改进鲸鱼优化算法的电力负荷短期预测模型

Power load short-term forecasting model based on improved whale optimization algorithm

鲸鱼优化算法是一种新兴的智能优化算法,针对其全局搜索能力差,易陷入局部最优等问题,提出了对标准鲸鱼优化算法改进的办法,然后对BP神经网络进行优化,得到一种新型的电力负荷短期预测模型(LWOABP),并通过收集到的电力负荷数据数据对本文建立的预测模型进行训练验证。研究结果表明,改进后的鲸鱼优化算法的预测模型的平均误差为:0.0181,标准鲸鱼算法优化BP神经网络得到的预测模型平均误差为:0.0275,其中前者的最大误差为:0.0272。综上所述,在对传统鲸鱼算法优化后,对于全局搜索能力和收敛性能都有显著的改善;且本文建立的新型电力负荷短期预测模型不仅收敛速度更快,预测结果也更加准确。

Whale algorithm is a new intelligent optimization algorithm.Aiming at its poor global search ability and easy to fall into local optimization,this paper puts forward an improved method for the standard whale optimization algorithm,and then optimizes the BP neural network to get a new power load short-term prediction model(lwoabp),Through the collected power load data,the prediction model established in this paper is trained and verified.The results show that the average error of the prediction model of the improved WOA and WOA are 0.0181 and 0.0275,of which the maximum error of the former is 0.0272.To sum up,after optimizing the traditional whale algorithm,the global search ability and convergence performance are significantly improved;The new power load short-term forecasting model established in this paper not only converges faster,but also the forecasting results are more accurate.