Research On The Optimal Configuration Of Construction Safety Investment Based On GWO-SVR

With the proposal of China’s "14th Five-Year Plan",the construction industry,as a pillar industry of the national economy,has officially entered a key period of transformation and development.Due to the unreasonable allocation of safety investment in the past,the safety production situation of construction companies has been severe,with frequent accidents resulting in serious casualties and economic losses.Therefore,how to optimize the allocation of safety investment in construction companies and improve the efficiency of safety investment has become an urgent issue to be addressed.This article is based on the research project "Assessment and Control of Construction Project Safety Risks and Optimization of Safety Investment" in cooperation with China Railway Construction Group Co.,Ltd.The article applies Cluster Analysis to establish a safety investment indicator system and a relationship model between safety investment and accident losses.Then,using the Order Relation(G1)-Entropy Weight(EWM)combination weighting method,the article determines safety investment constraints based on risk factors.Based on the relationship model and risk factors,the article establishes a safety investment optimization model with safety investment objectives as the basis and risk factors as constraints,in order to achieve a reasonable allocation of safety investment in construction and minimize accidents.The main contents of this article are as follows:(1)Establish a safety investment indicator system.By distinguishing between two types of factors that affect safety investment: policy environment factors and risk factors,the article initially selects nine safety investment indicators and optimizes them to five indicators using system cluster analysis.The safety investment indicator system includes five indicators:safety technology investment,safety management investment,safety protection investment,safety education investment,and safety emergency investment.(2)Construct the relationship model between construction safety investment and accident loss.The support vector machine(SVR)model is selected through model comparison,and each safety input index in the safety input index system is used as the model input index,and the accident loss is used as the output index.Taking the actual safety investment and accident loss data of Dalian Metro Control Center as the training set,the GWO algorithm and cross-validation method are used to optimize the parameters of the SVR model,and the relationship model between safety investment and accident loss is obtained.The determined coefficient of the tested model is 0.994826,which can provide an important basis for the optimization of safety investment.(3)Determine the security investment constraints based on risk factors.Apply Job Hazard Analysis(JHA)to carry out risk identification on the control center project.According to the identified 17 risk points and 156 hazard sources,a risk constraint index system was constructed.And apply the G1-EWM combination weighting method to quantitatively analyze the risk constraints of safety investment.(4)Construct safety investment optimization decision-making model.Firstly,the concept of generalized safety investment is introduced based on the relational model,and the objective function of safety investment is determined.Secondly,determine the specific constraints of security investment based on risk factors,and build a security investment optimization decision-making model based on the Gray Wolf Algorithm(GWO).Taking the actual safety investment of the control center project from October 2021 to March 2022 as the optimization object,the optimization plan of safety investment configuration for 6 months is obtained through model calculation.The safety input allocation scheme has obvious effect,which theoretically reduces the accident loss by 24.91%,and reduces the generalized safety input by5.037%.