Electric power engineering survey data processing method and related device

Abstract

The invention provides an electric power engineering survey data processing method and a related device, and the method comprises the steps: obtaining a multi-source survey data set of a target region, the multi-source survey data set comprises geological topographic data, electromagnetic interference data and equipment layout data, carrying out the joint feature mapping of the multi-source survey data set through a pre-trained feature extraction network, and obtaining a feature mapping result; a multi-dimensional survey feature vector is generated, the multi-dimensional survey feature vector is input into an optimization decision model for iterative parameter adjustment, an electric power engineering optimization parameter set of the target area is generated, and the optimization decision model is obtained through training based on a mapping relation between historical survey data and a verification optimization result; and generating an electric power facility deployment scheme of the target area according to the electric power engineering optimization parameter set, wherein the deployment scheme comprises a power transmission line path planning result, substation site selection coordinates and an electromagnetic interference suppression strategy. According to the method, the environmental adaptability, the global coordination and the construction performability of a power facility deployment scheme can be improved.

Description

Technical Field

[0001] The present invention relates to the field of data processing, and more particularly, to a method and related device for processing power engineering survey data. Background Art

[0002] The processing of power engineering survey data involves comprehensively analyzing a target area to formulate a power facility deployment plan. In the prior art, the transmission line path is usually planned based on terrain elevation data, or the substation site is selected according to the electromagnetic radiation intensity threshold, resulting in problems such as the overlap of path planning and electromagnetic interference areas, and the conflict between equipment layout and geological stability. In addition, the traditional deployment plan lacks a dynamic response mechanism for real-time terrain changes during the construction stage and electromagnetic fluctuations during the operation stage, resulting in a high construction rework rate, an increase in the cost of expanding shielding equipment in the later stage, poor global coordination of the deployment plan, and the inability to effectively handle multi-constraint conflicts in complex environments, limiting the overall reliability and economy of power engineering deployment. Summary of the Invention

[0003] The present invention provides a method and related device for processing power engineering survey data.

[0004] In a first aspect, an embodiment of the present invention provides a method for processing power engineering survey data, the method comprising: obtaining a multi-source survey data set of a target area, the multi-source survey data set including geological terrain data, electromagnetic interference data, and equipment layout data, wherein the electromagnetic interference data is used to describe the electromagnetic radiation intensity distribution of power facilities in the target area, and the equipment layout data is used to indicate the transmission line topology structure and substation location in the target area; performing joint feature mapping on the multi-source survey data set through a pre-trained feature extraction network to generate a multi-dimensional survey feature vector, wherein the multi-dimensional survey feature vector includes a terrain parameter sub-vector, an electromagnetic intensity sub-vector, and a layout topology sub-vector; inputting the multi-dimensional survey feature vector into an optimization decision model for iterative parameter adjustment to generate a power engineering optimization parameter set of the target area, wherein the optimization decision model is trained based on the mapping relationship between historical survey data and verified optimization results; generating a power facility deployment plan for the target area according to the power engineering optimization parameter set, the deployment plan including a transmission line path planning result, substation site selection coordinates, and an electromagnetic interference suppression strategy.

[0005] In a second aspect, an embodiment of the present invention provides a power engineering survey data processing device, the device comprising: a data acquisition module, used to acquire a multi-source survey data set of a target area, the multi-source survey data set comprising geological and topographic data, electromagnetic interference data and equipment layout data, wherein the electromagnetic interference data is used to describe the electromagnetic radiation intensity distribution of the power facilities in the target area, and the equipment layout data is used to indicate the transmission line topology and substation location in the target area; a feature mapping module, used to perform joint feature mapping on the multi-source survey data set through a pre-trained feature extraction network to generate a multi-dimensional survey feature vector, wherein the multi-dimensional survey feature vector comprises a terrain parameter sub-vector, an electromagnetic intensity sub-vector and a layout topology sub-vector; a parameter generation module, used to input the multi-dimensional survey feature vector into an optimization decision model for iterative parameter adjustment to generate a power engineering optimization parameter set for the target area, wherein the optimization decision model is trained based on the mapping relationship between historical survey data and verified optimization results; a solution generation module, used to generate a power facility deployment solution for the target area according to the power engineering optimization parameter set, the deployment solution comprising a transmission line path planning result, a substation site selection coordinate and an electromagnetic interference suppression strategy.

[0006] The power engineering survey data processing method provided by the present invention constructs a multi-source survey data set by acquiring geological and topographic data, electromagnetic interference data and equipment layout data of the target area, and uses a pre-trained feature extraction network to perform joint feature mapping to generate a multi-dimensional survey feature vector including a terrain parameter sub-vector, an electromagnetic intensity sub-vector and a layout topology sub-vector; iteratively adjusts the parameters of the multi-dimensional survey feature vector based on an optimization decision model to generate a power engineering optimization parameter set that integrates terrain constraints, electromagnetic safety and topological coverage requirements, and based on this, generates a transmission line path planning result, a substation site selection coordinate and a power facility deployment plan that is linked to an electromagnetic interference suppression strategy. This method breaks through the limitations of traditional single-dimensional data processing through collaborative feature extraction and dynamic optimization decision-making of multi-source survey data. It can simultaneously avoid geological risk areas and high electromagnetic radiation areas in path planning, balance coverage efficiency and electromagnetic suppression requirements in substation site selection, and realize dynamic adaptation of spatial parameters and real-time monitoring in equipment deployment, thereby significantly improving the environmental adaptability, global coordination and construction feasibility of power facility deployment plans; at the same time, through the spatial fusion of multi-dimensional feature vectors and the embedding of joint control logic, it ensures that the deployment plan can be dynamically optimized and adjusted according to real-time terrain changes, electromagnetic fluctuations and load conditions during the construction and operation stages, effectively reducing rework costs and operation risks, and improving the overall safety and economy of power engineering deployment. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] Figure 1It is a flowchart of a method for processing power engineering survey data provided by an embodiment of the present invention.

[0008] Figure 2 It is a schematic diagram of the composition of a device for processing power engineering survey data provided by an embodiment of the present invention. Detailed implementation manners

[0009] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0010] Please refer to Figure 1 , Figure 1 which is a flowchart of a method for processing power engineering survey data provided by an embodiment of the present invention. This method for processing power engineering survey data can be executed by a computer system and may include the following steps: Step S100: Obtain a multi-source survey data set of the target area. The multi-source survey data set includes geological and topographical data, electromagnetic interference data, and equipment layout data. Among them, the electromagnetic interference data is used to describe the electromagnetic radiation intensity distribution of power facilities in the target area, and the equipment layout data is used to indicate the transmission line topology and substation location in the target area.

[0011] In the embodiments of the present invention, the geological and topographical data is information about the geological structure, terrain undulation, etc. of the target area. This information can reflect the geological stability and terrain complexity of the area. For example, the geological and topographical data of a mountainous area will include the slope, height, rock type, etc. of the mountain. The electromagnetic interference data focuses on the electromagnetic radiation generated by power facilities in the target area. By collecting and analyzing these data, the distribution of electromagnetic radiation intensity at different positions and frequency bands can be clearly understood. For example, near a substation, the electromagnetic radiation intensity may be relatively high. The equipment layout data mainly records the connection method, orientation of the transmission lines in the target area, and the specific location of the substation. By analyzing these data, the power network topology of the area can be constructed.

[0012] In actual operation, geological and topographical data can be obtained through technical means such as satellite remote sensing and Geographic Information System (GIS). These technologies can accurately obtain information such as terrain elevation and geological stratification of the target area. For the acquisition of electromagnetic interference data, multiple electromagnetic monitoring points are usually set in the target area, and professional electromagnetic monitoring equipment is used to collect data at predefined time intervals to comprehensively and accurately grasp the distribution of electromagnetic radiation intensity. The equipment layout data can be obtained from the database of the power company, which details the laying of transmission lines and the construction information of substations.

[0013] Step S200: Perform joint feature mapping on the multi-source survey data set through a pre-trained feature extraction network to generate a multi-dimensional survey feature vector, where the multi-dimensional survey feature vector includes a terrain parameter sub-vector, an electromagnetic intensity sub-vector, and a layout topology sub-vector.

[0014] The pre-trained feature extraction network is a neural network model trained with historical data, and its role is to extract representative features from the input data. In the embodiment of the present invention, this network performs joint feature mapping on the multi-source survey data set, deeply fuses geological and topographical data, electromagnetic interference data, and equipment layout data, and generates a multi-dimensional survey feature vector containing multi-dimensional information.

[0015] The terrain parameter sub-vector is a vector obtained by extracting features from geological and topographical data. It contains key information such as terrain undulation characteristics and geological stability characteristics, which can reflect the terrain features and geological conditions of the target area. The electromagnetic intensity sub-vector is extracted from the electromagnetic interference data and contains the distribution information of electromagnetic radiation intensity in different frequency bands, which can help understand the complexity of the electromagnetic environment in the target area. The layout topology sub-vector is generated based on the equipment layout data and reflects information such as the topological structure of transmission lines and the coverage of substations.

[0016] In specific implementation, the pre-trained feature extraction network can adopt architectures such as Convolutional Neural Network (CNN) or Recurrent Neural Network (RNN). Taking CNN as an example, features can be extracted from the multi-source survey data through convolutional layers, and features can be reduced in dimension and compressed through pooling layers, finally generating a multi-dimensional survey feature vector. For example, for the survey of a power project in a mountainous area, the multi-dimensional survey feature vector generated after being processed by the pre-trained feature extraction network can accurately reflect the terrain, electromagnetic, and equipment layout and other multi-faceted feature information of the area.

[0017] As an implementation method, step S200, performing joint feature mapping on the multi-source survey data set through a pre-trained feature extraction network to generate a multi-dimensional survey feature vector, can specifically include the following steps: Step S210: Perform elevation gradient analysis on the geological terrain data, extract the terrain undulation characteristics and geological stability characteristics, and fuse the terrain undulation characteristics and geological stability characteristics into a terrain parameter sub-vector based on a preset terrain coding structure.

[0018] Elevation gradient analysis refers to processing and analyzing the elevation information in the geological terrain data. By calculating the numerical differences between adjacent elevation points, a slope gradient map is obtained, which contains the elevation change rate and direction angle of each coordinate point. The terrain undulation characteristics are obtained by further analyzing the slope gradient map, which reflects the undulation degree of the terrain in the target area. For example, the terrain undulation characteristics in mountainous areas are more obvious. The geological stability characteristics are obtained by performing a structural continuity analysis on the rock layer distribution data in the geological terrain data, which can reflect the geological stability of the target area. For example, the presence of fault lines or loose sediment areas will reduce the geological stability.

[0019] The preset terrain coding structure is a predefined rule for fusing the terrain undulation characteristics and geological stability characteristics. In practical applications, this rule can be implemented in the form of a feature fusion layer. The terrain undulation characteristics are converted into an undulation coding sequence, and the geological stability characteristics are converted into a stability coding sequence. Then, these two coding sequences are weighted and superimposed element by element according to the coordinate points to generate a terrain parameter sub-vector.

[0020] In specific operations, first obtain the elevation distribution matrix in the geological terrain data, and generate a slope gradient map by calculating the numerical differences between adjacent elevation points. Then, divide the terrain area types according to the elevation change rate in the slope gradient map, extract the boundary coordinate set of the continuously changing slope area, and generate the terrain undulation characteristics based on the maximum slope difference within the boundary coordinate set. At the same time, perform a structural continuity analysis on the rock layer distribution data in the geological terrain data, identify the range of fault lines or loose sediment areas, and generate the geological stability characteristics based on the proportion of the overlapping area between the range and the slope gradient map. Finally, call the feature fusion layer in the preset terrain coding structure to fuse the terrain undulation characteristics and geological stability characteristics to generate a terrain parameter sub-vector. For example, when conducting a power engineering survey in an area containing mountains and plains, the terrain undulation characteristics and geological stability characteristics of this area can be accurately extracted through the above steps, and the corresponding terrain parameter sub-vector can be generated.

[0021] As an implementation method, step S210, performing elevation gradient analysis on the geological terrain data, extracting the terrain undulation characteristics and geological stability characteristics, and fusing the terrain undulation characteristics and geological stability characteristics into a terrain parameter sub-vector, can specifically include the following steps: Step S211: Obtain the elevation distribution matrix in the geological terrain data, and calculate and generate a slope gradient map based on the numerical differences between adjacent elevation points. The slope gradient map includes the elevation change rate and direction angle of each coordinate point.

[0022] The elevation distribution matrix is a matrix representation of the elevation information in the geological terrain data, which records the elevation values of each coordinate point in the target area. By obtaining the elevation distribution matrix, a comprehensive understanding of the terrain elevation situation in the target area can be achieved. Calculating and generating a slope gradient map based on the numerical differences between adjacent elevation points is to calculate the elevation difference between adjacent elevation points and combine the distance between them to obtain the slope value and direction angle of each coordinate point.

[0023] The elevation change rate in the slope gradient map reflects the steepness of the terrain. The larger the elevation change rate, the steeper the terrain; the direction angle indicates the inclination direction of the terrain. Through the slope gradient map, the terrain undulation situation in the target area can be intuitively understood.

[0024] In actual operation, tools such as Geographic Information System (GIS) can be used to obtain the elevation distribution matrix in the geological terrain data. Then, by writing an algorithm to calculate the numerical differences between adjacent elevation points, a slope gradient map can be generated. For example, in the processing of geological terrain data in a mountainous area, by obtaining the elevation distribution matrix and calculating and generating a slope gradient map, the terrain undulation and slope situation in this mountainous area can be clearly understood.

[0025] Step S212: Divide the terrain area types according to the elevation change rate in the slope gradient map, extract the boundary coordinate set of the continuously changing slope area, and generate the terrain undulation feature based on the maximum slope difference within the boundary coordinate set.

[0026] Dividing the terrain area types according to the elevation change rate in the slope gradient map means dividing the target area into different terrain area types according to different elevation change rate ranges, such as plains, hills, mountainous areas, etc. The continuously changing slope area refers to the area where the slope values change continuously in the slope gradient map. Extracting the boundary coordinate sets of these areas can accurately determine the ranges of these areas. Generating the terrain undulation feature based on the maximum slope difference within the boundary coordinate set is obtained by calculating the difference between the maximum and minimum slope values within the boundary coordinate set, resulting in the terrain undulation feature. This feature can reflect the undulation degree of the terrain in the target area. The larger the maximum slope difference, the greater the terrain undulation. In specific implementation, first, according to the preset elevation change rate threshold, the areas in the slope gradient map are divided into different terrain area types. Then, through techniques such as image segmentation, the boundary coordinate sets of the continuously changing slope areas are extracted. Finally, the maximum slope difference within the boundary coordinate set is calculated to generate the terrain undulation feature. For example, when processing geological terrain data in an area with multiple terrain types, through the above steps, the terrain area types can be accurately divided, the boundary coordinate sets of the continuously changing slope areas can be extracted, and the terrain undulation feature can be generated.

[0027] Step S213: Conduct a structural continuity analysis on the rock layer distribution data in the geological terrain data, identify the ranges of fault lines or loose sediment areas, and generate a geological stability feature based on the proportion of the overlapping area between the range and the slope gradient map.

[0028] The rock layer distribution data records the distribution of rock layers in the target area, including information such as the type, thickness, and strike of the rock layers. Conducting a structural continuity analysis on the rock layer distribution data is to identify the ranges of fault lines or loose sediment areas by analyzing the structure and continuity of the rock layers. A fault line refers to the area where fractures occur in the rock layer, and a loose sediment area refers to the area where the rock layer structure is relatively loose. The existence of these areas will affect geological stability.

[0029] Generating a geological stability feature based on the proportion of the overlapping area between the range and the slope gradient map is obtained by calculating the proportion of the overlapping area between the range of the fault line or loose sediment area and the corresponding area in the slope gradient map, resulting in the geological stability feature. The larger the proportion of the overlapping area, the worse the geological stability. In actual operation, geological exploration and other methods can be used to obtain the rock layer distribution data in the geological terrain data. Then, through techniques such as geological modeling, a structural continuity analysis is conducted on the rock layer distribution data to identify the ranges of fault lines or loose sediment areas. Finally, the proportion of the overlapping area is calculated to generate the geological stability feature. For example, when processing geological terrain data in a region with complex geological conditions, through the above steps, the ranges of fault lines or loose sediment areas can be accurately identified, and the geological stability feature can be generated.

[0030] Step S214: Invoke the feature fusion layer in the preset terrain coding structure to convert the terrain undulation feature into an undulation coding sequence and convert the geological stability feature into a stability coding sequence.

[0031] The feature fusion layer in the preset terrain coding structure is a predefined layer structure used to encode and fuse the terrain undulation feature and the geological stability feature. Converting the terrain undulation feature into an undulation coding sequence means converting the terrain undulation feature according to the preset coding structure into a sequence-form coding. Similarly, converting the geological stability feature into a stability coding sequence is also to perform coding conversion on the geological stability feature.

[0032] In specific implementation, the feature fusion layer can adopt model structures such as neural networks. By training this model, it can accurately convert the terrain undulation feature and the geological stability feature into corresponding coding sequences. For example, use a multi-layer perceptron (MLP) as the feature fusion layer, take the terrain undulation feature and the geological stability feature as inputs, and after the processing of the MLP, output the undulation coding sequence and the stability coding sequence.

[0033] Step S215: Perform element-by-element weighted superposition on the undulation coding sequence and the stability coding sequence according to coordinate points to generate a terrain parameter sub-vector, where the weighting coefficient is dynamically adjusted according to the terrain region type.

[0034] Performing element-by-element weighted superposition on the undulation coding sequence and the stability coding sequence according to coordinate points means performing weighted summation on the elements at the corresponding coordinate points in the undulation coding sequence and the stability coding sequence to obtain a new vector. The weighting coefficient is dynamically adjusted according to the terrain region type, that is, different weight values are assigned to the elements of the undulation coding sequence and the stability coding sequence according to different terrain region types. In actual operation, first determine the weighting coefficient according to the terrain region type. Then perform element-by-element weighted superposition on the undulation coding sequence and the stability coding sequence according to coordinate points to generate a terrain parameter sub-vector. For example, when processing geological terrain data in an area containing mountains and plains, for the coordinate points in the mountainous area, the weighting coefficient of the undulation coding sequence can be appropriately increased; for the coordinate points in the plain area, the weighting coefficient of the stability coding sequence can be appropriately increased. In this way, the generated terrain parameter sub-vector can more accurately reflect the characteristics of different terrain regions.

[0035] Step S220: Perform spectral decomposition on the electromagnetic interference data to obtain electromagnetic radiation intensity distribution maps of different frequency bands, and extract the peak intensity features and spatial attenuation features of each frequency band by sliding a convolution kernel. Superpose the peak intensity features and the spatial attenuation features according to the frequency band to generate an electromagnetic intensity sub-vector.

[0036] Spectrum decomposition is to decompose electromagnetic interference data in the frequency domain and convert it into the electromagnetic radiation intensity distribution information of different frequency bands. Through spectrum decomposition, the distribution of electromagnetic radiation intensity in different frequency bands within the target area can be clearly understood. Convolution kernel sliding is a commonly used technique in image processing and signal processing. In the embodiments of the present invention, by sliding the convolution kernel on the electromagnetic radiation intensity distribution maps of different frequency bands, the peak intensity features and spatial attenuation features of each frequency band can be extracted.

[0037] The peak intensity feature refers to the maximum value of the electromagnetic radiation intensity within each frequency band, which reflects the strongest degree of electromagnetic radiation within that frequency band. The spatial attenuation feature describes the attenuation of the electromagnetic radiation intensity in space, that is, the variation law of the electromagnetic radiation intensity as the distance increases. The peak intensity feature and the spatial attenuation feature are superimposed by frequency band, that is, the peak intensity feature and the spatial attenuation feature of each frequency band are combined to generate an electromagnetic intensity sub-vector.

[0038] In actual operation, methods such as Fourier transform can be used to perform spectrum decomposition on the electromagnetic interference data to obtain the electromagnetic radiation intensity distribution maps of different frequency bands. Then, a suitable convolution kernel is designed and slid on the distribution map to extract the peak intensity features and spatial attenuation features of each frequency band. Finally, these two features are superimposed according to the frequency band to generate an electromagnetic intensity sub-vector. For example, when processing electromagnetic interference data in an area containing multiple substations and transmission lines, the electromagnetic radiation characteristics of different frequency bands in this area can be accurately extracted through the above steps to generate a representative electromagnetic intensity sub-vector.

[0039] Step S230: Perform topological structure analysis on the equipment layout data, identify the connection nodes of the transmission lines and the coverage radius of the substations, and generate a layout topology sub-vector based on the node connection density and the coverage radius weight.

[0040] Topological structure analysis is to deeply analyze the equipment layout data to identify the connection relationship of the transmission lines and the coverage range of the substations. The connection nodes of the transmission lines refer to the connection points between the transmission lines, and the distribution and connection methods of these nodes determine the topological structure of the transmission lines. The coverage radius of the substation refers to the range within which the substation can effectively supply power, which is related to the capacity of the substation and the layout of the transmission lines.

[0041] The node connection density refers to the number of connection nodes of the transmission lines in a certain area, which reflects the density of the transmission lines. The coverage radius weight is a weight value determined according to factors such as the coverage range and importance of the substation, and is used to weight the coverage radius when generating the layout topology sub-vector. Based on the node connection density and the coverage radius weight, these two factors are comprehensively considered to generate a layout topology sub-vector.

[0042] In specific implementation, methods such as graph theory can be used to analyze the topological structure of equipment layout data. The transmission lines and substations are abstracted as nodes and edges of a graph. By analyzing the structure of the graph, the connection nodes of the transmission lines and the coverage radius of the substations can be identified. Then, the node connection density is calculated and the coverage radius weight is determined, and these two factors are combined with weights to generate a layout topology sub-vector. For example, in the power network of a city, by analyzing the topological structure of the equipment layout data, the connection mode of the transmission lines and the coverage range of the substations can be clearly understood, and then an accurate layout topology sub-vector can be generated.

[0043] Step S240: Concatenate the terrain parameter sub-vector, the electromagnetic intensity sub-vector, and the layout topology sub-vector according to a preset dimension alignment rule to obtain a multi-dimensional survey feature vector.

[0044] The preset dimension alignment rule is a predefined rule used to ensure that the dimensions of the terrain parameter sub-vector, the electromagnetic intensity sub-vector, and the layout topology sub-vector are consistent when concatenating. In practical applications, this rule can be adjusted according to the dimension information of each sub-vector so that they can accurately correspond when concatenating.

[0045] Concatenating the terrain parameter sub-vector, the electromagnetic intensity sub-vector, and the layout topology sub-vector according to the preset dimension alignment rule means connecting these three sub-vectors together in a preset order to form a multi-dimensional survey feature vector containing multiple dimension information. This multi-dimensional survey feature vector synthesizes information in multiple aspects such as terrain, electromagnetism, and equipment layout, and can more comprehensively describe the characteristics of the target area.

[0046] In specific operations, first determine the preset dimension alignment rule, and adjust the dimensions of the terrain parameter sub-vector, the electromagnetic intensity sub-vector, and the layout topology sub-vector according to this rule. Then concatenate the three adjusted sub-vectors to obtain a multi-dimensional survey feature vector. For example, in the survey of a large-scale power project, by concatenating the terrain parameter sub-vector, the electromagnetic intensity sub-vector, and the layout topology sub-vector according to the preset dimension alignment rule, the obtained multi-dimensional survey feature vector can provide comprehensive and accurate feature information for subsequent power project planning.

[0047] Step S300: Input the multi-dimensional survey feature vector into an optimization decision model for iterative parameter adjustment to generate a set of optimized parameters for the power project in the target area, where the optimization decision model is trained based on the mapping relationship between historical survey data and verified optimization results.

[0048] The optimization decision-making model is a model based on machine learning or deep learning. By learning the mapping relationship between a large amount of historical survey data and the verified optimization results, it can generate a set of power engineering optimization parameters for the target area according to the input multi-dimensional survey feature vector. Iterative parameter adjustment refers to continuously adjusting the parameters of the model during the model training process to make the output result of the model gradually approach the optimal solution.

[0049] In specific implementation, the optimization decision-making model can adopt model structures such as neural networks and genetic algorithms. Taking the neural network as an example, it can process and analyze the multi-dimensional survey feature vector through the connection and calculation of multiple layers of neurons, and output a set of power engineering optimization parameters. During the training process, using the historical survey data as the input and the verified optimization results as the output, by continuously adjusting the weights and biases of the neural network, the model can learn the mapping relationship between the historical survey data and the verified optimization results.

[0050] For example, in the power engineering planning of a city, inputting the multi-dimensional survey feature vector of the city into the optimization decision-making model, after iterative parameter adjustment, the model can generate a set of power engineering optimization parameters for the city, including transmission line path planning parameters, substation location coordinate parameters, interference suppression parameters, etc. These parameters can provide scientific and reasonable guidance for the construction of power engineering.

[0051] As an implementation manner, the optimization decision-making model includes a parameter adjustment network and a verification feedback network. Based on this, in step S300, inputting the multi-dimensional survey feature vector into the optimization decision-making model for iterative parameter adjustment to generate a set of power engineering optimization parameters for the target area, which can specifically include the following steps: Step S310: Perform a non-linear transformation on the multi-dimensional survey feature vector through the parameter adjustment network to generate an initial optimization parameter set, where the initial optimization parameter set includes path planning parameters, location coordinate parameters, and interference suppression parameters.

[0052] The parameter adjustment network is an important part of the optimization decision-making model, which is used to perform a non-linear transformation on the multi-dimensional survey feature vector. Non-linear transformation refers to processing the input data through a non-linear function, so that the output result has more complex features and patterns. In the embodiment of the present invention, the parameter adjustment network performs a non-linear transformation on the multi-dimensional survey feature vector to generate an initial optimization parameter set.

[0053] The path planning parameters in the initial optimization parameter set are used to determine the path of the transmission line, the location coordinate parameters are used to determine the location of the substation, and the interference suppression parameters are used to control the influence of electromagnetic interference. These parameters are important bases for power engineering planning.

[0054] In specific implementation, the parameter adjustment network can adopt a fusion structure of a multi-layer perceptron (MLP) and an attention mechanism. A multi-layer perceptron is a common neural network structure, which consists of an input layer, a hidden layer, and an output layer. Through the connection and calculation of multiple layers of neurons, it realizes the non-linear transformation of input data. The attention mechanism can help the network pay more attention to the important features in the input data and improve the performance of the model.

[0055] For example, when inputting a multi-dimensional survey feature vector into the parameter adjustment network, through the calculation of the multi-layer perceptron and the action of the attention mechanism, the network can comprehensively analyze features such as terrain, electromagnetic, and equipment layout, and generate an initial set of optimized parameters. In the power engineering planning of a mountainous area, the parameter adjustment network can generate reasonable parameters for the transmission line path planning and substation site selection coordinates according to the terrain characteristics and electromagnetic environment of the mountainous area.

[0056] As an implementation method, the parameter adjustment network includes a fusion structure of a multi-layer perceptron and an attention mechanism. Based on this, in step S310, the multi-dimensional survey feature vector is non-linearly transformed through the parameter adjustment network to generate an initial set of optimized parameters, which can specifically include the following steps: Step S311: Input the terrain parameter sub-vector into the first hidden layer of the multi-layer perceptron for weight allocation to obtain the terrain influence weight coefficient.

[0057] A multi-layer perceptron is a neural network structure composed of multiple neuron layers, and the first hidden layer is the first hidden layer in the multi-layer perceptron. Inputting the terrain parameter sub-vector into the first hidden layer of the multi-layer perceptron for weight allocation means that the neurons in the first hidden layer perform weighted summation on each element in the terrain parameter sub-vector to obtain the terrain influence weight coefficient.

[0058] The terrain influence weight coefficient reflects the importance of terrain factors in power engineering planning, and it will affect subsequent decisions such as path planning and site selection. In actual operation, the first hidden layer of the multi-layer perceptron can obtain appropriate weight values through training, so that it can accurately calculate the terrain influence weight coefficient according to the terrain parameter sub-vector.

[0059] For example, in the power engineering planning of a mountainous area, the terrain parameter sub-vector contains information such as the terrain undulation and geological stability of the mountainous area. Inputting this sub-vector into the first hidden layer of the multi-layer perceptron, after the calculation of the neurons, the terrain influence weight coefficient is obtained. If the terrain of the mountainous area has large undulations and poor geological stability, then the terrain influence weight coefficient will be relatively large, indicating that terrain factors need to be considered key in power engineering planning.

[0060] Step S312: Perform cross-attention calculation on the electromagnetic intensity sub-vector and the layout topology sub-vector to generate electromagnetic-layout correlation features.

[0061] Cross-attention calculation is a calculation method for processing the relationships between multiple feature vectors, which can help the model better capture the correlation information between different feature vectors. In the embodiments of the present invention, performing cross-attention calculation on the electromagnetic intensity sub-vector and the layout topology sub-vector means that through the attention mechanism, the model focuses on the important features in the electromagnetic intensity sub-vector and the layout topology sub-vector to generate electromagnetic-layout correlation features.

[0062] The electromagnetic-layout correlation features reflect the mutual relationship between the electromagnetic environment and the device layout. For example, the layout of transmission lines may affect the distribution of electromagnetic radiation, and the intensity of electromagnetic radiation may also affect the operation of devices. By generating electromagnetic-layout correlation features, the roles of electromagnetic and layout factors in power engineering planning can be better comprehensively considered.

[0063] In specific implementation, methods such as dot-product attention in the attention mechanism can be used for cross-attention calculation. For example, the electromagnetic intensity sub-vector and the layout topology sub-vector are respectively used as the query vector and the key-value vector. By calculating the dot product between them, attention scores are obtained, and then the key-value vector is weighted and summed according to the attention scores to generate electromagnetic-layout correlation features.

[0064] Step S313: Based on the terrain influence weight coefficient, perform weighted aggregation on the electromagnetic-layout correlation features to generate comprehensive decision-making features.

[0065] Performing weighted aggregation on the electromagnetic-layout correlation features based on the terrain influence weight coefficient means multiplying the terrain influence weight coefficient by each element in the electromagnetic-layout correlation features, and then summing the multiplied results to obtain comprehensive decision-making features.

[0066] The comprehensive decision-making features integrate information from multiple aspects such as terrain, electromagnetic, and layout. It can more comprehensively reflect the characteristics of the target area and the requirements of power engineering planning. Through the weighted aggregation method, the terrain factor is reasonably reflected in the comprehensive decision-making features, enabling the model to make more scientific and accurate decisions during decision-making.

[0067] For example, in the power engineering planning of a city, the terrain influence weight coefficient reflects the degree of influence of the city's terrain on power engineering. Performing weighted aggregation of this weight coefficient with the electromagnetic-layout correlation features, the generated comprehensive decision-making features can comprehensively consider factors such as the city's terrain, electromagnetic environment, and device layout, providing a more comprehensive basis for subsequent path planning and site selection.

[0068] Step S314: Perform linear mapping on the comprehensive decision-making features through the output layer of the multi-layer perceptron to generate an initial optimization parameter set.

[0069] The output layer of the multi-layer perceptron is the last layer in the multi-layer perceptron, which is used to perform a linear mapping on the input feature vector to generate the final output result. In the embodiments of the present invention, performing a linear mapping on the comprehensive decision-making features through the output layer of the multi-layer perceptron means taking the comprehensive decision-making features as the input, and through the calculation of the neurons in the output layer, obtaining the initial optimization parameter set. The initial optimization parameter set includes information such as path planning parameters, site selection coordinate parameters, and interference suppression parameters. These parameters are the preliminary solutions for power engineering planning. Through the linear mapping of the output layer of the multi-layer perceptron, the comprehensive decision-making features can be converted into specific parameter values, providing a basis for subsequent simulation deployment verification and iterative adjustment.

[0070] In specific implementation, the output layer of the multi-layer perceptron can perform a linear calculation on the comprehensive decision-making features according to the trained weights and biases. For example, multiplying the comprehensive decision-making features by the weight matrix of the output layer and then adding the bias vector to obtain the initial optimization parameter set. In a large-scale power engineering planning, through the linear mapping of the output layer of the multi-layer perceptron on the comprehensive decision-making features, the generated initial optimization parameter set can provide specific parameter guidance for the preliminary planning of the project.

[0071] Step S320: Call the verification feedback network to perform simulation deployment verification on the initial optimization parameter set, and generate a parameter adjustment feedback signal, where the simulation deployment verification includes the calculation of the matching degree between the transmission line path and the terrain, the analysis of the overlapping rate of the substation coverage area, and the prediction of the electromagnetic interference suppression effect.

[0072] The verification feedback network is another component in the optimization decision-making model, which is used to perform simulation deployment verification on the initial optimization parameter set. Simulation deployment verification refers to simulating the implementation of the initial optimization parameter set in a virtual environment to evaluate its feasibility and effectiveness.

[0073] The calculation of the matching degree between the transmission line path and the terrain refers to calculating the matching degree between the path of the transmission line and the terrain of the target area, such as whether the transmission line passes through a steep slope area or a geological fault zone, etc. The analysis of the overlapping rate of the substation coverage area refers to analyzing the overlapping situation between the coverage area of the virtual substation and the coverage areas of existing power facilities to avoid waste of resources. The prediction of the electromagnetic interference suppression effect refers to predicting the attenuation of the electromagnetic radiation intensity in the target area after adopting the interference suppression parameters to evaluate the effectiveness of the interference suppression measures.

[0074] Based on the results of the simulation deployment verification, a parameter adjustment feedback signal is generated. The parameter adjustment feedback signal includes the path adjustment direction, the coordinate offset, and the interference suppression correction coefficient, etc. These signals are used to guide the parameter adjustment network to adjust the initial optimization parameter set.

[0075] In specific implementation, the verification feedback network can adopt a simulation model. By performing a simulated deployment on the initial optimized parameter set, it calculates indicators such as the matching degree between the transmission circuit path and the terrain, the overlap rate of the substation coverage area, and the electromagnetic interference suppression effect. Then, it generates a parameter adjustment feedback signal based on these indicators. For example, in the power engineering planning of a city, the verification feedback network can simulate the laying of transmission lines and the construction of substations, evaluate their matching conditions with the terrain and existing power facilities, and generate corresponding parameter adjustment feedback signals.

[0076] As an implementation method, in step S320, the verification feedback network is called to perform a simulated deployment verification on the initial optimized parameter set and generate a parameter adjustment feedback signal, which can specifically include the following steps: Step S321: Generate a virtual transmission circuit path according to the path planning parameters and calculate the matching degree score between the virtual transmission circuit path and the geological stability characteristics in the terrain parameter subvector.

[0077] Generating a virtual transmission circuit path according to the path planning parameters means simulating the path of the transmission line in a virtual environment according to the path planning parameters in the initial optimized parameter set. The geological stability characteristics are an important feature in the terrain parameter subvector, which reflects the geological stability of the target area.

[0078] Calculating the matching degree score between the virtual transmission circuit path and the geological stability characteristics is to obtain a matching degree score by evaluating whether the virtual transmission circuit path passes through geologically unstable areas and the degree of passing through these areas. The higher the matching degree score, the better the matching degree between the virtual transmission circuit path and the geological stability characteristics, and the higher the safety of the transmission line.

[0079] In specific implementation, spatial overlay analysis can be performed on the virtual transmission circuit path and the geological stability characteristics in the terrain parameter subvector, statistics can be made on indicators such as the length or area of the virtual transmission circuit path passing through geologically unstable areas, and then the matching degree score can be calculated according to the preset scoring rules. For example, in the power engineering planning of a mountainous area, a virtual transmission circuit path is generated according to the path planning parameters, and then the matching degree score between the path and the geological stability characteristics of the mountainous area is calculated. If the virtual transmission circuit path avoids fault lines and loose sediment areas, the matching degree score will be relatively high.

[0080] Step S322: Generate a virtual substation coverage area according to the site selection coordinate parameters and analyze the overlap rate between the virtual substation coverage area and the coverage range of existing power facilities.

[0081] Generating the coverage area of a virtual substation based on the site selection coordinate parameters means determining the location of the substation in a virtual environment according to the site selection coordinate parameters in the initial optimization parameter set, and calculating the coverage area of the substation based on the capacity of the substation and the layout of the transmission lines. The coverage range of existing power facilities refers to the coverage range of existing substations and transmission lines within the target area.

[0082] Analyzing the overlap rate between the coverage area of the virtual substation and the coverage range of existing power facilities is to obtain the overlap rate by calculating the ratio of the overlapping area between the coverage area of the virtual substation and the coverage range of existing power facilities to the area of the coverage area of the virtual substation. The higher the overlap rate, the greater the overlap between the coverage area of the virtual substation and the coverage area of existing power facilities, which may lead to waste of resources. In specific implementation, tools such as Geographic Information System (GIS) can be used to conduct spatial analysis on the coverage area of the virtual substation and the coverage range of existing power facilities to calculate the overlap rate. For example, in the power engineering planning of a city, the coverage area of a virtual substation is generated based on the site selection coordinate parameters, and then the overlap rate between this area and the coverage range of existing power facilities in the city is analyzed. If the overlap rate is high, it indicates that the site selection of the substation needs to be adjusted to avoid waste of resources.

[0083] Step S323: Predict the electromagnetic radiation intensity attenuation curve of the target area based on the interference suppression parameters, and calculate the deviation value between the attenuation curve and the preset electromagnetic safety threshold.

[0084] Predicting the electromagnetic radiation intensity attenuation curve of the target area based on the interference suppression parameters means predicting the attenuation of the electromagnetic radiation intensity in the target area after adopting interference suppression measures using methods such as electromagnetic propagation models according to the interference suppression parameters in the initial optimization parameter set, so as to obtain the electromagnetic radiation intensity attenuation curve. The preset electromagnetic safety threshold refers to the safety upper limit of the electromagnetic radiation intensity stipulated to ensure the safety of personnel and equipment.

[0085] Calculating the deviation value between the attenuation curve and the preset electromagnetic safety threshold is to obtain the deviation value by comparing the difference between each point on the attenuation curve and the preset electromagnetic safety threshold. The smaller the deviation value, the better the effect of the interference suppression measures, and the closer the electromagnetic radiation intensity is to the safety threshold. In specific implementation, tools such as electromagnetic simulation software can be used to predict the electromagnetic radiation intensity attenuation curve of the target area according to the interference suppression parameters. Then, by writing an algorithm, calculate the deviation value between the attenuation curve and the preset electromagnetic safety threshold. For example, in the power engineering planning near a substation, predict the electromagnetic radiation intensity attenuation curve of this area based on the interference suppression parameters, and then calculate the deviation value between this curve and the preset electromagnetic safety threshold. If the deviation value is small, it indicates that the interference suppression measures can effectively control the electromagnetic radiation intensity.

[0086] Step S324: Generate a parameter adjustment feedback signal based on the matching degree score, overlap rate, and deviation value. The parameter adjustment feedback signal includes a path adjustment direction, a coordinate offset, and an interference suppression correction coefficient. Generating a parameter adjustment feedback signal based on the matching degree score, overlap rate, and deviation value means generating a parameter adjustment feedback signal by comprehensively considering the impacts of the calculated matching degree score, overlap rate, and deviation value. The parameter adjustment feedback signal is used to guide the parameter adjustment network to adjust the initial optimization parameter set, making the final power engineering optimization parameter set more reasonable and effective.

[0087] The path adjustment direction refers to determining the direction in which the transmission line path needs to be adjusted according to the matching degree score, such as shifting left or right. The coordinate offset refers to determining the offset of the substation location coordinates that needs to be adjusted according to the overlap rate to avoid overlap of the coverage areas. The interference suppression correction coefficient refers to determining the coefficient by which the interference suppression parameters need to be adjusted according to the deviation value to improve the effect of the interference suppression measures. In specific implementation, the parameter adjustment feedback signal can be generated according to preset rules and algorithms based on the matching degree score, overlap rate, and deviation value. For example, if the matching degree score is low, it indicates that the transmission line path needs to be adjusted, and the path adjustment direction is determined according to the terrain and geological conditions; if the overlap rate is high, it indicates that the substation location needs to be adjusted, and the coordinate offset is calculated; if the deviation value is large, it indicates that the interference suppression parameters need to be adjusted, and the interference suppression correction coefficient is determined.

[0088] Step S330: Update the weight coefficients of the parameter adjustment network according to the parameter adjustment feedback signal, and repeat the non-linear transformation and simulation deployment verification until the initial optimization parameter set meets the preset deployment constraint conditions.

[0089] Updating the weight coefficients of the parameter adjustment network according to the parameter adjustment feedback signal means adjusting the weights of the parameter adjustment network according to the information in the parameter adjustment feedback signal, making the output result of the network more in line with the actual requirements. Repeating the non-linear transformation and simulation deployment verification means that the updated parameter adjustment network performs a non-linear transformation on the multi-dimensional survey feature vector again to generate a new initial optimization parameter set, and then the verification feedback network is called to perform a simulation deployment verification on the new initial optimization parameter set.

[0090] The preset deployment constraints refer to some conditions that need to be met in the power engineering planning, such as the safety of transmission lines, the coverage of substations, the limitation of electromagnetic interference, etc. Only when the initial optimization parameter set meets the preset deployment constraints will the iterative process stop. In specific implementation, optimization algorithms such as gradient descent can be used to update the weight coefficients of the parameter adjustment network according to the parameter adjustment feedback signal. For example, according to information such as the path adjustment direction and coordinate offset, the weight coefficients related to path planning and site selection in the parameter adjustment network are adjusted. By continuously repeating the iteration, the initial optimization parameter set gradually meets the preset deployment constraints.

[0091] Step S340: Determine the initial optimization parameter set that meets the deployment constraints as the power engineering optimization parameter set.

[0092] When, after multiple iterations, the initial optimization parameter set meets the preset deployment constraints, it is determined as the power engineering optimization parameter set. This set contains the optimal parameters required for the power engineering planning in the target area, including transmission line path planning parameters, substation site selection coordinate parameters, interference suppression parameters, etc.

[0093] The power engineering optimization parameter set is an important basis for power engineering construction. It can ensure the safe and efficient operation of transmission lines, the reasonable layout of substations, and the effective control of electromagnetic interference. For example, in the power engineering planning of a large industrial park, the determined power engineering optimization parameter set can guide the laying of transmission lines and the construction of substations, making the power supply in the park more stable and reliable.

[0094] As an implementation method, the pre-trained feature extraction network can be trained through the following steps: Step S10: Obtain a historical survey data set, which includes topographic data, electromagnetic data, and equipment layout data of multiple historical regions, as well as the verified optimization parameter set corresponding to each historical region.

[0095] The historical survey data set refers to the data set collected in past power engineering surveys. It contains topographic data, electromagnetic data, and equipment layout data of multiple historical regions, as well as the verified optimization parameter set corresponding to each historical region. The topographic data reflects the topographic characteristics of the historical region, the electromagnetic data records the electromagnetic environment of the historical region, the equipment layout data shows the layout of power facilities in the historical region, and the verified optimization parameter set is the power engineering optimization parameter that has been actually verified and is applicable to this historical region.

[0096] The historical survey dataset can be obtained in various ways, such as extracting from the power company's database, collecting from relevant survey reports, etc. These data are an important basis for the pre-trained feature extraction network. By learning these data, the network can master the relationships between data such as terrain, electromagnetic, and equipment layout and the optimization parameters of power engineering.

[0097] For example, collect historical survey data of multiple cities, including information such as terrain elevation, geological structure, electromagnetic radiation intensity distribution, transmission line topology, and substation locations in these cities, as well as the corresponding set of verified optimization parameters, such as transmission line path planning, substation site selection, and interference suppression measures. These data can provide rich training samples for the pre-trained feature extraction network.

[0098] Step S20: Perform data augmentation on the historical survey dataset to generate an augmented training sample set. Among them, the data augmentation includes randomly rotating terrain data, injecting noise into electromagnetic data, and perturbing nodes in layout data.

[0099] Data augmentation is a method of increasing the number and diversity of training samples by transforming and augmenting the original data. In the embodiment of the present invention, performing data augmentation on the historical survey dataset includes randomly rotating terrain data, injecting noise into electromagnetic data, and perturbing nodes in layout data.

[0100] Randomly rotating terrain data means randomly rotating information such as the elevation distribution matrix in the terrain data to simulate different terrain perspectives and directions. Injecting noise into electromagnetic data means adding noise to the electromagnetic data to simulate noise interference in the actual environment. Perturbing nodes in layout data means randomly disconnecting and reconnecting the transmission line nodes in the layout data to change the topology of the transmission line.

[0101] Through data augmentation, an augmented training sample set can be generated, increasing the diversity and complexity of training data and improving the generalization ability of the pre-trained feature extraction network. In actual operation, technologies such as image processing and data processing can be used to implement data augmentation. For example, use an image processing library to randomly rotate terrain data, use a noise generation algorithm to inject noise into electromagnetic data, and use graph theory algorithms to perturb nodes in layout data.

[0102] As an implementation method, in the above step S20, performing data augmentation on the historical survey dataset to generate an augmented training sample set can specifically include the following steps: Step S21: Perform a random rotation transformation on the terrain data to generate a rotated terrain elevation map, and add randomly generated geological fracture simulation features to the rotated terrain elevation map.

[0103] Performing a random rotation transformation on terrain data means performing a random rotation operation on the terrain elevation map in the terrain data to simulate different terrain perspectives and directions. The random rotation transformation can increase the diversity of the terrain data, making the training samples more representative.

[0104] After generating the rotated terrain elevation map, adding randomly generated geological fracture simulation features to it is to simulate the possible geological fracture conditions in the actual terrain. Geological fractures will affect the stability of the terrain and the construction of power projects. By adding geological fracture simulation features, the pre-trained feature extraction network can learn more complex terrain features.

[0105] In specific implementation, an image processing library such as OpenCV can be used to perform a random rotation transformation on the terrain elevation map. Then, a noise generation algorithm or image processing technology can be used to add randomly generated geological fracture simulation features to the rotated terrain elevation map. For example, use the Gaussian noise generation algorithm to generate a noise image of geological fractures, and then superimpose it on the rotated terrain elevation map to obtain a terrain elevation map containing geological fracture simulation features.

[0106] Step S22: Inject Gaussian white noise into the electromagnetic data to generate a noisy electromagnetic spectrogram, and perform smoothing processing on the noisy electromagnetic spectrogram through frequency domain filtering.

[0107] Injecting Gaussian white noise into the electromagnetic data means adding Gaussian white noise to the spectrogram of the electromagnetic data to simulate the noise interference in the actual environment. Gaussian white noise is a kind of noise with a uniform power spectral density, which can increase the complexity and diversity of the electromagnetic data.

[0108] After generating the noisy electromagnetic spectrogram, performing smoothing processing on it through frequency domain filtering is to remove the high-frequency components in the noise, making the electromagnetic spectrogram smoother and more stable. Frequency domain filtering can be implemented using methods such as Fourier transform. Convert the noisy electromagnetic spectrogram to the frequency domain, then use a filter to remove the high-frequency components, and finally convert it back to the time domain to obtain the smoothed electromagnetic spectrogram.

[0109] In specific implementation, a signal processing library such as SciPy can be used to inject Gaussian white noise into the electromagnetic data. Then, use Fourier transform and a filter to perform frequency domain filtering and smoothing processing on the noisy electromagnetic spectrogram. For example, use the fast Fourier transform (FFT) to convert the noisy electromagnetic spectrogram to the frequency domain, use a low-pass filter to remove the high-frequency components, and then use the inverse fast Fourier transform (IFFT) to convert it back to the time domain to obtain the smoothed electromagnetic spectrogram.

[0110] Step S23: Perform random disconnection and reconnection operations on the transmission line nodes in the device layout data to generate a device layout diagram with a mutated topological structure.

[0111] Performing random disconnection and reconnection operations on the transmission line nodes in the device layout data means randomly selecting transmission line nodes for disconnection or reconnection to change the topological structure of the transmission lines. This operation can simulate possible line faults or renovation situations in the actual power grid and increase the diversity of the device layout data.

[0112] Generating a device layout diagram with a mutated topological structure is the new device layout diagram obtained after performing random disconnection and reconnection operations on the transmission line nodes. This diagram shows the mutated topological structure of the transmission lines and provides more training samples for the pre-trained feature extraction network.

[0113] In specific implementation, graph theory algorithms can be used to perform random disconnection and reconnection operations on the transmission line nodes in the device layout data. For example, use the NetworkX library in Python to represent the graph structure of the device layout data, then randomly select nodes for disconnection or reconnection operations, update the graph structure, and obtain a device layout diagram with a mutated topological structure.

[0114] Step S24: Combine the rotated terrain elevation map, the noisy electromagnetic spectrum map, and the device layout diagram with a mutated topological structure into a new training sample and add it to the augmented training sample set.

[0115] Combining the rotated terrain elevation map, the noisy electromagnetic spectrum map, and the device layout diagram with a mutated topological structure into a new training sample means integrating the terrain data, electromagnetic data, and device layout data after data augmentation processing to form a new training sample. This sample contains more features and information and can improve the learning ability of the pre-trained feature extraction network.

[0116] Adding the new training sample to the augmented training sample set means adding the newly generated training sample to the augmented training sample set to increase the number and diversity of the training samples. By continuously performing data augmentation processing and sample combination, the augmented training sample set can contain more different types of training samples, enabling the pre-trained feature extraction network to learn more comprehensive features and patterns.

[0117] For example, combine the terrain elevation map that has undergone random rotation transformation and added geological fracture simulation features, the electromagnetic spectrum map that has undergone noise injection and frequency domain filtering processing, and the device layout diagram that has undergone topological structure mutation to form a new training sample. Then add this sample to the augmented training sample set for training the pre-trained feature extraction network.

[0118] Step S30: Construct an initial feature extraction network, and input the augmented training sample set into the initial feature extraction network for feature extraction to generate a set of predicted optimization parameters.

[0119] Constructing an initial feature extraction network means designing and building a neural network model for feature extraction. This model can adopt architectures such as convolutional neural networks (CNNs) and recurrent neural networks (RNNs), and is selected according to specific requirements and data characteristics.

[0120] Inputting the augmented training sample set into the initial feature extraction network for feature extraction means taking the terrain data, electromagnetic data, and equipment layout data in the augmented training sample set as inputs, and through the calculation and processing of the initial feature extraction network, extracting representative features to generate a set of predicted optimization parameters. The set of predicted optimization parameters is the preliminary optimization parameters generated by the initial feature extraction network based on the input data.

[0121] In specific implementation, deep learning frameworks such as TensorFlow and PyTorch can be used to construct the initial feature extraction network. For example, construct an initial feature extraction network based on CNN, input the augmented training sample set into this network, and through the calculation of convolutional layers, pooling layers, and fully connected layers, etc., generate a set of predicted optimization parameters.

[0122] Step S40: Calculate the mean squared error loss between the set of predicted optimization parameters and the set of verified optimization parameters, and update the parameters of the initial feature extraction network based on the gradient descent algorithm until the mean squared error loss converges to a preset threshold.

[0123] The mean squared error loss is a loss function used to measure the difference between predicted values and true values. In the embodiments of the present invention, calculating the mean squared error loss between the set of predicted optimization parameters and the set of verified optimization parameters is to calculate the average of the sum of the squares of the differences between each parameter in the set of predicted optimization parameters and the corresponding parameter in the set of verified optimization parameters to obtain the mean squared error loss.

[0124] Updating the parameters of the initial feature extraction network based on the gradient descent algorithm means adjusting the weights and biases of the initial feature extraction network according to the gradient information of the mean squared error loss, so that the mean squared error loss gradually decreases. The gradient descent algorithm is a commonly used optimization algorithm that continuously iteratively updates the parameters of the network to make the loss function reach the minimum value.

[0125] The preset threshold is a pre-set error value. When the mean squared error loss converges to the preset threshold, it indicates that the initial feature extraction network has learned sufficient information and the training process can stop. In actual operation, the gradient descent algorithm can be implemented using the optimizer provided by the deep learning framework. For example, use the Adam optimizer to update the parameters of the initial feature extraction network according to the mean squared error loss until the mean squared error loss converges to the preset threshold.

[0126] Step S50: Determine the initial feature extraction network with updated parameters as the pre-trained feature extraction network.

[0127] After the mean squared error loss converges to the preset threshold, determine the initial feature extraction network with updated parameters as the pre-trained feature extraction network. This network has learned the relationship between data such as terrain, electromagnetic, and equipment layout and the power engineering optimization parameters through learning the historical survey dataset, and can effectively extract features from new multi-source survey data.

[0128] The pre-trained feature extraction network can be used for subsequent power engineering survey data processing to support the generation of multi-dimensional survey feature vectors. For example, in a new power engineering survey project, input the multi-source survey data of this project into the pre-trained feature extraction network, and the network can extract representative features and generate multi-dimensional survey feature vectors to provide a basis for subsequent optimization decisions.

[0129] Step S400: Generate a power facility deployment plan for the target area according to the power engineering optimization parameter set. The deployment plan includes the transmission line path planning result, the substation location coordinates, and the electromagnetic interference suppression strategy.

[0130] The power engineering optimization parameter set contains information such as transmission line path planning parameters, substation location coordinate parameters, and interference suppression parameters. Generate a power facility deployment plan for the target area according to these parameters. The transmission line path planning result refers to the specific path of the transmission line determined according to the path planning parameters, which needs to consider various factors such as terrain, geology, and electromagnetism to ensure the safe and efficient operation of the transmission line. The substation location coordinates refer to the specific location of the substation determined according to the location coordinate parameters, which need to consider factors such as power load distribution, terrain conditions, and electromagnetic environment to achieve a reasonable layout of the substation. The electromagnetic interference suppression strategy refers to the measures to control electromagnetic interference formulated according to the interference suppression parameters, such as installing shielding equipment and adjusting the layout of the transmission line.

[0131] In specific implementation, tools such as Geographic Information System (GIS) can be used to generate a power facility deployment plan according to the power engineering optimization parameter set. For example, the transmission line path planning parameters and terrain data are input into the GIS, and through spatial analysis and path planning algorithms, the transmission line path planning result is generated. According to the substation site selection coordinate parameters and power load distribution data, the substation site selection coordinates are determined in the GIS. According to the interference suppression parameters and combined with the electromagnetic environment data, an electromagnetic interference suppression strategy is formulated.

[0132] As an implementation manner, step S400, generating a power facility deployment plan for the target area according to the power engineering optimization parameter set, may specifically include the following steps: Step S410: Extract an initial transmission line coordinate point set based on the path planning parameters in the power engineering optimization parameter set, and combine the elevation gradient distribution characteristics in the terrain parameter sub-vector to identify the coordinate points in the coordinate point set that overlap with the steep slope area or geological fault zone.

[0133] Extracting an initial transmission line coordinate point set based on the path planning parameters in the power engineering optimization parameter set means determining the general direction and position of the transmission line according to the path planning parameters and extracting the coordinate point set on the transmission line. This coordinate point set is the preliminary result of the transmission line path planning.

[0134] Combining the elevation gradient distribution characteristics in the terrain parameter sub-vector to identify the coordinate points in the coordinate point set that overlap with the steep slope area or geological fault zone means performing a spatial overlay analysis on the initial transmission line coordinate point set and the elevation gradient distribution characteristics in the terrain parameter sub-vector to find the coordinate points in the coordinate point set that are located in the steep slope area or geological fault zone. These coordinate points may affect the safety and stability of the transmission line and need further processing.

[0135] In specific implementation, tools such as Geographic Information System (GIS) can be used for spatial overlay analysis. The initial transmission line coordinate point set and the elevation gradient distribution characteristics in the terrain parameter sub-vector are imported into the GIS, and through the spatial analysis function, the coordinate points that overlap with the steep slope area or geological fault zone are identified. For example, in the power engineering planning of a mountainous area, an initial transmission line coordinate point set is extracted according to the path planning parameters, and then combined with the elevation gradient distribution characteristics in the terrain parameter sub-vector of the mountainous area, the coordinate points located in the steep slope area are identified.

[0136] Step S420: Verify the electromagnetic interference intensity of the overlapping coordinate points, call the peak intensity distribution data of the corresponding area in the electromagnetic intensity sub-vector, and screen the coordinate points with peak intensity exceeding the safety threshold as high-risk path nodes.

[0137] Verifying the electromagnetic interference intensity for overlapping coordinate points refers to checking whether the electromagnetic interference intensity of the coordinate points identified in step S410 that overlap with the steep slope area or geological fault zone exceeds the safety threshold. Retrieving the peak intensity distribution data for the corresponding area in the electromagnetic intensity sub-vector means obtaining the electromagnetic radiation peak intensity information for the areas where these coordinate points are located from the electromagnetic intensity sub-vector.

[0138] Selecting the coordinate points with peak intensity exceeding the safety threshold as high-risk path nodes means marking the coordinate points with electromagnetic radiation peak intensity exceeding the safety threshold as high-risk path nodes. These nodes may be subject to strong electromagnetic interference, affecting the normal operation of the transmission line, and corresponding measures need to be taken for treatment.

[0139] In specific implementation, an algorithm can be written to verify the electromagnetic interference intensity for overlapping coordinate points. According to the positions of the coordinate points, the peak intensity distribution data for the corresponding areas are extracted from the electromagnetic intensity sub-vector, and then compared with the safety threshold to select the coordinate points exceeding the safety threshold as high-risk path nodes. For example, in the power engineering planning near a substation, the electromagnetic interference intensity for the coordinate points overlapping with the steep slope area is verified, and the coordinate points with peak intensity exceeding the safety threshold are selected as high-risk path nodes.

[0140] Step S430: Perform path detour optimization on the high-risk path nodes, generate an avoidance path correction trajectory according to the geological stability characteristics in the terrain parameter sub-vector, and smoothly connect the endpoints of the correction trajectory with the unadjusted initial coordinate point set to form the transmission line path planning result.

[0141] Performing path detour optimization on the high-risk path nodes means taking measures to change the path of the transmission line to avoid these nodes for the high-risk path nodes selected in step S420. Generating an avoidance path correction trajectory according to the geological stability characteristics in the terrain parameter sub-vector means finding a reasonable path to avoid the high-risk path nodes according to the geological stability characteristics and generating an avoidance path correction trajectory.

[0142] Smoothly connecting the endpoints of the correction trajectory with the unadjusted initial coordinate point set means connecting the start and end points of the avoidance path correction trajectory with the unadjusted initial coordinate point set to make the path of the transmission line smoother and more continuous. Forming the transmission line path planning result is to obtain the final transmission line path planning result through path detour optimization and connection processing for the high-risk path nodes.

[0143] In specific implementation, a path planning algorithm and spatial analysis technology can be used to perform path detour optimization on high-risk path nodes. According to the geological stability characteristics in the terrain parameter sub-vector, a detour area is determined, and then an avoidance path correction trajectory is generated within the detour area. Finally, methods such as curve fitting are used to smoothly connect the endpoints of the correction trajectory with the unadjusted initial coordinate point set. For example, in the planning of a power project in a mountainous area, path detour optimization is performed on high-risk path nodes, an avoidance path correction trajectory is generated according to the geological stability characteristics of the mountainous area, and it is smoothly connected with the unadjusted initial coordinate point set to form the power transmission line path planning result.

[0144] As an implementation manner, in step S430, path detour optimization is performed on high-risk path nodes, and an avoidance path correction trajectory is generated according to the geological stability characteristics in the terrain parameter sub-vector. Specifically, it may include the following steps: Step S431: Extract the terrain undulation characteristics of the area where the high-risk path node is located, generate a contour map centered on this node, and determine the boundary of the detour area according to the changing direction of the contour curvature.

[0145] Extracting the terrain undulation characteristics of the area where the high-risk path node is located means obtaining the terrain undulation information of the area where the high-risk path node is located from the terrain parameter sub-vector. Generating a contour map centered on this node is to draw a contour map centered on the high-risk path node according to the terrain undulation characteristics, which can visually display the terrain undulation of this area.

[0146] Determining the boundary of the detour area according to the changing direction of the contour curvature means finding the areas with relatively gentle terrain and suitable for the detour of the power transmission line by analyzing the changing direction of the contour curvature, and determining the boundaries of these areas. The boundary of the detour area determines the detour range that the power transmission line can choose.

[0147] In specific implementation, tools such as Geographic Information System (GIS) can be used to extract the terrain undulation characteristics of the area where the high-risk path node is located and generate a contour map. Then, by analyzing the changing direction of the contour curvature, a spatial analysis algorithm is used to determine the boundary of the detour area. For example, in the planning of a power project in a mountainous area, for a high-risk path node, the terrain undulation characteristics of its location area are extracted, a contour map is generated, and the boundary of the detour area is determined according to the changing direction of the contour curvature.

[0148] Step S432: Call the peak intensity distribution data of the corresponding area in the electromagnetic intensity sub-vector, exclude the sub-areas with electromagnetic radiation intensity exceeding the preset threshold within the detour area, and generate an electromagnetic safety detour range.

[0149] Calling the peak intensity distribution data of the corresponding region in the electromagnetic intensity sub-vector means obtaining the electromagnetic radiation peak intensity information of the region where the high-risk path nodes are located and the detourable regions from the electromagnetic intensity sub-vector. Excluding the sub-regions with electromagnetic radiation intensity exceeding the preset threshold within the detourable regions means excluding the sub-regions with electromagnetic radiation intensity exceeding the threshold in the detourable regions according to the preset electromagnetic radiation safety threshold, and only retaining the regions where the electromagnetic radiation intensity is within the safe range.

[0150] Generating the electromagnetic safety detour range is the safe area available for the detour of the transmission line obtained by excluding the sub-regions with electromagnetic radiation intensity exceeding the preset threshold. This area takes into account both terrain factors and electromagnetic radiation factors to ensure that the transmission line will not be subject to excessive electromagnetic interference during the detour process.

[0151] In specific implementation, an algorithm can be written to call the peak intensity distribution data of the corresponding region in the electromagnetic intensity sub-vector and perform screening and exclusion operations. Import the detourable regions and electromagnetic intensity data into the algorithm, and exclude the sub-regions with electromagnetic radiation intensity exceeding the threshold according to the preset threshold to generate the electromagnetic safety detour range. For example, in the power engineering planning near a substation, call the peak intensity distribution data of the corresponding region in the electromagnetic intensity sub-vector, exclude the sub-regions with electromagnetic radiation intensity exceeding the preset threshold within the detourable regions, and generate the electromagnetic safety detour range. In specific implementation, an algorithm can be written to call the peak intensity distribution data of the corresponding region in the electromagnetic intensity sub-vector and perform screening and exclusion operations. Import the detourable regions and electromagnetic intensity data into the algorithm, and exclude the sub-regions with electromagnetic radiation intensity exceeding the threshold according to the preset threshold to generate the electromagnetic safety detour range. For example, in the power engineering planning near a substation, for the detourable regions of the high-risk path nodes, obtain the electromagnetic radiation peak intensity distribution of this region and its surrounding areas from the electromagnetic intensity sub-vector. Assume that the preset electromagnetic radiation safety threshold is a certain preset value, and the algorithm will detect the electromagnetic radiation intensity of each sub-region within the detourable regions. When it is detected that the electromagnetic radiation intensity of a certain sub-region exceeds the threshold, that sub-region will be excluded from the detourable regions. In this way, the final electromagnetic safety detour range can ensure that when the transmission line detours within this range, the electromagnetic radiation it receives is at a safe level, avoiding the impact on the normal operation of the transmission line caused by electromagnetic interference.

[0152] Step S433: Based on the superposition result of the terrain undulation characteristics and the electromagnetic safety detour range, construct a detour channel model allowing path correction. The model includes the maximum allowable slope and the electromagnetic radiation safety interval within the channel.

[0153] Constructing a detour channel model based on the superposition result of terrain undulation characteristics and electromagnetic safety detour range comprehensively considers terrain factors and electromagnetic safety factors. Terrain undulation characteristics reflect the terrain complexity of the area, while the electromagnetic safety detour range defines the feasible area under the premise of electromagnetic radiation safety. By superimposing these two, the area that meets both terrain conditions and electromagnetic safety requirements can be determined, and based on this, a detour channel model is constructed.

[0154] This model includes the maximum allowable slope within the channel and the electromagnetic radiation safety interval. The maximum allowable slope is determined according to the requirements of transmission line construction and terrain conditions. If the slope exceeds this value, it may increase the difficulty and cost of transmission line construction and even affect the stability of the line. The electromagnetic radiation safety interval is determined according to electromagnetic radiation safety standards. Within this interval, the electromagnetic interference received by the transmission line is within an acceptable range.

[0155] When constructing the model, the method of combining Geographic Information System (GIS) with mathematical modeling can be used. First, import the data of terrain undulation characteristics and electromagnetic safety detour range into GIS for superposition analysis to determine the boundary of the area that meets the requirements. Then, according to the specifications and experience of transmission line construction, determine the specific values of the maximum allowable slope and the electromagnetic radiation safety interval. Finally, use a mathematical model to integrate this information to construct a detour channel model that allows path correction. For example, in a power project in a mountainous area, by superimposing terrain undulation characteristics and electromagnetic safety detour range, it is found that although the terrain undulation is large in some areas, the electromagnetic radiation is within the safe range; while in some other areas, the terrain is relatively flat, but the electromagnetic radiation intensity is high. Through comprehensive analysis, a detour channel model that meets both terrain slope requirements and electromagnetic safety standards is determined.

[0156] Step S434: Generate multiple alternative path trajectories according to the detour channel model, and calculate the deviation angle, cumulative slope change amount between each trajectory and the original path, and the average electromagnetic radiation value of the area passed by.

[0157] Generating multiple alternative path trajectories according to the detour channel model uses the range and conditions determined by the detour channel model and uses path planning algorithms to generate multiple different transmission line paths. These paths are all within the range allowed by the detour channel model and meet the requirements of the maximum allowable slope and the electromagnetic radiation safety interval.

[0158] Calculating the deviation angle between each trajectory and the original path measures the deviation degree between the alternative path and the original transmission line path. The larger the deviation angle, the greater the difference between the alternative path and the original path. The cumulative slope change reflects the slope change of the alternative path over its entire length. The smaller the cumulative slope change, the flatter the path, and the lower the construction difficulty and cost may be. The average electromagnetic radiation value of the passing area refers to the average electromagnetic radiation intensity of the area through which the alternative path passes. The lower this value, the less electromagnetic interference the path is subject to.

[0159] In actual operation, path planning algorithms such as Dijkstra's algorithm or A* algorithm can be used to generate multiple alternative path trajectories within the detour channel model. For each alternative path, its spatial relationship with the original path is obtained through a Geographic Information System (GIS) to calculate the deviation angle; the cumulative slope change is calculated by analyzing the slope information of each point on the path; the average electromagnetic radiation value of the passing area is calculated by calling the data corresponding to the area in the electromagnetic intensity sub-vector. For example, in a power project in a city, multiple alternative path trajectories are generated for high-risk path nodes. Through calculation, it is found that for some paths, although the deviation angle from the original path is large, the cumulative slope change is small and the average electromagnetic radiation value of the passing area is low; while for other paths, the deviation angle is small, but the cumulative slope change is large and the electromagnetic radiation average value is high.

[0160] Step S435: Select the alternative trajectory with a deviation angle less than the preset tolerance, the smallest cumulative slope change, and the lowest average electromagnetic radiation value as the avoidance path correction trajectory, and continuously connect the endpoints of the correction trajectory with the unadjusted initial coordinate point set in terms of curvature.

[0161] Selecting the alternative trajectory with a deviation angle less than the preset tolerance, the smallest cumulative slope change, and the lowest average electromagnetic radiation value as the avoidance path correction trajectory is to screen among multiple alternative path trajectories to find the optimal path correction plan. The preset tolerance is a permitted deviation angle range determined according to the design requirements and actual situation of the transmission line. If the deviation angle of the alternative path exceeds this tolerance, it may have a greater impact on the overall layout and operation of the transmission line. The smallest cumulative slope change can ensure that the path is relatively flat, reducing construction costs and maintenance difficulties; the lowest average electromagnetic radiation value can reduce the impact of electromagnetic interference on the transmission line.

[0162] Continuously connecting the endpoints of the correction trajectory with the unadjusted initial coordinate point set in terms of curvature is to ensure the continuity and stability of the transmission line. Curvature continuous connection can avoid sudden changes in the line, reduce stress concentration on the line, and improve the safety and reliability of the transmission line.

[0163] In specific implementation, first, the alternative path trajectories are screened according to a preset tolerance, and the paths with deviation angles exceeding the tolerance are excluded. Then, the cumulative slope change and the average electromagnetic radiation of the screened paths are calculated, and the path with the smallest cumulative slope change and the lowest average electromagnetic radiation is selected as the avoidance path correction trajectory. Finally, the method of curve fitting is used to continuously connect the endpoints of the correction trajectory with the unadjusted initial coordinate point set in terms of curvature. For example, in a power project in a mountainous area, through screening and comparison, an alternative trajectory with a deviation angle within the preset tolerance range, the smallest cumulative slope change, and the lowest average electromagnetic radiation is selected as the avoidance path correction trajectory, and it is smoothly and continuously connected with the unadjusted initial coordinate point set in terms of curvature to ensure the smoothness and stability of the transmission line.

[0164] Step S436: Package the terrain constraint parameters and electromagnetic safety parameters involved in the avoidance path correction trajectory into path optimization rules and embed them into the dynamic adjustment logic of the transmission line path planning result to ensure that local correction instructions are generated based on real-time terrain data when encountering unsurveyed obstacles during the construction stage.

[0165] Packaging the terrain constraint parameters and electromagnetic safety parameters involved in the avoidance path correction trajectory into path optimization rules means organizing and packaging the terrain factors (such as the maximum allowable slope) and electromagnetic safety factors (such as the electromagnetic radiation safety range) considered during the generation of the avoidance path correction trajectory to form a set of rule systems. These rules can reflect the optimal conditions for transmission line path planning under predefined terrain and electromagnetic environments.

[0166] Embedding into the dynamic adjustment logic of the transmission line path planning result means integrating the packaged path optimization rules into the dynamic adjustment mechanism of the transmission line path planning result. During the construction stage, some unsurveyed obstacles may be encountered, such as sudden geological faults and newly discovered electromagnetic interference sources. At this time, based on the real-time terrain data obtained, these path optimization rules can be reused to locally correct the path of the transmission line.

[0167] In specific implementation, programming languages can be used to encapsulate the terrain constraint parameters and electromagnetic safety parameters into a rule object or class. Then, in the dynamic adjustment logic of the transmission line path planning, the call and application of this rule object are added. When real-time terrain data is obtained during the construction phase, the system will automatically determine whether the transmission line path needs to be corrected according to these data and the path optimization rules, and generate corresponding local correction instructions. For example, during the construction of a large-scale power project, when an un-surveyed steep slope area is encountered, the system will judge whether the area exceeds the maximum allowable slope according to the real-time terrain data and the encapsulated path optimization rules. If it exceeds, the system will automatically generate local correction instructions to adjust the path of the transmission line, avoiding the steep slope area while ensuring that the new path still meets the requirements of the electromagnetic radiation safety interval.

[0168] Step S440: Generate a set of candidate substation locations based on the site selection coordinate parameters in the power project optimization parameter set, and calculate the coverage efficiency of each candidate location for the surrounding lines and the electromagnetic radiation suppression requirement score by combining the coverage radius weight in the layout topology sub-vector and the spatial attenuation characteristics of the electromagnetic intensity sub-vector.

[0169] Generating a set of candidate substation locations based on the site selection coordinate parameters in the power project optimization parameter set means determining multiple possible substation locations according to the site selection coordinate parameters to form a set of candidate substation locations. Calculating the coverage efficiency of each candidate location for the surrounding lines and the electromagnetic radiation suppression requirement score by combining the coverage radius weight in the layout topology sub-vector and the spatial attenuation characteristics of the electromagnetic intensity sub-vector means evaluating the advantages and disadvantages of each candidate location considering the coverage range of the substation and the electromagnetic radiation impact.

[0170] The coverage efficiency refers to the degree of coverage of the substation for the surrounding transmission lines. The higher the coverage efficiency, the better the substation can supply power to the surrounding lines. The electromagnetic radiation suppression requirement score refers to evaluating the degree of electromagnetic radiation suppression requirement of the candidate location according to the spatial attenuation characteristics of the electromagnetic intensity sub-vector. The higher the score, the stronger the electromagnetic radiation suppression measures are required for this location.

[0171] In specific implementation, tools such as Geographic Information System (GIS) can be used to calculate the coverage efficiency of each candidate location for the surrounding lines and the electromagnetic radiation suppression requirement score. Import the set of candidate substation locations, the coverage radius weight in the layout topology sub-vector, and the spatial attenuation characteristics of the electromagnetic intensity sub-vector into GIS, and through spatial analysis and calculation, obtain the coverage efficiency and electromagnetic radiation suppression requirement score of each candidate location. For example, in the power project planning of a city, generate a set of candidate substation locations based on the site selection coordinate parameters, and then calculate the coverage efficiency and electromagnetic radiation suppression requirement score of each candidate location by combining the information of the layout topology sub-vector and the electromagnetic intensity sub-vector.

[0172] Step S450: Select the substation site coordinates according to the comprehensive weight value of the coverage efficiency and the suppression requirement score, and generate the installation density and azimuth parameters of the shielding equipment matching the site coordinates based on the interference suppression parameters in the power engineering optimization parameter set to form an electromagnetic interference suppression strategy.

[0173] Selecting the substation site coordinates according to the comprehensive weight value of the coverage efficiency and the suppression requirement score means comprehensively considering the coverage efficiency and the electromagnetic radiation suppression requirement score, assigning corresponding weights to each index, calculating the comprehensive weight value, and then selecting the candidate location with the highest comprehensive weight value as the substation site coordinates.

[0174] Generating the installation density and azimuth parameters of the shielding equipment matching the site coordinates based on the interference suppression parameters in the power engineering optimization parameter set means determining the installation density and azimuth of the shielding equipment according to the interference suppression parameters and the selected substation site coordinates to effectively suppress electromagnetic radiation. Forming an electromagnetic interference suppression strategy is to integrate information such as the installation density and azimuth parameters of the shielding equipment to form a complete electromagnetic interference suppression strategy.

[0175] In specific implementation, an algorithm can be written to select the substation site coordinates according to the comprehensive weight value of the coverage efficiency and the suppression requirement score. Then, according to the interference suppression parameters and the site coordinates, tools such as electromagnetic simulation software are used to calculate the installation density and azimuth parameters of the shielding equipment. For example, in the power engineering planning of an industrial park, the substation site coordinates are selected according to the comprehensive weight value of the coverage efficiency and the suppression requirement score, and then the installation density and azimuth parameters of the shielding equipment matching the site coordinates are generated based on the interference suppression parameters to form an electromagnetic interference suppression strategy.

[0176] Step S460: Perform spatial topological association on the transmission line path planning result, the substation site coordinates, and the electromagnetic interference suppression strategy to generate a power facility deployment plan including construction coordinate mapping rules and electromagnetic protection interlocking logic, where the interlocking logic is used to dynamically adjust the working parameters of the shielding equipment according to real-time electromagnetic data during the substation operation stage.

[0177] Performing spatial topological association on the transmission line path planning result, the substation site coordinates, and the electromagnetic interference suppression strategy means associating the path of the transmission line, the location of the substation, and the electromagnetic interference suppression measures in space to ensure their coordination and consistency. Generating a power facility deployment plan including construction coordinate mapping rules and electromagnetic protection interlocking logic is to integrate the results of the spatial topological association to form a complete power facility deployment plan including construction coordinate mapping rules and electromagnetic protection interlocking logic.

[0178] The construction coordinate mapping rule refers to converting the transmission line path planning result and the substation location coordinates into coordinate information in actual construction to guide construction personnel in construction. The electromagnetic protection joint control logic refers to dynamically adjusting the working parameters of shielding equipment according to real-time electromagnetic data during the operation stage of the substation to ensure that the electromagnetic radiation intensity always meets the safety standards.

[0179] In specific implementation, tools such as Geographic Information System (GIS) can be used for spatial topological association. Import the transmission line path planning result, substation location coordinates, and electromagnetic interference suppression strategy into GIS, and through spatial analysis and data processing, generate the construction coordinate mapping rule and the electromagnetic protection joint control logic. For example, in the planning of a large-scale power project, perform spatial topological association on the transmission line path planning result, substation location coordinates, and electromagnetic interference suppression strategy to generate a power facility deployment plan including the construction coordinate mapping rule and the electromagnetic protection joint control logic.

[0180] As an implementation manner, the method provided by the embodiment of the present invention further includes the step of updating the optimization decision model, which may specifically include the following steps: Step S500: Real-time collect the operation monitoring data set after the deployment of power facilities in the target area. The operation monitoring data set includes the load rate fluctuation curve of the transmission line, the operation efficiency index of substation equipment, and the real-time monitoring value of the electromagnetic radiation intensity.

[0181] Real-time collecting the operation monitoring data set after the deployment of power facilities in the target area is to timely understand the status and performance of power facilities during actual operation. The load rate fluctuation curve of the transmission line reflects the load conditions of the transmission line at different time periods. By analyzing this curve, the load change law of the transmission line can be understood, and whether there are abnormal situations such as overload or light load can be judged. The operation efficiency index of substation equipment reflects the operation status and efficiency of the equipment in the substation, such as the efficiency of the transformer, the number of operations of the circuit breaker, etc. These indexes can reflect the health status and operation quality of substation equipment. The real-time monitoring value of the electromagnetic radiation intensity can grasp the changes in the electromagnetic environment in the target area in real time to ensure that the electromagnetic radiation intensity is always within the safe range.

[0182] In actual operation, load sensors can be installed on the transmission line to real-time collect the current and voltage data of the transmission line, and calculate the load rate fluctuation curve through calculation. Install various monitoring devices in the substation, such as wattmeters, thermometers, etc., to collect the operation parameters of substation equipment and calculate the operation efficiency index. Set multiple electromagnetic monitoring points in the target area and use professional electromagnetic monitoring equipment to real-time collect the electromagnetic radiation intensity data. For example, in the power network of a city, through the monitoring devices distributed on each transmission line and substation, real-time collect the operation monitoring data set to provide data support for subsequent model update and optimization.

[0183] Step S600: Input the operation monitoring data set into the pre-trained feature extraction network to generate real-time multi-dimensional feature vectors, and call the optimization decision model to process the real-time multi-dimensional feature vectors to generate the current optimization parameter prediction set.

[0184] Input the operation monitoring data set into the pre-trained feature extraction network. The pre-trained feature extraction network will perform feature extraction and mapping on the operation monitoring data set. This network has learned the relationship between data such as terrain, electromagnetic, and equipment layout and features through previous training. Therefore, it can extract representative features from the operation monitoring data set to generate real-time multi-dimensional feature vectors. This vector contains feature information such as transmission line load, substation equipment operation status, and electromagnetic radiation.

[0185] Call the optimization decision model to process the real-time multi-dimensional feature vectors. The optimization decision model will generate the current optimization parameter prediction set according to the input real-time multi-dimensional feature vectors, combined with its internal parameters and algorithms. This set contains the predicted values of optimization parameters for transmission line path planning, substation site selection, and electromagnetic interference suppression in the current situation.

[0186] In specific implementation, format conversion and preprocessing are performed on the operation monitoring data set according to the input requirements of the pre-trained feature extraction network, and then it is input into the network. After calculation and processing by the network, real-time multi-dimensional feature vectors are output. This vector is input into the optimization decision model, and the model is processed through modules such as internal parameter adjustment network and verification feedback network to generate the current optimization parameter prediction set. For example, in a power project in an industrial park, the real-time collected operation monitoring data set is input into the pre-trained feature extraction network to generate real-time multi-dimensional feature vectors, and then this vector is input into the optimization decision model to obtain the current optimization parameter prediction set, providing a basis for the real-time optimization of power facilities.

[0187] Step S700: Compare the differences between the current optimization parameter prediction set and the power project optimization parameter set to generate model parameter deviation indicators, where the difference comparison includes path planning parameter offset degree, site selection coordinate distance difference, and interference suppression effect attenuation rate.

[0188] Comparing the current set of predicted optimization parameters with the set of optimization parameters for power engineering is to evaluate the degree of difference between the actual situation and the initial planning scheme during the operation of power facilities. The deviation degree of path planning parameters reflects the deviation degree of the currently predicted transmission line path from the initial planned path, which is calculated by comparing information such as their coordinates and orientations. The distance difference of site selection coordinates refers to the distance between the currently predicted substation site selection coordinates and the initially planned site selection coordinates. The larger this distance difference, the greater the difference between the actual and planned site selections of the substation. The attenuation rate of interference suppression effect reflects the attenuation situation between the current electromagnetic interference suppression effect and the initially planned interference suppression effect, which is calculated by comparing the electromagnetic radiation intensity control levels of the two.

[0189] Generating the model parameter deviation index involves comprehensively considering difference indicators such as the deviation degree of path planning parameters, the distance difference of site selection coordinates, and the attenuation rate of interference suppression effect to form an index that can reflect the degree of model parameter deviation. This index can help determine whether the optimization decision model needs to be updated and adjusted.

[0190] In specific implementation, an algorithm can be written to compare each element of the current set of predicted optimization parameters and the set of optimization parameters for power engineering. For path planning parameters, calculate the deviation values of their coordinates and orientations to obtain the deviation degree of path planning parameters; for site selection coordinates, calculate the Euclidean distance between the two to obtain the distance difference of site selection coordinates; for interference suppression effect, compare the electromagnetic radiation intensity control effects at different time points to calculate the attenuation rate of interference suppression effect. Finally, calculate the model parameter deviation index according to the weights of these indicators. For example, in the power network of a city, by comparing the current set of predicted optimization parameters and the set of optimization parameters for power engineering, it is found that the deviation degree of the transmission line path planning parameters is relatively large, the distance difference of site selection coordinates has also increased, and the attenuation rate of interference suppression effect has reached the preset level. The comprehensively calculated model parameter deviation index exceeds the preset threshold, indicating that the optimization decision model needs to be updated.

[0191] Step S800: If the model parameter deviation index exceeds the dynamic update threshold, then use the operation monitoring data set and the corresponding real-time multi-dimensional feature vectors as incremental training samples to perform local weight correction on the parameter adjustment network in the optimization decision model, generating an updated parameter adjustment network.

[0192] If the model parameter deviation index exceeds the dynamic update threshold, it indicates that there is a large difference between the actual operation situation of power facilities and the initial planning scheme, and the original optimization decision model may no longer be applicable and needs to be updated. Using the operation monitoring data set and the corresponding real-time multi-dimensional feature vectors as incremental training samples is because these data reflect the current actual operation state of power facilities and can provide new information for model update.

[0193] Performing local weight correction on the parameter adjustment network in the optimized decision-making model means adjusting some weights of the parameter adjustment network according to the incremental training samples without changing the overall structure of the model. This can enable the model to better adapt to the actual operation of power facilities and improve the prediction accuracy of the model.

[0194] In the specific implementation, first, it is judged whether the model parameter deviation index exceeds the dynamic update threshold. If it exceeds, the operation monitoring data set and the real-time multi-dimensional feature vector are combined to form an incremental training sample set. Then, the incremental training sample set is used to train the parameter adjustment network, and some weights of the network are updated through optimization algorithms such as gradient descent. For example, in a large-scale power project, when the model parameter deviation index exceeds the dynamic update threshold, the recently collected operation monitoring data set and the corresponding real-time multi-dimensional feature vector are used as incremental training samples to perform local weight correction on the parameter adjustment network. After multiple iterative trainings, an updated parameter adjustment network is generated.

[0195] Step S900: Re-process the multi-dimensional survey feature vector based on the updated parameter adjustment network to generate a corrected set of optimized parameters for the power project, and adjust the transmission line path planning result or substation site selection coordinates in the power facility deployment plan according to the corrected set of parameters.

[0196] Re-processing the multi-dimensional survey feature vector based on the updated parameter adjustment network. The updated parameter adjustment network has undergone local weight correction through incremental training samples and can better adapt to the actual operation of power facilities. Inputting the multi-dimensional survey feature vector into the updated parameter adjustment network, the network will perform a non-linear transformation on it to generate a corrected set of optimized parameters for the power project.

[0197] Adjusting the transmission line path planning result or substation site selection coordinates in the power facility deployment plan according to the corrected set of parameters means applying the corrected parameters to the actual power facility deployment plan. If the path planning parameters change, the path of the transmission line needs to be adjusted; if the site selection coordinate parameters change, the site selection of the substation needs to be re-determined.

[0198] In specific implementation, the multi-dimensional survey feature vector is input into the updated parameter adjustment network. Through the calculation and processing of the network, a corrected set of power engineering optimization parameters is obtained. According to the path planning parameters and site selection coordinate parameters in this set, tools such as Geographic Information System (GIS) are used to adjust the power facility deployment plan. For example, in a power project in a city, the multi-dimensional survey feature vector is reprocessed based on the updated parameter adjustment network to obtain a corrected set of power engineering optimization parameters. It is found that the path planning parameters have changed, so the GIS tool is used to adjust the path of the transmission line to make it more in line with the current actual situation.

[0199] Step S1000: Bind the adjusted deployment plan with the operation monitoring data set through a control logic to generate an update instruction set, which is used to automatically trigger the weight correction process of the optimization decision model according to the real-time monitoring value of the electromagnetic radiation intensity in subsequent real-time monitoring.

[0200] Binding the adjusted deployment plan with the operation monitoring data set through a control logic means associating the deployment plan information such as the adjusted transmission line path planning result, substation site selection coordinates, and electromagnetic interference suppression strategy with the operation monitoring data set collected in real time to establish a dynamic control logic. Through this binding, the deployment plan can be adjusted in a timely manner according to the changes in the operation monitoring data.

[0201] Generating an update instruction set means generating a series of instructions according to the control logic, which are used to automatically trigger the weight correction process of the optimization decision model according to the real-time monitoring value of the electromagnetic radiation intensity in subsequent real-time monitoring. When the real-time monitoring value of the electromagnetic radiation intensity is abnormal, the update instruction set can automatically start the weight correction process of the optimization decision model to update and adjust the model to ensure the safe and stable operation of the power facilities.

[0202] In specific implementation, programming languages can be used to write the control logic code to associate the adjusted deployment plan with the operation monitoring data set. Generate an update instruction set according to the control logic and store it in the system. During subsequent real-time monitoring, the system will monitor the real-time monitoring value of the electromagnetic radiation intensity in real time. When this value exceeds the preset safety threshold, the system will automatically execute the update instruction set to trigger the weight correction process of the optimization decision model. For example, in a power project near a substation, the adjusted deployment plan is bound with the operation monitoring data set through a control logic to generate an update instruction set. When it is detected in real time that the electromagnetic radiation intensity exceeds the safety threshold, the system automatically triggers the weight correction process of the optimization decision model to update the model and adjust the electromagnetic interference suppression strategy to ensure that the electromagnetic radiation intensity returns to the safe range.

[0203] As an implementation manner, after step S400 of generating a power facility deployment plan for the target area according to the power engineering optimization parameter set, the method further includes the step of dynamically adjusting the path during the construction phase: Step S1100: Real-time collect topographic scan data of the construction equipment in the target area. The topographic scan data includes the real-time elevation change value on the traveling path of the construction machinery and the detection result of geological rock layer fissures.

[0204] Real-time collecting the topographic scan data of the construction equipment in the target area is to timely understand the topographic changes in the target area during the construction process. The real-time elevation change value on the traveling path of the construction machinery can reflect the undulation of the terrain. By monitoring these values, it can be found whether unexpected topographic changes occur during the construction process, such as suddenly emerging steep slopes or low-lying areas. The detection result of geological rock layer fissures can help judge the geological stability of the construction area. If geological rock layer fissures are detected, it may be necessary to adjust the construction path to avoid potential geological disasters.

[0205] In actual operation, topographic scan equipment, such as laser scanners, radars, etc., can be installed on the construction equipment to real-time collect the elevation data on the traveling path of the construction machinery. At the same time, geological exploration equipment, such as ground-penetrating radars, drilling and sampling equipment, etc., are used to detect the geological rock layers of the construction area to obtain the detection result of geological rock layer fissures. For example, during the construction of a power project in a mountainous area, the elevation change values on the traveling path are real-time collected by the laser scanner installed on the construction machinery, and the ground-penetrating radar is used to detect the situation of geological rock layer fissures, providing accurate topographic data for the dynamic path adjustment during the construction phase.

[0206] Step S1200: Input the topographic scan data into a pre-trained feature extraction network to generate a real-time topographic feature vector of the construction area, and perform a difference comparison with the topographic parameter sub-vector in the deployment plan to identify the offset area between the construction path and the original topographic features.

[0207] Inputting the topographic scan data into a pre-trained feature extraction network, the pre-trained feature extraction network will perform feature extraction and mapping on the topographic scan data to generate a real-time topographic feature vector of the construction area. This vector contains the current topographic feature information of the construction area, such as terrain undulation degree, geological stability, etc.

[0208] Performing a difference comparison with the topographic parameter sub-vector in the deployment plan is to compare the real-time topographic feature vector of the construction area with the topographic parameter sub-vector used when initially generating the deployment plan to find the difference between the two. Through the difference comparison, the offset area between the construction path and the original topographic features, that is, the area where the terrain changes during the construction process, can be identified.

[0209] In specific implementation, the terrain scan data is subjected to format conversion and preprocessing according to the input requirements of the pre-trained feature extraction network, and then input into the network. After calculation and processing by the network, a real-time terrain feature vector of the construction area is output. The vector comparison algorithm is used to perform element-by-element comparison between the real-time terrain feature vector of the construction area and the terrain parameter sub-vector in the deployment plan, and the areas with large differences are found and marked as the offset areas between the construction path and the original terrain features. For example, in the construction of an urban power project, the real-time collected terrain scan data is input into the pre-trained feature extraction network to generate a real-time terrain feature vector of the construction area. By comparing the differences with the terrain parameter sub-vector in the deployment plan, it is found that the terrain undulation degree of a certain construction section has a large difference from the original plan, and this section is marked as an offset area.

[0210] Step S1300: Based on the position coordinates of the offset area, call the path planning parameters in the optimization decision model to generate a local path correction instruction, where the correction instruction includes a set of alternative path coordinates for bypassing the fissure area and an angle for slope adaptability adjustment.

[0211] Based on the position coordinates of the offset area, calling the path planning parameters in the optimization decision model to generate a local path correction instruction means obtaining the corresponding path planning parameters from the optimization decision model according to the position information of the identified offset area between the construction path and the original terrain features, and generating a local path correction instruction for this offset area.

[0212] The correction instruction includes a set of alternative path coordinates for bypassing the fissure area and an angle for slope adaptability adjustment. The set of alternative path coordinates for bypassing the fissure area refers to the coordinate set of the re-planned transmission line path in order to avoid the geological rock fissures in the offset area. The angle for slope adaptability adjustment refers to the angle value for adjusting the slope of the transmission line according to the terrain slope change in the offset area to ensure the safe and stable operation of the transmission line under the new terrain conditions.

[0213] In specific implementation, according to the position coordinates of the offset area, the corresponding path planning parameters are searched in the optimization decision model. Using the path planning algorithm, combined with the terrain scan data and the geological rock fissure detection results, a set of alternative path coordinates for bypassing the fissure area is generated. At the same time, according to the terrain slope change in the offset area, the angle for slope adaptability adjustment is calculated. The set of alternative path coordinates and the angle for slope adaptability adjustment are combined into a local path correction instruction. For example, in the construction of a mountain power project, when it is identified that there are geological rock fissures in a certain area, according to the position coordinates of this area, the path planning parameters in the optimization decision model are called to generate a set of alternative path coordinates for bypassing this fissure area, and the angle for slope adaptability adjustment is calculated to form a local path correction instruction.

[0214] Step S1400: Conduct a topological connection verification on the alternative path coordinate set and the transmission line path planning result in the deployment plan to ensure the curvature continuity and the consistency of the electromagnetic radiation superposition value between the corrected path endpoints and the unadjusted path segments.

[0215] Conducting a topological connection verification on the alternative path coordinate set and the transmission line path planning result in the deployment plan is to ensure the coherence and consistency between the corrected transmission line path and the original planned path. The topological connection verification mainly focuses on the curvature continuity and the consistency of the electromagnetic radiation superposition value between the corrected path endpoints and the unadjusted path segments.

[0216] Curvature continuity means that the curvature change at the connection point between the corrected path endpoints and the unadjusted path segments should be smooth, avoiding sudden turns or bends to ensure the stability and safety of the transmission line. The consistency of the electromagnetic radiation superposition value means that after the corrected path and the unadjusted path segments are connected, their electromagnetic radiation superposition value should meet the requirements of the original plan to ensure the stability of the electromagnetic environment.

[0217] In specific implementation, use geographic information system (GIS) and electromagnetic simulation software to conduct a topological connection verification on the alternative path coordinate set and the transmission line path planning result in the deployment plan. Analyze the spatial relationship between the corrected path endpoints and the unadjusted path segments through GIS, calculate the curvature change situation to ensure curvature continuity. Use electromagnetic simulation software to simulate the electromagnetic radiation superposition situation after the corrected path and the unadjusted path segments are connected to verify the consistency of the electromagnetic radiation superposition value. For example, in the construction of a power project in a city, conduct a topological connection verification on the generated alternative path coordinate set and the original transmission line path planning result. Through GIS analysis, it is found that the curvature change at the path connection point is smooth, and using electromagnetic simulation software to simulate and verify that the electromagnetic radiation superposition value meets the requirements, indicating that the corrected path meets the topological connection requirements.

[0218] Step S1500: Update the construction coordinate mapping rule in the power facility deployment plan according to the verification result, and synchronize the updated rule to the navigation control system of the construction equipment to drive the construction machinery to perform high-precision laying operations according to the corrected path.

[0219] Updating the construction coordinate mapping rule in the power facility deployment plan according to the verification result means that if the topological connection verification passes, update the corrected path coordinate information to the construction coordinate mapping rule in the power facility deployment plan. The construction coordinate mapping rule is used to convert the planned path of the transmission line into coordinate information in actual construction to guide the construction machinery for construction.

[0220] Synchronizing the updated rules to the navigation control system of the construction equipment means transmitting the updated construction coordinate mapping rules to the navigation control system of the construction equipment, enabling the construction equipment to obtain the latest construction path information. Driving the construction machinery to perform high-precision laying operations according to the corrected path means that the construction equipment automatically adjusts its travel route according to the updated rules in the navigation control system and lays the transmission line according to the corrected path to ensure the high precision and accuracy of the construction.

[0221] In specific implementation, update the verified set of alternative path coordinates into the construction coordinate mapping rules to generate an updated rule file. Use wireless communication technology to transmit the updated rule file to the navigation control system of the construction equipment. After receiving the updated rules, the navigation control system of the construction equipment automatically adjusts the navigation route and drives the construction machinery to lay the transmission line according to the corrected path. For example, in the construction of a large-scale power project, after the topological connection verification, update the corrected path coordinate information into the construction coordinate mapping rules, and synchronize the updated rules to the navigation control system of the construction equipment through wireless communication. The construction machinery accurately lays the transmission line according to the corrected path, improving the construction efficiency and quality.

[0222] As an implementation manner, the method provided by the embodiment of the present invention further includes the step of optimizing the electromagnetic joint control during the operation phase, specifically including: Step S1600: After the power facilities in the target area are put into operation, continuously collect the electromagnetic radiation monitoring data around the substation and the load fluctuation data of the transmission line to generate a real-time monitoring data set during the operation phase.

[0223] After the power facilities in the target area are put into operation, continuously collecting the electromagnetic radiation monitoring data around the substation and the load fluctuation data of the transmission line is to grasp the electromagnetic environment and load conditions during the operation of the power facilities in real time. The electromagnetic radiation monitoring data around the substation can reflect the change of the electromagnetic radiation intensity generated during the operation of the substation. By monitoring these data, abnormal electromagnetic radiation conditions can be detected in time and corresponding measures can be taken for control. The load fluctuation data of the transmission line reflects the load change of the transmission line at different time periods. Understanding the load fluctuation law helps to reasonably allocate power resources and improve the operation efficiency of the transmission line.

[0224] Generating a real-time monitoring data set during the operation phase means integrating the continuously collected electromagnetic radiation monitoring data around the substation and the load fluctuation data of the transmission line to form a data set containing various operation information. This data set can provide data support for the subsequent electromagnetic joint control optimization.

[0225] In actual operation, multiple electromagnetic monitoring points can be set around the substation, and professional electromagnetic monitoring equipment is used to collect electromagnetic radiation intensity data in real time. Load sensors are installed on the transmission line to collect current and voltage data of the transmission line in real time, and load fluctuation data is obtained through calculation. The collected electromagnetic radiation monitoring data and transmission line load fluctuation data are sorted and stored to generate a real-time monitoring data set for the operation stage. For example, in the power grid of a city, through the electromagnetic monitoring points around each substation and the load sensors on the transmission line, data is continuously collected to generate a real-time monitoring data set for the operation stage, providing comprehensive and accurate data for the electromagnetic joint control optimization in the operation stage of power facilities.

[0226] Step S1700: Input the real-time monitoring data set into the pre-trained feature extraction network to extract the electromagnetic intensity sub-vector and load distribution sub-vector in the operation stage, and perform dynamic matching analysis with the electromagnetic interference suppression strategy in the deployment plan.

[0227] Input the real-time monitoring data set into the pre-trained feature extraction network, and the pre-trained feature extraction network will perform feature extraction and mapping on the real-time monitoring data set. This network can extract the electromagnetic intensity sub-vector and load distribution sub-vector in the operation stage from the data set. The electromagnetic intensity sub-vector in the operation stage contains the characteristic information of the electromagnetic radiation intensity around the substation, and the load distribution sub-vector reflects the load distribution of the transmission line.

[0228] Performing dynamic matching analysis with the electromagnetic interference suppression strategy in the deployment plan is to compare and analyze the extracted electromagnetic intensity sub-vector and load distribution sub-vector in the operation stage with the initially formulated electromagnetic interference suppression strategy. Through dynamic matching analysis, it can be judged whether the current electromagnetic environment and load conditions meet the requirements of the electromagnetic interference suppression strategy and whether the strategy needs to be adjusted.

[0229] In specific implementation, the real-time monitoring data set is subjected to format conversion and preprocessing according to the input requirements of the pre-trained feature extraction network, and then input into the network. After calculation and processing by the network, the electromagnetic intensity sub-vector and load distribution sub-vector in the operation stage are output. A matching algorithm is used to compare and analyze these two sub-vectors with the electromagnetic interference suppression strategy in the deployment plan to find differences and potential problems. For example, in the power project of an industrial park, the real-time monitoring data set is input into the pre-trained feature extraction network to extract the electromagnetic intensity sub-vector and load distribution sub-vector in the operation stage. Through dynamic matching analysis with the electromagnetic interference suppression strategy in the deployment plan, it is found that the electromagnetic radiation intensity exceeds the range specified by the strategy during a certain period, and the strategy needs to be adjusted.

[0230] Step S1800: Identify the frequency band peak distribution area in the electromagnetic intensity sub-vector that exceeds the preset safety range, and generate an incremental shielding device deployment instruction based on the interference suppression parameters in the optimization decision model. The instruction includes the installation coordinates of the newly added shielding device and the working frequency band adjustment parameters.

[0231] Identifying the frequency band peak distribution area in the electromagnetic intensity sub-vector that exceeds the preset safety range is achieved by analyzing the electromagnetic intensity sub-vector during the operation stage to find the frequency bands and corresponding distribution areas where the electromagnetic radiation intensity exceeds the preset safety range. The preset safety range is an electromagnetic radiation intensity range determined according to electromagnetic radiation safety standards and the operation requirements of power facilities. Exceeding this range may cause harm to personnel and equipment.

[0232] Generating an incremental shielding device deployment instruction based on the interference suppression parameters in the optimization decision model is to generate an instruction for deploying incremental shielding devices according to the interference suppression parameters in the optimization decision model, combined with the information on the frequency band peak distribution area that exceeds the preset safety range. The instruction includes the installation coordinates of the newly added shielding device and the working frequency band adjustment parameters. The installation coordinates of the newly added shielding device refer to the specific location where the shielding device needs to be installed to suppress the electromagnetic radiation that exceeds the safety range. The working frequency band adjustment parameters refer to adjusting the working frequency band of the shielding device according to the frequency band that exceeds the safety range to make it more effective in suppressing electromagnetic radiation.

[0233] In specific implementation, a data analysis algorithm is used to process the electromagnetic intensity sub-vector to identify the frequency band peak distribution area that exceeds the preset safety range. According to the interference suppression parameters in the optimization decision model and the location information of this area, the installation coordinates of the newly added shielding device and the working frequency band adjustment parameters are calculated to generate an incremental shielding device deployment instruction. For example, in a power project near a substation, through the analysis of the electromagnetic intensity sub-vector during the operation stage, it is found that the electromagnetic radiation intensity in a certain frequency band exceeds the preset safety range and is mainly distributed in a certain area of the substation. Based on the interference suppression parameters in the optimization decision model, an incremental shielding device deployment instruction is generated to determine the installation of a newly added shielding device in this area and adjust its working frequency band to effectively suppress electromagnetic radiation.

[0234] Step S1900: Adjust the weight of the substation coverage radius in the deployment plan according to the line load rate fluctuation characteristics in the load distribution sub-vector, and recalculate the topological connection priority of the surrounding transmission lines.

[0235] Adjusting the weight of the substation coverage radius in the deployment plan according to the line load rate fluctuation characteristics in the load distribution sub-vector means judging whether the coverage range of the substation needs to be adjusted by analyzing the fluctuation of the transmission line load rate in the load distribution sub-vector. If the load rates of some transmission lines are continuously high, it may be necessary to increase the weight of the coverage radius of the corresponding substation to ensure that the substation can better supply power to these lines.

[0236] Recalculating the topological connection priority of the surrounding transmission lines means, after adjusting the weight of the substation coverage radius, determining the connection priority of the surrounding transmission lines again according to the new weight and the topological structure of the transmission lines. Transmission lines with a higher connection priority have higher priority in aspects such as power distribution and fault handling.

[0237] In specific implementation, statistical analysis methods are used to analyze the line load rate fluctuation characteristics in the load distribution sub-vector, and the weight of the substation coverage radius in the deployment plan is adjusted according to the analysis results. Then, graph theory algorithms and power system analysis methods are used to recalculate the topological connection priority of the surrounding transmission lines in combination with the adjusted weight and the topological structure of the transmission lines. For example, in the power network of a city, it is found through analyzing the load distribution sub-vector that the load rate of the transmission lines in a certain area continues to increase, so the weight of the coverage radius of the corresponding substation in this area is increased. After recalculating the topological connection priority of the surrounding transmission lines, it is determined that the connection priority of some lines needs to be improved to ensure the reasonable distribution of power and the stable operation of the system.

[0238] Step S2000: Integrate the incremental shielding device deployment instruction and the updated topological connection priority into an operation optimization parameter set, synchronize it to the substation control center and the line dispatching system, and adjust the working mode of the shielding device and the line load distribution strategy in real time to ensure that the electromagnetic radiation intensity and the load efficiency continuously meet the constraint conditions of the power engineering optimization parameter set.

[0239] Integrating the incremental shielding device deployment instruction and the updated topological connection priority into an operation optimization parameter set means integrating the instruction for deploying the incremental shielding device and the recalculated topological connection priority of the transmission lines to form a parameter set containing various optimization information. This parameter set combines the optimization strategies for electromagnetic interference suppression and line load distribution.

[0240] Synchronizing to the substation control center and the line dispatching system means transmitting the set of operation optimization parameters to the substation control center and the line dispatching system, enabling these systems to obtain the latest optimization information. Real-time adjustment of the working mode of the shielding device and the line load distribution strategy means that the substation control center adjusts the working mode of the shielding device in real time according to the incremental shielding device deployment instruction in the set of operation optimization parameters, such as turning on or off the shielding device, adjusting the power of the shielding device, etc.; the line dispatching system adjusts the line load distribution strategy in real time according to the updated topological connection priority to reasonably allocate power resources.

[0241] By real-time adjusting the working mode of the shielding device and the line load distribution strategy, it is ensured that the electromagnetic radiation intensity and the load efficiency continuously meet the constraint conditions of the power engineering optimization parameter set. This means that during the operation of power facilities, the electromagnetic radiation intensity is always controlled within a safe range, and at the same time, the load distribution of the transmission line is reasonable, which can improve the operation efficiency and stability of the power system. For example, in a large-scale power project, the incremental shielding device deployment instruction and the updated topological connection priority are integrated into the set of operation optimization parameters and synchronized to the substation control center and the line dispatching system. The substation control center adjusts the working mode of the shielding device according to the instruction, effectively suppressing the electromagnetic radiation; the line dispatching system adjusts the line load distribution strategy according to the updated priority, improving the load efficiency of the power system and ensuring that the electromagnetic radiation intensity and the load efficiency continuously meet the constraint conditions of the power engineering optimization parameter set.

[0242] Based on the foregoing embodiments, an embodiment of the present invention provides a power engineering survey data processing device. Each unit included in the device and each module included in each unit can be implemented by a processor in a computer device; of course, it can also be implemented by specific logic circuits; during the implementation process, the processor can be a central processing unit (CPU), a microprocessor unit (MPU), a digital signal processor (DSP), or a field programmable gate array (FPGA), etc.

[0243] Figure 2 As shown in the schematic diagram of the composition structure of a power engineering survey data processing device provided by an embodiment of the present invention, Figure 2 as shown, the power engineering survey data processing device 200 includes: A data acquisition module 210, configured to acquire a multi-source survey data set of a target area. The multi-source survey data set includes geological terrain data, electromagnetic interference data, and equipment layout data. Among them, the electromagnetic interference data is used to describe the electromagnetic radiation intensity distribution of power facilities in the target area, and the equipment layout data is used to indicate the transmission line topology and substation locations in the target area; A feature mapping module 220, configured to perform joint feature mapping on the multi-source survey data set through a pre-trained feature extraction network to generate a multi-dimensional survey feature vector. Among them, the multi-dimensional survey feature vector includes a terrain parameter sub-vector, an electromagnetic intensity sub-vector, and a layout topology sub-vector; A parameter generation module 230, configured to input the multi-dimensional survey feature vector into an optimization decision model for iterative parameter adjustment to generate a power engineering optimization parameter set of the target area. Among them, the optimization decision model is trained based on the mapping relationship between historical survey data and verified optimization results; A scheme generation module 240, configured to generate a power facility deployment scheme for the target area according to the power engineering optimization parameter set. The deployment scheme includes a transmission line path planning result, substation location coordinates, and an electromagnetic interference suppression strategy.

[0244] The description of the above device embodiments is similar to the description of the above method embodiments and has similar beneficial effects to the method embodiments. In some embodiments, the functions or modules included in the device provided by the embodiments of the present invention can be used to execute the methods described in the above method embodiments. For the technical details not disclosed in the device embodiments of the present invention, please refer to the description of the method embodiments of the present invention for understanding.

Claims

1. A method for processing electric power engineering survey data, characterized in that: The method comprises: Acquire a multi-source survey data set of a target area, the multi-source survey data set comprising geological and topographic data, electromagnetic interference data, and equipment layout data, wherein the electromagnetic interference data is used to describe the electromagnetic radiation intensity distribution of the power facilities in the target area, and the equipment layout data is used to indicate the topological structure of the transmission lines and the location of the substation in the target area; Performing joint feature mapping on the multi-source survey data set through a pre-trained feature extraction network to generate a multi-dimensional survey feature vector, wherein the multi-dimensional survey feature vector includes a terrain parameter sub-vector, an electromagnetic intensity sub-vector, and a layout topology sub-vector; Inputting the multi-dimensional survey feature vector into an optimization decision model for iterative parameter adjustment to generate a set of power engineering optimization parameters for the target area, wherein the optimization decision model is trained based on a mapping relationship between historical survey data and verified optimization results; A power facility deployment plan for the target area is generated according to the power engineering optimization parameter set, and the deployment plan includes a transmission line path planning result, a substation site selection coordinate, and an electromagnetic interference suppression strategy.

2. The method according to claim 1, characterized in that The method of performing joint feature mapping on the multi-source survey data set through a pre-trained feature extraction network to generate a multi-dimensional survey feature vector includes: Performing elevation gradient analysis on the geological terrain data, extracting terrain relief characteristics and geological stability characteristics, and fusing the terrain relief characteristics and geological stability characteristics into the terrain parameter sub-vector based on a preset terrain coding structure; Performing spectrum decomposition on the electromagnetic interference data to obtain electromagnetic radiation intensity distribution diagrams of different frequency bands, and extracting peak intensity features and spatial attenuation features of each frequency band by sliding the convolution kernel, and superimposing the peak intensity features and spatial attenuation features according to the frequency band to generate the electromagnetic intensity sub-vector; Performing topological structure analysis on the equipment layout data, identifying connection nodes of the transmission line and coverage radius of the substation, and generating the layout topology subvector based on node connection density and coverage radius weight; The terrain parameter sub-vector, the electromagnetic intensity sub-vector and the layout topology sub-vector are spliced ​​according to a preset dimensional alignment rule to obtain the multi-dimensional survey feature vector.

3. The method according to claim 2, characterized in that The optimization decision model includes a parameter adjustment network and a verification feedback network. The multi-dimensional survey feature vector is input into the optimization decision model for iterative parameter adjustment to generate a set of power engineering optimization parameters for the target area, including: Performing nonlinear transformation on the multi-dimensional survey feature vector through the parameter adjustment network to generate an initial optimization parameter set, wherein the initial optimization parameter set includes path planning parameters, site selection coordinate parameters and interference suppression parameters; Calling the verification feedback network to perform simulated deployment verification on the initial optimization parameter set to generate a parameter adjustment feedback signal, wherein the simulated deployment verification includes calculation of the matching degree between the transmission path and the terrain, analysis of the overlap rate of the substation coverage, and prediction of the electromagnetic interference suppression effect; updating the weight coefficient of the parameter adjustment network according to the parameter adjustment feedback signal, and repeatedly performing the nonlinear transformation and the simulated deployment verification until the initial optimization parameter set meets the preset deployment constraint condition; An initial optimization parameter set satisfying the deployment constraint condition is determined as the power engineering optimization parameter set.

4. The method according to claim 3, characterized in that The parameter adjustment network includes a fusion structure of a multi-layer perceptron and an attention mechanism. The parameter adjustment network is used to perform a nonlinear transformation on the multi-dimensional survey feature vector to generate an initial optimization parameter set, including: Inputting the terrain parameter subvector into the first hidden layer of the multilayer perceptron for weight assignment to obtain a terrain influence weight coefficient; Performing cross-attention calculation on the electromagnetic intensity sub-vector and the layout topology sub-vector to generate an electromagnetic-layout association feature; Performing weighted aggregation on the electromagnetic-layout association features based on the terrain influence weight coefficient to generate a comprehensive decision feature; Linearly mapping the comprehensive decision features through the output layer of the multilayer perceptron to generate the initial optimization parameter set; The calling of the verification feedback network to perform simulated deployment verification on the initial optimization parameter set to generate a parameter adjustment feedback signal includes: generating a virtual power transmission path according to the path planning parameters, and calculating a matching score between the virtual power transmission path and the geological stability characteristics in the terrain parameter subvector; Generate a virtual substation coverage area according to the site selection coordinate parameters, and analyze the overlap rate between the virtual substation coverage area and the coverage range of existing power facilities; Predicting an electromagnetic radiation intensity attenuation curve of the target area based on the interference suppression parameter, and calculating a deviation value between the attenuation curve and a preset electromagnetic safety threshold; The parameter adjustment feedback signal is generated according to the matching score, the overlap rate and the deviation value, wherein the parameter adjustment feedback signal includes a path adjustment direction, a coordinate offset and an interference suppression correction coefficient.

5. The method according to claim 1, characterized in that The pre-trained feature extraction network is trained by the following steps: Acquire a historical survey data set, wherein the historical survey data set includes topographic data, electromagnetic data, and equipment layout data of multiple historical areas, and a verified optimization parameter set corresponding to each historical area; Performing data enhancement processing on the historical survey data set to generate an expanded training sample set, wherein the data enhancement processing includes random rotation of terrain data, injection of electromagnetic data noise, and perturbation of layout data nodes; Constructing an initial feature extraction network, and inputting the expanded training sample set into the initial feature extraction network to perform feature extraction, and generating a prediction optimization parameter set; Calculating the mean square error loss between the predicted optimization parameter set and the verified optimization parameter set, and updating the parameters of the initial feature extraction network based on a gradient descent algorithm until the mean square error loss converges to a preset threshold; The initial feature extraction network after parameter update is determined as the pre-trained feature extraction network.

6. The method according to claim 5, characterized in that The performing data enhancement processing on the historical survey data set to generate an expanded training sample set includes: Performing a random rotation transformation on the terrain data to generate a rotated terrain elevation map, and adding randomly generated geological fracture simulation features to the rotated terrain elevation map; Injecting Gaussian white noise into the electromagnetic data to generate an electromagnetic spectrum with noise, and smoothing the electromagnetic spectrum with noise by frequency domain filtering; Randomly disconnecting and reconnecting the transmission line nodes in the equipment layout data to generate an equipment layout diagram after the topology structure is mutated; The rotated terrain elevation map, the noisy electromagnetic spectrum map and the equipment layout map after topological structure variation are combined into new training samples and added into the expanded training sample set.

7. The method according to claim 1, characterized in that The generating the electric power facility deployment plan of the target area according to the electric power engineering optimization parameter set includes: Extracting an initial transmission line coordinate point set based on the path planning parameters in the power engineering optimization parameter set, and identifying coordinate points in the coordinate point set that overlap with steep slope areas or geological fault zones in combination with the elevation gradient distribution characteristics in the terrain parameter subvector; Perform electromagnetic interference intensity verification on the overlapping coordinate points, call the peak intensity distribution data of the corresponding area in the electromagnetic intensity sub-vector, and select the coordinate points whose peak intensity exceeds the safety threshold as high-risk path nodes; Performing path detour optimization on the high-risk path nodes, generating an avoidance path correction trajectory according to the geological stability characteristics in the terrain parameter subvector, and smoothly connecting the endpoints of the correction trajectory with the unadjusted initial coordinate point set to form the transmission line path planning result; Generate a candidate substation location set based on the site selection coordinate parameters in the power engineering optimization parameter set, and calculate the coverage efficiency of each candidate location for the surrounding lines and the electromagnetic radiation suppression demand score by combining the coverage radius weight in the layout topology subvector and the spatial attenuation characteristics of the electromagnetic intensity subvector; The substation site selection coordinates are selected according to the comprehensive weight value of the coverage efficiency and the suppression demand score, and the shielding equipment installation density and orientation parameters matching the site selection coordinates are generated based on the interference suppression parameters in the power engineering optimization parameter set to form the electromagnetic interference suppression strategy; The transmission line path planning results, substation site selection coordinates and electromagnetic interference suppression strategies are spatially topologically associated to generate a power facility deployment plan that includes construction coordinate mapping rules and electromagnetic protection joint control logic, wherein the joint control logic is used to dynamically adjust the shielding equipment working parameters according to real-time electromagnetic data during the substation operation phase.

8. The method according to claim 7, characterized in that The performing of path detour optimization on the high-risk path node and generating an avoidance path correction trajectory according to the geological stability characteristics in the terrain parameter subvector includes: Extracting the terrain undulation characteristics of the area where the high-risk path node is located, generating a contour distribution map centered on the node, and determining the boundary of the detourable area according to the direction of change of the contour curvature; Calling the peak intensity distribution data of the corresponding area in the electromagnetic intensity sub-vector, excluding the sub-area whose electromagnetic radiation intensity exceeds a preset threshold in the detourable area, and generating an electromagnetic safety detour range; Based on the superposition result of the terrain undulation characteristics and the electromagnetic safety detour range, a detour channel model that allows path correction is constructed, wherein the model includes the maximum allowable slope and electromagnetic radiation safety range in the channel; Generate multiple candidate path trajectories according to the detour channel model, and calculate the deviation angle of each trajectory from the original path, the cumulative slope change and the mean electromagnetic radiation of the passing area; Select the alternative trajectory with a deviation angle less than the preset tolerance, the smallest cumulative slope change and the lowest electromagnetic radiation mean as the avoidance path correction trajectory, and connect the endpoint of the correction trajectory with the unadjusted initial coordinate point set in a curvature continuous manner; The terrain constraint parameters and electromagnetic safety parameters involved in the avoidance path correction trajectory are encapsulated as path optimization rules and embedded in the dynamic adjustment logic of the transmission line path planning result to ensure that when unsurveyed obstacles are encountered during the construction phase, the rules are reused based on real-time terrain data to generate local correction instructions.

9. The method according to claim 1, characterized in that: The method further comprises the step of updating the optimization decision model, including: Real-time collection of operation monitoring data sets after the deployment of the power facilities in the target area, the operation monitoring data sets including the load rate fluctuation curve of the transmission line, the operation efficiency index of the substation equipment and the real-time monitoring value of the electromagnetic radiation intensity; Inputting the operation monitoring data set into the pre-trained feature extraction network to generate a real-time multi-dimensional feature vector, and calling the optimization decision model to process the real-time multi-dimensional feature vector to generate a current optimization parameter prediction set; Compare the difference between the current optimization parameter prediction set and the power engineering optimization parameter set to generate a model parameter deviation index, wherein the difference comparison includes path planning parameter deviation, site selection coordinate distance difference and interference suppression effect attenuation rate; If the model parameter deviation index exceeds the dynamic update threshold, the operation monitoring data set and the corresponding real-time multi-dimensional feature vector are used as incremental training samples to perform local weight correction on the parameter adjustment network in the optimization decision model to generate an updated parameter adjustment network; Reprocessing the multidimensional survey feature vector based on the updated parameter adjustment network to generate a revised power engineering optimization parameter set, and adjusting the transmission line path planning result or substation site selection coordinates in the power facility deployment plan according to the revised parameter set; The adjusted deployment plan is logically bound to the operation monitoring data set to generate an update instruction set, which is used to automatically trigger the weight correction process of the optimization decision model according to the real-time monitoring value of the electromagnetic radiation intensity in subsequent real-time monitoring.

10. A power engineering survey data processing device, characterized in that: The device comprises: A data acquisition module, used to acquire a multi-source survey data set of a target area, wherein the multi-source survey data set includes geological and topographic data, electromagnetic interference data, and equipment layout data, wherein the electromagnetic interference data is used to describe the electromagnetic radiation intensity distribution of the power facilities in the target area, and the equipment layout data is used to indicate the topological structure of the transmission lines and the location of the substation in the target area; A feature mapping module, configured to perform joint feature mapping on the multi-source survey data set through a pre-trained feature extraction network to generate a multi-dimensional survey feature vector, wherein the multi-dimensional survey feature vector includes a terrain parameter sub-vector, an electromagnetic intensity sub-vector, and a layout topology sub-vector; A parameter generation module, used for inputting the multi-dimensional survey feature vector into an optimization decision model for iterative parameter adjustment to generate a set of power engineering optimization parameters for the target area, wherein the optimization decision model is trained based on a mapping relationship between historical survey data and verified optimization results; A scheme generation module is used to generate a power facility deployment scheme for the target area based on the power engineering optimization parameter set, wherein the deployment scheme includes transmission line path planning results, substation site selection coordinates and electromagnetic interference suppression strategy.

Images