“According to a report by a technology media reporter, in order to address the issue of environmental noise affecting the speech system of exoskeleton equipment, a speech feature based on discrete orthogonal Stockwell transform and gamma pass filter frequency cepstral coefficients is proposed. Combined with discrete path transform to represent the time-domain information of speech signal energy and zero crossing rate, a mixed feature is formed. Under low signal-to-noise ratio conditions, an improved correlation fast filtering feature selection algorithm is used to obtain the optimal feature subset, which is then applied to the speech system of exoskeleton device control commands. The experimental results show that this method is more robust and robust at low signal-to-noise ratios. Under pink noise with zero signal-to-noise ratio, the recognition rate of Mel frequency cepstral coefficients is improved by about 20% compared to traditional methods.”