基于改进麻雀算法的短期负荷预测研究

Short Term Load Forecasting Based on Improved Sparrow Algorithm

摘要: 针对麻雀算法(SSA)早熟、易陷入局部最优解的问题，采用混沌策略反向学习初始化种群代替随机初始化种群，增强种群多样性；同时引入黄金正弦因子更新发现者位置以平衡局部和全局搜索能力；引入自适应t分布改进跟随者位置。与粒子群算法、灰狼算法作对比，在测试函数上验证改进麻雀算法的优越性。采用改进后的麻雀算法优化BP神经网络模型(GSASSA-BP)对短期负荷问题进行预测，结果表明，提出的GSASSA-BP模型在短期负荷问题上预测效果更好。

Abstract: In order to solve the problem of being premature and easy to fall into the local optimal solution of sparrow algorithm, chaos strategy is used to replace random initialization population to enhance the diversity of population. At the same time, the golden sine factor is introduced to update the location of the discoverer to balance the local and global search ability. Adaptive t-distribution is introduced to improve the follower position. Compared with particle swarm optimization algorithm and gray wolf algorithm, the superiority of the improved sparrow algorithm is verified in the test function. The improved sparrow algorithm is used to optimize the BP neural network model to predict the short-term load problem. The results show that the proposed GSASSA-BP model has better prediction effect on the short-term load problem.