Method and device for evaluating new energy consumption proportion contribution, electronic equipment and medium

Abstract

The application provides a new energy consumption proportion contribution evaluation method and device, electronic equipment and medium, the method comprises the following steps: constructing a target scheduling plan meeting the annual new energy consumption proportion of a comprehensive energy base to determine the historical new energy consumption proportion contribution score of the historical day-ahead scheduling plan data of the comprehensive energy base; generating a training sample set according to the historical day-ahead scheduling plan data, the historical day-ahead power prediction data corresponding to the historical day-ahead scheduling plan data and the historical new energy consumption proportion contribution score; training the CLA model by using the training sample set to obtain a new energy consumption proportion contribution evaluation model; obtaining the day-ahead scheduling plan data and the corresponding day-ahead power prediction data of the comprehensive energy base and obtaining the new energy consumption proportion contribution score of the day-ahead scheduling plan data based on the new energy consumption proportion contribution evaluation model. The accuracy of the contribution proportion evaluation of the day-ahead scheduling plan data is improved, and energy waste is avoided.

Description

Technical Field

[0001] This application relates to the field of data processing, and in particular to a method, apparatus, electronic device, and medium for assessing the contribution of new energy consumption ratio. Background Technology

[0002] With the development of technology, energy bases can collect wind power, solar power and thermal power and transmit them via DC. There is a set requirement for the proportion of new energy power consumed by the energy base in the total annual power consumption of the energy base.

[0003] In this scenario, energy bases need to formulate dispatch plans for wind power, solar power, and thermal power in advance. However, the dispatch plans formulated in related technologies do not take into account the proportion of thermal power and new energy power, including wind power and solar power, in the dispatch plan, or do not take into account the seasonal characteristics of new energy power to formulate an annual dispatch plan for thermal power and new energy power, including wind power and solar power. This results in low accuracy of dispatch plan formulation and leads to resource waste. Summary of the Invention

[0004] The purpose of this application is to at least partially solve one of the technical problems in the aforementioned technologies.

[0005] The first aspect of this application provides a method for evaluating the contribution of renewable energy consumption ratio, comprising: constructing a target scheduling plan that meets the annual renewable energy consumption ratio of a comprehensive energy base; determining a historical renewable energy consumption ratio contribution score based on the target scheduling plan and the historical day-ahead scheduling plan data of the comprehensive energy base; generating a training sample set based on the historical day-ahead scheduling plan data, the historical day-ahead power prediction data corresponding to the historical day-ahead scheduling plan data, and the historical renewable energy consumption ratio contribution score; training a preset hybrid CLA model using the training sample set to obtain a renewable energy consumption ratio contribution evaluation model; wherein, the CLA model includes an input layer, a convolutional neural network (CNN) layer, a long short-term memory (LSTM) artificial neural network layer, an attention layer, and an output layer, wherein the input layer outputs the input vector corresponding to each training sample, the CNN layer generates a feature vector which is then input into the LSTM layer, the LSTM layer and the attention layer learn the feature vector, and the output layer outputs a prediction result; acquiring the day-ahead scheduling plan data and the corresponding day-ahead power prediction data of the comprehensive energy base; inputting the day-ahead scheduling plan data and the corresponding day-ahead power prediction data into the renewable energy consumption ratio contribution evaluation model to output the renewable energy consumption ratio contribution score of the day-ahead scheduling plan data.

[0006] The method for assessing the contribution of new energy consumption ratio provided in the first aspect of this application also has the following technical features, including:

[0007] According to an embodiment of this application, training a preset hybrid CLA model using the training sample set to obtain a new energy consumption ratio contribution evaluation model includes: inputting training samples from the training sample set into the input layer to obtain input vectors corresponding to each training sample through the input layer; inputting the input vectors into the CNN layer and performing feature extraction on the input vectors to filter out target feature vectors; obtaining the target feature vectors and inputting the target feature vectors into the LSTM layer to obtain first output vectors corresponding to each target feature vector; inputting the first output vectors corresponding to each target feature vector into the attention layer and filtering the first output vectors according to the attention weight parameter values ​​of the first output vectors in the attention layer to obtain second output vectors; inputting the second output vectors into the output layer to determine the historical new energy consumption ratio contribution prediction score; adjusting the model parameters of the CLA model according to the historical new energy consumption ratio contribution score and the historical new energy consumption ratio contribution prediction score in the training samples, and returning to use the next training sample to continue training the CLA model with adjusted model parameters until the training ends and the new energy consumption ratio contribution evaluation model is obtained.

[0008] According to one embodiment of this application, the CNN layer includes a convolutional layer and a dropout layer. The input vector is input into the CNN layer, and feature extraction is performed on the input vector to filter out target feature vectors. This includes: inputting the input vector into the convolutional layer to obtain multiple feature vectors of the input vector extracted by the convolutional layer; and inputting the multiple feature vectors of the input vector into the dropout layer to filter out target feature vectors from the multiple features.

[0009] According to one embodiment of this application, after inputting the second output vector into the output layer to determine the historical new energy consumption ratio contribution prediction score, the method further includes: obtaining the historical new energy consumption ratio contribution prediction score and performing inverse normalization processing on the historical new energy consumption ratio contribution prediction score.

[0010] According to an embodiment of this application, the step of determining the historical renewable energy consumption ratio contribution score of the historical day-ahead scheduling plan data of the integrated energy base based on the target scheduling plan includes: obtaining historical day-ahead power prediction data samples corresponding to multiple historical sampling time points of the integrated energy base, each historical day-ahead power prediction data sample including: historical day-ahead photovoltaic power prediction data and historical day-ahead wind power prediction data; obtaining historical day-ahead scheduling plan data samples for the next day corresponding to each of the historical day-ahead power prediction data samples, each historical day-ahead scheduling plan sample including: historical day-ahead scheduling plan photovoltaic ratio, historical day-ahead scheduling plan wind power ratio, and historical day-ahead scheduling plan thermal power ratio; and determining the historical renewable energy consumption ratio contribution score corresponding to each of the historical day-ahead scheduling plan samples based on the renewable energy consumption ratio of the day under the target scheduling plan and the renewable energy consumption ratio corresponding to each of the historical day-ahead scheduling plan samples.

[0011] According to one embodiment of this application, before training a preset hybrid CLA model using the training sample set, the method further includes: constructing a loss function to represent the difference between the predicted score of the contribution of new energy consumption ratio output by the CLA model and the score of the contribution of new energy consumption ratio; and training the CLA model based on the training sample set with the goal of minimizing the loss function.

[0012] According to an embodiment of this application, the historical renewable energy consumption ratio contribution score is determined using the following formula: Based on a preset renewable energy consumption ratio contribution score quantification formula, the historical renewable energy consumption ratio contribution score corresponding to each of the historical day-ahead scheduling plan samples is determined; wherein, the renewable energy consumption ratio contribution score quantification formula includes:

[0013] (1)

[0014] In formula (1), The score represents the contribution of the proportion of new energy consumption. This indicates the proportion of new energy consumption corresponding to each of the historical pre-dated scheduling plan samples; This indicates the historical percentage of renewable energy consumption on a given day under the target scheduling plan.

[0015] A second aspect of this application provides a device for evaluating the contribution of renewable energy consumption ratio, comprising: a construction module for constructing a target scheduling plan that satisfies the annual renewable energy consumption ratio of a comprehensive energy base; a determination module for determining, based on the target scheduling plan, a historical renewable energy consumption ratio contribution score of the historical day-ahead scheduling plan data of the comprehensive energy base; a generation module for generating a training sample set based on the historical day-ahead scheduling plan data, historical day-ahead power prediction data corresponding to the historical day-ahead scheduling plan data, and the historical renewable energy consumption ratio contribution score; and a training module for training a preset hybrid CLA model using the training sample set to obtain a renewable energy consumption ratio contribution evaluation model; wherein, the CLA... Model A includes an input layer, a convolutional neural network (CNN) layer, a long short-term memory (LSTM) layer, an attention layer, and an output layer. The input layer outputs the input vectors corresponding to each training sample. These vectors are then processed by the CNN layer to generate feature vectors, which are then input into the LSTM layer. The LSTM layer and the attention layer learn from these feature vectors, and the output layer outputs the prediction results. A module is used to acquire the day-ahead scheduling plan data and the corresponding day-ahead power prediction data of the integrated energy base. A scoring module is used to input the day-ahead scheduling plan data and the corresponding day-ahead power prediction data into the renewable energy consumption ratio contribution evaluation model to output a renewable energy consumption ratio contribution score for the day-ahead scheduling plan data.

[0016] The new energy consumption ratio contribution assessment device provided in the second aspect of this application also has the following technical features, including:

[0017] According to an embodiment of this application, the training module is further configured to: input training samples from the training sample set into the input layer, and obtain input vectors corresponding to each training sample through the input layer; input the input vectors into the CNN layer, and perform feature extraction on the input vectors to filter out target feature vectors; obtain the target feature vectors, input the target feature vectors into the LSTM layer to obtain first output vectors corresponding to each target feature vector; input the first output vectors corresponding to each target feature vector into the attention layer, and filter the first output vectors according to the attention weight parameter values ​​of the first output vectors in the attention layer to obtain second output vectors; input the second output vectors into the output layer to determine the historical new energy consumption ratio contribution prediction score; adjust the model parameters of the CLA model according to the historical new energy consumption ratio contribution score and the historical new energy consumption ratio contribution prediction score in the training samples, and return to use the next training sample to continue training the CLA model after adjusting the model parameters until the training ends and the new energy consumption ratio contribution evaluation model is obtained.

[0018] According to one embodiment of this application, the CNN layer includes a convolutional layer and a dropout layer. The training module is further configured to: input the input vector into the convolutional layer to obtain multiple feature vectors of the input vector extracted by the convolutional layer; and input the multiple feature vectors of the input vector into the dropout layer to filter out a target feature vector from the multiple features.

[0019] According to one embodiment of this application, the training module is further configured to: obtain the historical new energy consumption ratio contribution prediction score, and perform inverse normalization processing on the historical new energy consumption ratio contribution prediction score.

[0020] According to an embodiment of this application, the determining module is further configured to: acquire historical day-ahead power prediction data samples corresponding to multiple historical sampling time points of the integrated energy base, each historical day-ahead power prediction data sample including: historical day-ahead photovoltaic power prediction data and historical day-ahead wind power prediction data; acquire historical day-ahead scheduling plan data samples corresponding to each of the historical day-ahead power prediction data samples for the next day, each historical day-ahead scheduling plan sample including: historical day-ahead scheduling plan photovoltaic proportion, historical day-ahead scheduling plan wind power proportion, and historical day-ahead scheduling plan thermal power proportion; and determine the historical new energy consumption ratio contribution score corresponding to each of the historical day-ahead scheduling plan samples based on the new energy consumption ratio of the target scheduling plan on the current day and the new energy consumption ratio corresponding to each of the historical day-ahead scheduling plan samples.

[0021] According to one embodiment of this application, the training module is further configured to: construct a loss function to represent the difference between the predicted score of the contribution of new energy consumption ratio output by the CLA model and the score of the contribution of new energy consumption ratio; and train the CLA model based on the training sample set with the minimum value of the loss function as the objective.

[0022] According to an embodiment of this application, the historical renewable energy consumption ratio contribution score is determined using the following formula: Based on a preset renewable energy consumption ratio contribution score quantification formula, the historical renewable energy consumption ratio contribution score corresponding to each of the historical day-ahead scheduling plan samples is determined; wherein, the renewable energy consumption ratio contribution score quantification formula includes:

[0023] (1)

[0024] In formula (1), The score represents the contribution of the proportion of new energy consumption. This indicates the proportion of new energy consumption corresponding to each of the historical pre-dated scheduling plan samples; This indicates the historical percentage of renewable energy consumption on a given day under the target scheduling plan.

[0025] A third aspect of this application provides an electronic device, including: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the new energy consumption ratio contribution assessment method provided in the first aspect of this application.

[0026] A fourth aspect of this application provides a non-transitory computer-readable storage medium storing computer instructions, the computer instructions being used to cause the computer to execute the new energy consumption ratio contribution assessment method provided in the first aspect of this application.

[0027] The fifth aspect of this application provides a computer program product, which, when executed by an instruction processor, performs the new energy consumption ratio contribution assessment method provided in the first aspect of this application.

[0028] The renewable energy consumption ratio contribution assessment method and apparatus provided in this application construct a target scheduling plan that meets the annual renewable energy consumption ratio of a comprehensive energy base, and determine the historical renewable energy consumption ratio contribution score of the comprehensive energy base based on the historical day-ahead scheduling plan data. Further, based on the historical day-ahead scheduling plan data, the corresponding historical day-ahead power prediction data, and the corresponding historical renewable energy consumption ratio contribution score, a training sample set is generated. A preset hybrid CLA model is trained based on the generated training sample set to obtain a trained renewable energy consumption ratio contribution assessment model. Further, the day-ahead scheduling plan data and corresponding day-ahead power prediction data corresponding to the target scheduling plan of the comprehensive energy base are obtained and input into the trained renewable energy consumption ratio contribution assessment model. Based on the model output, the renewable energy consumption ratio contribution score corresponding to the day-ahead scheduling plan data of the target scheduling plan of the comprehensive energy base is obtained. In this application, based on a trained new energy consumption ratio contribution assessment model, the contribution ratio of the day-ahead dispatch plan data corresponding to the target dispatch plan of the integrated energy base is scored, which improves the accuracy of the contribution ratio assessment of the new energy power consumed by the integrated energy base based on the day-ahead dispatch plan data in the total power consumption, thereby improving the accuracy of the day-ahead dispatch plan data corresponding to new energy and traditional energy in the integrated energy base and avoiding energy waste caused by unreasonable dispatch plans.

[0029] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0030] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0031] Figure 1 A flowchart illustrating the contribution assessment method for the proportion of new energy consumption according to an embodiment of this application;

[0032] Figure 2 A flowchart illustrating the contribution assessment method for the proportion of new energy consumption according to another embodiment of this application;

[0033] Figure 3 This is a schematic diagram of the structure of a new CLA model according to an embodiment of this application;

[0034] Figure 4 This is a schematic diagram of the dropout layer structure according to an embodiment of this application;

[0035] Figure 5 A schematic diagram of the structure of a new energy consumption ratio contribution assessment device according to an embodiment of this application;

[0036] Figure 6 This is a block diagram of an electronic device according to an embodiment of this application. Detailed Implementation

[0037] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.

[0038] The following description, with reference to the accompanying drawings, describes a method, apparatus, electronic device, and medium for assessing the contribution of new energy consumption ratio according to embodiments of this application.

[0039] Figure 1 This is a flowchart illustrating a method for assessing the contribution of renewable energy consumption ratio according to an embodiment of this application, as shown below. Figure 1 As shown, the method includes:

[0040] S101, to construct a target scheduling plan that meets the annual renewable energy consumption ratio of the comprehensive energy base.

[0041] In practice, integrated energy bases have set percentage requirements for the consumption of new energy sources each year. This can be understood as the total amount of new energy sources such as wind power and photovoltaic power consumed by the integrated energy base each year needing to reach the set percentage of the total electricity consumption for the corresponding year.

[0042] This set percentage can be defined as the annual renewable energy consumption percentage of the integrated energy base.

[0043] In this embodiment of the application, a corresponding new energy dispatch plan can be constructed for the integrated energy base according to the set annual new energy consumption ratio, and the new energy dispatch plan can be determined as the target dispatch plan to meet the annual new energy consumption ratio of the integrated energy base.

[0044] This can be understood as follows: based on the target scheduling plan, the proportion of renewable energy power consumed by the integrated energy base within the annual time frame corresponding to the target scheduling plan, in the total electricity consumed by the integrated energy base within that annual time frame, can meet the preset annual renewable energy consumption ratio.

[0045] As an example, the target scheduling plan for an integrated energy base can be as follows:

[0046]

[0047] In the above expression, The target scheduling plan for the integrated energy base , as well as Let J represent the output percentages of wind power, photovoltaic power, and thermal power at time j on day i, where j = 1, 2, ..., N.

[0048] S102, Based on the target scheduling plan, determine the historical new energy consumption ratio contribution score of the historical day-ahead scheduling plan data of the integrated energy base.

[0049] In this embodiment of the application, the target scheduling plan is the annual new energy scheduling plan of the integrated energy base. In this scenario, the integrated energy base also needs to construct a daily new energy scheduling plan. The daily new energy scheduling plan constructed by the new energy base based on the target scheduling plan can be determined as the day-ahead scheduling plan of the integrated energy base.

[0050] In order to achieve a reasonable construction of the day-ahead dispatch plan, the amount of renewable energy power consumed by the integrated energy base based on the day-ahead dispatch plan can be obtained, and the contribution of the renewable energy power to the total renewable energy power to be consumed by the integrated energy base in a year can be obtained. Then, the contribution score of the day-ahead dispatch plan can be calculated based on the contribution level.

[0051] Optionally, in order to achieve an accurate contribution score for the day-ahead scheduling plan corresponding to the target scheduling plan, a scoring model for contribution scoring can be obtained, and a trained scoring model can be obtained based on the scoring model to achieve an accurate contribution score for the day-ahead scheduling plan corresponding to the target scheduling plan of the integrated energy base.

[0052] In this embodiment of the application, historical day-ahead scheduling plan data of the integrated energy base can be obtained, and training samples for the scoring model can be constructed based on the historical day-ahead scheduling plan data. The contribution score of the historical day-ahead scheduling plan data in the target scheduling plan can be determined based on the historical day-ahead scheduling plan data and the target scheduling plan.

[0053] Among them, the contribution score can be determined as the historical renewable energy consumption ratio contribution score of the historical day-ahead dispatch plan data of the integrated energy base.

[0054] S103. Generate a training sample set based on historical day-ahead scheduling plan data, historical day-ahead power prediction data corresponding to the historical day-ahead scheduling plan data, and historical renewable energy consumption ratio contribution scores.

[0055] In this embodiment of the application, training samples for the scoring model can be constructed based on the obtained historical day-ahead scheduling plan data and the historical renewable energy consumption contribution score corresponding to the historical day-ahead scheduling plan data.

[0056] Optionally, the predicted power data corresponding to the historical day-ahead scheduling plan data can also be obtained, and the predicted power data can be identified as the historical day-ahead power prediction data corresponding to the historical day-ahead scheduling plan data.

[0057] Furthermore, training samples for the contribution scoring model can be constructed based on historical day-ahead scheduling plan data, historical day-ahead power prediction data corresponding to the historical day-ahead scheduling plan data, and historical contribution scores of renewable energy consumption ratio.

[0058] As an example, the training samples for a contribution score model constructed based on historical day-ahead scheduling plan data, corresponding historical day-ahead power prediction data, and historical renewable energy consumption contribution scores can be represented by the following formula:

[0059]

[0060] In the above expression, For wind power forecast data in historical daytime power forecast data, Photovoltaic power forecast data from historical day-ahead power forecast data, The wind power percentage data in the historical day-ahead dispatch plan data, The photovoltaic percentage data in the historical pre-schedule scheduling data, This data represents the proportion of thermal power generation in the historical pre-scheduling data. The historical renewable energy consumption ratio is contributed to the score based on the scheduling plan data of this historical date.

[0061] S104: The pre-defined hybrid CLA model is trained using a training sample set to obtain a contribution assessment model for the proportion of new energy consumption. The CLA model includes an input layer, a convolutional neural network (CNN) layer, a long short-term memory (LSTM) layer, an attention layer, and an output layer. The input layer outputs the input vectors corresponding to the training samples. These vectors are then processed by the CNN layer to generate feature vectors, which are then input into the LSTM layer. The LSTM layer and the attention layer learn from the feature vectors, and the output layer outputs the prediction results.

[0062] In this embodiment of the application, a preset hybrid CLA model can be used as the contribution scoring model to be trained. In this scenario, the hybrid CLA model to be trained can be trained based on the historical day-ahead scheduling plan data, the historical day-ahead power prediction data corresponding to the historical day-ahead scheduling plan data, and the training samples of the contribution scoring model constructed by the historical new energy consumption ratio contribution score.

[0063] Optionally, the training termination condition of the hybrid CLA model can be set based on the training rounds. The training rounds of the hybrid CLA model can be monitored and recorded. When the training round of a certain round is completed and the set training termination condition is met, the training of the hybrid CLA model can be terminated. The hybrid CLA model obtained in the last training round is taken as the trained model and is determined as the new energy consumption ratio contribution assessment model.

[0064] Optionally, the training termination condition of the hybrid CLA model can be set based on the output results of the model training. When the output results of the model training in a certain round meet the set model training termination condition, the model training of the hybrid CLA model can be terminated, and the hybrid CLA model obtained in the last training round can be used as the trained model. The trained model can then be determined as the new energy consumption ratio contribution assessment model.

[0065] In this embodiment of the application, the hybrid CLA model may include an input layer, a convolutional neural network (CNN) layer, a long short-term memory (LSTM) artificial neural network layer, an attention layer, and an output layer. The input layer of the hybrid CLA model can process the training samples of the input model to output the corresponding input vectors of the training samples.

[0066] Furthermore, the feature vectors generated and output by the CNN layer are input into the LSTM layer, and the feature vectors are learned through the LSTM layer and the attention layer, and then the prediction results are output through the output layer.

[0067] The output of the hybrid CLA model is the historical contribution score of renewable energy consumption to the historical day-ahead scheduling plan data corresponding to the training samples of the input model.

[0068] As an example, the training samples input to the hybrid CLA model can be represented by the following formula:

[0069]

[0070] In the above expression, For wind power forecast data in historical daytime power forecast data, Photovoltaic power forecast data from historical day-ahead power forecast data, The wind power percentage data in the historical day-ahead dispatch plan data, The photovoltaic percentage data in the historical pre-schedule scheduling data, This data represents the proportion of thermal power generation in the historical pre-scheduling data. The historical renewable energy consumption ratio of the scheduling plan data before this historical date contributes to the score, which can include... , , , and As input to the hybrid CLA model, As the output of the hybrid CLA model.

[0071] S105: Obtain the day-ahead dispatch plan data and corresponding day-ahead power forecast data of the integrated energy base.

[0072] In this embodiment of the application, the training samples for training the hybrid CLA model are obtained through historical day-ahead scheduling plan data, historical day-ahead power prediction data corresponding to the historical day-ahead scheduling plan data, and historical renewable energy consumption contribution scores.

[0073] Optionally, in order to achieve an accurate contribution score for the day-ahead scheduling plan corresponding to the target scheduling plan of the integrated energy base, the day-ahead scheduling plan data corresponding to the target scheduling plan of the integrated energy base, and the day-ahead power prediction data corresponding to the day-ahead scheduling plan data of the integrated energy base can be obtained.

[0074] In this embodiment of the application, the energy dispatch plan data of the integrated energy base for each natural day can be obtained based on the target dispatch plan, thereby obtaining the day-ahead dispatch plan data of the integrated energy base.

[0075] Optionally, the new energy dispatch data and traditional energy dispatch data of the integrated energy base for each natural day can be obtained to obtain the energy dispatch plan data of the integrated energy base for each natural day, and then the day-ahead dispatch plan data of the integrated energy base can be obtained.

[0076] Specifically, based on the target scheduling plan, the scheduling data of new energy sources such as wind power and photovoltaic power, as well as the scheduling data of traditional energy sources such as thermal power, can be obtained for each natural day, so as to obtain the energy scheduling plan data of the integrated energy base for each natural day, and then obtain the day-ahead scheduling plan data of the integrated energy base.

[0077] Furthermore, based on the acquired day-ahead scheduling plan data of the integrated energy base for each natural day, the power of the integrated energy base for each natural day is predicted, thereby obtaining the power prediction data of the integrated energy base for each natural day, and thus obtaining the day-ahead power prediction data of the integrated energy base for each natural day.

[0078] Specifically, based on the day-ahead scheduling plan data, wind power forecast data and photovoltaic power forecast data of the integrated energy base can be obtained for each natural day, thereby obtaining the day-ahead power forecast data of the integrated energy base for each natural day.

[0079] S106. Input the day-ahead scheduling plan data and the corresponding day-ahead power forecast data into the renewable energy consumption ratio contribution assessment model to output the renewable energy consumption ratio contribution score of the day-ahead scheduling plan data.

[0080] In this embodiment of the application, the day-ahead scheduling plan data and the corresponding day-ahead power prediction data corresponding to the target scheduling plan of the integrated energy base can be input into the trained new energy consumption ratio contribution evaluation model.

[0081] Furthermore, based on the trained new energy consumption ratio contribution assessment model, the day-ahead dispatch plan data of the integrated energy base is evaluated to obtain the new energy consumption ratio contribution score of the day-ahead dispatch plan data corresponding to the target dispatch plan of the integrated energy base.

[0082] The proposed method for assessing the contribution of renewable energy consumption ratio in this application constructs a target scheduling plan that meets the annual renewable energy consumption ratio of a comprehensive energy base, and determines the historical renewable energy consumption ratio contribution score based on the historical day-ahead scheduling plan data of the comprehensive energy base. Further, based on the historical day-ahead scheduling plan data, the corresponding historical day-ahead power prediction data, and the corresponding historical renewable energy consumption ratio contribution score, a training sample set is generated. A pre-defined hybrid CLA model is then trained using this training sample set to obtain a trained renewable energy consumption ratio contribution assessment model. Further, the day-ahead scheduling plan data and corresponding day-ahead power prediction data corresponding to the target scheduling plan of the comprehensive energy base are obtained and input into the trained renewable energy consumption ratio contribution assessment model. Based on the model output, the renewable energy consumption ratio contribution score corresponding to the day-ahead scheduling plan data of the target scheduling plan of the comprehensive energy base is obtained. In this application, based on a trained new energy consumption ratio contribution assessment model, the contribution ratio of the day-ahead dispatch plan data corresponding to the target dispatch plan of the integrated energy base is scored, which improves the accuracy of the contribution ratio assessment of the new energy power consumed by the integrated energy base based on the day-ahead dispatch plan data in the total power consumption, thereby improving the accuracy of the day-ahead dispatch plan data corresponding to new energy and traditional energy in the integrated energy base and avoiding energy waste caused by unreasonable dispatch plans.

[0083] In the above embodiments, the assessment model for the contribution of new energy consumption ratio can be combined with... Figure 2 To understand further, Figure 2 This is a flowchart illustrating another embodiment of the renewable energy consumption ratio contribution assessment method of this application, as shown below. Figure 2 As shown, the method includes:

[0084] S201, input the training samples in the training sample set into the input layer, and obtain the input vector corresponding to each training sample through the input layer.

[0085] like Figure 3 As shown, training samples from the training sample set can be input into the input layer of the hybrid CLA model, and the input vector corresponding to each data item carried in the training samples can be obtained through the input layer.

[0086] Based on the above examples, it can be seen that , , , and For training samples of the hybrid CLA model, it is possible to , , , and Input the input layer of the hybrid CLA model, and then obtain... , , , and Each has its corresponding input vector.

[0087] In some implementations, training samples can be fed into the CLA model through the input layer to obtain an input vector transformed from the training samples. For example, if the length of the batch input training samples is set to m, then the input vector is... .

[0088] It should be noted that before training the pre-defined hybrid CLA model using the training sample set, a loss function can be constructed to represent the difference between the predicted score of the contribution ratio of new energy consumption and the score of the contribution ratio of new energy consumption output by the CLA model. This loss function can be as follows:

[0089]

[0090] In this scenario, the CLA model is trained based on the training sample set with the goal of minimizing the loss function.

[0091] S202, the input vector is fed into the CNN layer, and features are extracted from the input vector to select the target feature vector.

[0092] Optionally, the input vectors corresponding to the training samples obtained from the input layer can be input into the... Figure 3 The CNN layer shown.

[0093] like Figure 3 As shown, the CNN layer includes convolutional layers and dropout layers. In this scenario, the input vector can be fed into the convolutional layer to obtain multiple feature vectors of the input vector extracted by the convolutional layer.

[0094] The input vector can be fed into a convolutional layer to obtain multiple feature vectors of the input vector extracted by the convolutional layer, and then the multiple feature vectors of the input vector can be fed into a dropout layer to filter out the target feature vector from multiple features.

[0095] Optionally, in a scenario where the data dimension of solar power generation is 1-dimensional, the convolutional layers included in the CNN layer can be 1-dimensional convolutions with a kernel size of 3. In this scenario, the ReLU activation parameter can be used to obtain multiple feature vectors of the input vector.

[0096] Furthermore, multiple feature vectors of the input vector can be fed into the dropout layer to filter out the target feature vector from multiple features.

[0097] In some implementations, the dropout layers included in a CNN layer can be as follows: Figure 4 As shown, the dropout layer can be set to 0.2. In this scenario, half of the hidden neurons in the dropout layer can be temporarily and randomly removed while the input and output neurons remain unchanged. Figure 4 As shown, Figure 4 The circles in the middle represent neurons that were not deleted, while the circles with crosses represent neurons that were deleted.

[0098] like Figure 4 As shown, the input neurons can be propagated forward through the modified network, and the resulting loss can be propagated backward through the modified network. After a small batch of training samples has completed this process, the parameters (w, b) of the neurons that have not been deleted are updated according to the stochastic gradient descent method.

[0099] It should be noted that after obtaining the updated parameters (w, b), in order to avoid overfitting of the trained model, some implementations can repeatedly perform the following process: restore the deleted neurons, where the deleted neurons remain unchanged, while the neurons that were not deleted have been updated; temporarily delete a subset of half the size of the hidden neurons, and back up the parameters of the deleted neurons; for a small batch of training samples, perform forward propagation and then backward propagation of the loss, and update the parameters (w, b) according to the stochastic gradient descent method to solve the overfitting problem of different networks.

[0100] Where w represents the parameter weights in the neural network, and b represents the bias in the neural network.

[0101] In other implementations, with the dropout layer set to 0.2, if the number of neurons is n, then 0.2n neurons can be deleted. One way to delete neurons is to change the activation function value of the neuron in the network to 0 with probability p. If the output vector length is i at this time, then the target feature vector is... In this context, dropout neurons compute activation function values ​​within the network. One calculation method is as follows:

[0102]

[0103] The Bernoulli function generates a probability vector r, which is a randomly generated vector of 0s and 1s.

[0104] S203, Obtain the target feature vector, input the target feature vector into the LSTM layer to obtain the first output vector corresponding to each target feature vector.

[0105] Optionally, the acquired target feature vector can be input into an LSTM layer. Through the LSTM layer and a bidirectional long short-term memory (biLSTM) artificial neural network structure, the photovoltaic power generation behavior characteristics can be learned. If the length of the first output vector is j, then the first output vector of the LSTM layer is... ,calculate One calculation method is as follows:

[0106]

[0107] The LSTM layer needs to be connected to the dropout layer and the max pooling layer. Max is the maximum value function in the max pooling layer, br is the bias of the pooling layer, and L is the output of the LSTM layer, where are the weights and biases of the LSTM layer, respectively.

[0108] S204, input the first output vector corresponding to each of the target feature vectors into the attention layer, and filter the first output vectors according to the attention weight parameter values ​​of the first output vectors in the attention layer to obtain the second output vector.

[0109] In this embodiment, the first output vector corresponding to each target feature vector can be input into the attention layer. Then, according to the weight allocation principle in the attention layer, the attention weight parameter value of the first output vector is allocated to obtain the attention weight parameter value of the first output vector. Based on the attention weight parameter value of the first output vector, the first output vector is filtered to obtain the second output vector.

[0110] In some implementations, if the length of the second output vector is k, then the second output vector... for .

[0111] S205, input the second output vector into the output layer to determine the historical contribution prediction score of renewable energy consumption ratio.

[0112] In this embodiment, after the second output vector is input to the output layer, the output layer obtains the historical renewable energy consumption ratio contribution prediction score through a fully connected layer. Assuming the compensation predicted by the output layer is n, the historical renewable energy consumption ratio contribution prediction score Y is... An exemplary way to calculate Y is as follows:

[0113]

[0114] in, For output layer weights, f is the output layer bias, and f is the activation function of the fully connected layer.

[0115] In other implementations, after inputting the second output vector into the output layer to determine the historical renewable energy consumption ratio contribution prediction score, the historical renewable energy consumption ratio contribution prediction score can also be obtained and inversely normalized.

[0116] In this embodiment of the application, after obtaining the historical contribution prediction score of the renewable energy consumption ratio, in order to accurately train the model, the historical contribution prediction score of the renewable energy consumption ratio can be denormalized to obtain the actual prediction value after denormalization.

[0117] One exemplary method for inverse normalization is as follows:

[0118]

[0119] in, y represents the historical renewable energy consumption ratio contribution prediction score before inverse normalization, obtained by the CLA model, while y represents the historical renewable energy consumption ratio contribution prediction score after inverse normalization. These are the minimum and maximum values ​​in the historical output data before normalization.

[0120] It should be noted that the historical contribution score of renewable energy consumption ratio is determined using the following formula:

[0121] Based on the preset formula for quantifying the contribution of historical renewable energy consumption ratio, the contribution score of historical renewable energy consumption ratio corresponding to each historical day-ahead scheduling plan sample is determined.

[0122] The quantitative formula for the historical contribution of renewable energy consumption includes:

[0123]

[0124] In the above formula, This indicates the contribution score of the historical proportion of new energy consumption. This indicates the proportion of new energy consumption corresponding to the dispatch plan samples for each historical date. This indicates the historical percentage of renewable energy consumption on a given day under the target scheduling plan.

[0125] S206. Based on the historical renewable energy consumption ratio contribution score and the historical renewable energy consumption ratio contribution prediction score in the training samples, adjust the model parameters of the CLA model, and then return to use the next training sample to continue training the CLA model with adjusted model parameters until the training ends and the renewable energy consumption ratio contribution evaluation model is obtained.

[0126] Optionally, the training loss of the CLA model in the current round can be obtained based on the historical renewable energy consumption ratio contribution score and the historical renewable energy consumption ratio contribution prediction score in the training samples, and the model parameters of the CLA model can be adjusted based on the obtained training loss.

[0127] Furthermore, the CLA model with adjusted parameters is trained on the next training sample in the training sample set until the training termination condition is met, at which point the training of the CLA model ends, and a well-trained new energy consumption ratio contribution assessment model is obtained.

[0128] Optionally, the training termination condition of the hybrid CLA model can be set based on the training rounds. The training rounds of the hybrid CLA model can be monitored and recorded. When the training round of a certain round is completed and the set training termination condition is met, the training of the hybrid CLA model can be terminated. The hybrid CLA model obtained in the last training round is taken as the trained model and is determined as the new energy consumption ratio contribution assessment model.

[0129] Optionally, the training termination condition of the hybrid CLA model can be set based on the output results of the model training. When the output results of the model training in a certain round meet the set model training termination condition, the model training of the hybrid CLA model can be terminated, and the hybrid CLA model obtained in the last training round can be used as the trained model. The trained model can then be determined as the new energy consumption ratio contribution assessment model.

[0130] The proposed method for evaluating the contribution of renewable energy consumption ratio involves inputting training samples from a training sample set into an input layer. The input layer generates input vectors for each training sample. Optionally, these input vectors are input into a CNN layer, where feature extraction is performed to obtain target feature vectors. Further, the target feature vectors are input into an LSTM layer to obtain first output vectors for each target feature vector. These first output vectors are then input into an attention layer, where they are filtered based on weight parameters to obtain second output vectors. The second output vectors are then input into an output layer to determine historical renewable energy consumption ratio prediction scores. Based on these historical renewable energy consumption ratio prediction scores and the historical renewable energy consumption ratio prediction scores from the training samples, the model parameters of the CLA model are adjusted. The model is then trained again using the next training sample with adjusted parameters until the training is complete, resulting in a well-trained renewable energy consumption ratio evaluation model.

[0131] In this application, a CLA model is obtained based on training samples and a multi-model fusion algorithm, which improves the flexibility of the attention mechanism in selecting weights. This allows the new energy consumption ratio contribution assessment model trained on the CLA model to more accurately score the contribution ratio of the day-ahead scheduling plan data corresponding to the target scheduling plan of the integrated energy base. This improves the accuracy of the assessment of the contribution ratio of new energy power consumed by the integrated energy base based on the day-ahead scheduling plan data in the total consumed power, thereby improving the accuracy of the day-ahead scheduling plan data corresponding to new energy and traditional energy in the integrated energy base and avoiding energy waste caused by unreasonable scheduling plans.

[0132] Corresponding to the new energy consumption ratio contribution assessment methods proposed in the above embodiments, one embodiment of this application also proposes a new energy consumption ratio contribution assessment device. Since the new energy consumption ratio contribution assessment device proposed in this application corresponds to the new energy consumption ratio contribution assessment methods proposed in the above embodiments, the implementation methods of the above-mentioned new energy consumption ratio contribution assessment methods are also applicable to the new energy consumption ratio contribution assessment device proposed in this application, and will not be described in detail in the following embodiments.

[0133] Figure 5 This is a schematic diagram of the structure of a new energy consumption ratio contribution assessment device according to an embodiment of this application, as shown below. Figure 5 As shown, the renewable energy consumption ratio contribution assessment device 500 includes a construction module 51, a determination module 52, a generation module 53, a training module 54, an acquisition module 55, and a scoring module 56, wherein:

[0134] Module 51 is used to construct a scheduling plan that meets the annual target for the proportion of new energy consumption in the integrated energy base.

[0135] Module 52 is used to determine the historical renewable energy consumption ratio contribution score of the historical day-ahead scheduling plan data of the integrated energy base based on the target scheduling plan;

[0136] The generation module 53 is used to generate a training sample set based on historical day-ahead scheduling plan data, historical day-ahead power prediction data corresponding to the historical day-ahead scheduling plan data, and historical renewable energy consumption ratio contribution scores.

[0137] Training module 54 is used to train a pre-defined hybrid CLA model using a training sample set to obtain a contribution evaluation model for the proportion of new energy consumption. The CLA model includes an input layer, a convolutional neural network (CNN) layer, a long short-term memory (LSTM) artificial neural network layer, an attention layer, and an output layer. The input layer outputs the input vectors corresponding to the training samples, which are then processed by the CNN layer to generate feature vectors and input into the LSTM layer. The LSTM layer and the attention layer learn the feature vectors, and the output layer outputs the prediction results.

[0138] The acquisition module 55 is used to acquire the day-ahead scheduling plan data and the corresponding day-ahead power forecast data of the integrated energy base;

[0139] The scoring module 56 is used to input the day-ahead scheduling plan data and the corresponding day-ahead power prediction data into the renewable energy consumption ratio contribution assessment model, so as to output the renewable energy consumption ratio contribution score of the day-ahead scheduling plan data.

[0140] In this embodiment, the training module 54 is further configured to: input training samples from the training sample set into the input layer, and obtain the input vectors corresponding to each training sample through the input layer; input the input vectors into the CNN layer, and perform feature extraction on the input vectors to filter out target feature vectors; obtain the target feature vectors, input the target feature vectors into the LSTM layer to obtain the first output vectors corresponding to each target feature vector; input the first output vectors corresponding to each target feature vector into the attention layer, and filter the first output vectors according to the attention weight parameter values ​​of the first output vectors in the attention layer to obtain the second output vectors; input the second output vectors into the output layer to determine the historical new energy consumption ratio contribution prediction score; adjust the model parameters of the CLA model according to the historical new energy consumption ratio contribution score and the historical new energy consumption ratio contribution prediction score in the training samples, and return to use the next training sample to continue training the CLA model after adjusting the model parameters, until the training ends and the new energy consumption ratio contribution evaluation model is obtained.

[0141] In this embodiment of the application, the CNN layer includes a convolutional layer and a dropout layer. The training module 54 is further configured to: input the input vector into the convolutional layer to obtain multiple feature vectors of the input vector extracted by the convolutional layer; and input the multiple feature vectors of the input vector into the dropout layer to filter out the target feature vector from the multiple features.

[0142] In this embodiment of the application, the training module 54 is further used to: obtain the historical new energy consumption ratio contribution prediction score, and perform inverse normalization processing on the historical new energy consumption ratio contribution prediction score.

[0143] In this embodiment of the application, the determining module 52 is further configured to: obtain historical day-ahead power prediction data samples corresponding to multiple historical sampling time points of the integrated energy base, each historical day-ahead power prediction data sample including: historical day-ahead photovoltaic power prediction data and historical day-ahead wind power prediction data; obtain historical day-ahead scheduling plan data samples corresponding to each historical day-ahead power prediction data sample for the next day, each historical day-ahead scheduling plan sample including: historical day-ahead scheduling plan photovoltaic proportion, historical day-ahead scheduling plan wind power proportion, and historical day-ahead scheduling plan thermal power proportion; and determine the historical new energy consumption ratio contribution score corresponding to each historical day-ahead scheduling plan sample based on the new energy consumption ratio of the target scheduling plan on the current day and the new energy consumption ratio corresponding to each historical day-ahead scheduling plan sample.

[0144] In this embodiment of the application, the training module 54 is further configured to: construct a loss function to represent the difference between the predicted score of the contribution of new energy consumption ratio output by the CLA model and the score of the contribution of new energy consumption ratio; and train the CLA model based on the training sample set with the goal of minimizing the loss function.

[0145] In this embodiment of the application, the historical renewable energy consumption ratio contribution score is determined using the following formula: Based on a preset renewable energy consumption ratio contribution score quantification formula, the historical renewable energy consumption ratio contribution score corresponding to each historical day-ahead scheduling plan sample is determined; wherein, the renewable energy consumption ratio contribution score quantification formula includes:

[0146] (1)

[0147] In formula (1), The score represents the contribution of the proportion of new energy consumption. This indicates the proportion of new energy consumption corresponding to the dispatch plan samples for each historical date; This indicates the historical percentage of renewable energy consumption on a given day under the target scheduling plan.

[0148] The renewable energy consumption ratio contribution assessment device proposed in this application constructs a target scheduling plan that meets the annual renewable energy consumption ratio of a comprehensive energy base, and determines the historical renewable energy consumption ratio contribution score of the comprehensive energy base based on the historical day-ahead scheduling plan data. Further, based on the historical day-ahead scheduling plan data, the corresponding historical day-ahead power prediction data, and the corresponding historical renewable energy consumption ratio contribution score, a training sample set is generated. A preset hybrid CLA model is trained based on the generated training sample set to obtain a trained renewable energy consumption ratio contribution assessment model. Further, the day-ahead scheduling plan data and corresponding day-ahead power prediction data corresponding to the target scheduling plan of the comprehensive energy base are obtained and input into the trained renewable energy consumption ratio contribution assessment model. Based on the model output, the renewable energy consumption ratio contribution score corresponding to the day-ahead scheduling plan data of the target scheduling plan of the comprehensive energy base is obtained. In this application, based on a trained new energy consumption ratio contribution assessment model, the contribution ratio of the day-ahead dispatch plan data corresponding to the target dispatch plan of the integrated energy base is scored, which improves the accuracy of the contribution ratio assessment of the new energy power consumed by the integrated energy base based on the day-ahead dispatch plan data in the total power consumption, thereby improving the accuracy of the day-ahead dispatch plan data corresponding to new energy and traditional energy in the integrated energy base and avoiding energy waste caused by unreasonable dispatch plans.

[0149] To achieve the above embodiments, this application also provides an electronic device, a computer-readable storage medium, and a computer program product.

[0150] Figure 6 This is a block diagram of an electronic device according to an embodiment of this application, such as... Figure 6 As shown, device 600 includes a memory 61, a processor 62, and a computer program stored on the memory 61 and executable on the processor 62. When the processor 62 executes program instructions, it performs... Figures 1 to 4 The embodiment of the new energy consumption ratio contribution assessment method.

[0151] To implement the above embodiments, this application also provides a non-transitory computer-readable storage medium storing computer instructions for causing a computer to execute... Figures 1 to 4 The embodiment of the new energy consumption ratio contribution assessment method.

[0152] To implement the above embodiments, this application also provides a computer program product that, when the instruction processor in the computer program product is executed, performs... Figures 1 to 4 The embodiment of the new energy consumption ratio contribution assessment method.

[0153] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0154] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0155] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.

[0156] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0157] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0158] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0159] Furthermore, the functional units in the various embodiments of this application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0160] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of this application.

Claims

1. A method for assessing the contribution of new energy consumption ratio, characterized in that, The method includes: Construct a scheduling plan that meets the annual target for the proportion of new energy consumption in the comprehensive energy base; Based on the target scheduling plan, determine the historical renewable energy consumption ratio contribution score of the historical day-ahead scheduling plan data of the integrated energy base; A training sample set is generated based on the historical day-ahead scheduling plan data, the historical day-ahead power prediction data corresponding to the historical day-ahead scheduling plan data, and the historical renewable energy consumption ratio contribution score. The training sample set is used to train a preset hybrid CLA model to obtain a contribution evaluation model for the proportion of new energy consumption. The CLA model includes an input layer, a convolutional neural network (CNN) layer, a long short-term memory (LSTM) artificial neural network layer, an attention layer, and an output layer. The input layer outputs the input vectors corresponding to the training samples. The CNN layer generates feature vectors, which are then input into the LSTM layer. The LSTM layer and the attention layer learn the feature vectors, and the output layer outputs the prediction results. Obtain the day-ahead scheduling plan data and the corresponding day-ahead power forecast data of the integrated energy base; The day-ahead scheduling plan data and the corresponding day-ahead power prediction data are input into the renewable energy consumption ratio contribution evaluation model to output the renewable energy consumption ratio contribution score of the day-ahead scheduling plan data. The historical contribution score of renewable energy consumption ratio is determined using the following formula: Based on the preset historical renewable energy consumption ratio contribution score quantification formula, the historical renewable energy consumption ratio contribution score corresponding to each of the historical day-ahead scheduling plan samples is determined. The quantitative formula for the historical renewable energy consumption ratio contribution score includes: In formula (1), This indicates the contribution score of the historical proportion of new energy consumption; This indicates the proportion of new energy consumption corresponding to each of the historical pre-dated scheduling plan samples; This indicates the historical percentage of renewable energy consumption on a given day under the target scheduling plan.

2. The method for assessing the contribution of new energy consumption ratio according to claim 1, characterized in that, The step of training a pre-defined hybrid CLA model using the training sample set to obtain a new energy consumption ratio contribution assessment model includes: The training samples in the training sample set are input into the input layer, and the input vector corresponding to each training sample is obtained through the input layer. The input vector is fed into the CNN layer, and features are extracted from the input vector to filter out the target feature vector; The target feature vector is obtained, and the target feature vector is input into the LSTM layer to obtain the first output vector corresponding to each of the target feature vectors; The first output vector corresponding to each of the target feature vectors is input into the attention layer, and the first output vector is filtered according to the attention weight parameter value of the first output vector in the attention layer to obtain the second output vector; The second output vector is input into the output layer to determine the historical new energy consumption ratio contribution prediction score; Based on the historical renewable energy consumption ratio contribution score and the historical renewable energy consumption ratio contribution prediction score in the training samples, the model parameters of the CLA model are adjusted, so that the CLA model with the adjusted model parameters is used again to continue training with the next training sample until the training ends and the renewable energy consumption ratio contribution evaluation model is obtained.

3. The method for assessing the contribution of new energy consumption ratio according to claim 2, characterized in that, The CNN layer includes convolutional layers and dropout layers. The input vector is fed into the CNN layer, and feature extraction is performed on the input vector to filter out target feature vectors, including: The input vector is fed into the convolutional layer to obtain multiple feature vectors of the input vector extracted by the convolutional layer; Multiple feature vectors of the input vector are input into the dropout layer to filter out the target feature vector from the multiple features.

4. The method according to claim 2, characterized in that, After inputting the second output vector into the output layer to determine the historical renewable energy consumption ratio contribution prediction score, the method further includes: Obtain the predicted score of the historical renewable energy consumption ratio contribution, and perform inverse normalization on the predicted score of the historical renewable energy consumption ratio contribution.

5. The method for assessing the contribution of new energy consumption ratio according to any one of claims 1-4, characterized in that, The historical renewable energy consumption ratio contribution score, determined based on the historical day-ahead scheduling plan data of the integrated energy base according to the target scheduling plan, includes: Obtain historical day-ahead power prediction data samples corresponding to multiple historical sampling time points of the integrated energy base. Each historical day-ahead power prediction data sample includes: historical day-ahead photovoltaic power prediction data and historical day-ahead wind power prediction data. Obtain the historical day-ahead scheduling plan data sample corresponding to each of the historical day-ahead power prediction data samples for the next day. Each historical day-ahead scheduling plan sample includes: the historical day-ahead scheduling plan photovoltaic ratio, the historical day-ahead scheduling plan wind power ratio, and the historical day-ahead scheduling plan thermal power ratio. Based on the proportion of new energy consumption on the current day under the target scheduling plan and the proportion of new energy consumption corresponding to each of the historical day-ahead scheduling plan samples, the contribution score of the historical new energy consumption ratio corresponding to each of the historical day-ahead scheduling plan samples is determined.

6. The method for assessing the contribution of new energy consumption ratio according to any one of claims 1-4, characterized in that, Before training the preset hybrid CLA model using the training sample set, the method further includes: Construct a loss function to represent the difference between the predicted score of the contribution ratio of new energy consumption output by the CLA model and the score of the contribution ratio of new energy consumption. The CLA model is trained based on the training sample set with the goal of minimizing the loss function.

7. A device for assessing the contribution of new energy consumption ratio, characterized in that, include: The module is used to build a scheduling plan that meets the annual target for the proportion of new energy consumption in the integrated energy base; The determination module is used to determine the historical new energy consumption ratio contribution score of the historical day-ahead scheduling plan data of the integrated energy base based on the target scheduling plan; The generation module is used to generate a training sample set based on the historical day-ahead scheduling plan data, the historical day-ahead power prediction data corresponding to the historical day-ahead scheduling plan data, and the historical renewable energy consumption ratio contribution score. The training module is used to train a preset hybrid CLA model using the training sample set to obtain a new energy consumption ratio contribution evaluation model. The CLA model includes an input layer, a convolutional neural network (CNN) layer, a long short-term memory (LSTM) artificial neural network layer, an attention layer, and an output layer. The input layer outputs the input vectors corresponding to the training samples, which are then processed by the CNN layer to generate feature vectors and input into the LSTM layer. The LSTM layer and the attention layer learn the feature vectors, and the output layer outputs the prediction results. The acquisition module is used to acquire the day-ahead scheduling plan data and the corresponding day-ahead power forecast data of the integrated energy base; The scoring module is used to input the day-ahead scheduling plan data and the corresponding day-ahead power prediction data into the renewable energy consumption ratio contribution assessment model, so as to output the renewable energy consumption ratio contribution score of the day-ahead scheduling plan data; The historical contribution score of renewable energy consumption ratio is determined using the following formula: Based on the preset formula for the contribution score of renewable energy consumption ratio, the historical renewable energy consumption ratio contribution score corresponding to each historical day-ahead scheduling plan sample is determined. The quantitative formula for the contribution score of the new energy consumption ratio includes: （1） In formula (1), The score represents the contribution of the proportion of new energy consumption. This indicates the proportion of new energy consumption corresponding to each of the historical pre-dated scheduling plan samples; This indicates the historical percentage of renewable energy consumption on a given day under the target scheduling plan.

8. The new energy consumption ratio contribution assessment device according to claim 7, characterized in that, The training module is also used for: The training samples in the training sample set are input into the input layer, and the input vector corresponding to each training sample is obtained through the input layer. The input vector is fed into the CNN layer, and features are extracted from the input vector to filter out the target feature vector; The target feature vector is obtained, and the target feature vector is input into the LSTM layer to obtain the first output vector corresponding to each of the target feature vectors; The first output vector corresponding to each of the target feature vectors is input into the attention layer, and the first output vector is filtered according to the attention weight parameter value of the first output vector in the attention layer to obtain the second output vector; The second output vector is input into the output layer to determine the historical new energy consumption ratio contribution prediction score; Based on the historical renewable energy consumption ratio contribution score and the historical renewable energy consumption ratio contribution prediction score in the training samples, the model parameters of the CLA model are adjusted, so that the CLA model with the adjusted model parameters is used again to continue training with the next training sample until the training ends and the renewable energy consumption ratio contribution evaluation model is obtained.

9. The new energy consumption ratio contribution assessment device according to claim 8, characterized in that, The CNN layer includes convolutional layers and dropout layers, and the training module is further used for: The input vector is fed into the convolutional layer to obtain multiple feature vectors of the input vector extracted by the convolutional layer; Multiple feature vectors of the input vector are input into the dropout layer to filter out the target feature vector from the multiple features.

10. The apparatus according to claim 8, characterized in that, The training module is also used for: Obtain the predicted score of the historical renewable energy consumption ratio contribution, and perform inverse normalization on the predicted score of the historical renewable energy consumption ratio contribution.

11. The new energy consumption ratio contribution assessment device according to any one of claims 7-10, characterized in that, The determining module is further configured to: Obtain historical day-ahead power prediction data samples corresponding to multiple historical sampling time points of the integrated energy base. Each historical day-ahead power prediction data sample includes: historical day-ahead photovoltaic power prediction data and historical day-ahead wind power prediction data. Obtain the historical day-ahead scheduling plan data sample corresponding to each of the historical day-ahead power prediction data samples for the next day. Each historical day-ahead scheduling plan sample includes: the historical day-ahead scheduling plan photovoltaic ratio, the historical day-ahead scheduling plan wind power ratio, and the historical day-ahead scheduling plan thermal power ratio. Based on the proportion of new energy consumption on the current day under the target scheduling plan and the proportion of new energy consumption corresponding to each of the historical day-ahead scheduling plan samples, the contribution score of the historical new energy consumption ratio corresponding to each of the historical day-ahead scheduling plan samples is determined.

12. The new energy consumption ratio contribution assessment device according to any one of claims 7-10, characterized in that, The training module is also used for: Construct a loss function to represent the difference between the predicted score of the contribution ratio of new energy consumption output by the CLA model and the score of the contribution ratio of new energy consumption. The CLA model is trained based on the training sample set with the goal of minimizing the loss function.

13. An electronic device, characterized in that, include: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-6.

14. A non-transitory computer-readable storage medium storing computer instructions, characterized in that, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-6.

Images