基于双变权缓冲算子的多变量灰色电量预测

Multi- variable grey prediction based on double-variable weight buffer operator

针对多变量灰色模型预测电力系统用电量的局限性问题,笔者根据变量与预测精度之间的关系,借助双变权缓冲算子对初始变量进行预处理,建立了基于双变权缓冲算子的多变量灰色模型,采用粒子群算法配置该模型的全局最优参数,并将该模型用于算例分析。分析结果表明,基于双变权缓冲算子的多变量灰色模型在中长期电量预测中,其预测精度和稳定性能优于传统多变量模型。

Aiming at the limit of multivariable grey model in forecasting power system electricity demand,the author pretreated the initial variables based on double variable weight buffer operator,according to the relation between variables and prediction accuracy,established the multivariable grey model based on double variable weight buffer operator,and analyzed the example by the model,of which the global optimum parameters were allocated by particle swarm optimization. The analysis result shows that the prediction accuracy and stability performance of the multivariable grey model based on double variable weight buffer operator are superior to thoses of traditional multivariable models in medium and long- term load forecasting.