Lower limb motion recognition method based on improved sparrow search algorithm optimized KELM

The invention provides a lower limb motion recognition method based on an improved sparrow search algorithm optimization KELM, and the method comprises the steps: carrying out the lower limb motion recognition of an electromyographic signal after collection and feature extraction through employing an improved sparrow search algorithm optimization kernel extreme learning machine; the improved sparrow search algorithm is used for optimizing a regularization coefficient C and an optimal kernel parameter sigma of the KELM neural network, constructing a KELM optimal model, using Tent chaotic mapping to replace a random generation mode in SSA to initialize a sparrow population, and using a warning person number decreasing strategy to adjust the number of warning persons in an iteration process. The method has the advantages of high calculation speed, good stability, high accuracy and the like. 本发明提出一种基于改进麻雀搜索算法优化KELM的下肢运动识别方法，对采集与特征提取后的肌电信号，采用改进麻雀搜索算法优化核极限学习机，进行下肢运动识别；所述改进麻雀搜索算法用于对KELM神经网络的正则化系数C和最优核参数σ进行优化，构造KELM最