| Safety performance evaluation is a comprehensive concept describing the overall operation state of safety system, which is also an important way to know better about the effects of security system in every aspect. With further research and development in the safety concept and theories, How to establish an scientific model has become a hotspot, a large number of application researches were emerged, the definition category, related factors, modeling structures are all covered in the research content, which is helpful in enriching the theory and practical value of safety performance.In order to assess what have achieved in a company after all the investment in safety work,Enterprises themselves were also seeking for measurement during the production process. Nevertheless, theoretical research in this area is not mature, the modeling work often met with some difficulties due to lack of theoretical guidance. According to this situation, this paper has established a safety performance evaluation model based on DPSIR and BP neutral network by learning from the previous studies. and then put it into practice in S company to verify the feasibility of this model. The contents are mainly about the following aspects:(1) In order to lay a sound theoretical foundation for later studies, a careful research on the related literature both at home and abroad was made, to fully understand the definition of safety performance, the influential factors that determine the safety performance level and how to measure those indicators,what kind of method can be used for evaluation etc. and finally decide to evaluate the safety performance from two aspects, accidents and operation state of safety system.(2) Analyzed the situations about studies in safety performance home and abroad. By learning from the DPSIR(Driving force-Pressure-State-Impact-Response) model in environmental evaluation system, and filtering the related indicators, a evaluation system was obtained, which include 5 first indicators and second level indicators of a total of 25.(3) By comparing different evaluation methods that have been commonly used, BP neural network was chose to assess the safety performance which can reduce the influence of subjective factors. The principle and evaluating process of this method were described, and finally built a safety performance evaluation model based on DPSIR and BP neural network.(4) Use this model to evaluate the safety performance of a turbocharger assembly company. The result shows that the performance was level three, which also proved the feasibility of this model. |
