【摘要】 该文针对传统电力负荷预报方法的不足以及标准遣传算法用于人工神经网络时遇到的编码长度和表达精度之间的两难矛盾，提出了一种基于多分辨率遗传算法的神经网络，并将其应用在短期电力负荷预报系统中。运行实例表明，该预报方法具有预报精度高、训练速度快、鲁棒性和适应性强等特点。更多还原

【Abstract】 Aiming at the drawbacks of the conventional electric load forecasting methods and the dilemma be- tween encoding length and representing precision in standard GA,this paper proposes a multi -resolution GA based on artificial neural network and applies it in short term electric load forecasting system .The practice shows that the method has the superiority of high forecasting accuracy,fast traning speed.satisfied robustness and fitness.更多还原