River slope ecological restoration method
Through the combination of deep learning model and evaluation model, the optimal river slope ecological restoration solution is generated, which solves the problem of fixed or single river slope ecological restoration solution in the existing technology, and achieves multi-index optimization and improvement of ecological restoration effect.
Patent Information
- Application Number
- CN202410919119.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-10
- Publication Date
- 2024-11-08
- Estimated Expiration
- 2044-07-10
AI Technical Summary
The existing ecological restoration methods of river slopes are usually copied by fixed or single solutions, and failing to scientifically take into account multiple aspects of river slopes, resulting in poor ecological restoration results.
By obtaining the river slope ecological restoration scheme for encoding, building deep learning models and evaluation models, using the encoder-decoder structure generation network and convolutional neural network, training and optimization, iteratively generate the optimal ecological restoration scheme, combined with Expert scoring method ensures multi-index optimization of the plan.
Effectively find the optimal river slope ecological restoration plan with non-fixed solutions, improves the effectiveness and practicality of ecological restoration, takes into account a variety of indicators, and guides actual work.
Smart Images

Figure CN118917712A_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of slope restoration, and in particular relates to a river slope ecological restoration method. Background Art
[0002] The river slope ecological restoration method is a comprehensive technical means, which aims to restore and improve the ecological environment of the river slope and improve the stability and anti-erosion ability of the slope. However, in the scientific river slope ecological restoration method, relevant professionals are required to carry out statistical calculations and relevant scheme designs for the river slope according to local conditions. However, in the existing river slope ecological restoration methods, relevant personnel generally use fixed river slope ecological restoration schemes to copy or directly adopt corresponding single schemes to reinforce the river slopes, and do not carry out scientific river slope ecological restoration. At the same time, the construction of ecological restoration schemes cannot take into account multiple aspects of river slopes. Summary of the invention
[0003] In order to solve the above technical problems, the present invention proposes a river slope ecological restoration method to solve the problems existing in the above prior art.
[0004] To achieve the above object, the present invention provides a river slope ecological restoration method, comprising:
[0005] Obtaining a river slope ecological restoration plan, encoding the river slope ecological restoration plan, and obtaining a slope matrix;
[0006] Constructing a deep learning model, and training the deep learning model through a slope matrix to obtain a trained decoder, wherein the deep learning model adopts a generative network of an encoder-decoder structure;
[0007] An evaluation model is constructed, and the river slope ecological restoration plan is evaluated based on the current river slope through the expert scoring method to obtain the evaluation score. The evaluation model is optimized according to the slope matrix and the evaluation score to obtain the optimized evaluation model.
[0008] The trained decoder is used to iteratively generate an initial slope matrix, and the iteratively generated initial slope matrix is evaluated by the optimized evaluation model, and the optimal slope matrix is obtained according to the evaluation result;
[0009] The optimal slope matrix is decoded to obtain the optimal ecological restoration plan for the current river slope. According to the optimal river slope ecological restoration plan, the current river slope is set up to achieve ecological restoration.
[0010] Optionally, the process of encoding the river slope ecological restoration plan includes:
[0011] Obtain the materials and locations used in the river slope ecological restoration plan;
[0012] Construct the numbering relationship between the initial coding sequence and the material and the label, wherein the initial coding sequence is set with corresponding coding positions according to the types of different materials. Fill in the initial coding sequence according to the type of material and the corresponding label, generate the coding result, and construct the initial slope matrix. Fill the coding result into the initial slope matrix according to the setting position of the material in the river slope scheme to obtain the slope matrix.
[0013] Optionally, the deep learning model adopts a variational autoencoder, wherein the variational autoencoder adopts an encoder-decoder structure, the slope matrix serves as the input of the encoder, the output of the decoder is the slope matrix, and there is a latent space between the encoder and the decoder, wherein the latent space contains latent variables, and the input space, latent space, and output space all belong to integer space.
[0014] Optionally, the structure of the encoder includes: four sequentially connected convolution modules, a Flatten layer, a Linear layer, an LReLU layer and a Linear layer, wherein the convolution module includes a sequentially connected Conv layer, a BN layer and an LReLU layer,
[0015] Optionally, the decoding structure includes an Unflatten layer, a deconvolution Deconv layer, a BN layer, an LReLU layer, two upsampling modules connected in sequence, a convolution module, a convolution layer and an output layer, wherein the upsampling module includes an Upsample layer, a Conv layer, a BN layer, and an LReLU layer connected in sequence.
[0016] Optionally, the evaluation model adopts a convolutional neural network.
[0017] Optionally, the structure of the convolutional neural network includes: an input layer, three convolutional layers, two fully connected layers, and an output layer.
[0018] Optionally, the process of evaluating river slope ecological restoration options includes:
[0019]
[0020] Among them, J p represents the final score of the expert scoring for the pth repair solution, J pq represents the scoring result of the qth expert for the pth river slope ecological restoration plan, p represents the number of different indicators, q represents the number of different experts, n represents the total number of experts, and J o,pq represents the score corresponding to the oth indicator among the qth expert in the pth river slope ecological restoration plan, w oIt represents the weight of the indicator in the oth position, which is set according to manual experience. o represents the indicator number, and m represents the number of indicators. The indicators include ecological restoration capacity, anti-solidification capacity, appearance score, cost score, construction difficulty, and scheme risk. The ecological restoration capacity includes the restoration capacity of soil and water loss, vegetation destruction, soil erosion, etc. in the current river slope ecology.
[0021] Optionally, the process of obtaining the optimal slope matrix includes:
[0022] A population is constructed and initialized according to the input of a trained decoder, wherein the population includes a plurality of individuals, wherein the individuals include different hidden parameters; an objective function, a fitness and a population update method are set, wherein the objective function is to maximize the evaluation score of a slope matrix generated by the trained decoder; and an optimization algorithm is used to search for the best result according to the objective function, the fitness and the population update method to generate an optimal slope matrix.
[0023] Optionally, the optimization algorithm adopts an improved moth-to-flame algorithm.
[0024] Compared with the prior art, the present invention has the following advantages and technical effects:
[0025] Through the above technical scheme, the present invention statistics the existing river slope ecological restoration schemes, and trains the deep learning model according to the statistical results, so that the deep learning model learns and can regenerate new river slope ecological restoration schemes. On this basis, by scoring the new river slope ecological restoration schemes and iterative optimization based on this, the optimal river slope ecological restoration scheme that is not a fixed scheme can be effectively found. Through the above technical scheme, the present invention can find or predict the optimal river slope ecological restoration scheme to improve the effectiveness of the river slope ecological restoration scheme. At the same time, the scoring process takes into account multiple indicators to ensure that the scheme under each indicator is optimal. Based on this, it can effectively guide the actual work of river slope ecological restoration, and has strong practicality. BRIEF DESCRIPTION OF THE DRAWINGS
[0026] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:
[0027] Figure 1 The figure is a schematic diagram of a method flow of an embodiment of the present invention. DETAILED DESCRIPTION
[0028] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.
[0029] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.
[0030] like Figure 1 As shown, this embodiment provides a river slope ecological restoration method, including:
[0031] Obtain a river slope ecological restoration plan, encode the river slope ecological restoration plan, encode the river slope ecological restoration plan into a matrix form to obtain a slope matrix, continuously transform the slope matrix through a deep learning model, and score and evaluate the transformed plan matrix, and obtain the optimal plan matrix according to the evaluation results, decode the optimal plan matrix to obtain the optimal river slope ecological restoration plan, and set the river slope according to the optimal river slope ecological restoration plan to restore the river slope ecology.
[0032] As some embodiments, the river slope ecological restoration plan selects as many existing river slope ecological restoration plans under different environments as possible, wherein the river slope ecological restoration plan is a restoration system setting plan including functional layers of different materials and material setting locations, for example, a setting plan for different river slope ecological restoration systems including clay layers, gravel layers, soil layers, organic matter layers and vegetation planting layers. In the above content, different functional layers are encoded, and the main material types used in the functional layers are first determined, wherein the material types include three types: vegetation materials, natural materials and prefabricated materials. Vegetation materials include vegetation of different types and varieties. Natural materials include the above-mentioned clay, gravel, soil, fertilizers, organic matter and the like of different textures. Prefabricated materials include different types of concrete blocks, cells and other industrial products used for slope ecological restoration or slope fixation. During the coding process, an initial coding sequence is constructed, where the coding sequence is based on six digits, and each two digits correspond to the types of vegetation materials, natural materials, and prefabricated materials. The materials in different material types (the materials with the largest proportion in different types) are numbered, such as the turf in vegetation materials is numbered 01, a certain soil in natural materials is numbered 20, and a certain concrete block in prefabricated materials is numbered 09, and the final number is 012009. The slope is divided into different grids according to the region, and the main materials in the grid are counted. The initial matrix is constructed according to the position of the grid, where the number of rows and columns of the initial matrix corresponds to the number of rows and columns of the grid, and the elements of the slope matrix are assigned according to the final code corresponding to the main material in the grid to generate the slope matrix. Each slope matrix corresponds to a river slope ecological restoration plan.
[0033] As some embodiments, existing ecological restoration plans for river slopes are collected from constructed ecological restoration plans for river slopes or from relevant databases or institutional data. Different ecological restoration plans are encoded according to the above-mentioned initial matrix method to generate a historical slope matrix. After the encoding is completed, the encoding matrix of the above-mentioned ecological restoration plan is integrated into a data set to generate a historical slope data set. The historical slope data set is used as a training set and a test set to train and test subsequent deep learning models.
[0034] As some embodiments, the deep learning model used in this embodiment is a variational autoencoder, wherein the variational autoencoder adopts an encoder-decoder structure. At the same time, in order to facilitate the use of the deep learning model, a two-dimensional deep learning model is adopted. The above-mentioned slope matrix is used as the input of the encoder. There is a latent space between the encoder and the decoder, wherein the latent space contains latent variables, wherein the input space, latent space, and output space all belong to integer space, and the encoder output is a latent variable and a latent variable distribution parameter. At the same time, the latent variable is used as the input of the decoder, so that the decoder can output the corresponding slope matrix, wherein the encoder represents the mapping relationship between the slope matrix and the latent variable, and the slope matrix can be mapped to the latent variable, and the decoder represents the associative mapping of the conversion of the latent variable to the slope matrix. Through training, the decoder learns this mapping relationship, and when used alone, it can generate different and more slope matrices through latent variables corresponding to different existing slope matrices. The potential distribution of latent variables in the latent space is common knowledge in the art, and the structure of the variational autoencoder can be set through conventional structures in the art, which will not be elaborated here.
[0035] As some embodiments, the variational autoencoder used in the present invention is different from the conventional structure of the convolutional neural network architecture, and the structure of the encoder includes: four sequentially connected convolution modules, Flatten layer, Linear layer, LReLU layer and Linear layer, wherein the convolution module includes a sequentially connected Conv layer, BN layer and LReLU layer, and the decoding structure includes an Unflatten layer, a deconvolution Deconv layer, a BN layer, an LReLU layer, two sequentially connected upsampling modules, a convolution module, a convolution layer and an output layer, wherein the upsampling module includes a sequentially connected Upsample layer, a Conv layer, a BN layer and a LReLU layer. Among the above contents, the Conv layer is a two-dimensional convolution layer; the Deconv layer is a two-dimensional deconvolution layer; the Linear layer is a fully connected layer; the BN layer is a two-dimensional batch normalization layer; the LReLU layer indicates the use of the LeakyReLU activation function; Upsample indicates a two-dimensional bilinear interpolation upsampling operation; Flatten indicates expanding the tensor format data into a vector, Unflatten is the reverse process of the above operation, and the output layer uses the tanh function. The above model parameters can be determined according to actual needs and will not be described in detail here.
[0036] As some embodiments, the above deep learning model is trained by using a historical slope data set, the training set trains a variational autoencoder, i.e., a deep learning model, and the deep learning model is trained using an Adam optimizer. The loss function of the VAE model training is expressed as follows:
[0037] min L=λ1NMSE(x)+λ2KL(x)
[0038]
[0039] Where NMSE is the normalized root mean square error, which is used to measure the difference between the slice matrix generated by the decoder, that is, the reconstructed sample x, and the original slice matrix, that is, the original sample x0; KL represents the Kullback-Leibler divergence, which is used as a regular term in the loss function to ensure the generation ability of the VAE model. λ1 and λ2 represent the corresponding item weight coefficients, μ and σ represent the mean and standard deviation of each component of the latent variable in the latent space, respectively. Represents the expected function of the encoder input data operation result, Represents the expected function of the encoder output data operation result.
[0040] As some embodiments, after training, the deep learning model uses a decoder to generate a slope matrix, wherein hidden variables are randomly adjusted and selected in the latent space, and the corresponding slope matrix is generated by the decoder using the hidden variables, and the generated slope matrix belongs to the law that conforms to the existing scheme, but the existing scheme does not provide a related scheme. By measuring the maximization of the slope matrix score generated by the decoder as the goal, the hidden variables are optimized and selected, and the final slope matrix is generated as a redesigned or currently designed slope scheme. In the above content, the decoder can store the relationship between the hidden variables and the slope matrix, and the slope matrix finally generated by the decoder can simulate the ability to generate the existing slope matrix, and conform to the actual setting rules of the non-implementable river slope ecological restoration plan, providing a relevant data basis for subsequent optimization.
[0041] The expert scoring method is used to score the river slope ecological restoration plan corresponding to the existing slope matrix based on the current river slope environment, or to evaluate the impact of the existing river slope ecological restoration plan on the current river slope environment under the current river slope environment. The final score is determined by weighted sum, where:
[0042]
[0043] Among them, J p represents the final score of the expert scoring for the pth repair solution, J pq represents the scoring result of the qth expert for the pth river slope ecological restoration plan, p represents the number of different indicators, q represents the number of different experts, n represents the total number of experts, and J o,pq represents the score corresponding to the oth indicator among the qth expert in the pth river slope ecological restoration plan, w o It represents the weight of the indicator in the oth position, which is set according to manual experience. o represents the indicator number, and m represents the number of indicators. The indicators include ecological restoration capacity (such as the ability to repair soil and water loss, vegetation destruction, soil erosion, etc. in the current river slope ecology), anti-solidification capacity, appearance score, cost score, construction difficulty, scheme risk, etc., which are set according to expert scoring.
[0044] It should be noted that the evaluation model is constructed for a single environment, not a specific area. For an environment similar to the current river slope environment (such as slope, height, soil quality, ecological problems, etc.), the above-trained evaluation model can also be used directly. As an extension, the corresponding evaluation models can be pre-built and stored for different environmental categories as a preparatory work for the technical solution, rather than a real-time process. In this way, different river slope environments can be classified into different environmental categories, and the evaluation model can be directly extracted and used later.
[0045] Through a simple convolutional neural network to fit the correlation between the above scheme and the score, the above existing repair scheme and the corresponding expert score are used as the fitting or training data of the convolutional neural network, and the convolutional neural network is trained to construct a scoring model, i.e., an evaluation model, in which the input of the convolutional neural network is the slope matrix, and the output is the corresponding score. The structure of the convolutional neural network is exemplified by an input layer, three convolutional layers, two fully connected layers, and an output layer. The functions used in each layer are common sense in the field and are not described here. The score data of the above convolutional neural network is used as the evaluation basis. In the optimization process, the main optimization goal is to maximize the score of the repair scheme generated by the encoder, and the optimization is carried out on this basis.
[0046] As some embodiments, in the process of finding the optimal solution, the optimization algorithm is combined with the above-mentioned deep learning model and evaluation model to search. In this embodiment, the improved moth-to-fire algorithm is used as the optimization algorithm to implement the above content. This embodiment is mainly aimed at finding the slope matrix or the river slope ecological restoration plan. The relevant content of the optimization algorithm includes:
[0047] The hidden parameters are used to generate the initial population, which contains different individuals, and the individuals contain different hidden parameters. The data encoded by the slope matrix represented by the hidden parameters can generate different slope matrices by adjusting and decoding the hidden parameters; its goal is to maximize the score of the ecological restoration plan corresponding to the slope matrix. The objective function is:
[0048] f(x)=maxJ(B(x)) 2 =maxJ 2
[0049] Where J represents the score of the slope matrix corresponding scheme. J() represents the output score of the convolutional neural network, and B(x) represents the processing of the implicit function x by the decoder to generate the slope matrix. Set the population update method and the iteration stop condition (satisfy the maximum number of iterations), and after setting the above content, optimize the pollution source.
[0050] As some embodiments, the specific steps of the optimization algorithm are:
[0051] 1) Initialize the population, where the population contains a number of different individuals or moth positions, each of which contains the value of a hidden parameter, and the position selected according to artificial experience or randomly is used as the initialization individual.
[0052] Set the relevant objective function, constraints and corresponding fitness, where the objective function is:
[0053] f(x)=maxJ(B(x)) 2 =maxJ 22
[0054] Where J represents the score of the slope matrix corresponding solution. J() represents the output score of the convolutional neural network, and B(x) represents the processing of the implicit function x by the decoder to generate the slope matrix. The fitness function is J(B(x)) 2 .
[0055] 2) In the improved moth-to-fire algorithm, the moth needs to update and iterate the flames around it. The flames represent the local optimal solution, and the number of flames is determined. At the same time, during the iteration process, the number of flames needs to be adaptively reduced. The adaptive reduction mechanism formula is as follows:
[0056] FN = round(Nt*(N-1) / T)
[0057] Where: N is the initial number of flames; t is the current number of iterations; T is the maximum number of iterations, and FN represents the number of flames during the tth iteration.
[0058] 3) The fitness of different individuals is calculated according to the objective function. The greater the fitness, the better the quality of the individual. During the calculation process, the data corresponding to the individual is decoded as a latent variable through a decoder to generate a slope matrix, and the corresponding score is generated through the evaluation model according to the slope matrix.
[0059] The fitness is sorted from large to small. After sorting, the corresponding number of flames is selected according to the sorting result, and the individuals in the fitness sorting are used as flames. The number of flames is calculated according to the above fitness reduction mechanism formula, and the current best flame is stored. The best flame is the individual with the highest number of flames in the fitness sorting.
[0060] 4) According to the above individual update method, or the moth position update method, the update is performed. In the update process, the initial position update method is first used for updating. In the initial position update method, the above moth position update method is adaptively updated through the inertia weight:
[0061] Mi=Di·e bθ ·cos(2πθ)+ω·Fj
[0062] Among them, b is a constant used to define the shape of the logarithmic spiral; θ is a random number between [-1,1]; Mi represents the i-th moth, that is, the individual; Fj represents the j-th flame; Di represents the distance between the i-th moth and the j-th flame, and the formula is the constant of the logarithmic spiral shape; the formula is: Di=|Fj-Mi|, ω is the inertia weight, which refers to the inertia weight idea of the particle swarm algorithm. The larger the ω, the greater the search intensity in the global range of the algorithm, and the smaller the ω, the higher the search accuracy in the local range of the algorithm.
[0063]
[0064] Where t is the current iteration number; T is the maximum iteration number; a and d are constants respectively.
[0065] 5) After the update, update the moth position again:
[0066]
[0067] Among them, MI represents the label of the moth individual that is updated again, rand() represents a random number, and Levy represents the corresponding Levy random step.
[0068]
[0069] Among them, β is a constant, which has been set to 1.5, and the distribution number μ~N(0,σ 2 ), σ 2 represents the upper limit; random distribution number v~N(0,1);
[0070]
[0071] Where Г() represents the standard gamma function.
[0072] 6) Determine whether the flame data satisfies the requirement of only one flame remaining. If only one flame remains, directly output the final result;
[0073] 8) Determine whether the maximum number of iterations has been reached. If not, repeat the above steps 2-6. If reached, output the optimal moth individual, that is, the slope matrix corresponding to the hidden parameters finally found.
[0074] The optimal slope matrix is decoded according to the above coding rules to generate the corresponding optimal river slope ecological restoration plan.
[0075] Through the above technical scheme, the present invention statistics the existing river slope ecological restoration schemes, and trains the deep learning model according to the statistical results, so that the deep learning model learns and can regenerate new river slope ecological restoration schemes. On this basis, by scoring the new river slope ecological restoration schemes and iterative optimization based on this, the optimal river slope ecological restoration scheme can be effectively found. Through the above technical scheme, the present invention can find or predict the optimal river slope ecological restoration scheme to improve the effectiveness of the river slope ecological restoration scheme. At the same time, the scoring process takes into account multiple indicators to ensure that the scheme under each indicator is optimal. Based on this, it can effectively guide the actual work of river slope ecological restoration, and has strong practicality.
[0076] The above are only preferred specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.
Claims
1. A method for ecological restoration of river slopes, characterized in that: include: Obtaining a river slope ecological restoration plan, encoding the river slope ecological restoration plan, and obtaining a slope matrix; Constructing a deep learning model, and training the deep learning model through a slope matrix to obtain a trained decoder, wherein the deep learning model adopts a generative network of an encoder-decoder structure; An evaluation model is constructed, and the river slope ecological restoration plan is evaluated based on the current river slope through the expert scoring method to obtain the evaluation score. The evaluation model is optimized according to the slope matrix and the evaluation score to obtain the optimized evaluation model. The trained decoder is used to iteratively generate an initial slope matrix, and the iteratively generated initial slope matrix is evaluated by the optimized evaluation model, and the optimal slope matrix is obtained according to the evaluation result; The optimal slope matrix is decoded to obtain the optimal ecological restoration plan for the current river slope. According to the optimal river slope ecological restoration plan, the current river slope is set up to achieve ecological restoration.
2. The river slope ecological restoration method according to claim 1, characterized in that: The process of coding the river slope ecological restoration plan includes: Obtain the materials and locations used in the river slope ecological restoration plan; Construct the numbering relationship between the initial coding sequence and the material and the label, wherein the initial coding sequence is set with corresponding coding positions according to the types of different materials. Fill in the initial coding sequence according to the type of material and the corresponding label, generate the coding result, and construct the initial slope matrix. Fill the coding result into the initial slope matrix according to the setting position of the material in the river slope scheme to obtain the slope matrix.
3. The river slope ecological restoration method according to claim 1, characterized in that: The deep learning model adopts a variational autoencoder, wherein the variational autoencoder adopts an encoder-decoder structure, the slope matrix is used as the input of the encoder, the output of the decoder is the slope matrix, there is a latent space between the encoder and the decoder, wherein the latent space contains latent variables, wherein the input space, latent space, and output space all belong to integer space.
4. The method for ecological restoration of river slopes according to claim 3, characterized in that: The structure of the encoder includes: four sequentially connected convolution modules, a Flatten layer, a Linear layer, an LReLU layer and a Linear layer, wherein the convolution module includes a sequentially connected Conv layer, a BN layer and an LReLU layer.
5. The method for ecological restoration of river slopes according to claim 3, characterized in that: The decoding structure includes an Unflatten layer, a deconvolution Deconv layer, a BN layer, an LReLU layer, two upsampling modules connected in sequence, a convolution module, a convolution layer and an output layer, wherein the upsampling module includes an Upsample layer, a Conv layer, a BN layer, and an LReLU layer connected in sequence.
6. The method for ecological restoration of river slopes according to claim 1, characterized in that: The evaluation model adopts a convolutional neural network.
7. The river slope ecological restoration method according to claim 6, characterized in that: The structure of the convolutional neural network includes: input layer, three convolutional layers, two fully connected layers, and output layer.
8. The method for ecological restoration of river slopes according to claim 1, characterized in that: The process of evaluating river slope ecological restoration plans includes: Among them, J p represents the final score of the expert scoring for the pth repair solution, J pq represents the scoring result of the qth expert for the pth river slope ecological restoration plan, p represents the number of different indicators, q represents the number of different experts, n represents the total number of experts, and J o,pq represents the score corresponding to the oth indicator among the qth expert in the pth river slope ecological restoration plan, w o It represents the weight of the indicator in the oth position, which is set according to manual experience. o represents the indicator number, and m represents the number of indicators. The indicators include ecological restoration capacity, anti-solidification capacity, appearance score, cost score, construction difficulty, and scheme risk. The ecological restoration capacity includes the restoration capacity of soil and water loss, vegetation destruction, soil erosion, etc. in the current river slope ecology.
9. The method for ecological restoration of river slopes according to claim 1, characterized in that: The process of obtaining the optimal slope matrix includes: A population is constructed and initialized according to the input of a trained decoder, wherein the population includes a plurality of individuals, wherein the individuals include different hidden parameters; an objective function, a fitness and a population update method are set, wherein the objective function is to maximize the evaluation score of a slope matrix generated by the trained decoder; and an optimization algorithm is used to search for the best result according to the objective function, the fitness and the population update method to generate an optimal slope matrix.
10. The method for ecological restoration of river slopes according to claim 1, characterized in that: The optimization algorithm adopts the improved moth-to-fire algorithm.
Citation Information
Patent Citations
Dynamic feedback adjustment method for integrated evaluation and guidance of artificial regulation of slope ecological restoration by using slope ecological restoration project
CN106651138A
Ecological restoration method for mining area
CN116090623A
CGAN-based gas turbine sealing structure shape optimization method
CN117493772A
Wind-solar active power output scene generation method and device based on improved VAE model
CN118095052A
Device and method for estimating the oil consumption of a vehicle engine
EP1956203A1
