Research On The Safety Pre-warning For Hoisting Operation In Fabricated Building Project Based On RVM

The development of fabricated building is beneficial to saving resource and energy,reducing the pollution caused by construction,improving productivity and the level of safety in production,as well as fully enhancing the construction standard and building quality.It’s the stage of comprehensive development of Chinese fabricated building,where policy support has been constantly being intensified and standard technique system has been more perfected.No matter technical support,market cultivation,building subsidiary facilities and management transformation are less proficient and there exists great distance to developed country.There are some shortcomings,like antiquated construction organization and technology,backward machinery and equipment,coexistence of cast-in-place and fabricated,which make the low level of fabricated construction,potential risks highlighting and facing great challenges in construction safety management.Safety management is a series of risk prevention and management procedure completed by following the law of information,where the key and difficult point is the safety pre-warning of accident risk.The correlated research indicated,traditional warning methods are less effective which based on evaluation and analysis model,while machine learning like relevance vector machine is more applicable in data mining and forewarning prediction.Accordingly,from the hoisting procedures in construction site of fabricated building project,this paper tries to achieve the safety pre-warning analyze and expound for hoisting of prefabricated components by tower crane by introducing RVM,in order to take effective measures to solve potential problems and risks for accident prevention.The safety pre-warning for hoisting operation in fabricated building projects was discussed through literature study,theoretical and methodological analysis,industry status investigation.First,comparing with the lifting of tower crane in conventional the cast-in-site technique,the problems and risk of the hoisting operation in fabricated building projects were analyzed,combined the process flow and the main pointsabout hoisting operation in fabricated building projects.The safety pre-warning indicator system was established in accordance with the“Manpower-Machine-Material-Method-Environments”,which developing the safety pre-warning 4M1 E elements framework.Then,the feasibility of the achieving the safety pre-warning by machine learning was analyzed.And the values and pretreatment of input variables,the option and principle of the algorithm,the determining of parameters and the working principle of model were elaborated.Rough set theory and improved particle swarm optimization were used to improved model algorithm,and the safety pre-warning model for hoisting operation in fabricated building project based on RS-IPSO-RVM was established.At last,the paper verified the validity and the advantage of the model used the dormitory building project of Xiangya campus of Central South University as an example.Some proposals and suggestions were provided to reducing the accident rate by the effective safety pre-warning.