CAJ | 학술논문

电力网线损率是综合反映电力网规划设计、生产运行和经营管理的一项重要指标。如何有效地降低电力网电能损耗、提高企业经济效益已成为迫切需要解决的问题。文中提出一种基于遗传算法的RBF神经网络对线损进行计算，并且针对RBF神经网络的隐含层与输出层相互独立和输出结果容易陷入局部最小等缺点，运用遗传算法对整个网络进行优化，另外对遗传算法进行了相应的改进。为了验证算法的可行性，文中以大庆油田某一地区的67条线路为样本进行仿真计算。实验结果证明了遗传算法优化的RBF神经网络在计算精度和训练速度上都有较大提高。
전력망선손솔시종합반영전력망규화설계、생산운행화경영관리적일항중요지표。여하유효지강저전력망전능손모、제고기업경제효익이성위박절수요해결적문제。문중제출일충기우유전산법적RBF신경망락대선손진행계산，병차침대RBF신경망락적은함층여수출층상호독립화수출결과용역함입국부최소등결점，운용유전산법대정개망락진행우화，령외대유전산법진행료상응적개진。위료험증산법적가행성，문중이대경유전모일지구적67조선로위양본진행방진계산。실험결과증명료유전산법우화적RBF신경망락재계산정도화훈련속도상도유교대제고。
Lineloss rate is an important index reflecting the power network planning and design,producting operation and management. How to effectively reduce the power loss and improve the economic profit of enterprises have became an urgent problem. Present a RBF neural network based on genetic algorithm for line loss calculation and in view of the defects of RBF neural network hidden layer and out-put layer are independent of each other and the output is likely to fall into local minimum,apply genetic algorithm to optimize the whole network,in addition to the genetic algorithm for the corresponding improvements. In order to verify the feasibility of the algorithm,select 67 line in a certain area of Daqing oil field as samples for simulation and experimental results indicate that the the RBF neural network op-timized by genetic algorithm in the calculation precision and speed of train has increased greatly.