【摘要】 负荷短期预测是电力系统运行和调度每年的重要工作,尤其在市场环境下负荷短期预测更显重要。对于电力系统短期负荷的随机性和突变性的特点,提出了应用遗传算法(GA)和模拟退火(SA)优化的Elman神经网络的短期负荷预测模型。其特点是模型简单、运算效率高,并具有较好的全局最优性能,从而很好地克服了传统BP算法容易陷入局部极小点的缺陷。利用改进Elman网络的良好学习能力,同时利用遗传和模拟退火优化算法对Elman动态递归网络的前馈和反馈值进行优化,实现全局最优的拟合结果。比较了Elman网络和BP网络结构的建模效果,仿真实验证明了利用遗传和模拟退火算法优化的Elman神经网络具有动态特性好、逼近速度快、精度高等特点,说明Elman网络是一种新颖、可靠的负荷预测方法。更多还原

【Abstract】 A method for short-term load forecasting based on genetic algorithm(GA) and simulated annealing(SA) optimized Elman neural networks is proposed.An Elman neural network based model is proposed to construct a basic forecasting model for short-term load of power systems,and a hybrid algorithm of genetic algorithm(GA) and simulated annealing(SA) is introduced in the forecasting model for obtaining some coefficients changed by some factors leading to the load change of power systems.Some of these factors always are stochastic.The proposed model is simple,more accurate of forecasting results,and it is easier to obtain a global optimal solution in the computation process.更多还原