Infrared spectrum multivariate correction detection method and system for SF6 gas leakage in closed space

The invention discloses an infrared spectrum multivariate correction detection method and system for SF6 gas leakage in a closed space, belongs to the technical field of SF6 gas leakage detection in the closed space, and solves the problems of complex data processing algorithm and long response time in SF6 gas leakage detection in a closed space scene. The method comprises the following steps: introducing an atmospheric background interference infrared spectrum in an electric power typical scene, completing SF6 infrared spectrum radiation correction and spectral feature extraction, removing background noise interference and improving a signal-to-noise ratio through multivariate regression correction training, and optimizing the wavelength by adopting a spectral wavelength based on a variable interval after the training is completed to obtain an SF6 gas multivariate correction coefficient; according to the method, wavelength optimization is carried out on the input to-be-detected SF6 gas infrared spectrum, the