Research On Low-Order Equivalent Systems In Aircraft Handling Qualities

Aircraft handling qualities are closely related to the aircraft flying safety, while theresearch on low-order equivalent systems is an important basis for the assessment on theaircraft handling qualities. Whether or not the low-order equivalent systems can wellmatch with the high-order systems in a limited frequency, will affect the analysis ofaircraft handling qualities. Consequently, it is very significant to investigate thelow-order equivalent systems.This research focuses on low-order equivalent systems and develops algorithms tooptimize or identify parameters in the low-order equivalent systems models. Therefore,the low-order equivalent systems are very closely related to system identification. Basedon the means of the low-order equivalent systems, this paper presents the fundamentalknowledge about aircraft and describes the principle, contents and methods of systemidentification. Then it mainly introduces the least squares method, the maximumlikelihood method as well as the frequency domain recognition algorithm-Fouriertransform, period diagram, and the average period diagram method.The principle of low-order equivalent systems in frequency-domain is the basis tomatch the low-order equivalent systems. Analysis in frequency-domain is a fundamentalmethod to obtain the parameters of the low-order equivalent systems, and widely used.It emphasizes the characteristics of systems in the main frequency band. This thesisanalyzes the low-order equivalent systems models, and mainly adopts two kinds ofoptimization algorithms to match the low-order equivalent systems. One is the patternsearch algorithm to optimize the parameters. Furthermore some examples are given.The other uses damping least squares method to optimize part of the parameters, andalso some simulation have been done. The method is improved through the conclusionfrom the relationship curve between the mismatch function and delay-time. Finally, thetwo algorithms are summarized, and the applicabilitiy of the two algorithms is analyzed.