一种改进果蝇算法优化神经网络短期负荷预测模型  

An improved fruit fly optimization algorithm to optimize neural network short term load forecasting model

作　　者：王亚琴 王耀力[1] 王力波 常青[1] Wang Yaqin;Wang Yaoli;Wang Libo;Chang Qing(School of Information Engineering,Taiyuan University of Technology,Jinzhong 030600,Shanxi,China)

机构地区：[1]太原理工大学信息与工程学院,山西晋中030600

出　　处：《电测与仪表》2018年第22期13-18,24,共7页Electrical Measurement & Instrumentation

基　　金：全国工程专业学位研究生教育指导委员会立项项目(2016-ZX-095)

摘　　要：针对基本BP神经网络用于电力负荷短期预测时存在预测精度较低等问题,提出一种改进的直连BP神经网络初始连接权值果蝇优化算法(IFOA),用于优化负荷预测模型。首先,采用在BP神经网络中加入从输入到输出连接的网络(BPNN-DIOC,直连BP神经网络)构建负荷预测模型,以减少隐含层所需的神经元个数,降低网络在训练过程中调整的参数个数,提高负荷预测精度。此外,由于BP算法随机初始化网络参数使得算法收敛速度慢、容易陷入局部极值,提出一种改进的果蝇优化算法(IFOA)用于优化神经网络的初始连接权值和阈值,以实现全局优化。然后,综合IFOA和BPNN-DIOC,构建了基于IFOA优化的BPNN-DIOC负荷预测模型。最后,文中以AEMO中新南威尔士州2015年9月份的数据为例进行了仿真验证,IFOA-BPNN-DIOC模型的预测平均绝对误差百分比为0. 635 7%,均方根误差为0. 011 8,并将该结果与文中其它模型的负荷预测结果进行比较。结果表明,文中负荷预测模型是一种更加有效的短期负荷预测方法。Aiming at the low prediction accuracy of the basic BP neural network for short-term load forecasting, an improved fruit fly optimization algorithm （IFOA） for initial connection weight of direct connected BP neural network to optimize load forecasting model. Firstly, the load forecasting model is built by adding connections from input-to-output based on BP neural network （ BPNN-DIOC, direct BP neural network）, to reduce the number of neurons required by the hidden layer, reduce the number of parameters that the network adjusts during training and improve the load forecasting accuracy. In addition, since the BP algorithm randomly initializes the network parameters to make the algorithm converge slowly and easily fall into the local extreme, an improved fruit fly optimization algorithm （IFOA） was p to optimize the initial connection weights and thresholds of the neural network to achieve global optimization. Then, combining IFOA and BPNN-DIOC, a load forecasting model based on IFOA optimized BPNN-DIOC was constructed. Finally, a simulation verification was conducted using the data of AEMO in New South Wales on September 2015 as example, the average absolute error percentage of IFOA-BPNN-DIOC model was 0. 635 7% and the root mean square error was 0.0118, and the result was compared with the prediction results of other models in this paper. The results showed that this model is a more effective method for short-term load forecasting.

关 键 词：BP神经网络 改进果蝇优化算法 输入到输出连接 负荷预测模型 预测精度 

分 类 号：TM93[电气工程—电力电子与电力传动]