论文部分内容阅读
为了改善传统的电价预测灰色模型GM(1,1)的预测精度,提出一种内变量参数辨识的电价预测模型——PSOGM(1,1)模型.首先采用灰色微分方程建立模型内变量(发展系数、灰作用量、背景值权重系数、边值)与预测值之间的非线性内涵表达式,然后采用粒子群算法(PSO)对内变量参数进行辨识,得到问题的最优解,建立PSOGM(1,1)模型.与GM(1,1)模型相比较,PSOGM(1,1)模型具有较快的收敛速度和更好的预测精度.对北欧NORDPOOL电力市场历史电价数据的分析实验表明,PSOGM(1,1)模型的短期电价平均预测精度为94%,较已有的几种典型改进GM(1,1)模型预测精度提高了1%~3%.
In order to improve the prediction accuracy of the traditional gray model GM (1,1), a model of PSOGM (1,1), which is a model for predicting the internal variable, is proposed.It is first established by using the gray differential equation Coefficient, the amount of ash, the background value weight coefficient, the value of the boundary value) and the predicted value of non-linear connotation expression, and then use Particle Swarm Optimization (PSO) to identify the internal variable parameters, the optimal solution to the problem, the establishment of PSOGM (1,1) model.Compared with the GM (1,1) model, the PSOGM (1,1) model has faster convergence rate and better prediction accuracy.Experiments on the historical price data of the Nordic NORDPOOL electricity market . The average prediction accuracy of short-term electricity price of PSOGM (1,1) model is 94%, which is 1% -3% higher than that of several typical improved GM (1,1) models.