A Research On Safety Assessment Model Of Construction Site Based On Rough Set And Ann

The construction industry is one of the principle industries of national economy. Hence, the healthy development of construction industry is important to healthy development the national economy. However, the inherent features in the construction industry production lead to the security accidents occurred frequently, and the quantities of construction accidents and death toll showed a rising trend continuously, which affect badly the report of construction industry and its attraction. Although the government has taken a number of countermeasures and measures, it is difficult to effectively control the construction accidents occurred frequently. Therefore, according to the basic national conditions of China, we should establish the safety evaluation control system of construction site grounding on advanced international methods of safety management, which has great theoretical and practical significance to prevent and reduce security accidents.Aiming at the existent problem of the safety assessment technique on construction site in China, this paper researches the safety assessment deeply and systematically with the feature of construction sites, combing security system project, rough-set theory, modern artificial intelligence and computer software engineering, and it establishes the safety assessment model of construction site based on rough-set and artificial neural network.The main contents are as follows: First, this paper points out the limitations of safety assessment methods on construction building, summarizes recent safety management theories, and analyzes the status of construction safety management at home and abroad. Second, according to the relevant state regulations and the actual status of construction site, this paper establishes the safety assessment indexes system of construction site, and designs safety check table for each index. Third, combining the characteristics of rough-set and artificial neural network, this paper proposes a safety assessment hybrid model of construction site based on rough-set and artificial neural network. The comparative analysis between conventional artificial neural network model and hybrid model carries out through the empirical studies. The results indicate that hybrid model has higher forecast precision and higher efficiency than conventional artificial neural network model. Fourth, In order to distinguish the importance of the indexes in the safety assessment system, sensitive analysis conducts on the indexes of rough-set and artificial neural network model. According to the different importance of each index, we should take corresponding response in the safety monitor process of construction site, for improving the efficiency of safety control effectively.