基于粒子群算法和BP神经网络改进的灰色电力负荷预测研究

灰色预测模型被广泛运用于电力负荷预测中，取得了较好的效果，但是灰色预测模型在实际应用中的缺点和局限性导致其预测精度有待提高，存在改进的必要。本文对于灰色预测模型的改进，分别从优化初值和改进模型等方面进行，从而提高普通灰色GM（1，1）模型的预测精度。对初值的处理可以削弱异常值的影响，强化趋势，从而避免由于初值选择不当而造成预测误差。本文中对模型的改进主要通过建立等维新息预测模型、灰色粒子群组合预测模型和灰色BP神经网络组合预测模型来实现。通过这些对灰色预测模型的修正和改进，进一步提高了灰色预测模型的适用性．最大限唐妯提高了灰乍．GM（1，1）模型的预测精唐．

Grey Power Load Forecasting Based on Particle Swarm Optimization and BP Neural Network

Grey forecasting model is applied to power load forecasting, but prediction accuracy of grey forecasting model needs improvements due to defects and limitations in practical application. Grey forecasting model improvements are from initial value optimization and model improvement to improve prediction accuracy of general grey GM （ 1,1 ） model in the presented work. Initial value processing reduces abnormal value impact, avoiding forecast errors due to improper initial value. The model improvements are through equal-dimension and new-information forecasting model, grey particle swarm ensemble forecasting model and grey BP neural network ensemble forecasting model. Grey forecasting model improvements enhance applicability of grey forecasting model and prediction accuracy of grey GM （ 1,1 ） model.