The invention belongs to the field of helicopter model identification, and provides a
flight experiment data rapid
frequency domain identification method suitable for a
telex flight control helicopter, which comprises the following steps of: 1, eliminating and correcting a data
outlier, performing low-pass filtering, correcting a sensor position, checking data compatibility and reconstructing data; 2, converting the
flight experiment data into a
frequency domain, constructing an identification model, and adopting a completely linearized
state space model; 3, fast initial value
estimation is carried out, and initial value
estimation theta of the to-be-identified parameters is obtained; 4, model structure identification is carried out, and before the next step of identification is carried out, the parameters are selected and used; And 5, outputting an error in a fast
frequency domain. According to the method, the accelerated optimization
algorithm is designed according to the characteristic that the convergence rates of the parameters to be identified are different, so that the calculation efficiency is further improved, the relatively detailed description of the
frequency spectrum is realized, and meanwhile, the calculation amount is ensured.