Power grid prospective risk assessment method and device and medium

Abstract

The invention discloses a power grid prospective risk assessment method and device and a medium, relates to the technical field of power grid dispatching optimization, and solves the problems of unstable wind speed, low fan efficiency, difficult power grid dispatching and poor safety of wind power generation in the prior art. According to the method, the target fan power time sequence is constructed according to the fan power data of the historical time period, then the extreme learning machine model, the long-short term memory neural network model and the time convolutional network model are trained, and then the prediction result of the short-term wind power is obtained, so that the accuracy and the flexibility of the prediction result of the short-term wind power are improved. In the off-line stage of the power grid, a generative adversarial network model is constructed according to the power grid section data set, and at least one power grid prospective scene sequence is generated; in the online operation stage of the power grid, the prospective operation scene sequence of the power grid is obtained according to the preset risk assessment index and the current operation scene information of the power grid, and the accuracy and timeliness of operation risk assessment of the power grid are enhanced.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid dispatch optimization, and in particular to a power grid forward-looking risk assessment method, device and medium. Background Art

[0002] In recent years, with the rapid development of the whole network economy and power grid construction, the power system, especially the ultra-high voltage AC and DC projects, has been continuously put into operation, and the main grid of 220kV and above has been continuously strengthened, resulting in an increasingly complex grid structure. In addition, as a clean and renewable energy source, the development and utilization of wind energy is of great significance to promoting energy structure adjustment and reducing environmental pollution. However, the large-scale access of wind power to the grid increases the risk of grid operation. Therefore, it is necessary to improve the accuracy and reliability of wind power forecasting, and then perceive and evaluate the forward-looking operation risks of the grid.

[0003] At present, a single prediction model is often used to predict short-term wind power. However, a single prediction model often fails to take into account the complex interactions between various factors, and the accuracy and flexibility of the prediction results for short-term wind power are low. The research methods and business application technologies for grid operation risk perception and identification are mainly based on risk indicators for section assessment and plan control, and a grid fault outage model is constructed. The risk transfer process is characterized by flow tracking, and the risk node set is optimized and solved by section risk indicators. However, in operation scenarios such as extreme weather and the coupling of multiple risks in the grid, the accuracy and timeliness of the grid operation risk assessment need to be enhanced.

[0004] Therefore, there is an urgent need for a power grid operation risk assessment method, device, computer equipment and storage medium to improve the accuracy of short-term wind power forecast results and enhance the accuracy and timeliness of power grid operation risk assessment. Summary of the invention

[0005] In view of the problems that wind power generation in the prior art still has, such as unstable wind speed, low wind turbine efficiency, difficult grid dispatching and poor safety, the present invention provides a grid forward-looking risk assessment method, device and medium, which can enhance the accuracy and timeliness of grid operation risk assessment. The specific technical solution is as follows:

[0006] A method for forward-looking risk assessment of a power grid comprises the following steps:

[0007] Obtain power grid operation status data, meteorological monitoring data and wind turbine power data for historical periods; construct a wind turbine power time series based on the wind turbine power data for historical periods;

[0008] Obtain missing data in the wind turbine power time series, and fill the missing data through a piecewise linear regression model to obtain a target wind turbine power time series;

[0009] According to the target wind turbine power time series, respectively training an extreme learning machine model, a long short-term memory neural network model and a temporal convolutional network model;

[0010] Training a multi-objective prediction model according to the extreme learning machine model, the long short-term memory neural network model and the temporal convolutional network model;

[0011] Using a multi-objective Harris Hawk algorithm, obtaining a Pareto optimal solution set of a combined weight vector of the multi-objective prediction model;

[0012] Using a deep deterministic policy gradient method, according to the Pareto optimal solution set, to obtain the dynamic optimal weight of each model in the multi-objective prediction model;

[0013] According to each model in the multi-objective prediction model and the corresponding dynamic optimal weight, a prediction result of wind power is obtained.

[0014] Generate a power grid section data set according to the power grid operation status data, the meteorological monitoring data and the wind power prediction result;

[0015] In the offline stage of the power grid, a generative adversarial network model is constructed according to the power grid section data set;

[0016] Obtaining at least one type of power grid operation risk data through the generative adversarial network model;

[0017] Generate a grid operation risk scenario set according to the at least one type of grid operation risk data, wherein the grid operation risk scenario set includes at least one grid operation risk scenario;

[0018] Using a time series operation simulation method, at least one power grid operation risk scenario is simulated and calculated to generate at least one power grid forward-looking scenario sequence;

[0019] During the online operation phase of the power grid, obtain the preset risk assessment indicators and current power grid operation scenario information;

[0020] According to the preset risk assessment index and the current power grid operation scenario information, a power grid operation risk assessment result is obtained, and a corresponding power grid forward-looking operation scenario sequence is matched and obtained.

[0021] Preferably, the obtaining of missing value data in the wind turbine power time series includes:

[0022] The wind turbine power data in the wind turbine power time series that meets the missing value preset condition is determined as the missing value data; the missing value preset condition includes that the value of the wind turbine power data is zero, and the change value of the wind turbine power data relative to the fan power data of the previous time series is greater than a preset threshold.

[0023] Preferably, before filling the missing data by the piecewise linear regression model, the method further comprises:

[0024] Interpolating and filling the missing data to obtain an intermediate wind turbine power time series;

[0025] An initial piecewise linear regression model is trained according to the intermediate wind turbine power time series to optimize the parameters of the initial piecewise linear regression model until the trained piecewise linear regression model is obtained.

[0026] Preferably, the training of the initial piecewise linear regression model according to the intermediate wind turbine power time series includes:

[0027] Obtaining exponentially weighted moving average characteristics and moving median characteristics of the intermediate wind turbine power time series;

[0028] Processing the intermediate wind turbine power time series according to the BiLSTM model to obtain meta-features of the intermediate wind turbine power time series;

[0029] The initial piecewise linear regression model is trained according to the meta-feature, the exponentially weighted moving average feature and the moving median feature to optimize the parameters of the initial piecewise linear regression model until the trained piecewise linear regression model is obtained.

[0030] Preferably, according to the target wind turbine power time series, an extreme learning machine model, a long short-term memory neural network model and a time convolutional network model are trained respectively, including:

[0031] Using a sequence decomposition method CEEMDAN, the time series data of the historical wind power is decomposed to obtain a plurality of first subsequence data;

[0032] The extreme learning machine model, the long short-term memory neural network model and the temporal convolutional network model are trained simultaneously through the multiple first subsequence data.

[0033] Preferably, according to each model in the multi-objective prediction model and the corresponding dynamic optimal weight, a prediction result of short-term wind power is obtained, including:

[0034] Obtain time series data of target wind power;

[0035] Using a sequence decomposition method CEEMDAN, the time series data of the target wind power is decomposed to obtain a plurality of second subsequence data;

[0036] By using each model in the multi-objective prediction model, the plurality of second subsequence data are traversed and predicted at the same time to obtain corresponding first prediction results, second prediction results and third prediction results;

[0037] According to the dynamic optimal weights corresponding to each model, weighted calculation is performed on the first prediction result, the second prediction result and the third prediction result to obtain a prediction result of short-term wind power.

[0038] Preferably, training a multi-objective prediction model according to the extreme learning machine model, the long short-term memory neural network model and the temporal convolutional network model includes:

[0039] Obtain time series data of test wind power;

[0040] Using the sequence decomposition method CEEMDAN, the time series data of the test wind power is decomposed to obtain a plurality of third subsequence data;

[0041] According to the extreme learning machine model, the long short-term memory neural network model and the temporal convolutional network model, the plurality of third subsequence data are traversed and tested at the same time to obtain corresponding first test results, second test results and third test results;

[0042] Obtaining mean square errors and standard deviations between the actual short-term wind power and the first test result, the second test result, and the third test result, respectively;

[0043] The minimum mean square error and the minimum standard deviation are used as optimization objectives to train the multi-objective prediction model.

[0044] Preferably, iteratively updating the weights and biases of the multi-objective prediction model through the action vector of the agent comprises:

[0045] Selecting a static weight coefficient from the Pareto optimal solution set based on the minimum mean absolute percentage error principle through the action vector of the agent;

[0046] Obtaining a reference error according to the static weight coefficient and the actual short-term wind power;

[0047] Analyzing the correlation between the minimum value of the mean square error, the minimum value of the standard deviation, and the reference error;

[0048] According to the correlation, obtaining a reward function;

[0049] The weights and biases of the multi-objective prediction model are iteratively updated through the reward function.

[0050] A power grid operation risk assessment device, comprising:

[0051] Data acquisition module, used to obtain power grid operation status data and meteorological monitoring data;

[0052] A data integration module, used to generate a power grid section data set according to the power grid operation status data and the meteorological monitoring data;

[0053] A modeling module is used to construct a generative adversarial network model according to the power grid section data set during the offline phase of the power grid; and obtain at least one type of power grid operation risk data through the generative adversarial network model;

[0054] A processing module, configured to generate a set of power grid operation risk scenarios according to the at least one type of power grid operation risk data, wherein the set of power grid operation risk scenarios includes at least one power grid operation risk scenario;

[0055] A deduction and evaluation module is used to use a time series operation simulation method to perform deduction and calculation on at least one power grid operation risk scenario and generate at least one power grid forward-looking scenario sequence;

[0056] The data acquisition module is also used to obtain preset risk assessment indicators and current grid operation scenario information during the online operation stage of the grid;

[0057] The online application module is used to obtain the power grid operation risk assessment result according to the preset risk assessment index and the current power grid operation scenario information, and match and obtain the corresponding power grid forward-looking operation scenario sequence.

[0058] A computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the power grid forward-looking risk assessment method as described above.

[0059] Compared with the prior art, the present invention has the following beneficial effects:

[0060] The present invention constructs a wind turbine power time series based on wind turbine power data of historical periods; obtains missing data in the wind turbine power time series, and fills the missing data through a piecewise linear regression model to obtain a target wind turbine power time series; then trains an extreme learning machine model, a long short-term memory neural network model, and a time convolutional network model; then trains a multi-objective prediction model; adopts a multi-objective Harris Eagle algorithm to obtain the Pareto optimal solution set of the combined weight vector of the multi-objective prediction model; adopts a deep deterministic policy gradient method to obtain the dynamic optimal weights of each model in the multi-objective prediction model according to the Pareto optimal solution set; obtains the prediction result of short-term wind power according to each model in the multi-objective prediction model and the corresponding dynamic optimal weights. It can improve the accuracy and flexibility of the prediction result of short-term wind power. In the offline stage of the power grid, a generative adversarial network model is constructed based on the power grid section data set; at least one type of power grid operation risk data is obtained through the generative adversarial network model to generate a set of power grid operation risk scenarios; a time series operation simulation method is used to deduce and calculate at least one power grid operation risk scenario to generate at least one power grid forward-looking scenario sequence; in the online operation stage of the power grid, the power grid operation risk assessment results are obtained based on the preset risk assessment indicators and the current power grid operation scenario information, and the corresponding power grid forward-looking operation scenario sequence is matched. It can enhance the accuracy and timeliness of the power grid operation risk assessment. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following is a brief introduction to the drawings required for the specific embodiments or the description of the prior art. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn according to the actual scale.

[0062] Figure 1 An application environment diagram of a power grid operation risk assessment method in one embodiment;

[0063] Figure 2 A schematic diagram of a flow chart of a method for assessing power grid operation risk in one embodiment;

[0064] Figure 3 A schematic flow chart of a method for assessing power grid operation risk in another embodiment;

[0065] Figure 4 is a flow chart of a method for assessing power grid operation risk in the most detailed embodiment;

[0066] Figure 5 A schematic diagram of IEEE39 node power grid risk assessment in an embodiment of an embodiment;

[0067] Figure 6is a structural block diagram of a power grid operation risk assessment device in one embodiment;

[0068] Figure 7 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0069] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0070] It should be understood that when used in this specification and the appended claims, the terms "include" and "comprises" indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0071] It should also be understood that the terms used in the present specification are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the present specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0072] It should be further understood that the term "and / or" used in the present description and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0073] The wind power prediction method provided in the embodiment of the present application can be applied to Figure 1 In the application environment shown, the terminal 102 communicates with the server 104 through a network. The data storage system can store data that the server 104 needs to process. The data storage system can be integrated on the server 104, or placed on the cloud or other network servers.

[0074] Server 104 obtains the time series data of power grid operation status data, meteorological monitoring data and historical wind power; trains the extreme learning machine model, the long short-term memory neural network model and the time convolution network model respectively according to the time series data of historical wind power; trains the multi-objective prediction model according to the extreme learning machine model, the long short-term memory neural network model and the time convolution network model; adopts the multi-objective Harris Eagle algorithm to obtain the Pareto optimal solution set of the combined weight vector of the multi-objective prediction model; adopts the deep deterministic policy gradient method to obtain the dynamic optimal weights of each model in the multi-objective prediction model according to the Pareto optimal solution set; obtains the prediction result of wind power according to each model in the multi-objective prediction model and the corresponding dynamic optimal weights. A power grid section data set is generated based on power grid operation status data, meteorological monitoring data and wind power prediction results; in the offline stage of the power grid, a generative adversarial network model is constructed based on the power grid section data set; at least one type of power grid operation risk data is obtained through the generative adversarial network model; based on at least one type of power grid operation risk data, a power grid operation risk scenario set is generated, and the power grid operation risk scenario set includes at least one power grid operation risk scenario; using a time series operation simulation method, at least one power grid operation risk scenario is deduced and calculated to generate at least one power grid forward-looking scenario sequence; in the online operation stage of the power grid, preset risk assessment indicators and current power grid operation scenario information are obtained; based on the preset risk assessment indicators and the current power grid operation scenario information, a power grid operation risk assessment result is obtained, and the corresponding power grid forward-looking operation scenario sequence is obtained by matching.

[0075] The terminal 102 may be, but is not limited to, various personal computers, laptops, smart phones, tablet computers, IoT devices, and portable wearable devices. The IoT devices may be smart speakers, smart TVs, smart air conditioners, smart car-mounted devices, etc. The portable wearable devices may be smart watches, smart bracelets, head-mounted devices, etc. The server 104 may be implemented as an independent server or a server cluster consisting of multiple servers.

[0076] In an exemplary embodiment, Figure 2 As shown in the figure, a method for evaluating the risk of power grid operation is provided. Figure 1 The server in the example is used to illustrate, including the following steps S202 to S228. Among them:

[0077] Step S202: construct a wind turbine power time series based on wind turbine power data in a historical period.

[0078] The wind turbine power data is the power generated by the wind turbines in the wind farm, and can also be referred to as wind power output data. In some embodiments, the wind turbine power data is collected at preset time intervals during the operation of the wind farm, for example, the wind turbine power data is collected every 15 minutes to construct a wind turbine power time series.

[0079] In some embodiments, the constructed wind turbine power time series can be expressed as {P T , P 2 , ..., P n}, P is the power value, n represents the number of elements in the sequence, for example, a wind turbine power time series can be {100, 102, 98, 100, 98, ...}. Since the power of each wind turbine in the wind turbine power time series is collected at a preset time interval, taking the collection every 15 minutes as an example, the wind turbine power time series can be a corresponding time feature sequence {T 1 , T 2 , ..., T n}, P n With T n Corresponding to, for example, {2021-01-0100:00, 2021-01-0100:15, 2021-01-0100:30, …}.

[0080] Step S204, obtaining missing data in the wind turbine power time series, and filling the missing data through a piecewise linear regression model to obtain a target wind turbine power time series;

[0081] Among them, missing data refers to data that was not collected at the time of collection. It is understandable that the data may be interfered by external factors during collection, or abnormal situations such as sudden power outages or shutdowns may occur. Therefore, there may be some missing data in the wind turbine power time series, which need to be automatically found and completed. Since the missing data is generally saved as 0, it may be confused with the normal 0 value of wind turbine power data during the actual search process, and it needs to be accurately distinguished.

[0082] In some embodiments, the piecewise linear regression model can be used to capture local features and regularities in the target wind turbine power time series. It can be understood that the piecewise linear regression model is not just a simple linear regression of the entire wind turbine power time series, but can also divide the time series into different segments according to the local characteristics of the data, and perform linear regression in each segment. This processing method makes the model more adaptable to changes and fluctuations in the data, so that the target wind turbine power can be predicted more accurately.

[0083] Through the above method, the piecewise linear regression model is used to fill the missing data and obtain the complete target wind turbine power time series. The data filling process helps to retain the local characteristics in the data, thereby improving the accuracy of filling the missing values ​​and making the final time series more complete and reliable.

[0084] Step S206: training an extreme learning machine model, a long short-term memory neural network model and a temporal convolutional network model respectively according to the time series data of historical wind power.

[0085] Specifically, based on the time series data of historical wind power, the following are the general steps for training wind power prediction using three models: Extreme Learning Machine (ELM), Long Short-Term Memory (LSTM) and Time Convolutional Neural Network (TCN).

[0086] Extreme Learning Machine (ELM) Model:

[0087] Prepare the data set: Divide the historical wind power time series data into training set and test set, and normalize the data.

[0088] Constructing the ELM model: ELM is a single hidden layer feedforward neural network with randomly initialized weight matrix and bias vector. The wind power data is input into the input layer, and the hidden layer output is calculated through the random weight matrix and bias vector.

[0089] Training model: Adjust the hidden layer weight matrix and bias vector through least squares method or regularization to make the model output as close to the actual wind power value as possible.

[0090] Model evaluation: Use the test set data to evaluate the prediction performance of the model, such as calculating indicators such as the root mean square error (RMSE).

[0091] Long Short-Term Memory Neural Network (LSTM) Model:

[0092] Prepare the data set: Convert the historical wind power time series data into input series and target series, and normalize the data.

[0093] Build an LSTM model: LSTM is a recurrent neural network suitable for sequence data. It has memory units and gating mechanisms to capture long-term dependencies in sequences. The model includes an input layer, an LSTM layer, and an output layer.

[0094] Training model: Through the back-propagation algorithm and gradient descent optimizer, the weights and biases of the model are adjusted according to the input sequence and the target sequence, so that the model can accurately predict wind power.

[0095] Model evaluation: Use the test set data to evaluate the prediction performance of the model, such as calculating indicators such as the root mean square error (RMSE).

[0096] Temporal Convolutional Network (TCN) Model:

[0097] Prepare the dataset: Convert the historical wind power time series data into input series and target series, and perform normalization.

[0098] Build TCN model: TCN is a sequence modeling method based on convolutional neural network, which captures long-term dependencies in the sequence through one-dimensional convolutional layers and residual blocks. The model includes input layer, convolutional layer, residual block and output layer.

[0099] Training model: Through the back-propagation algorithm and gradient descent optimizer, the weights and biases of the model are adjusted according to the input sequence and the target sequence, so that the model can accurately predict wind power.

[0100] When training the extreme learning machine model, the long short-term memory neural network model, and the temporal convolutional network model, some parameters of each model will be set based on empirical values, such as:

[0101] ELM model: The number of neurons in the hidden layer is set to 16;

[0102] LSTM model; the number of neurons in the hidden layer is set to 100; the learning rate is set to 0.001; the maximum number of rounds is set to 100; the number of convolution kernels is set to 64;

[0103] TCN model: the number of convolution kernels is set to 64; the number of training batches is set to 32; the number of iterations is set to 50; and the inactivation rate is set to 0.2.

[0104] Step S208, training a multi-target prediction model based on the extreme learning machine model, the long short-term memory neural network model and the temporal convolutional network model.

[0105] Specifically, the entire multi-objective model is trained using back propagation and optimization algorithms, with the goal of minimizing the model's error on the target variable. The model can be trained using gradient descent or other optimization algorithms. The trained multi-objective prediction model is evaluated using a test data set. The evaluation metric can be selected based on the type of target variable, such as root mean square error (RMSE) or mean absolute error (MAE). Once the model is trained and evaluated, it can be used to predict new target variable values.

[0106] Among them, the expression of the multi-objective prediction model is:

[0107]

[0108] In the formula, y i is the measured power at time i, Represents the prediction of a single model at time i, w i,1 ,w i,2 ,w i,3 is the combined weight at time i.

[0109] Step S210, using the multi-objective Harris Hawk algorithm to obtain the Pareto optimal solution set of the combined weight vector of the multi-objective prediction model.

[0110] Specifically, the multi-objective Harris Hawk algorithm is a commonly used method for solving multi-objective optimization problems. The multi-objective prediction task is transformed into a multi-objective optimization problem, in which each target variable can be obtained by different prediction models, and an objective function can be defined to measure the optimization target of each target. This objective function may include indicators such as prediction error, generalization ability, and complexity. Determine the feasible domain of the Pareto optimal solution set. The feasible domain determines the value range of the weight vector, and constraints can be set according to actual needs. A group of eagles is generated using a random initialization method. The individual eagle represents a possible weight vector. Through continuous iteration and optimization, the group gradually converges to the optimal solution. Calculate the fitness of each eagle individual, and the fitness will evaluate the quality of each individual according to the value of the objective function, usually based on the definition of the Pareto optimal solution. Update the position of each eagle individual according to the selected update strategy. The update strategy can adopt the search strategy in the standard Harris Hawk algorithm, such as exploration and tracking, population search, and encounter and fusion. Check whether the eagle group converges to a stable Pareto optimal solution set. If the stopping condition is not met, continue iterative optimization. The final stable group is the Pareto optimal solution set, which contains the optimal solutions of multiple weight vector combinations.

[0111] Among them, the following multi-objective optimization is defined:

[0112] minf(x)=[f 1 (x),f 2 (x),...f n (x)]

[0113]

[0114] where g i (x), h j (x) are all constraints. Find the minimum value of f(x) under the conditions of satisfying g(x) and h(x).

[0115] Given a vector and

[0116] If and only if When [f(x i )≥f(y i )]∧[i∈1,2,...,k:f(x i )];

[0117] and is a Pareto optimal solution if The set of Pareto optimal solutions is recorded as

[0118] Furthermore, P f ={f(x)|x∈P set All Pareto optimal solutions constitute the Pareto optimal solution set. These solutions, through the mapping of the objective function, constitute the Pareto optimal frontier or Pareto frontier surface of the problem. That is, the objective function value corresponding to the Pareto optimal solution is the Pareto optimal frontier.

[0119] Step S212, using a deep deterministic policy gradient method, obtains the dynamic optimal weight of each model in the multi-objective prediction model according to the Pareto optimal solution set.

[0120] Among them, Deep Deterministic Policy Gradient (DDPG) is a commonly used reinforcement learning algorithm that can be used to learn dynamic optimal weights.

[0121] Specifically, the state space and action space of the problem are defined. In the multi-objective prediction model, the state space can be indicators such as the prediction error and generalization ability of each prediction model, and the action space can be each weight in the weight vector. Construct a DDPG network consisting of two neural networks, one is the Actor network for learning the action strategy, and the other is the Critic network for evaluating the value of the action. The Actor network outputs the action vector of each weight in the current state, and the Critic network outputs a Q value to evaluate the value of the action based on the Actor's action vector and state information. Randomly initialize the parameters of the Actor network and the Critic network. According to the DDPG strategy, use the Actor network to select and execute actions to obtain reward signals and the next state. Use these samples to update the parameters of the Critic network, and use the gradient descent method to minimize the loss function of the Critic network. At the same time, update the parameters of the Actor network according to the output of the Critic network, and use the gradient ascent method to maximize the objective function of the Actor network. Repeat sampling, optimization, and updating of network parameters until the network converges or reaches the preset number of iterations. After the DDPG training is completed, the dynamic optimal weights of each model in the multi-objective prediction model can be obtained through the Actor network. Use the Actor network to input the current state and get the output weight vector as the weight of the model.

[0122] Step S214, obtaining the wind power prediction result according to each model in the multi-objective prediction model and the corresponding dynamic optimal weight.

[0123] According to the current wind input and the characteristics of the prediction model, use each prediction model to obtain its own prediction results. Apply the corresponding dynamic optimal weight to the prediction results of each prediction model. Multiply each prediction result with the corresponding dynamic optimal weight to obtain a weighted prediction result. Add the weighted prediction results to obtain the final wind power prediction result.

[0124] In the above-mentioned wind power prediction method, by adopting a combination of multiple prediction models for prediction, the complex interactions between different models can be taken into account, thereby improving the accuracy of short-term wind power prediction results. By using multiple prediction models and adjusting the dynamic optimal weights, the weights of each model can be flexibly adjusted according to actual conditions, thereby improving the adaptability and flexibility of the prediction method. By using the multi-objective Harris Eagle algorithm to obtain the Pareto optimal solution set of the combined weight vector of the prediction model, multiple prediction strategies can be provided for selection while ensuring the accuracy of the prediction. By using the deep deterministic policy gradient method, the optimal weights of each model can be dynamically adjusted according to the real-time Pareto optimal solution set, further improving the accuracy and adaptability of the prediction. This method can improve the accuracy and flexibility of short-term wind power prediction results, and provide important support for wind energy utilization and the stability of the power system.

[0125] In an exemplary embodiment, according to the time series data of historical wind power, an extreme learning machine model, a long short-term memory neural network model and a time convolutional network model are trained respectively, including:

[0126] The sequence decomposition method CEEMDAN is used to decompose the time series data of historical wind power to obtain multiple first subsequence data. The extreme learning machine model, the long short-term memory neural network model and the temporal convolutional network model are trained simultaneously through multiple second subsequence data.

[0127] Specifically, the sequence decomposition method CEEMDAN (Complete Ensemble Empirical Mode Decomposition with Adaptive Noise) is used to decompose the time series data of historical wind power to obtain multiple first subsequence data. CEEMDAN can effectively decompose the original sequence into multiple subsequences, each of which represents a different frequency component or mode. The multiple first subsequence data obtained by decomposition are used as input, and the corresponding training set, validation set, and test set are set according to the needs to ensure that the data division is reasonable.

[0128] Compared with traditional decomposition methods (EMD, VMD and WPD), CEEMDAN overcomes the problem of modal aliasing in empirical mode decomposition and the problem of random residual noise in aggregated empirical mode decomposition. Gaussian white noise is added at each stage of decomposition before EMD decomposition, which solves the problem of residual noise and can fully decompose time series with complex changing characteristics such as wind power.

[0129] In this embodiment, the ELM model, LSTM model and TCN model are trained respectively according to the multiple first subsequence data decomposed by the CEEMDAN method, so that the most suitable model can be selected according to the needs, and the wind power can be predicted according to the historical wind power data.

[0130] Step S216, generating a power grid section data set according to the power grid operation status data, meteorological monitoring data and wind power prediction results.

[0131] Specifically, the collected power grid operation status data and meteorological monitoring data are preprocessed and integrated, including data cleaning, denoising, correction, etc. Ensure the accuracy and consistency of the data. Select specific section ranges and parameters according to the specific purpose of power grid analysis and assessment. These parameters can be power grid components or specific power grid characteristics in a specific area, such as transmission line length, substation capacity, etc. Utilize the preprocessed and integrated data to generate a power grid section data set, which is then labeled and classified. Labeling can be based on indicators such as power grid operation risk level and failure probability for subsequent risk assessment and analysis. Next, the generated power grid section data set is verified and validated to ensure the integrity and accuracy of the data. This can include comparison and comparative analysis with actual observed data.

[0132] The power grid section data set described in step S202 is specifically represented as follows:

[0133] N=[n 1 ,n 2 ,...n i ,...,n m ];

[0134]

[0135] Where n i is the characteristic information vector of the i-th section, s mi is the mth electrical measurement information of the ith section, w mi is the mth meteorological measurement information of the ith section.

[0136] Step S218: During the offline phase of the power grid, a generative adversarial network model is constructed based on the power grid section data set.

[0137] Specifically, in the offline stage of the power grid, building a generative adversarial network (GAN) model based on the power grid section dataset means utilizing the generation ability of the GAN model to train the GAN model using the power grid section dataset as input, so that it can learn and generate virtual power grid section data with similar characteristics.

[0138] Generative adversarial networks are a neural network model consisting of a generator and a discriminator. Through adversarial training, the generator can gradually generate virtual data similar to real data, while the discriminator attempts to accurately distinguish between real data and generated data. This adversarial training gradually improves the performance of the generator, and the generated data becomes closer and closer to the distribution of real data.

[0139] Based on the construction of the power grid section data set, the GAN model can be used to generate more virtual power grid section data. The GAN model can simulate a new synthetic data set by learning the association and distribution between power grid section data. In order to train this GAN model, it is necessary to define a suitable loss function to evaluate the difference between the data generated by the generator and the real data, and update the parameters of the model through the back propagation algorithm.

[0140] The constructed GAN model can be used to generate virtual grid section data similar to the grid section data set. Such virtual data can be used to supplement the deficiencies of the existing data set, expand the scale of the training set, and improve the generalization ability of the model. At the same time, the generated virtual data can also be used to simulate various risk scenarios in the operation of the power grid for risk assessment and decision support.

[0141] By building a generative adversarial network model and training it with a power grid section data set, we can obtain a model that can generate virtual power grid section data similar to real power grid data. This model can learn the distribution and characteristics of power grid data and generate data with similar characteristics and statistical properties, providing more data resources for power grid operation risk assessment and related applications.

[0142] Step S220: obtaining at least one type of power grid operation risk data by generating an adversarial network model.

[0143] Specifically, a batch of new grid operation risk data is generated using the trained GAN model. The generated data will have similar characteristics and distribution as the training data. Feature extraction is performed on the generated data and the original data to obtain a high-dimensional risk feature vector. Various feature extraction methods can be used, such as principal component analysis (PCA) or other dimensionality reduction techniques.

[0144] Step S222: Generate a grid operation risk scenario set based on at least one type of grid operation risk data, wherein the grid operation risk scenario set includes at least one grid operation risk scenario, and calculate and screen high-dimensional risk features using a K-Means clustering method.

[0145] Specifically, a common scenario for understanding grid operation risks is power load fluctuations. During grid operation, load fluctuations may cause power system overload, voltage fluctuations, and other potential risks. The following is an example scenario set that includes grid operation risk scenarios caused by load fluctuations:

[0146] Scenario 1: Voltage drop during high load period

[0147] Description: During high load periods, the power system faces huge load demands that exceed supply capacity. This causes the supply voltage to drop, which may cause user equipment failure or power outage.

[0148] Characteristics: High load, supply voltage drop, potential user equipment failure or risk of power outage.

[0149] Solutions: Increase power generation capacity, improve power system dispatch and management strategies, and optimize power supply networks.

[0150] Scenario 2: Sudden load increase

[0151] Description: Due to certain events (such as sudden weather, equipment failure, etc.), the power system suffers a sudden increase in load, which exceeds the capacity of the system and may cause overload and voltage drop.

[0152] Characteristics: sudden load increase, power system capacity overload, potential overload and voltage drop risk.

[0153] Solution: Implement load management strategies, respond to load changes in a timely manner, and improve the dispatchability of power supply and distribution equipment.

[0154] Scenario 3: Unbalanced load distribution

[0155] Description: Due to uneven load distribution or power supply system failure, some areas or lines carry too much load, which may cause some areas of the power grid to be overloaded while other parts are underloaded.

[0156] Characteristics: Unbalanced load distribution, overloaded areas or lines, and light loads in other areas.

[0157] Solution: Optimize load distribution, improve power supply network planning, and expand transmission and distribution capabilities in some areas.

[0158] These risk scenarios provide examples of the risks that may be caused by load fluctuations in grid operation and their potential solutions. Note that actual grid risk scenarios may be more complex and need to be determined based on specific grid data and conditions.

[0159] The extracted high-dimensional risk feature vector is input into the K-Means clustering algorithm to divide the risk data into different clusters. K-Means is a commonly used unsupervised learning algorithm that can automatically cluster similar data points together. Calculating and screening high-dimensional risk features through the K-Means clustering method can help identify and analyze risk patterns and abnormal situations in power grid operation scenarios. This can provide valuable information for power system managers so that they can take appropriate measures to address potential risk issues.

[0160] Step S224, using a time series operation simulation method to deduce and calculate at least one power grid operation risk scenario, and generate at least one power grid forward-looking scenario sequence.

[0161] Specifically, the time series operation simulation method is a commonly used power grid risk assessment tool. By simulating the operation of the power grid, the forward-looking scenario sequence of the power grid is deduced and calculated. First, one or more power grid operation risk scenarios are selected, such as load surge, transmission line failure, etc., and the parameters required for the simulation are determined, such as the probability of failure and repair time. Then, the power grid operation simulation calculation is performed according to the determined risk scenarios and parameters through the time series operation simulation method. The simulation can take into account the changes in load, the start and stop of the generator set, the occurrence and repair of failures, etc. According to the results of the simulation calculation, the state of the power grid at different time points is obtained, and the information on the reliability, stability and vulnerability of the power grid can be analyzed. Possible problems can be identified and the future operation of the power grid can be predicted. Based on the analysis results, a forward-looking scenario sequence of the power grid is generated. These sequences can help power grid operators formulate reasonable operation strategies to deal with possible risk scenarios in the future.

[0162] Step S226, during the online operation phase of the power grid, obtaining preset risk assessment indicators and current power grid operation scenario information.

[0163] Specifically, during the online operation phase of the power grid, real-time power grid operation data, including parameters such as power load, generator output, voltage and current, are obtained through monitoring equipment such as sensors and monitoring devices. These data can be used to evaluate the operation status and performance of the power grid. Relevant environmental data, such as weather data, temperature, humidity, etc., are obtained. These data can affect the operation of the power grid, such as the intensity of wind in wind power generation scenarios. According to the safety and reliability requirements of the power grid, a series of risk assessment indicators are formulated. These indicators can be used to assess the risk level of the power grid in different scenarios, such as voltage stability, frequency stability, capacity utilization, etc. These indicators can be set and modified according to actual conditions. The operation of the power grid is analyzed and reviewed using historical data and fault records to obtain past power grid operation scenario information. This information can be used to improve the accuracy of risk assessment and identify possible problems and risks. According to the preset risk assessment indicators and the collected power grid operation scenario information, the corresponding risk assessment models and algorithms are used to calculate and evaluate the risks of the power grid. These models and algorithms can be selected and adjusted according to actual needs and scenarios to achieve accurate risk assessment.

[0164] Step S228, obtaining the grid operation risk assessment result according to the preset risk assessment index and the current grid operation scenario information, and matching and obtaining the corresponding grid forward-looking operation scenario sequence.

[0165] Specifically, the operation risk of the power grid is assessed by using the collected grid operation scenario information and preset risk assessment indicators, and applying the corresponding risk assessment models and algorithms. The current risk level of the power grid is determined based on the preset indicators and assessment results. According to the risk assessment results, the corresponding grid forward-looking operation scenario sequence is matched. These sequences can be pre-planned strategies and measures to deal with grid operation conditions at different risk levels. For example, in the case of a sudden increase in power load, there can be corresponding generator start-up and stop strategies; in the case of a transmission line failure, there can be corresponding line switching and repair strategies. According to the grid forward-looking operation scenario sequence, the operation strategy of the power grid is further analyzed and optimized. This can include adjusting the output of the generator set, optimizing the power load scheduling, formulating a backup power supply strategy, etc., to improve the safety and reliability of the power grid.

[0166] In the above-mentioned power grid operation risk assessment method, by obtaining the power grid operation status data and meteorological monitoring data, a power grid section data set is generated, and then the power grid operation risk data is generated by using the generative adversarial network model. This method takes advantage of the GAN model, can generate power grid operation risk data more accurately, and further improve the accuracy of risk assessment. Using the power grid operation risk data generated by the generative adversarial network model, a variety of power grid operation risk scenarios can be generated. Then, the time series operation simulation method is used to deduce and calculate these risk scenarios to generate a power grid forward-looking scenario sequence. In this way, the operation of the power grid under different risk scenarios can be better evaluated, and possible problems can be predicted in advance. In the online operation stage of the power grid, the preset risk assessment indicators and the current power grid operation scenario information are obtained, and then the power grid operation risk assessment results are obtained based on this information, and the corresponding power grid forward-looking operation scenario sequence is matched. This process can be carried out in real time, providing real-time risk assessment results, and helping decision makers to make quick adjustments during operation. In summary, the new power grid operation risk assessment method utilizes the GAN model and the time series operation simulation method, which can improve the accuracy and timeliness of power grid risk assessment and make the power grid operation more stable and reliable.

[0167] In an exemplary embodiment, Figure 3 As shown, the steps adopt a time series operation simulation method to deduce and calculate at least one power grid operation risk scenario to generate at least one power grid forward-looking scenario sequence, including steps S302 to S306. Among them: the time series operation simulation method includes a Monte Carlo decision tree;

[0168] Step S302: Using the weighted cross entropy function as the loss function of the Monte Carlo decision tree, and obtaining an evaluation value of at least one power grid operation risk scenario through the Monte Carlo decision tree.

[0169] Step S304: dividing and processing the evaluation value of at least one power grid operation risk scenario to obtain a processing result.

[0170] Step S306, adopting a parallel deduction architecture, performing multi-threaded parallel deduction calculations on the processing results through the Monte Carlo decision tree, and generating at least one grid forward-looking scenario sequence. Provide threaded deduction calculations for multi-branch scenarios, and output grid operation deduction scenario sequences under different objectives such as new energy consumption, economic optimization, and safety restrictions.

[0171] Specifically, a weighted cross entropy function can be used as the loss function of the Monte Carlo decision tree to evaluate the power grid operation risk scenario. The Monte Carlo decision tree is a probability-based decision tree model that can fully consider uncertainty factors and weights. When constructing a Monte Carlo decision tree, the nodes and boundary conditions of the decision tree can be set according to the characteristics of the power grid and the risk scenarios that need to be evaluated. These conditions can include power load levels, generator status, transmission line parameters, etc. Ensure the rationality and operability of the decision tree. Use the Monte Carlo method to construct and optimize the decision tree. Through random sampling and simulation evaluation using the Monte Carlo method, a set of sampled power grid operation risk scenarios and the operation information of each scenario are obtained, and the risk evaluation values ​​under different decision paths are calculated.

[0172] Specifically, a weighted cross entropy function is used to calculate the risk assessment value, and the result is adjusted according to the weight. Finally, one or more power grid operation risk scenarios with higher assessment values ​​are obtained. For power grid operation risk scenarios with higher assessment values, they can be divided and processed according to actual needs and priorities. For example, the assessment values ​​can be divided into different priorities for better response and control.

[0173] Specifically, a parallel deduction architecture is adopted to perform multi-threaded parallel deduction calculations on the processing results through the Monte Carlo decision tree to generate at least one grid forward-looking scenario sequence. These forward-looking scenario sequences can be used to formulate grid emergency response strategies, adjust power load scheduling, etc., to deal with different grid operation risks.

[0174] In this embodiment, through the evaluation of the Monte Carlo decision tree and multi-threaded parallel deduction calculation, a probability and weight-based power grid forward-looking scenario sequence can be provided to assist decision-making and planning measures and improve the safety and reliability of power grid operation.

[0175] In an exemplary embodiment, the evaluation value of at least one power grid operation risk scenario is divided and processed to obtain a processing result, including:

[0176] Through the preset evaluation value calculation formula, the maximum deduced expected evaluation value of each power grid operation risk scenario is obtained;

[0177] According to the maximization of the expected evaluation value, at least one power grid operation risk scenario is divided and processed to obtain a processing result having a sequence of multiple power grid operation risk scenarios;

[0178] The evaluation value calculation formula is:

[0179]

[0180] Where, Ε is the expected operator; T is the access step length; Q(v t ) is to access the scene vt The reward value; N(v p ) is the number of visits to the parent scene; N(v c ) is the number of visits to the sub-scene; V pen is the penalty term for violating the constraint conditions; γ is a constant coefficient, and γ>0.

[0181] In this embodiment, each power grid operation risk scenario is deduced and calculated according to a given evaluation value calculation formula. According to the expectation operator and other relevant parameters, the maximized deduced expected evaluation value of each scenario is calculated. According to the maximized deduced expected evaluation value, each power grid operation risk scenario is divided and processed. The power grid operation risk scenario can be divided into multiple sections according to different ranges of evaluation values ​​or set thresholds, and each section represents a different risk level or processing priority. According to the results of the division processing, the power grid operation risk scenario sequences of each section are combined to form a processing result with multiple sections of power grid operation risk scenario sequences. Each scenario sequence can represent different risk levels, so that the corresponding countermeasures can be adjusted according to different risk levels.

[0182] In the process of division, appropriate evaluation value ranges and thresholds, as well as corresponding penalty items, can be set according to specific needs and related constraints. In this way, the power grid operation risk scenario can be divided into multiple sections according to different evaluation values ​​and constraints, so as to better manage risks and plan corresponding control strategies.

[0183] In an exemplary embodiment, the loss function of the Monte Carlo decision tree is:

[0184]

[0185] Where m is the number of scenes; when the current scene is in the target scene set, then l i is 1, the current scene is not in the target scene set, then l i is 0; p 1i 、p 2i are the probability values ​​of the current scene being in the target scene set; r key is the weight of the number of scenes in the target scene set; is the weight of the number of scenes that are not in the target scene set.

[0186] In this embodiment, the first part of the loss function is used to measure the model's ability to correctly classify scenes as being in the target scene set, and the second part measures the model's ability to correctly classify scenes as not being in the target scene set. By weighing these two parts, the loss function weights the penalties for correct classification and incorrect classification. In the Monte Carlo decision tree, the optimization algorithm attempts to minimize the loss function to find the best decision tree model. Based on the specific weights and scene probability values, the algorithm can determine the best decision points and partitioning strategies to minimize misclassification and missed target scenes, thereby improving the model's prediction accuracy and decision-making effectiveness.

[0187] In an exemplary embodiment, the evaluation value of at least one power grid operation risk scenario is divided and processed to obtain a processing result, including:

[0188] Constraints are set to constrain the Monte Carlo decision tree, and the evaluation value of at least one power grid operation risk scenario is divided and processed by the constrained Monte Carlo decision tree to obtain a processing result, wherein the constraints include an evaluation penalty item, and the expression of the evaluation penalty item is:

[0189] ∑V pen,i =V pen,bl +V pen,op +V pen,pf ;

[0190] Where V pen,bl 、V pen,op 、V pen,pf They are operation constraint penalty term, safety and stability penalty term, and power balance penalty term.

[0191] In this embodiment, the optimization algorithm can use this objective function, combined with other constraints, to adjust the partitioning strategy and decision points of the Monte Carlo decision tree to minimize the value of the objective function. Through this constraint and optimization method, a processing result can be obtained, which better divides and evaluates the target power grid operation risk scenario while satisfying the constraints.

[0192] In an exemplary embodiment, obtaining a preset risk value assessment indicator includes:

[0193] Obtain the load loss risk value, voltage over-limit risk value and line over-limit risk value of the power grid;

[0194] According to the load loss risk value, voltage over-limit risk value and line over-limit risk value, the preset risk value assessment index is obtained according to the preset weight ratio; among which, the expression of the load loss risk value is:

[0195]

[0196]

[0197] Where K is the number of simulation states; P(E k ) is the system state E k The probability of S C (E k ) is the system state E k The severity of the consequences of load loss; P ci is the load shedding amount at load node i, ND is the number of load nodes;

[0198] The expression of voltage over-limit risk value is:

[0199]

[0200]

[0201] Where K is the number of simulation states; P(E k ) is the system state E k The probability of S v (E k ) is the system state E k The severity of the voltage limit consequences under NI is the number of system nodes;

[0202] The expression of line crossing risk value is:

[0203]

[0204]

[0205] Where K is the number of simulation states; P(E k ) is the system state E k The probability of S s (E k ) is the system state E k The severity of the voltage limit consequences under the condition; NL is the number of system nodes, P Li is the active power flowing through branch i, P Limax is the maximum active power allowed to flow through branch i.

[0206] In this embodiment, the technical effect of obtaining the load loss risk value, voltage over-limit risk value and line over-limit risk value of the power grid, and calculating the preset risk value assessment index based on these risk values ​​is to improve the stability and reliability of the power grid.

[0207] The load loss risk value, voltage over-limit risk value and line over-limit risk value of the power grid are important indicators in the operation of the power grid. They reflect the stability and reliability of the power grid under specific circumstances. By calculating and analyzing these risk values, possible problems in the power grid can be discovered in time, and corresponding measures can be taken to solve them, thereby reducing the probability of power grid failure and improving the stability and reliability of the power grid.

[0208] The preset risk value assessment indicators are calculated based on the actual operation of the power grid and historical data, and can comprehensively reflect the risk level of the power grid in different aspects. By comparing the preset risk value assessment indicators with the actual risk value, the operation status and reliability of the power grid can be evaluated, providing a reference for the operation management and maintenance of the power grid, thereby better ensuring the safe and stable operation of the power grid.

[0209] In summary, obtaining the load loss risk value, voltage over-limit risk value and line over-limit risk value of the power grid, and calculating the preset risk value assessment index according to the preset weight ratio, can improve the stability and reliability of the power grid and provide an important reference basis for the operation management and maintenance of the power grid.

[0210] In the most specific embodiment, reference is made to Figure 4 , Figure 4 Schematic diagram of the most specific embodiment.

[0211] Obtain the grid operation status data and meteorological monitoring data, generate the grid section data set, build the generative adversarial network model, generate the grid operation risk scenario set, the current number of iterations is 1, then cluster each operation risk scenario and perform discriminant analysis, then sample through the Monte Carlo method, and then use the decision tree with weighted cross entropy function to screen through the evaluation value. Set the constraints of the Monte Carlo method, constrain when performing the deduction calculation, and then verify the accuracy of the Monte Carlo method. When the accuracy meets the conditions, perform parallel deduction calculations to obtain the grid operation risk assessment results and the grid forward-looking operation scenario sequence. When the accuracy does not meet the conditions, re-add constraints or reduce constraints. After the parallel deduction calculation is completed, determine whether all scenarios have been traversed. If not, reiterate the above steps. If it has been completed, generate a comprehensive grid risk assessment model based on the results of the parallel deduction calculation, and then process the model offline. Then, in the online operation stage of the grid, obtain the preset risk assessment indicators and the current grid operation scenario information, perform real-time risk assessment through the model, obtain a more accurate risk assessment result, and then roll back the risk assessment result to the model for deep learning.

[0212] This embodiment performs a grid operation risk assessment test based on IEEE39 nodes, with 39 nodes, 38 lines, 10,000 grid operation simulation scenarios, a time resolution of 15 minutes for timing simulation calculation, and a simulation time window of 24 hours.

[0213] Test results such as Figure 5 As shown in the figure, the calculation example is based on the power system risk assessment calculation test of the generative adversarial network and the time series deduction method. The assessment results are expressed in a graded manner. The identified risk warning lines include 1-4, 11-15, 25-36, and the risk warning nodes are 15, 17, and 23. The risk value of the comprehensive risk assessment is 0.731, which belongs to the risk warning level of the entire network. The single section risk assessment calculation takes 1.39 seconds.

[0214] The test results of the embodiment show that the risk assessment method provided by the present invention can accurately and quickly assess the risk lines and nodes of the power grid, solve the comprehensive operation risk level of the power grid, and meet the accuracy and timeliness requirements of the risk identification of the dispatching business.

[0215] It should be understood that, although the steps in the flowcharts involved in the above embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above embodiments may include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0216] Based on the same inventive concept, the embodiment of the present application also provides a power grid operation risk assessment device for implementing the power grid operation risk assessment method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more power grid operation risk assessment device embodiments provided below can refer to the limitations of the power grid operation risk assessment method above, and will not be repeated here.

[0217] In an exemplary embodiment, referring to Figure 6 , Figure 7 , a power grid operation risk assessment device is provided, comprising: a data acquisition module 702, for acquiring power grid operation status data and meteorological monitoring data;

[0218] The data integration module 704 is used to generate a power grid section data set according to the power grid operation status data and the meteorological monitoring data;

[0219] The modeling module 706 is used to construct a generative adversarial network model according to the power grid section data set during the offline phase of the power grid; and obtain at least one type of power grid operation risk data through the generative adversarial network model;

[0220] A processing module 708 is used to generate a grid operation risk scenario set according to at least one type of grid operation risk data, wherein the grid operation risk scenario set includes at least one grid operation risk scenario;

[0221] A deduction and evaluation module 710 is used to perform deduction and calculation on at least one power grid operation risk scenario by using a time series operation simulation method to generate at least one power grid forward-looking scenario sequence;

[0222] The data acquisition module 702 is also used to obtain preset risk assessment indicators and current grid operation scenario information during the online operation stage of the grid;

[0223] The online application module 712 is used to obtain the grid operation risk assessment result according to the preset risk assessment index and the current grid operation scenario information, and to match and obtain the corresponding grid forward-looking operation scenario sequence.

[0224] In an exemplary embodiment, the time series operation simulation method includes a Monte Carlo decision tree; the deduction evaluation module 710 is also used to:

[0225] The weighted cross entropy function is used as the loss function of the Monte Carlo decision tree, and the evaluation value of at least one power grid operation risk scenario is obtained through the Monte Carlo decision tree;

[0226] Dividing and processing the evaluation value of at least one power grid operation risk scenario to obtain a processing result;

[0227] A parallel deduction architecture is adopted to perform multi-threaded parallel deduction calculations on the processing results through the Monte Carlo decision tree to generate at least one grid forward-looking scenario sequence.

[0228] In an exemplary embodiment, the deduction and evaluation module 710 is further configured to:

[0229] Through the preset evaluation value calculation formula, the maximum deduced expected evaluation value of each power grid operation risk scenario is obtained;

[0230] According to the maximization of the expected evaluation value, at least one power grid operation risk scenario is divided and processed to obtain a processing result having a sequence of multiple power grid operation risk scenarios;

[0231] The evaluation value calculation formula is:

[0232]

[0233] Where, Ε is the expected operator; T is the access step length; Q(v t ) is to access the scene v t The reward value; N(v p ) is the number of visits to the parent scene; N(v c ) is the number of visits to the sub-scene; V pen is the penalty term for violating the constraint conditions; γ is a constant coefficient, and γ>0.

[0234] In an exemplary embodiment, the loss function of the Monte Carlo decision tree is:

[0235]

[0236] Where m is the number of scenes; when the current scene is in the target scene set, then l i is 1, the current scene is not in the target scene set, then l i is 0; p 1i 、p 2i are the probability values ​​of the current scene being in the target scene set; r key is the weight of the number of scenes in the target scene set; is the weight of the number of scenes that are not in the target scene set.

[0237] In an exemplary embodiment, the deduction and evaluation module 710 is further used to: set constraints to constrain the Monte Carlo decision tree, and divide the evaluation value of at least one power grid operation risk scenario through the constrained Monte Carlo decision tree to obtain a processing result, wherein the constraints include an evaluation penalty item, and the expression of the evaluation penalty item is:

[0238] ∑V pen,i =V pen,bl +V pen,op +V pen,pf ;

[0239] Where V pen,bl 、V pen,op 、V pen,pf They are operation constraint penalty term, safety and stability penalty term, and power balance penalty term.

[0240] In an exemplary embodiment, the data acquisition module 702 is further configured to:

[0241] Obtain the load loss risk value, voltage over-limit risk value and line over-limit risk value of the power grid;

[0242] According to the load loss risk value, voltage over-limit risk value and line over-limit risk value, the preset risk value assessment index is obtained according to the preset weight ratio; among which, the expression of the load loss risk value is:

[0243]

[0244]

[0245] Where K is the number of simulation states; P(E k ) is the system state E k The probability of S C (E k ) is the system state E k The severity of the consequences of load loss; P ci is the load shedding amount at load node i, ND is the number of load nodes;

[0246] The expression of voltage over-limit risk value is:

[0247]

[0248]

[0249] Where K is the number of simulation states; P(E k ) is the system state E k The probability of S v (E k ) is the system state E k The severity of the voltage limit consequences under NI is the number of system nodes;

[0250] The expression of line crossing risk value is:

[0251]

[0252]

[0253] Where K is the number of simulation states; P(E k ) is the system state E k The probability of S s (E k ) is the system state E k The severity of the voltage limit consequences under the condition; NL is the number of system nodes, P Li is the active power flowing through branch i, P Limax is the maximum active power allowed to flow through branch i.

[0254] Each module in the above-mentioned power grid operation risk assessment device can be implemented in whole or in part by software, hardware and a combination thereof. Each of the above-mentioned modules can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute the operations corresponding to each of the above modules.

[0255] In an exemplary embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. Figure 7 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. Among them, the processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store power grid operation status data and meteorological monitoring data. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a power grid operation risk assessment method is implemented.

[0256] Those skilled in the art will understand that Figure 7 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0257] In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the following steps are implemented:

[0258] Obtain power grid operation status data and meteorological monitoring data;

[0259] Generate a power grid section data set according to the power grid operation status data and the meteorological monitoring data;

[0260] In the offline stage of the power grid, a generative adversarial network model is constructed according to the power grid section data set;

[0261] Obtaining at least one type of power grid operation risk data through the generative adversarial network model;

[0262] Generate a grid operation risk scenario set according to the at least one type of grid operation risk data, wherein the grid operation risk scenario set includes at least one grid operation risk scenario;

[0263] Using a time series operation simulation method, at least one power grid operation risk scenario is simulated and calculated to generate at least one power grid forward-looking scenario sequence;

[0264] During the online operation phase of the power grid, obtain the preset risk assessment indicators and current power grid operation scenario information;

[0265] According to the preset risk assessment index and the current power grid operation scenario information, a power grid operation risk assessment result is obtained, and a corresponding power grid forward-looking operation scenario sequence is matched and obtained.

[0266] In one embodiment, when the processor executes the computer program, the following steps are also implemented:

[0267] The time series operation simulation method includes a Monte Carlo decision tree; the time series operation simulation method is used to deduce and calculate at least one power grid operation risk scenario to generate at least one power grid forward-looking scenario sequence, including:

[0268] The weighted cross entropy function is used as the loss function of the Monte Carlo decision tree, and the evaluation value of at least one power grid operation risk scenario is obtained through the Monte Carlo decision tree;

[0269] Dividing and processing the evaluation value of at least one power grid operation risk scenario to obtain a processing result;

[0270] A parallel deduction architecture is adopted to perform multi-threaded parallel deduction calculations on the processing results through the Monte Carlo decision tree to generate at least one grid forward-looking scenario sequence.

[0271] In one embodiment, when the processor executes the computer program, the following steps are also implemented:

[0272] The evaluation value of at least one power grid operation risk scenario is divided and processed to obtain processing results, including:

[0273] Through the preset evaluation value calculation formula, the maximum deduced expected evaluation value of each power grid operation risk scenario is obtained;

[0274] According to the maximization of the expected evaluation value, at least one power grid operation risk scenario is divided and processed to obtain a processing result having a sequence of multiple power grid operation risk scenarios;

[0275] The evaluation value calculation formula is:

[0276]

[0277] Where, Ε is the expected operator; T is the access step length; Q(v t ) is to access the scene v t The reward value; N(v p ) is the number of visits to the parent scene; N(v c ) is the number of visits to the sub-scene; V pen is the penalty term for violating the constraint conditions; γ is a constant coefficient, and γ>0.

[0278] In one embodiment, when the processor executes the computer program, the following steps are also implemented:

[0279] The loss function of the Monte Carlo decision tree is:

[0280]

[0281] Where m is the number of scenes; when the current scene is in the target scene set, then l i When it is 1, the current scene is not in the target scene set, then l i is 0; p 1i 、p 2i are the probability values ​​of the current scene being in the target scene set; r key is the weight of the number of scenes in the target scene set; is the weight of the number of scenes that are not in the target scene set.

[0282] In one embodiment, when the processor executes the computer program, the following steps are also implemented:

[0283] The evaluation value of at least one power grid operation risk scenario is divided and processed to obtain processing results, including:

[0284] Constraints are set to constrain the Monte Carlo decision tree, and the evaluation value of at least one power grid operation risk scenario is divided and processed by the constrained Monte Carlo decision tree to obtain a processing result, wherein the constraints include an evaluation penalty item, and the expression of the evaluation penalty item is:

[0285] ∑V pen,i =V pen,bl +V pen,op +V pen,pf ;

[0286] Where V pen,bl 、V pen,op 、V pen,pf They are operation constraint penalty term, safety and stability penalty term, and power balance penalty term.

[0287] In one embodiment, when the processor executes the computer program, the following steps are also implemented:

[0288] Obtain preset risk value assessment indicators, including:

[0289] Obtain the load loss risk value, voltage over-limit risk value and line over-limit risk value of the power grid;

[0290] According to the load loss risk value, voltage over-limit risk value and line over-limit risk value, the preset risk value assessment index is obtained according to the preset weight ratio; among which, the expression of the load loss risk value is:

[0291]

[0292]

[0293] Where K is the number of simulation states; P(E k ) is the system state E k The probability of S C (E k ) is the system state E k The severity of the consequences of load loss; P ci is the load shedding amount at load node i, ND is the number of load nodes;

[0294] The expression of voltage over-limit risk value is:

[0295]

[0296]

[0297] Where K is the number of simulation states; P(E k ) is the system state E k The probability of S v (E k ) is the system state E k The severity of the voltage limit consequences under NI is the number of system nodes;

[0298] The expression of line crossing risk value is:

[0299]

[0300]

[0301] Where K is the number of simulation states; P(E k ) is the system state E k The probability of S s (E k ) is the system state E k The severity of the voltage limit consequences under the condition; NL is the number of system nodes, P Li is the active power flowing through branch i, P Limaxis the maximum active power allowed to flow through branch i.

[0302] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0303] Obtain power grid operation status data and meteorological monitoring data;

[0304] Generate a power grid section data set according to the power grid operation status data and the meteorological monitoring data;

[0305] In the offline stage of the power grid, a generative adversarial network model is constructed according to the power grid section data set;

[0306] Obtaining at least one type of power grid operation risk data through the generative adversarial network model;

[0307] Generate a grid operation risk scenario set according to the at least one type of grid operation risk data, wherein the grid operation risk scenario set includes at least one grid operation risk scenario;

[0308] Using a time series operation simulation method, at least one power grid operation risk scenario is simulated and calculated to generate at least one power grid forward-looking scenario sequence;

[0309] During the online operation phase of the power grid, obtain the preset risk assessment indicators and current power grid operation scenario information;

[0310] According to the preset risk assessment index and the current power grid operation scenario information, a power grid operation risk assessment result is obtained, and a corresponding power grid forward-looking operation scenario sequence is matched and obtained.

[0311] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0312] The time series operation simulation method includes a Monte Carlo decision tree; the time series operation simulation method is used to deduce and calculate at least one power grid operation risk scenario to generate at least one power grid forward-looking scenario sequence, including:

[0313] The weighted cross entropy function is used as the loss function of the Monte Carlo decision tree, and the evaluation value of at least one power grid operation risk scenario is obtained through the Monte Carlo decision tree;

[0314] Dividing and processing the evaluation value of at least one power grid operation risk scenario to obtain a processing result;

[0315] A parallel deduction architecture is adopted to perform multi-threaded parallel deduction calculations on the processing results through the Monte Carlo decision tree to generate at least one grid forward-looking scenario sequence.

[0316] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0317] The evaluation value of at least one power grid operation risk scenario is divided and processed to obtain processing results, including:

[0318] Through the preset evaluation value calculation formula, the maximum deduced expected evaluation value of each power grid operation risk scenario is obtained;

[0319] According to the maximization of the expected evaluation value, at least one power grid operation risk scenario is divided and processed to obtain a processing result having a sequence of multiple power grid operation risk scenarios;

[0320] The evaluation value calculation formula is:

[0321]

[0322] Where, Ε is the expected operator; T is the access step length; Q(v t ) is to access the scene v t The reward value; N(v p ) is the number of visits to the parent scene; N(v c ) is the number of visits to the sub-scene; V pen is the penalty term for violating the constraint conditions; γ is a constant coefficient, and γ>0.

[0323] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0324] The loss function of the Monte Carlo decision tree is:

[0325]

[0326] Where m is the number of scenes; when the current scene is in the target scene set, then l i When it is 1, the current scene is not in the target scene set, then l i is 0; p 1i 、p 2i are the probability values ​​of the current scene being in the target scene set; r key is the weight of the number of scenes in the target scene set; is the weight of the number of scenes that are not in the target scene set.

[0327] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0328] The evaluation value of at least one power grid operation risk scenario is divided and processed to obtain processing results, including:

[0329] Constraints are set to constrain the Monte Carlo decision tree, and the evaluation value of at least one power grid operation risk scenario is divided and processed by the constrained Monte Carlo decision tree to obtain a processing result, wherein the constraints include an evaluation penalty item, and the expression of the evaluation penalty item is:

[0330] ∑V pen,i =V pen,bl +V pen,op +V pen,pf ;

[0331] Where V pen,bl 、V pen,op 、V pen,pf They are operation constraint penalty term, safety and stability penalty term, and power balance penalty term.

[0332] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0333] Obtain preset risk value assessment indicators, including:

[0334] Obtain the load loss risk value, voltage over-limit risk value and line over-limit risk value of the power grid;

[0335] According to the load loss risk value, voltage over-limit risk value and line over-limit risk value, the preset risk value assessment index is obtained according to the preset weight ratio; among which, the expression of the load loss risk value is:

[0336]

[0337]

[0338] Where K is the number of simulation states; P(E k ) is the system state E k The probability of S C (E k ) is the system state E k The severity of the consequences of load loss; P ci is the load shedding amount at load node i, ND is the number of load nodes;

[0339] The expression of voltage over-limit risk value is:

[0340]

[0341]

[0342] Where K is the number of simulation states; P(E k ) is the system state E k The probability of S v (E k ) is the system state Ek The severity of the voltage limit consequences under NI is the number of system nodes;

[0343] The expression of line crossing risk value is:

[0344]

[0345]

[0346] Where K is the number of simulation states; P(E k ) is the system state E k The probability of S s (E k ) is the system state E k The severity of the voltage limit consequences under the condition; NL is the number of system nodes, P Li is the active power flowing through branch i, P Limax is the maximum active power allowed to flow through branch i.

[0347] In one embodiment, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements the following steps:

[0348] Obtain power grid operation status data and meteorological monitoring data;

[0349] Generate a power grid section data set according to the power grid operation status data and the meteorological monitoring data;

[0350] In the offline stage of the power grid, a generative adversarial network model is constructed according to the power grid section data set;

[0351] Obtaining at least one type of power grid operation risk data through the generative adversarial network model;

[0352] Generate a grid operation risk scenario set according to the at least one type of grid operation risk data, wherein the grid operation risk scenario set includes at least one grid operation risk scenario;

[0353] Using a time series operation simulation method, at least one power grid operation risk scenario is simulated and calculated to generate at least one power grid forward-looking scenario sequence;

[0354] During the online operation phase of the power grid, obtain the preset risk assessment indicators and current power grid operation scenario information;

[0355] According to the preset risk assessment index and the current power grid operation scenario information, a power grid operation risk assessment result is obtained, and a corresponding power grid forward-looking operation scenario sequence is matched and obtained.

[0356] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0357] The time series operation simulation method includes a Monte Carlo decision tree; the time series operation simulation method is used to deduce and calculate at least one power grid operation risk scenario to generate at least one power grid forward-looking scenario sequence, including:

[0358] The weighted cross entropy function is used as the loss function of the Monte Carlo decision tree, and the evaluation value of at least one power grid operation risk scenario is obtained through the Monte Carlo decision tree;

[0359] Dividing and processing the evaluation value of at least one power grid operation risk scenario to obtain a processing result;

[0360] A parallel deduction architecture is adopted to perform multi-threaded parallel deduction calculations on the processing results through the Monte Carlo decision tree to generate at least one grid forward-looking scenario sequence.

[0361] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0362] The evaluation value of at least one power grid operation risk scenario is divided and processed to obtain processing results, including:

[0363] Through the preset evaluation value calculation formula, the maximum deduced expected evaluation value of each power grid operation risk scenario is obtained;

[0364] According to the maximization of the expected evaluation value, at least one power grid operation risk scenario is divided and processed to obtain a processing result having a sequence of multiple power grid operation risk scenarios;

[0365] The evaluation value calculation formula is:

[0366]

[0367] Where, Ε is the expected operator; T is the access step length; Q(v t ) is to access the scene v t The reward value; N(v p ) is the number of visits to the parent scene; N(v c ) is the number of visits to the sub-scene; V pen is the penalty term for violating the constraint conditions; γ is a constant coefficient, and γ>0.

[0368] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0369] The loss function of the Monte Carlo decision tree is:

[0370]

[0371] Where m is the number of scenes; when the current scene is in the target scene set, then l i When it is 1, the current scene is not in the target scene set, then l i is 0; p 1i 、p 2i are the probability values ​​of the current scene being in the target scene set; r key is the weight of the number of scenes in the target scene set; is the weight of the number of scenes that are not in the target scene set.

[0372] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0373] The evaluation value of at least one power grid operation risk scenario is divided and processed to obtain processing results, including:

[0374] Constraints are set to constrain the Monte Carlo decision tree, and the evaluation value of at least one power grid operation risk scenario is divided and processed by the constrained Monte Carlo decision tree to obtain a processing result, wherein the constraints include an evaluation penalty item, and the expression of the evaluation penalty item is:

[0375] ∑V pen,i =V pen,bl +V pen,op +V pen,pf ;

[0376] Where V pen,bl 、V pen,op 、V pen,pf They are operation constraint penalty term, safety and stability penalty term, and power balance penalty term.

[0377] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0378] Obtain preset risk value assessment indicators, including:

[0379] Obtain the load loss risk value, voltage over-limit risk value and line over-limit risk value of the power grid;

[0380] According to the load loss risk value, voltage over-limit risk value and line over-limit risk value, the preset risk value assessment index is obtained according to the preset weight ratio; among which, the expression of the load loss risk value is:

[0381]

[0382]

[0383] Where K is the number of simulation states; P(E k ) is the system state E kThe probability of S C (E k ) is the system state E k The severity of the consequences of load loss; P ci is the load shedding amount at load node i, ND is the number of load nodes;

[0384] The expression of voltage over-limit risk value is:

[0385]

[0386]

[0387] Where K is the number of simulation states; P(E k ) is the system state E k The probability of S v (E k ) is the system state E k The severity of the voltage limit consequences under NI is the number of system nodes;

[0388] The expression of line crossing risk value is:

[0389]

[0390]

[0391] Where K is the number of simulation states; P(E k ) is the system state E k The probability of S s (E k ) is the system state E k The severity of the voltage limit consequences under the condition; NL is the number of system nodes, P Li is the active power flowing through branch i, P Limax is the maximum active power allowed to flow through branch i.

[0392] Those of ordinary skill in the art will appreciate that the units of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition of each example has been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0393] In the embodiments provided by the present invention, it should be understood that the division of units is only a logical function division, and there may be other division methods in actual implementation, for example, multiple units can be combined into one unit, one unit can be split into multiple units, or some features can be ignored, etc.

[0394] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0395] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-0nlyMemory), random access memory (RAM, RandomAccessMemory), mobile hard disk, magnetic disk or optical disk, etc., which can store program code.

[0396] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. These modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and specification of the present invention.

Claims

1. A method for forward-looking risk assessment of a power grid, characterized in that: The following steps are involved: Obtain power grid operation status data, meteorological monitoring data and wind turbine power data for historical periods; construct a wind turbine power time series based on the wind turbine power data for historical periods; Obtain missing data in the wind turbine power time series, and fill the missing data through a piecewise linear regression model to obtain a target wind turbine power time series; According to the target wind turbine power time series, the extreme learning machine model, the long short-term memory neural network model and the temporal convolutional network model are trained respectively; Training a multi-objective prediction model according to the extreme learning machine model, the long short-term memory neural network model and the temporal convolutional network model; Using a multi-objective Harris Hawk algorithm, obtaining a Pareto optimal solution set of a combined weight vector of the multi-objective prediction model; Using a deep deterministic policy gradient method, according to the Pareto optimal solution set, to obtain the dynamic optimal weight of each model in the multi-objective prediction model; Obtaining a wind power prediction result according to each model in the multi-objective prediction model and the corresponding dynamic optimal weight; Generate a power grid section data set according to the power grid operation status data, the meteorological monitoring data and the wind power prediction result; In the offline stage of the power grid, a generative adversarial network model is constructed according to the power grid section data set; Obtaining at least one type of power grid operation risk data through the generative adversarial network model; Generate a grid operation risk scenario set according to the at least one type of grid operation risk data, wherein the grid operation risk scenario set includes at least one grid operation risk scenario; Using a time series operation simulation method, at least one power grid operation risk scenario is simulated and calculated to generate at least one power grid forward-looking scenario sequence; During the online operation phase of the power grid, obtain the preset risk assessment indicators and current power grid operation scenario information; According to the preset risk assessment index and the current power grid operation scenario information, a power grid operation risk assessment result is obtained, and a corresponding power grid forward-looking operation scenario sequence is matched and obtained.

2. A method for forward-looking risk assessment of a power grid according to claim 1, characterized in that: Obtain missing data in the wind turbine power time series, including: The wind turbine power data that meets the missing value preset condition in the wind turbine power time series is determined as missing value data; the missing value preset condition includes that the value of the wind turbine power data is zero, and the change value of the wind turbine power data relative to the fan power data of the previous time series is greater than a preset threshold.

3. A method for forward-looking risk assessment of a power grid according to claim 1, characterized in that: Before filling the missing data by the piecewise linear regression model, the method further includes: Interpolating and filling the missing data to obtain an intermediate wind turbine power time series; The initial piecewise linear regression model is trained according to the intermediate wind turbine power time series to optimize the parameters of the initial piecewise linear regression model until a trained piecewise linear regression model is obtained.

4. A method for forward-looking risk assessment of a power grid according to claim 3, characterized in that: The initial piecewise linear regression model is trained based on the intermediate wind turbine power time series, including: Obtain the exponentially weighted moving average and moving median characteristics of the intermediate wind turbine power time series; The intermediate wind turbine power time series is processed according to the BiLSTM model to obtain the meta-features of the intermediate wind turbine power time series; The initial piecewise linear regression model is trained according to the meta-features, the exponentially weighted moving average features and the moving median features to optimize the parameters of the initial piecewise linear regression model until a trained piecewise linear regression model is obtained.

5. A method for forward-looking risk assessment of a power grid according to claim 1, characterized in that: According to the target wind turbine power time series, the extreme learning machine model, the long short-term memory neural network model and the time convolution network model are trained respectively, including: The sequence decomposition method CEEMDAN is used to decompose the time series data of historical wind power to obtain multiple first subsequence data; The extreme learning machine model, the long short-term memory neural network model and the temporal convolutional network model are trained simultaneously through the multiple first subsequence data.

6. A method for forward-looking risk assessment of a power grid according to claim 1, characterized in that: According to each model in the multi-objective prediction model and the corresponding dynamic optimal weight, a short-term wind power prediction result is obtained, including: Obtain time series data of target wind power; The sequence decomposition method CEEMDAN is used to decompose the time series data of the target wind power to obtain multiple second subsequence data; By using each model in the multi-objective prediction model, the plurality of second subsequence data are traversed and predicted at the same time to obtain corresponding first prediction results, second prediction results and third prediction results; According to the dynamic optimal weights corresponding to each model, weighted calculation is performed on the first prediction result, the second prediction result and the third prediction result to obtain a prediction result of short-term wind power.

7. A method for forward-looking risk assessment of a power grid according to claim 1, characterized in that: According to the extreme learning machine model, the long short-term memory neural network model and the temporal convolutional network model, a multi-objective prediction model is trained, including: Obtain time series data of test wind power; The sequence decomposition method CEEMDAN is used to decompose the time series data of the test wind power to obtain multiple third subsequence data; According to the extreme learning machine model, the long short-term memory neural network model and the temporal convolutional network model, the plurality of third subsequence data are traversed and tested at the same time to obtain corresponding first test results, second test results and third test results; Obtaining mean square errors and standard deviations between the actual short-term wind power and the first test result, the second test result, and the third test result, respectively; The minimum mean square error and the minimum standard deviation are used as optimization objectives to train the multi-objective prediction model.

8. A method for forward-looking risk assessment of a power grid according to claim 1, characterized in that: The weights and biases of the multi-target prediction model are iteratively updated through the agent’s action vector, including: Through the action vector of the agent, the static weight coefficient is selected from the Pareto optimal solution set based on the minimum mean absolute percentage error principle; Obtain reference error based on static weight coefficient and actual short-term wind power; Analyze the correlation between the minimum value of mean square error, the minimum value of standard deviation and reference error; According to the correlation, obtain the reward function; Through the reward function, the weights and biases of the multi-objective prediction model are iteratively updated.

9. A power grid operation risk assessment device, characterized in that: The device includes: Data acquisition module, used to obtain power grid operation status data and meteorological monitoring data; A data integration module is used to generate a power grid section data set based on power grid operation status data and meteorological monitoring data; A modeling module is used to construct a generative adversarial network model based on the power grid section data set during the offline phase of the power grid; Obtaining at least one type of power grid operation risk data by generating an adversarial network model; A processing module, configured to generate a set of power grid operation risk scenarios according to at least one type of power grid operation risk data, wherein the set of power grid operation risk scenarios includes at least one power grid operation risk scenario; A deduction and evaluation module is used to use a time series operation simulation method to perform deduction and calculation on at least one power grid operation risk scenario and generate at least one power grid forward-looking scenario sequence; The data acquisition module is also used to obtain preset risk assessment indicators and current grid operation scenario information during the online operation stage of the grid; The online application module is used to obtain the grid operation risk assessment results based on the preset risk assessment indicators and the current grid operation scenario information, and match the corresponding grid forward-looking operation scenario sequence.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein when the program is running, the device where the computer-readable storage medium is located is controlled to execute the power grid forward-looking risk assessment method according to any one of claims 1 to 8.

Images