Enhanced fusion load prediction method, system and equipment based on parallel variable component sub-circuit and multilayer perception neural network, and medium

Abstract

The invention is suitable for the field of power load prediction, and discloses an enhanced fusion load prediction method, system and device based on a parallel variable component sub-circuit and a multilayer perception neural network, and a medium, and the method comprises the steps: obtaining historical load time sequence data, and setting a historical sequence length to obtain an original sequence; decomposing the original sequence through a moving average decomposition method to obtain a trend sequence and a residual sequence; an enhanced prediction network is constructed through a variable component sub-circuit and a multi-layer perception neural network which are arranged in parallel; and respectively inputting the trend sequence and the residual sequence into an enhanced prediction network for prediction, and fusing a generated trend prediction result and a residual prediction result to obtain a power load prediction result. According to the method, after the sequence is decomposed through the moving average method, the parallel variable component sub-circuit and the multi-layer perception neural network are adopted, the power load result is obtained, and therefore the prediction precision and the generalization ability of the model are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of power load forecasting, and in particular to an enhanced fusion load forecasting method, system, device, and medium based on parallel variable quantum circuits and multilayer sensing neural networks. Background Technology

[0002] Power load forecasting is a crucial component of power system planning and operation management, primarily addressing issues related to power supply balance, grid planning, economic dispatch, energy cost management, system reliability, and stability. Power load forecasting plays a vital role in ensuring the economic efficiency, reliability, and environmental friendliness of the power system. With technological advancements, load forecasting methods are continuously evolving to adapt to the ever-changing demands of the power system. Currently, the DLinear model not only outperforms existing complex Transformer-based models but also demonstrates significant advantages in many situations. However, DLinear employs only a single layer of a multilayer perceptual MLP model in the time dimension. While this model may be suitable for some autoregressive tasks with fixed time dependencies, it degrades considerably when learning higher-level representations.

[0003] Therefore, an improved single-layer linear layer is proposed, incorporating variational quantum circuits (VQCs) to adjust the parameters of the quantum gates through variational training, thereby optimizing the predicted output. Compared to classical methods, VQCs can explore the solution space more efficiently, and are particularly suitable for capturing complex nonlinear relationships in the remaining sequences. This approach not only increases the interpretability of the model but also demonstrates the superiority of quantum computing in handling large-scale data. Summary of the Invention

[0004] In view of the aforementioned problems with the capture of nonlinear features and the limitations in learning complex patterns in existing technologies, this invention is proposed.

[0005] Therefore, this invention provides an enhanced fusion load prediction method, system, device, and medium based on parallel variable quantum circuits and multilayer perceptron neural networks to address the performance bottleneck of existing DLinear models when processing complex nonlinear load data; the remaining sequence is difficult to predict: the "remaining sequence" after moving average decomposition usually contains high-frequency fluctuations and nonlinear noise, which traditional methods are difficult to fit.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, the present invention provides an enhanced fusion load prediction method based on parallel variable quantum circuits and multilayer sensing neural networks, comprising: Acquire historical load time-series data and construct the original sequence based on the preset historical sequence length; The original sequence is decomposed using the moving average decomposition method to obtain the trend sequence and the residual sequence; An enhanced prediction network is constructed by using parallel variable quantum circuits and multilayer perceptual neural networks; The trend sequence and the residual sequence are input into the enhanced prediction network for prediction. The generated trend prediction results and residual prediction results are then fused to obtain the power load prediction results.

[0007] As a preferred embodiment of the enhanced fusion load prediction method based on parallel variable quantum circuits and multilayer perceptron neural networks described in this invention, the original sequence is decomposed using a moving average decomposition method to obtain a trend sequence and a residual sequence, including: The moving average decomposition method is used to calculate the moving average of the original sequence and extract the trend sequence. The remaining sequence is obtained by calculating the difference between the original sequence and the trend sequence.

[0008] As a preferred embodiment of the enhanced fusion load prediction method based on parallel variable quantum circuits and multilayer perceptron neural networks described in this invention, wherein: both the trend sequence and the residual sequence are recorded as input sequences, which are then input into the enhanced prediction network for prediction, including: The input sequence is normalized to obtain a multi-feature representation; Based on the multi-feature representation, the outputs are respectively fed to the parallel variable quantum circuit VQC and the multilayer perceptron MLP to obtain the output features of the variable quantum circuit VQC and the output features of the multilayer perceptron MLP. The output features of VQC and MLP are weighted and combined to obtain the first prediction result, which is either the trend prediction result or the residual prediction result.

[0009] As a preferred embodiment of the enhanced fusion load prediction method based on parallel variable quantum circuits and multilayer perceptron neural networks described in this invention, the variable quantum circuit (VQC) includes: a feature encoding layer, trainable parameters and entanglement layers, and a measurement layer, comprising: The multi-feature representation is encoded using a feature coding layer through unitary transformation. The encoded quantum state is evolved using multiple trainable parameters and entanglement layers; The output characteristics of the variable quantum circuit VQC are obtained by measuring the evolved quantum state through the measurement layer.

[0010] As a preferred embodiment of the enhanced fusion load prediction method based on parallel variable quantum circuits and multilayer perceptron neural networks described in this invention, wherein: the multilayer perceptron neural network (MLP) comprises: The multi-feature representation is multiplied by the weight matrix and then passed to the hidden neuron layer; By applying a bias to the values ​​of the hidden neuron layer and scaling them using an activation function, the output features of the multilayer perceptron neural network (MLP) are obtained.

[0011] As a preferred embodiment of the enhanced fusion load prediction method based on parallel variable quantum circuits and multilayer perceptron neural networks described in this invention, wherein: the weighted combination of the output features of VQC and MLP to obtain the first prediction result includes: The output features of the variable quantum circuit (VQC) are combined with the output features of the multilayer perceptron (MLP), and a two-to-one linear weight layer is used with trainable weighted parameters to perform weighted summation, thereby obtaining the first prediction result of the enhanced prediction network.

[0012] As a preferred embodiment of the enhanced fusion load forecasting method based on parallel variable quantum circuits and multilayer sensing neural networks described in this invention, the method involves: inputting trend sequences and residual sequences into the enhanced forecasting network for prediction; fusing the generated trend forecasting results and residual forecasting results to obtain the power load forecasting result, including: Based on the enhanced prediction network, a trend sequence enhanced prediction network is constructed to receive the trend sequence and obtain the trend prediction result; Based on the enhanced prediction network, a residual sequence enhanced prediction network is constructed to receive the residual sequence and obtain the residual prediction results; The power load forecast result is obtained by adding and fusing the trend forecast result and the residual forecast result.

[0013] Secondly, the present invention provides an enhanced fusion load prediction system based on parallel variable quantum circuits and multilayer sensing neural networks, comprising: The acquisition module is used to acquire historical load time series data and obtain the original sequence by setting the length of the historical sequence. The decomposition module is used to decompose the original sequence using a moving average decomposition method to obtain a trend sequence and a residual sequence. The building blocks are used to construct augmented prediction networks using parallel-configured variable quantum circuits and multilayer perceptual neural networks. The prediction module is used to input the trend sequence and the residual sequence into the enhanced prediction network for prediction, and then fuse the generated trend prediction results and residual prediction results to obtain the power load prediction results.

[0014] Thirdly, the present invention provides an electronic device, comprising: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions, which, when executed by the processor, implement the steps of the enhanced fusion load prediction method based on parallel variable quantum circuits and multilayer perceptual neural networks.

[0015] Fourthly, the present invention provides a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the enhanced fusion load prediction method based on parallel variable quantum circuits and multilayer perceptron neural networks.

[0016] Compared with existing technologies, the beneficial effects of this invention are as follows: By combining a multilayer perceptron neural network (MLP) and a variable quantum circuit (VQC), an enhanced prediction network is constructed. VQC excels at capturing smooth periodic features, while MLP excels at capturing abrupt nonlinear features. The two complement each other, improving the model's ability to fit complex power load patterns. Furthermore, the enhanced prediction network can process multiple data points in parallel. When processing large-scale load data, compared to simply increasing the number of layers in the MLP, it achieves better results with fewer parameters and accelerates training convergence. Simultaneously, by explicitly decomposing the sequence into trend and residual parts and modeling them separately, it significantly improves prediction accuracy and the model's generalization ability. This technology is not only applicable to load forecasting but also demonstrates excellent performance in multiple fields such as finance, climate, and energy management, enabling more accurate prediction of future trends, reducing computational resource consumption, and improving prediction efficiency. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic diagram of the overall process of the enhanced fusion load prediction method based on parallel variable quantum circuits and multilayer sensing neural networks according to an embodiment of the present invention.

[0019] Figure 2 This is a schematic diagram of the overall framework of the enhanced fusion load prediction method based on parallel variable quantum circuits and multilayer perceptual neural networks according to an embodiment of the present invention.

[0020] Figure 3 This is a schematic diagram of the enhanced prediction network for an enhanced fusion load prediction method based on parallel variable quantum circuits and multilayer perceptrons, as described in an embodiment of the present invention.

[0021] Figure 4This is a variable quantum circuit structure diagram of the enhanced fusion load prediction method based on parallel variable quantum circuits and multilayer perceptual neural networks according to an embodiment of the present invention.

[0022] Figure 5 This is a block diagram of an enhanced fusion load prediction method based on parallel variable quantum circuits and multilayer perceptrons, according to an embodiment of the present invention.

[0023] Figure 6 This is a structural block diagram of a computer device for an enhanced fusion load prediction method based on parallel variable quantum circuits and multilayer perceptual neural networks, according to an embodiment of the present invention. Detailed Implementation

[0024] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0025] Example 1, referring to Figure 1 As an embodiment of the present invention, an enhanced fusion load prediction method based on parallel variable quantum circuits and multilayer perceptron neural networks is provided, comprising: S100: Acquire historical load time-series data and construct the original sequence according to the preset historical sequence length; S200: The original sequence is decomposed using the moving average decomposition method to obtain the trend sequence and the residual sequence; S300: An enhanced prediction network is constructed by using parallel variable quantum circuits and multilayer perceptual neural networks; S400: The trend sequence and residual sequence are input into the enhanced prediction network for prediction. The generated trend prediction results and residual prediction results are then fused to obtain the power load prediction results.

[0026] It should be noted that electricity load data is a typical multivariate non-stationary time series. During its changes, it is influenced by multiple coupled factors, resulting in a complex pattern of strong nonlinearity, volatility, and intertwined long-term and short-term dependencies in the load curve. While the existing linear forecasting model DLinear performs well in extracting long-term trends, it struggles to capture the high-dimensional nonlinear features and abrupt change patterns hidden deep within the data when processing the decomposed residual series. This is because it uses only a single-layer linear structure, making it difficult to capture these patterns. Consequently, the model's generalization ability decreases under complex operating conditions, and its prediction accuracy often reaches a bottleneck. Therefore, developing a high-precision load forecasting technique that can balance long-term trend linear fitting with complex residual nonlinear mining is also crucial.

[0027] Therefore, to address the aforementioned problems, steps S100-S400 are used to decompose the original sequence using the moving average decomposition method to obtain the trend sequence and residual sequence, reducing the difficulty of model learning and avoiding mutual interference between information of different frequencies. An enhanced prediction network with a parallel architecture of VQC and MLP is established, which enables the simultaneous utilization of the linear mapping capability of VQC and the high-dimensional nonlinear mapping capability of MLP at the same level. Through a weighted fusion mechanism, the outputs of VQC and MLP are adaptively combined, and the predicted values ​​of the trend term and residual term are finally superimposed to obtain the final power load prediction value, thereby significantly improving the prediction accuracy and the model's generalization ability.

[0028] Example 2, refer to Figures 1-4 As an embodiment of the present invention, based on the above embodiment, an enhanced fusion load prediction method based on parallel variable quantum circuits and multilayer perceptron neural networks is provided.

[0029] In this embodiment of the application, step S100, which involves acquiring historical load time-series data and constructing an original sequence based on a preset historical sequence length, includes: Specifically, for a time series containing variable C, given historical load time series data... The original sequence is obtained by constructing it, where It is the size of the look-back history sequence. It is a variable exist The value at any given moment.

[0030] The task of time series forecasting is to predict... In the future The value of a time step.

[0031] In this embodiment of the application, step S200 involves decomposing the original sequence using a moving average decomposition method to obtain a trend sequence and a residual sequence, including: Specifically, the moving average decomposition method is used to extract the trend sequence of the time series, and the difference between the original sequence and the trend component is the remaining sequence.

[0032] It should be noted that the moving average method is a statistical analysis method that smooths time series data by calculating the average of a set of data, eliminating random fluctuations and thus more clearly showing the long-term trend of the data. It is suitable for short-term forecasting, especially when product demand is stable and there are no significant seasonal changes, and can effectively improve the accuracy of forecasts.

[0033] In this embodiment of the application, step S300 constructs an enhanced prediction network using parallel variable quantum circuits and multilayer perceptrons, including the following steps A1-A3: A1: Both the trend sequence and the residual sequence are recorded as input sequences, which are then fed into the enhanced prediction network for prediction. Specifically, constructing an enhanced prediction network involves using a weighted sum operation to regress historical sequences in order to predict the future. Its mathematical expression is as follows: in It is a linear layer along the time axis. These are the input value and the predicted value, respectively.

[0034] Both the trend sequence and the residual sequence are recorded as input sequences, which are then fed into the augmented prediction network for prediction.

[0035] For example, such as Figure 2 The diagram shows the overall framework. The system is decomposed into residual and trend sequences through moving averages, which are then fed into enhanced trend forecasting for prediction. The results are then fused to obtain the final power load forecast.

[0036] A2: Based on multi-feature representation, the outputs are respectively fed to the parallel variable quantum circuit VQC and the multilayer perceptron MLP to obtain the output features of the variable quantum circuit VQC and the output features of the multilayer perceptron MLP. Specifically, the features of the input sequence are normalized to obtain a multi-feature representation. This multi-feature representation is then passed through a variable quantum circuit (VQC) and a multilayer perceptron (MLP) to obtain corresponding output features. These features are then weighted and combined to obtain the first prediction result, which includes trend prediction and residual prediction results. The weighted combination is a trainable weighted linear addition, where each weight determines the contribution of each network to the final output feature.

[0037] Wherein, the variable quantum circuit VQC: Specifically, such as Figure 4The diagram shows the structure of the Variable Quantum Circuit (VQC). Based on multi-feature representation, the VQC first initializes the L qubits, and then processes the initialized features... The quantum state is encoded by unitary transformation after passing through a feature encoding layer. Next, N trainable parameters and an entanglement layer are used to evolve the encoded quantum state. Finally, the evolved quantum state is measured through a measurement layer to obtain the output characteristics of the variable quantum circuit VQC. The measurement layer is where quantum information collapses into T (T and L values ​​are not necessarily the same) classical output values.

[0038] Wherein, the multilayer perceptron neural network (MLP); Specifically, a multilayer perceptron (MLP) also accepts multiple feature representations and passes them to a single hidden neuron layer of size F by multiplying the feature vectors by a weight matrix of size L×F. Then, biases are applied to the values ​​of the hidden neurons, and they are scaled using an activation function. The activation function is necessary to add nonlinearity to a linear system. The neurons then pass through their respective biases and activation functions (denoted as L×F). The output features are propagated to T MLP output neurons to obtain the MLP output features.

[0039] For example, such as Figure 3 The image shows an enhanced prediction network, including VQC and MLP.

[0040] It should be noted that an MLP with a single hidden layer creates a non-harmonic function for the dataset; while VQC creates a truncated Fourier series, a harmonic function; therefore, a network combining these results can map the smooth sine portion through VQC and fill in the spurs through MLP.

[0041] A3: The output features of VQC and MLP are weighted and combined to obtain the first prediction result, which is either the trend prediction result or the residual prediction result. Specifically, the output features of MLP and VQC are combined using a 2-to-1 linear weight layer to form the first prediction result, i.e., the output of the enhanced prediction network. This includes trend prediction results or residual prediction results. The final layer combines the first output of VQC with the first output of MLP. The same applies to all T outputs, where These are trainable parameters.

[0042] In this embodiment of the application, step S400 involves inputting the trend sequence and the residual sequence into the enhanced prediction network for prediction, and then fusing the generated trend prediction results and residual prediction results to obtain the power load prediction result, including: Specifically, a trend sequence augmented prediction network is constructed based on the augmented prediction network. Receive the trend sequence and obtain the trend prediction result; Construct a residual sequence enhanced prediction network based on the enhanced prediction network. Receive the remaining sequence and obtain the remaining prediction results; The power load forecast result is obtained by adding and fusing the trend forecast result and the residual forecast result.

[0043] In summary, this invention utilizes moving average decomposition technology to decouple complex power load time-series data into trend sequences and residual sequences, effectively reducing the difficulty of prediction. For the decomposed sequences, an innovative enhanced prediction network is designed, employing a multilayer perceptron (MLP) to handle abrupt nonlinear features, while utilizing a variable quantum circuit (VQC) to handle smooth periodic features and high-dimensional nonlinear relationships. The advantages of both approaches are complemented and integrated through a trainable linear weight layer. This not only overcomes the limitations of traditional DLinear models in nonlinear feature extraction but also fully leverages the advantages of quantum computing in exploring complex solution spaces, significantly improving the accuracy and robustness of power load prediction.

[0044] Example 3 illustrates an illustrative scheme of an enhanced fusion load forecasting method based on parallel variable quantum circuits and multilayer perceptrons. It should be noted that the technical solution of this enhanced fusion load forecasting system based on parallel variable quantum circuits and multilayer perceptrons belongs to the same concept as the aforementioned enhanced fusion load forecasting method based on parallel variable quantum circuits and multilayer perceptrons. Details not described in detail in this embodiment can be found in the description of the aforementioned enhanced fusion load forecasting method based on parallel variable quantum circuits and multilayer perceptrons.

[0045] This embodiment also provides an enhanced fusion load forecasting system based on parallel variable quantum circuits and multilayer sensing neural networks, such as... Figure 5 As shown, it includes: The acquisition module is used to acquire historical load time series data and obtain the original sequence by setting the length of the historical sequence. The decomposition module is used to decompose the original sequence using a moving average decomposition method to obtain a trend sequence and a residual sequence. The building blocks are used to construct augmented prediction networks using parallel-configured variable quantum circuits and multilayer perceptual neural networks. The prediction module is used to input the trend sequence and the residual sequence into the enhanced prediction network for prediction, and then fuse the generated trend prediction results and residual prediction results to obtain the power load prediction results.

[0046] Figure 6 A structural block diagram of a computer device illustrated in an embodiment of this application is shown. This computer device can be implemented as a server as described in the above-described scheme of this application. The computer device includes a Central Processing Unit (CPU), a system memory including Random Access Memory (RAM) and Read-Only Memory (ROM), and a system bus connecting the system memory and the CPU. The computer device also includes a mass storage device for storing the operating system, applications, and other program modules. The mass storage device is connected to the CPU via a mass storage controller (not shown) connected to the system bus. The mass storage device and its associated computer-readable medium provide non-volatile storage for the computer device. That is, the mass storage device may include computer-readable media (not shown) such as a hard disk or a Compact Disc Read-Only Memory (CD-ROM) drive. The aforementioned system memory and mass storage device can be collectively referred to as memory.

[0047] This embodiment also provides an electronic device suitable for enhanced fusion load prediction based on parallel variable quantum circuits and multilayer perceptron neural networks, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the enhanced fusion load prediction method based on parallel variable quantum circuits and multilayer perceptron neural networks as proposed in the above embodiment.

[0048] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the enhanced fusion load prediction method based on parallel variable quantum circuits and multilayer perceptron neural networks as proposed in the above embodiments.

[0049] The storage medium proposed in this embodiment belongs to the same inventive concept as the enhanced fusion load prediction method based on parallel variable quantum circuits and multilayer perceptual neural networks proposed in the above embodiments. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0050] Based on the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0051] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. An enhanced fusion load prediction method based on parallel variable quantum circuits and multilayer sensing neural networks, characterized in that, include: Acquire historical load time-series data and construct the original sequence based on the preset historical sequence length; The original sequence is decomposed using the moving average decomposition method to obtain the trend sequence and the residual sequence; An enhanced prediction network is constructed by using parallel variable quantum circuits and multilayer perceptual neural networks; The trend sequence and the residual sequence are input into the enhanced prediction network for prediction. The generated trend prediction results and residual prediction results are then fused to obtain the power load prediction results.

2. The enhanced fusion load prediction method based on parallel variable quantum circuits and multilayer perceptron neural networks as described in claim 1, characterized in that, The original sequence is decomposed using the moving average decomposition method to obtain a trend sequence and a residual sequence, including: The moving average decomposition method is used to calculate the moving average of the original sequence and extract the trend sequence. The remaining sequence is obtained by calculating the difference between the original sequence and the trend sequence.

3. The enhanced fusion load prediction method based on parallel variable quantum circuits and multilayer perceptrons as described in claim 2, characterized in that, Both the trend sequence and the residual sequence are recorded as input sequences, which are then fed into the enhanced prediction network for prediction, including: The input sequence is normalized to obtain a multi-feature representation; Based on the multi-feature representation, the outputs are respectively fed to the parallel variable quantum circuit VQC and the multilayer perceptron MLP to obtain the output features of the variable quantum circuit VQC and the output features of the multilayer perceptron MLP. The output features of VQC and MLP are weighted and combined to obtain the first prediction result, which is either the trend prediction result or the residual prediction result.

4. The enhanced fusion load prediction method based on parallel variable quantum circuits and multilayer perceptron neural networks as described in claim 3, characterized in that, The variable quantum circuit (VQC) includes: a feature encoding layer, trainable parameters and entanglement layers, and a measurement layer, including: The multi-feature representation is encoded using a feature coding layer through unitary transformation. The encoded quantum state is evolved using multiple trainable parameters and entanglement layers; The output characteristics of the variable quantum circuit VQC are obtained by measuring the evolved quantum state through the measurement layer.

5. The enhanced fusion load prediction method based on parallel variable quantum circuits and multilayer perceptrons as described in claim 4, characterized in that, The multilayer perceptron neural network (MLP) includes: The multi-feature representation is multiplied by the weight matrix and then passed to the hidden neuron layer; By applying a bias to the values ​​of the hidden neuron layer and scaling them using an activation function, the output features of the multilayer perceptron neural network (MLP) are obtained.

6. The enhanced fusion load prediction method based on parallel variable quantum circuits and multilayer perceptron neural networks as described in claim 5, characterized in that, The weighted combination of the output features of VQC and MLP to obtain the first prediction result includes: The output features of the variable quantum circuit (VQC) are combined with the output features of the multilayer perceptron (MLP), and a two-to-one linear weight layer is used with trainable weighted parameters to perform weighted summation, thereby obtaining the first prediction result of the enhanced prediction network.

7. The enhanced fusion load prediction method based on parallel variable quantum circuits and multilayer perceptron neural networks as described in claim 6, characterized in that: The trend sequence and residual sequence are input into the enhanced prediction network for prediction. The generated trend prediction results and residual prediction results are then fused to obtain the power load prediction results, including: Based on the enhanced prediction network, a trend sequence enhanced prediction network is constructed to receive the trend sequence and obtain the trend prediction result; Based on the enhanced prediction network, a residual sequence enhanced prediction network is constructed to receive the residual sequence and obtain the residual prediction results; The power load forecast result is obtained by adding and fusing the trend forecast result and the residual forecast result.

8. An enhanced fusion load forecasting system based on parallel variable quantum circuits and multilayer sensing neural networks, using the method of any one of claims 1-7, characterized in that, include: The acquisition module is used to acquire historical load time series data and obtain the original sequence by setting the length of the historical sequence. The decomposition module is used to decompose the original sequence using a moving average decomposition method to obtain a trend sequence and a residual sequence. The building blocks are used to construct augmented prediction networks using parallel-configured variable quantum circuits and multilayer perceptual neural networks. The prediction module is used to input the trend sequence and the residual sequence into the enhanced prediction network for prediction, and then fuse the generated trend prediction results and residual prediction results to obtain the power load prediction results.

9. An electronic device, characterized in that, include: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, they implement the steps of the enhanced fusion load prediction method based on parallel variable quantum circuits and multilayer perceptron neural networks as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, It includes storing computer-executable instructions that, when executed by a processor, implement the steps of the enhanced fusion load prediction method based on parallel variable quantum circuits and multilayer perceptron neural networks as described in any one of claims 1 to 7.

Images