Font Size: a A A

Based On The Dpsir - Oil And Gas Production Enterprises Performance Evaluation Model Of Svm Research

Posted on:2013-10-30Degree:MasterType:Thesis
Country:ChinaCandidate:D S ChenFull Text:PDF
GTID:2249330377958020Subject:Management Science and Engineering
Abstract/Summary:PDF Full Text Request
The characteristics of the production risk and risk complexity decide the oil and gas production companies should keep alert to prevent the production safety accidents. It may bring on serious results once some accidents happened, besides, it will even cause inestimable disaster. With the ceaseless development of the oil industry, safety performance assessment plays an important rule in the production and management of the oil and gas production enterprises. The choice of assessment model directly affects the objectivity and impartiality of the performance appraisal result. And it is also the key of the safety performance appraisal. It has a very important practical significance to improve oil and gas production enterprise’s safety performance assessment results to ensure the safety production in enterprises.Based on the research object of oil and gas production enterprises safety performance, with the performance assessment, safety accident causing theory and combining with DPSIR model, This paper comprehensively analyzes China’s oil and gas production enterprise safety situation from the five aspects such as driving force, pressure, station, impaction and response. On These Foundations, the twenty key indicators were selected to establish oil gas safety performance evaluation index system by the using of two eight rule. To deal with the problems of multi-dimension and nonlinear in safety performance appraisal index system, the support vector machine model has been introduced, which has the ability of classification and regression prediction to reflect oil and gas production enterprise safety performance results objectively and accurately. The model of safety performance assessment based on DPSIR-SVM in oil and gas production enterprise is constructed by cross validation method to optimize Support vector machine’s parameters. When compared the assessment results with neural network results, it is proved that the model is more effective and feasible. Finally, based on the training and testing samples, this paper chooses two oil and gas production enterprises, subordinated to CNPC, to evaluate safety performance level, empirically research the actual effect of the model.
Keywords/Search Tags:DPSIR, SVM, Safety performance, Evaluation model, Neural network
PDF Full Text Request
Related items