Study of Analytical Model for Risk Evaluation in Interconnected Composite System
|School||Hefei University of Technology|
|Course||Proceedings of the|
|Keywords||reliability theory risk evaluation interconnected power systems analytical model Back-Propagation neural network sensitivityanalysis|
With the continuous expansion of power systems, regional interconnection hasbeen an important measure to improve power system reliability and optimizeallocation of resources. However, for the complex structure and numerouscomponents, the calculate disaster of risk evaluation remains a challenge. Therefore,based on reliability theory, this article took interconnection composite system asthe research object, trying to search new ways to solve this problem.First of all, this paper presented the analytical expressions of several riskindices with respect to components reliability parameters, on the basis of improvedequivalent reliability model for interconnected bulk power systems. Utilizing thismethord, the risk indices of bulk power system can be gotten quickly and accuratelyafter component reliability parameters have been changed. The curves and tableswere obtained too, through which one can find the different kind of effect ofcomponent parameters on system risk indices. In addition, the article also explainedand demonstrated the importance of high level contingency to the interconnectedbulk power system reliability by quantitative analysis of the different contingencylevels. Eventually, the proposed method was validated by taking the interconnectedRBTS and IEEE-RTS96test systems as examples.Secondly, on based of the analytical model, this paper established the trendanalysis model and the quantitative caculation method of risk indices whenconsiderates the different influent factors. Combining the mathematical reliabilitytheory and artificial intelligence algorithms, a prediction method of interconnectioncomposite system based on improved BP artificial neural network was established.Through training the existing samples, the prediction can be obtained directly whenthe influent factors changes simultaneously. By the sensitivity analysis on theinfluent factors, the quantitative results of its impact on the system reliability wereavailable. Taking interconnected RBTS system as an example to verify theeffectiveness of the model and method.