CN-121979930-A - Intelligent recommendation method, system, equipment and medium for ecological protection scheme of rock slope

CN121979930ACN 121979930 ACN121979930 ACN 121979930ACN-121979930-A

Abstract

The application relates to an intelligent recommendation method, system, equipment and medium for a rock slope ecological protection scheme. The method comprises the steps of obtaining a rock slope ecological protection theory and a historical rock slope ecological protection case, constructing an ecological protection knowledge graph according to the ecological protection theory and the historical rock slope ecological protection case, obtaining slope basic data of a target rock slope area, obtaining candidate protection schemes by matching from the knowledge graph through a semantic matching algorithm, inputting the slope basic data and parameters of the candidate protection schemes into a protection scheme feasibility assessment machine learning model, carrying out feasibility quantitative assessment and generating assessment values, setting genetic fitness function values based on the assessment values, optimizing the candidate protection schemes by adopting a genetic algorithm, selecting the candidate protection scheme with the highest feasibility quantitative assessment value as a recommended ecological protection scheme when the condition of genetic termination is met, realizing intelligent recommendation of the rock slope ecological protection scheme, and improving scientificity and efficiency of scheme recommendation.

Inventors

ZHAO MEI
MA JIANRONG
LI CHUANHONG
ZHANG DONG
SUN BOWEI
ZHAO SEN

Assignees

交通运输部公路科学研究所

Dates

Publication Date: 20260505
Application Date: 20260210

Claims (10)

1. An intelligent recommendation method for a rock slope ecological protection scheme is characterized by comprising the following steps: Acquiring a rock slope ecological protection theory and a historical rock slope ecological protection case, and constructing an ecological protection knowledge map according to the rock slope ecological protection theory and the historical rock slope ecological protection case; Acquiring slope basic data of a target rock slope region, and matching from the ecological protection knowledge graph through a semantic matching algorithm based on the slope basic data to obtain a candidate protection scheme; Inputting the slope basic data and the protection scheme parameters of the candidate protection scheme into a protection scheme feasibility evaluation machine learning model, performing feasibility quantitative evaluation, and generating a feasibility quantitative evaluation value; And setting a genetic fitness function value of the candidate protection scheme based on the feasibility quantification evaluation value, carrying out genetic optimization on the candidate protection scheme according to a genetic algorithm, and selecting the candidate protection scheme with the highest feasibility quantification evaluation value as a recommended ecological protection scheme when the candidate protection scheme meets a genetic termination condition.
2. The method according to claim 1, wherein constructing an ecological protection knowledge-graph based on the rock slope ecological protection theory and the historical rock slope ecological protection cases comprises: Based on a geotechnical engineering ontology framework, carrying out entity recognition on the geotechnical slope ecological protection theory by adopting a natural language algorithm to obtain a geotechnical slope ecological protection theoretical entity, and carrying out relation extraction on the geotechnical slope ecological protection theoretical entity to obtain a theoretical triplet; Extracting engineering parameters, a protection scheme and effect data from the historical rock slope ecological protection case, and constructing and obtaining case feature nodes based on the engineering parameters, the protection scheme and the effect data; embedding the case feature nodes into the theoretical triples by adopting a graph embedding algorithm, and constructing a first knowledge graph; Performing multi-hop reasoning optimization on the case feature nodes and the theoretical triples in the first knowledge graph to obtain a second knowledge graph; and carrying out knowledge module division on the second knowledge graph to obtain the ecological protection knowledge graph.
3. The method according to claim 2, wherein the matching, based on the slope base data, by a semantic matching algorithm, from the ecological protection knowledge-graph to obtain a candidate protection scheme includes: based on the slope basic data, extracting geological parameters, environmental characteristics and engineering requirements from the ecological protection knowledge graph; based on the geological parameters, the environmental characteristics and the engineering requirements, constructing and obtaining a module query vector by taking each knowledge module in the ecological protection knowledge graph as a query standard, wherein the knowledge modules comprise a stability adaptation module, an ecological adaptation module and an engineering implementation adaptation module; Fusing the module query vectors based on a graph attention mechanism, generating a comprehensive query vector, and calculating to obtain a multi-dimensional semantic similarity score of the comprehensive query vector based on the comprehensive query vector and the case feature nodes in the ecological protection knowledge graph; And selecting a protection scheme corresponding to the case feature nodes with the sum of the dimension semantic similarity scores being N before ranking according to the dimension semantic similarity scores and the dimension score threshold, wherein each dimension semantic similarity score in the dimension semantic similarity scores exceeds each dimension score threshold in the dimension score threshold.
4. The method of claim 3, wherein the multi-dimensional semantic similarity score is calculated as: Wherein, the For the semantic weighting coefficients of the kth knowledge module, An engineering fitness decay factor for the kth of the knowledge modules, For the projection of the comprehensive query vector in the characteristic dimension space corresponding to the kth knowledge module, For the transpose operation, For the embedded vector of the case feature node corresponding to the candidate protection scheme in the dimension corresponding to the kth knowledge module, For the engineering parameter feature vector in the case feature node corresponding to the kth knowledge module, An ideal feature vector corresponding to the engineering parameter feature vector in the case feature node of the kth knowledge module, And the tolerance coefficient of the engineering parameter of the dimension corresponding to the kth knowledge module.
5. A method according to claim 3, wherein the protection scheme feasibility assessment machine learning model is a random forest model; the step of inputting the slope basic data and the protection scheme parameters of the candidate protection scheme into a protection scheme feasibility evaluation machine learning model, and performing feasibility quantitative evaluation, and generating a feasibility quantitative evaluation value comprises the following steps: Determining an evaluation index based on engineering requirements in the slope base data and the candidate protection scheme; Constructing a candidate scheme feasibility evaluation data set based on the evaluation index and the slope basic data; and inputting the feasibility evaluation data set of the candidate scheme into the random forest model, generating a multi-dimensional evaluation index parameter, and calculating to obtain the feasibility quantification evaluation value according to the multi-dimensional evaluation index parameter and the evaluation index weight distribution rule.
6. The method of claim 5, wherein the evaluation metrics include stability fitness, ecological fitness, engineering implementation difficulty, and economic fitness; The constructing a candidate scheme feasibility evaluation data set based on the evaluation index and the slope basic data comprises the following steps: extracting index feature data corresponding to each evaluation index from the candidate protection schemes based on the evaluation index; Determining the association relationship between the slope basic data and the evaluation index by taking the theoretical triples in each knowledge module in the ecological protection knowledge graph as the basis; calculating the correlation strength of the association relation, and screening out key feature attributes according to a correlation threshold based on the correlation strength; Constructing an initial scheme feasibility evaluation data set based on the key feature attributes and the association relation; And carrying out equalization processing on the initial scheme feasibility evaluation data set to obtain a candidate scheme feasibility evaluation data set.
7. The method according to claim 5, wherein the setting the genetic fitness function value of the candidate protection scheme based on the feasibility quantization evaluation value, performing genetic optimization on the candidate protection scheme according to a genetic algorithm, and selecting the candidate protection scheme with the highest feasibility quantization evaluation value as a recommended ecological protection scheme when it is identified that a genetic termination condition is satisfied, comprises: constructing an adaptability function matrix of the genetic algorithm according to the feasibility quantization evaluation value; based on the fitness function matrix and iteration termination conditions, carrying out iterative computation by adopting the genetic algorithm to obtain an iterative computation result; The iteration termination condition comprises that the iteration number exceeds the upper limit of the iteration number and/or the fitness fluctuation value of continuous iteration is smaller than the fitness fluctuation threshold; and selecting a gene sequence with the highest fitness value in the population based on the iterative calculation result, and decoding the gene sequence to obtain the recommended ecological protection scheme.
8. An intelligent recommendation system for a rock slope ecological protection scheme, which is characterized by comprising: the knowledge map construction module is used for acquiring a rock slope ecological protection theory and a historical rock slope ecological protection case, and constructing an ecological protection knowledge map according to the rock slope ecological protection theory and the historical rock slope ecological protection case; the candidate scheme matching module is used for acquiring slope basic data of the target rock slope region, and matching the ecological protection knowledge graph to obtain a candidate protection scheme through a semantic matching algorithm based on the slope basic data; the scheme feasibility evaluation module is used for inputting the slope basic data and the protection scheme parameters of the candidate protection scheme into a protection scheme feasibility evaluation machine learning model, carrying out feasibility quantitative evaluation and generating a feasibility quantitative evaluation value; and the scheme optimization recommendation module is used for setting the genetic fitness function value of the candidate protection scheme based on the feasibility quantification evaluation value, carrying out genetic optimization on the candidate protection scheme according to a genetic algorithm, and selecting the candidate protection scheme with the highest feasibility quantification evaluation value as a recommended ecological protection scheme when the candidate protection scheme meeting the genetic termination condition is identified.
9. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor implements the steps of the method of any of claims 1 to 7 when the computer program is executed.
10. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 7.

Description

Intelligent recommendation method, system, equipment and medium for ecological protection scheme of rock slope Technical Field The invention belongs to the field of geotechnical engineering and ecological protection, and particularly relates to an intelligent recommendation method, system, equipment and medium for a geotechnical slope ecological protection scheme. Background Along with the development of the technology in the geotechnical engineering and ecological protection fields, the technology for ecologically protecting the rock slope has the dual characteristics of slope structural stability reinforcement and surrounding ecological environment restoration, is an important development direction in the geotechnical engineering field, and most of the industries currently rely on the experience of engineering technicians and combine basic investigation data to formulate a protection scheme. In the traditional technology, technicians firstly perform geological investigation on a target rock slope area, obtain basic data such as slope geology, topography and the like, refer to part of similar historical cases by combining engineering experience, select a protection scheme in a manual analysis mode, and determine the scheme after part of scenes are simply compared with the applicability of a plurality of conventional schemes. The existing mode has the following problems that the mode is affected by personal experience of technicians and limitation of knowledge reserve, the matching degree of a protection scheme and actual conditions of a slope is insufficient, the geological characteristics of the slope, ecological restoration requirements and engineering implementation conditions are difficult to consider, systematic excavation and utilization of historical engineering cases are lacked, the scheme formulation efficiency is low, a standardized quantitative evaluation system is not available, the scheme feasibility judgment subjectivity is high, poor protection effect or engineering cost waste is easily caused, and the scientific and intelligent development requirements of the current rock slope ecological protection engineering cannot be met. . Disclosure of Invention Based on the above, it is necessary to provide a method, a system, a device and a medium for intelligent recommendation of a rock slope ecological protection scheme, which can solve the above problems. In a first aspect, the application provides an intelligent recommendation method for a rock slope ecological protection scheme, which comprises the following steps: acquiring a rock slope ecological protection theory and a historical rock slope ecological protection case, and constructing an ecological protection knowledge map according to the rock slope ecological protection theory and the historical rock slope ecological protection case; Acquiring slope basic data of a target rock slope region, and matching from an ecological protection knowledge graph by a semantic matching algorithm based on the slope basic data to obtain a candidate protection scheme; inputting the slope basic data and the protection scheme parameters of the candidate protection scheme into a protection scheme feasibility evaluation machine learning model, performing feasibility quantitative evaluation, and generating a feasibility quantitative evaluation value; and setting a genetic fitness function value of the candidate protection scheme based on the feasibility quantization evaluation value, carrying out genetic optimization on the candidate protection scheme according to a genetic algorithm, and selecting the candidate protection scheme with the highest feasibility quantization evaluation value as a recommended ecological protection scheme when the genetic termination condition is identified. In one embodiment, constructing an ecological protection knowledge graph according to a rock slope ecological protection theory and a historical rock slope ecological protection case includes: Based on a geotechnical engineering ontology framework, carrying out entity recognition on a geotechnical slope ecological protection theory by adopting a natural language algorithm, recognizing to obtain a geotechnical slope ecological protection theoretical entity, and carrying out relation extraction on the geotechnical slope ecological protection theoretical entity to obtain a theoretical triplet; Extracting engineering parameters, a protection scheme and effect data from the historical rock slope ecological protection case, and constructing case feature nodes based on the engineering parameters, the protection scheme and the effect data; Embedding case feature nodes into theoretical triples by adopting a graph embedding algorithm, and constructing to obtain a first knowledge graph; Carrying out multi-hop reasoning optimization on the case feature nodes and the theoretical triples in the first knowledge graph to obtain a second knowledge graph; and carrying out knowledge module division