基于改进秃鹰算法优化极限学习机的谐波发射水平估计

Harmonic emission level estimation method based on an improved bald eagle search optimized extreme learning machine

针对目前电力系统谐波发射水平难以直接测量的问题,提出了一种基于改进秃鹰算法(improved bald eagle search,IBES)优化极限学习机(extreme learning machine,ELM)的谐波发射水平估计方法。首先,在传统秃鹰搜索算法中引入Tent混沌映射和柯西变异算子,利用IBES算法对ELM模型的输入权重和阈值进行寻优。其次,输入公共连接点(point of common coupling,PCC)处谐波电压和谐波电流,代入IBES-ELM模型,估计用户侧和系统侧谐波发射水平。最后进行仿真和工程实例分析,并与其他算法的估计结果进行对比。结果表明,所提IBES-ELM方法估计精度优于长短期记忆网络(long short-term memory,LSTM)、卷积神经网络(convolution neural network,CNN)、反向传播神经网络(back propagation neural network,BP)和CNN-LSTM算法模型,验证了该方法的有效性和稳定性。

A harmonic emission level estimation method based on an improved bald eagle search(IBES)optimized extreme learning machine(ELM)is proposed to address the problem that it is difficult to measure that level directly.First,the Tent chaotic mapping and the Cauchy variant operator are introduced into the traditional bald eagle search algorithm,and the input weights and thresholds of the ELM model are optimized using the IBES algorithm.Second,the harmonic voltage and current at the point of common coupling(PCC)are input and substituted into the IBES-ELM model to estimate the customer-side and system-side harmonic emission levels.Finally,simulations and engineering examples are analyzed and the estimation results are compared with those of other algorithms.The results show that the estimation accuracy of the proposed IBES-ELM method is better than that of long short-term memory(LSTM),convolution neural network(CNN),the back propagation neural network(BP)and CNN-LSTM algorithm models.This verifies the effectiveness and stability of the method.