摘要
采用支持向量机预测光伏电池功率时,其预测精确度与惩罚参数c和核函数参数g的取值有关。提出一种改进粒子群算法优化支持向量机并预测光伏电池输出功率。改进粒子群算法通过控制粒子的初始速度和位移并限制粒子速度的最大值和最小值,保证粒子不超出规定的速度边界,且具有自主调节功能;增加自适应粒子变异,避免粒子群算法陷入局部最优而导致搜索精度下降。结果表明,采用改进粒子群算法优化支持向量机预测光伏电池输出功率的精度更高,误差更小。
When the support vector machine is used to predict the power of the photovoltaic cell,the prediction accuracy is related to the values of penalty parameter c and kernel function parameter g.An improved particle swarm algorithm was proposed to optimize the support vector machine in prediction of the output power of the photovoltaic cell.The improved particle swarm controlled the initial velocity and displacement of the particles and limited their maximal and minimal velocities to ensure that the particles would not exceed the specified velocity boundary,and self regulation function was maintained.Adaptive particle variation was enhanced to prevent partial optimization of the particle swarm algorithm,thus avoiding decline of searching precision.The results indicated that the improved particle swarm optimization support vector machine could achieve higher precision of prediction of the output power of the photovoltaic cells,with less error.
作者
汪友明
穆恒星
徐国宁
Wang Youming;Mu Hengxing;Xu Guoning(College of Automation,Xi'an University of Posts and Telecommunications,Xi'an Shaanxi 710121,China;Research Institute of Opto-electronics,Chinese Academy of Sciences,Beijing 100094,China)
出处
《电气自动化》
2019年第3期63-65,91,共4页
Electrical Automation
基金
北京市自然科学基金资助项目(4164106)
关键词
支持向量机
粒子群算法
功率预测
仿真
光伏发电
support vector machine
particle swarm optimization
power forecasting
simulation
photovoltaic power generation
