基于IBAS-BP神经网络的风功率预测

Wind Power Prediction Based on IBAS-BP Neural Network

风电功率预测精度对风电场调度有着重要意义。针对目前预测方法中模型预测精度不高且训练速度较慢等问题,提出一种改进天牛须算法优化神经网络参数的短期风电功率预测模型。采用群体寻优的方式改进BAS算法,并在信息选取利用时提出精英个体概念,对现有精英个体继续进行寻优,从而改善了原始模型在权重更新的过程中易出现局部极值的问题。

Wind power prediction accuracy is of great significance to wind farm schedule. Aiming at the problems of low prediction accuracy and slow training speed in current prediction methods, a short-term wind power prediction model based on improved beetle antennae algorithm to optimize neural network parameters is proposed. The BAS algorithm is improved by using the method of group optimization, and the concept of elite individual is proposed in the process of information selection and utilization. The existing elite individuals continue to be optimized, thus improving the problem that the original model is prone to local extremum in the process of weight update.