基于BP神经网络的工业园区废弃物物流系统模糊评价研究

Researchon Fuzzy Evaluation of Waste Logistics System in Industrial Park Based on BP Neural Network

依据评价指标设计原则,结合系统运作效率与效益、发展生态城市和生态工业园区的要求,从系统投入、系统能力、运作效率、处理效率和环境绩效5个方面提出22个二级指标的园区废弃物物流系统评价指标体系,并利用层次分析法合理确定指标权重,选择12个工业园区为对象进行实例分析,再利用提出的指标体系开展模糊评价,获得模糊评价的输入和评价结果。然后,根据模糊评价过程构建三层结构的BP神经网络,以模糊评价的输入和输出训练BP网络,获得可执行模糊评价的BP网络。实例分析表明:BP网络对各个园区的评价结果与模糊评价结果相同,是评价园区废弃物物流系统绩效便捷、可靠而有效的科学工具。

Based on the index design principles,combined with the system operation efficiency and benefit requirements,as well as the development requirements of ecological city and ecological industrial park,the evaluation index system of park waste logistics system consisted of 22 indexes with two level is put forward from the five aspects including system input,system capacity,operation efficiency,processing efficiency and environmental performance,meanwhile the index weights are determined reasonably by using the method of analytic hierarchy process. The 12 industrial parks are selected as the object of case analysis to be evaluated by the fuzzy evaluation method and the proposed index system,and the inputs and evaluation results of the fuzzy evaluation are obtained. Then according to the fuzzy evaluation process,the three layered BP neural network is constructed,trained by the input and outputs of the fuzzy evaluation,and the BP network is obtained to perform fuzzyevaluation. The example analysis shows that the evaluation result of the BP network is the same as the fuzzy evaluation, which is a convenient, reliable and effective scientific tool to evaluate the performance of waste logistics system.