毕业设计（论文）-基于粒子群优化最小二乘支持向量机的短期风速预测.docVIP

基于粒子群优化最小二乘支持向量机的短期风速预测 摘 要 为了能够减少或消除风电开发并网带来的对电网的稳定性的不良影响，风电场风速的短期预测已经成为各个国家共同关注的问题。风电场风速的准确预测，对风电场的规划计划设计、大型风场中风电机组开停机计划的安排、保持电网的安全稳定性、提高经济效益和社会效益都有很重要的意义。本文的历史风速数据来自我校校史馆处的风速采集器，模拟风电场风速进行短期的风速预测。本文采用粒子群优化最小二乘支持向量机(PSO-LSSVM)方法对风电场进行短期风速预测。并与支持向量机(SVM)、最小二乘支持向量机(LS-SVM)进行分析对比，体现经过粒子群优化后预测准确度的优势。粒子群优化算法(PSO)分别对LS-SVM的超参数和核函数进行优化，从而使最小二乘支持向量机(LS-SVM)对短期风速预测的结果更加准确。本文的三种模型支持向量机(SVM)、最小二乘支持向量机(LS-SVM)、基于粒子群优化最小二乘支持向量机(PSO-LSSVM),利用实际数据对模型进行训练和测试，提前一步（即一个小时）对风速进行预测，并把三种模型的预测值与下一时刻的实际风速值进行比较，体现粒子群优化最小二乘支持向量机(PSO-LSSVM)模型的准确度高、收敛性好。为PSO-LSSVM模型实际运用提供理论支持。 关键词：风力发电、风速预测、粒子群优化算法、支持向量机、最小二乘支持向量机、PSO-LSSVM Particle swarm optimization based on LS-SVM prediction of short-term wind speed Abstract In order to reduce or eliminate the development of wind power grid stability of the grid caused by the adverse effects of short-term forecast wind speed has become a common concern of all countries. Accurate forecasts of wind speed, wind farm planning program on the design of large-scale wind turbine open stroke down to planned arrangements and to maintain security and stability of the grid and improve the economic and social benefits are very important significance. This historical wind speed data from the wind speed at my school History Museum collection, short-term simulated wind speed wind speed forecast. In this paper, particle swarm optimization support vector machine (PSO-LSSVM) method of short-term wind speed prediction. And with the support vector machine (SVM), least squares support vector machines (LS-SVM) for analysis and comparison of expression through particle swarm optimization after prediction accuracy advantage. Particle swarm optimization (PSO) on the LS-SVM, respectively, the hyper parameters and to optimize the kernel function, so that the least squares support vector machines (LS-SVM) the results of short-term wind speed forecasting more accurate. Three models of this paper support vector machine (SVM), least squares support vector machines (LS-SVM), based on pa