基于蒙特卡洛模拟和主客观综合权重的化工园区配电系统模糊综合评价

Research on fuzzy comprehensive assessment for chemical industry park on fuzzy comprehensive assessment for chemical industry park power supply and distribution system based on Monte-Carlo simulation

化工园区配电系统网络结构及装置配备复杂,因配电系统故障导致的用户停电故障占总故障的大多数,合理的配电系统评价是化工园区正常运转的保障。采用模糊综合评价法,专家打分阶段选择"给出得分区间"、"给出得分分布"及"精确打分"等任意一种方法实现打分,结合蒙特卡洛模拟产生服从给定区间或分布的随机得分,并据此求取随机模糊评判矩阵;依据蒙特卡洛可随机产生多组数据的特点,求各指标得分的标准差以反映得分的分散情况,并据此求算客观权重,将客观权重作为指标层次结构中的指标层权重,应用AHP方法求取主客观综合权重;在以上工作基础上,迭代求取并汇总各次模糊评价结果,最后得出的综合评价结论与实际情况相符。

The present paper is dedicated to a fuzzy comprehensive assessment for chemical industry park power supply and＂ distribution system based on Monte-Carlo simulation. We have chosen the topic for the reason that the network structure and configuration system of power supply and distribution are very complicated in the chemical industry park. Most of the power apparatus and equipment prove to fail to satisfy the ever increasing need due to incomprehensive and in- complete extent in power supply and distribution. Therefore, it is of great urgency to make a proper evaluation for the system of power supply and distribution so as to be able to ensure the proper function-ing of the chemical industry parks. In this paper, we would like to make a fuzzy comprehensive assessment as the basic evaluation method. According to the cognitive identification of the experts on the evaluation target, we would like to introduce such methods, as ＂the score interval＂, the ＂score distribution＂, and ＂the exact scoring＂ as the reference for evaluating the system. Then, we have introduced the random score for the given interval or distribution by using the Monte-Carlo simulation. And, based on the random score, we have attained the random fuzzy assessment matrix and the final valuation results, which are made of a group of numbers, instead of just a few numbers. Through the analysis of the range of the group numbers, the overall operation condition of the system of power supply and dis- tribution in the chemical industry parks can be determined. In addi- tion, since multiple groups of data could be generated by Monte-Carlo simulation, the standard deviation for the scores of indexes could show the dispersion situation of index scores, and the objective weight could be calculated. Monte Carlo random number generation is based on pre-set index score distribution. For normal distribution, its stan- dard deviation is pre-set. But, since Weibull distribution whose stan- dard deviation is not pre-settled, the standard deviation can only be obtained through application probability and statistics theory. Take the objective weight as the index weights in index layers, we have ap- plied the AHP method to the calculation of the subjective and objec- tive comprehensive weights. And, finally, the evaluation index sys- tem of power supply in chemical industry park has been established, whose guidelines layers have been found consisting of the supply safe- ty, the reliability, the economy, and the harmony of the different sections of the enterprise. Then, the model in this paper has been applied to the fuzzy comprehensive evaluation on the power supply and distribution system in the chemical industry park. The evaluation results have been mainly settled up between 73.5-75.5 whose univer- sal comment is ＇fine＇, with the results of evaluation considered as down-to-earth and objective.