and perform feature correction on the new power grid data points; The constraints of the generator and discriminator of the loss function to optimize the generative adversarial network are: In the above formula, G is the loss function of the generative adversarial network. c is the generator, D c is the discriminator, is the loss of the discriminator, is the loss of the generator, D c (x) is the score of the discriminator for the data point x in the original power grid data training set, G c (z) is the power grid data generated by the generator according to the random noise z, and the power grid data generated by the generator according to the random noise z is p data (x) is the distribution of data points in the original power grid data training set, p z (z) is the probability distribution of input noise; The method of performing feature correction on the new power grid data point is: In the above formula, is the corrected power grid data, Projs() is the operation of projecting the power grid data onto the Riemann manifold, η c is the learning rate, is the gradient of the t+1th iteration, RS() is the regularization function of the correction, and λ c，i To calibrate the intensity parameters, is the i-th grid data point in the t+1-th iteration, is the average value of all grid data points in the t+1th iteration; The quality scoring unit is used to calculate the quality score of the corrected new power grid data point, using the discriminator D c Evaluate the similarity between the new power grid data and the real power grid data. When the quality score of the new power grid data point is greater than the preset similarity threshold, retain the generated new power grid data point; otherwise, discard the new power grid data point. The quality score of the corrected new grid data point is calculated using the following formula: In the above formula, is the quality score of the new power grid data point, d() is the distance measurement function between the means of two power grid data point sets, X real is the real power grid data, |||| is the Euclidean distance; The meaning adjustment unit is used to adopt an adaptive meaning space mapping strategy to adjust the new power grid data point G of the generator retained by the quality scoring unit. c (z) Perform meaning adjustment to obtain new power grid data points after meaning adjustment: S c (G c (z))=G c (z)+β c ·tanh(Ω c ·BERT(G c (z))); In the above formula, S c () is the meaning adjustment function to adapt to the meaning space mapping module, β c is the weight parameter, tanh() is the hyperbolic tangent activation function, Ω c is a learnable transformation matrix, BERT (G c (z)) is the deep meaning feature extracted using the pre-trained BERT model; The secondary adjustment unit is used to adopt a dynamic meaning perception adjustment mechanism to perform secondary adjustment on the new power grid data point after the meaning adjustment, and dynamically adjust the generation process of the new power grid data point: In the above formula, X c，adjusted is the new sample after dynamic meaning perception adjustment, that is, the final sample generated, γ c To adjust the weight, soffmax() is the Softmax function, Θ c is a learnable transformation matrix, Attn() is the self-attention function, is the before and after data vector of the power grid data; The iteration unit is used to increase the number of iterations by 1, return the new power grid data point generation unit, input the second adjusted power grid data point into the generator of the generative adversarial network, repeat the above operation steps until the preset maximum number of iterations is reached, the data expansion model training is completed, and the generated new power grid data point is superimposed on the original power grid data training set to obtain the expanded power grid data training set.

7. A grid stability assessment system for large-scale electric vehicle access according to claim 5, characterized in that: The power grid stability assessment model training module includes a random initialization unit, an output calculation unit, a weight and bias dynamic adjustment unit, a weight and bias update unit, and a circulation unit; The random initialization unit is used to randomly initialize the hidden layer neurons of the extreme learning machine: W u large(N u ,D u ); b u =rande(N u ,1); In the above formula, W u is the weight matrix of the neurons in the hidden layer of the extreme learning machine, N u is the number of neurons in the hidden layer of the extreme learning machine, D u is the dimension of the feature input to the extreme learning machine, b u is the bias vector of the hidden layer neuron of the extreme learning machine, rande(N u , D u ) function represents the generation of an N u ×D u A matrix whose elements are random numbers, rande(N u , 1) The function represents the generation of an N u ×1 vector, where the elements are random numbers; The output calculation unit is used to input the expanded power grid data training set into the randomly initialized extreme learning machine hidden layer neurons, and calculate the output of the extreme learning machine hidden layer neurons: y u，j =f u (W u，j ·x u，j +b u，j ); In the above formula, y u，j is the output of the jth hidden neuron of the extreme learning machine, f u () is the activation function of the extreme learning machine, W u，j is the weight of the jth hidden neuron of the extreme learning machine, x u，j is the power grid data input into the jth hidden neuron of the extreme learning machine, b u，j is the bias vector of the jth hidden neuron of the extreme learning machine; The weight and bias dynamic adjustment unit is used to calculate the error adaptive threshold according to the error between the output of the extreme learning machine hidden layer neuron and the target output, and dynamically adjust the learning rate of the extreme learning machine through the adaptive threshold, thereby dynamically adjusting the weight and bias of the extreme learning machine hidden layer neuron; The adaptive threshold of the error is calculated using the following formula: The learning rate of the extreme learning machine is calculated using the following formula: The weights and biases of the neurons in the hidden layer of the extreme learning machine are calculated using the following formula: In the above formula, is the adaptive threshold of the t-th iteration error, β is the smoothing coefficient, std() is the function for calculating the standard deviation, E (t-1) is the error set of all samples in the previous iteration, η(t) is the learning rate of the extreme learning machine in the tth iteration, η0 is the initial learning rate, is the average absolute value of the error at the tth iteration, is the weight of the extreme learning machine at the tth iteration, is the gradient of the loss function L() with respect to the weight, L() is the loss function of the extreme learning machine, X u , Y u are the input and output of the extreme learning machine respectively, μ is the influence factor of the change of the extreme learning machine, is the weight change of the extreme learning machine between this iteration and the previous iteration, The bias of the extreme learning machine at the tth iteration, is the gradient of the loss function L() with respect to the bias, is the change in bias of the extreme learning machine between this iteration and the previous iteration, M u is the number of samples input to the current iteration batch of hidden layer neurons in the extreme learning machine, is the cross entropy loss of a single sample, λ is the regularization coefficient of the extreme learning machine, is the regularization term of the extreme learning machine; The weight and bias updating unit is used to calculate the gradient of each neuron in the hidden layer of the extreme learning machine through the reversible neuron based on the weight and bias of the neurons in the hidden layer of the extreme learning machine after the dynamic adjustment, and further update the weight and bias of the neurons in the hidden layer of the extreme learning machine: In the above formula, ← is the parameter update operation, α is the learning rate of the reversible neuron, and δ u，j is the gradient of the jth hidden neuron of the extreme learning machine, is the transpose of the input grid data of the jth hidden neuron of the extreme learning machine, is the predicted label of the extreme learning machine for the jth hidden neuron sample, f′ u () is the derivative of the extreme learning machine activation function, z u，j is the activation function feature vector input to the jth hidden neuron of the extreme learning machine; The circulation unit is used to use the updated weights and biases of the hidden layer neurons of the extreme learning machine to return to the output calculation unit to recalculate the outputs of the hidden layer neurons of the extreme learning machine, and define a convergence function of the output power grid data, and determine whether the convergence function is less than a set threshold. If so, it is considered that the power grid stability assessment model has stabilized and the training is terminated. If not, it returns to the weight and bias dynamic adjustment unit to readjust and update the weights and biases of the hidden layer neurons of the extreme learning machine until the convergence function of the output power grid data is less than the set threshold, thereby obtaining a trained power grid stability assessment model; The convergence function of the output power grid data is: In the above formula, C u is the convergence function of the output power grid data.

8. A grid stability assessment system for large-scale electric vehicle access according to claim 5, characterized in that: In the training set data acquisition module, the grid data set related to electric vehicles in the grid includes: data on the use frequency and driving mode of electric vehicles obtained from the transportation department, data on the distribution of electric vehicles obtained from the electric vehicle registration database, and data on the geographical location and operation of charging stations obtained from the charging infrastructure database; The grid is classified and labeled, and the labeled category is the stability assessment level, which is specifically divided into 5 levels, including unstable, less stable, relatively stable, stable, and high stability. The higher the level, the more stable the grid.

9. A device for evaluating the stability of a power grid for large-scale electric vehicle access, characterized in that: including a processor and a memory; The memory is used to store computer program code and transmit the computer program code to the processor; The processor is used to execute the grid stability assessment method for large-scale electric vehicle access as described in any one of claims 1-4 according to the instructions in the computer program code.

10. A computer storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of a method for evaluating the stability of a power grid with large-scale electric vehicle access as described in any one of claims 1-4 are implemented.

Images