基于混沌理论的鸡群改进算法及其在风电功率区间预测中的应用

APPLICATION OF CCSO IN WIND POWER INTERVAL PREDICTION

提出一种基于混沌鸡群优化和极限学习机(CCSO-ELM)的风电功率区间预测模型。首先,针对传统优化算法种群多样性低、易陷入局部极值的缺点,采用混沌理论改进鸡群算法可提高算法的寻优性能和效率。然后,针对传统代价函数未体现区间外点的偏离程度的问题,提出考虑区间外点相对偏差的评价指标。最后,以新评价指标作为适应度函数,以CCSO优化ELM的输出层权值,采用上下限估计的方法直接输出预测区间。通过实例进行仿真验证,结果表明所提方法在保证区间覆盖率的前提下,可有效降低平均带宽、提高区间预测质量。

wind power interval prediction model based on chaotic chicken swarm optimization and extreme learning machine(CCSOELM)is proposed.First,to avoid the disadvantages of low population diversity and easy to fall into local minima of the traditional optimization,chaos theory is adopted in the chicken swarm optimization,which improves its performance and efficiency.In addition,traditional cost function does not reflect the deviation degree of off-interval points,then an evaluation index considering the relative deviation of off-interval points is proposed in this paper.Finally,the new cost function is taken as the fitness function,the output layer weight of ELM is optimized by CCSO,the lower upper bound estimation(LUBE)is adopted to output the prediction interval directly.The simulation result shows that the proposed method can effectively reduce the average bandwidth and improve the quality of interval prediction under the premise of guaranteeing the interval coverage.