Research On Risk Early Warning Of Safety Accidents Of Construction Tower Cranes

Building construction tower crane is one of the indispensable machinery and equipment in the construction site,its use has greatly improved the efficiency of the project construction.However,due to the complex operating environment and operating conditions,tower crane safety accidents occur from time to time,tower crane safety accidents not only cause large casualties but also generate huge economic losses.In order to reduce the probability of tower crane safety accidents and reduce their impact on the project,it is particularly important to conduct early warning research on tower crane safety accidents.However,the current tower crane safety accident risk early warning system is not perfect,for the tower crane safety accident early warning research is still small.Therefore,carrying out tower crane safety accident risk early warning system research,for tower crane safety accident prevention and control is of great significance.In this thesis,from the perspective of building construction tower crane safety accident analysis,the tower crane safety accident risk warning system has been studied.firstly,literature analysis and case study method are used to identify the causal factors of tower crane safety accidents,and on this basis,the index system of tower crane safety accident risk warning model is determined by using grey correlation method.Secondly,a BP neural network(BPNN)is used for the early warning of tower crane safety accidents,and the genetic algorithm(GA)and particle swarm algorithm(PSO),which have better global search capability,is used to optimize the BP neural network.Finally,the GA-BPNN model with a better prediction effect is applied to the JL Building project to verify the effectiveness of the tower crane safety accident risk warning model.With the above research contents,the final main research results of this thesis are as follows:(1)Through literature analysis and case study method,the main causal factors of tower crane safety accidents were summarized.The grey correlation method is also used to analyze the factors that lead to the high incidence of tower crane collapse,scattering,and accidental falls from height.21 key causes of tower crane safety accidents in four aspects,namely people,equipment,environment,and management,are finally identified in the tower crane safety accident risk warning model index system.(2)The GA and PSO algorithms were used to optimize the BP neural network to establish the tower crane safety accident risk warning model,and the GA-BPNN model was found to have better results than the PSO-BPNN model for tower crane safety accident risk warning through the testing and comparison of the two warning models,and the accident safety state prediction accuracy reached over 80%.(3)The GA-BPNN model is used for early warning of tower crane safety problems in the JL building project.The analysis of 50 sets of operation results classified that the four accident prediction scores are all floating up and down in their respective safety state intervals without exceeding the other intervals,which verifies that the GA-BPNN model has good generalization performance and can accurately predict tower crane safety problems.The research shows that the GA-BPNN model established in this thesis has a better effect in the early warning of tower crane safety accident risk,with higher prediction accuracy,which provides a theoretical basis and reference method for the research of tower crane safety accident risk warning using intelligent algorithms and enriches the means of tower crane safety accident prevention and control.