变电站选址是电网规划的重要组成,介绍了针对伴随科学技术和人类文明发展进步出现的新问题建立的变电站选址评价指标体系。运用加入动量项和基于SSE变学习速率的改进BP神经网络建立了变电站选址评价模型,这种智能评价方法摆脱了传统方法评价过程中人为因素及模糊随机性的影响,同时能保证评价结果的准确性,减少了变电站选址的所需经费和时间,实例运算得出令人满意的评价结果,证明了这种方法的有效性。
Substation site selection is an important part of grid planning. To deal with some new problems emerged with the development of science and technology and human civilization, the assessment index system for substation site selection was established. The BP neural network was improved by taking in the additional momentum and by the method of SSE-based variable leaning rate, and the assessment model for substation site selection was developed. The intelligent evaluation method gets rid of the influences of human factors and fuzzy randomicity which exist in the tradi- tional evaluation process, and ensures the accuracy of the assessment result, and the expense and time needed for substation site selection can be reduced. Calculation sample proves the method effective.