Flight Control Actuator Test Method And Model Identification Based On Video

Flight control actuator is an important part of a missile and an UAV. An accurate model is the basis for design and simulation of the precise guidance control system. Recently the visual theory has been developed rapidly, so it is a trend to combine the image tracking and various technologies. This paper, which applies the visual technology to the actuator test to get the model of steering gear with no contact and without any change of the original system, mainly researches on the methods of actuator tracking and the methods of parameter identification. The main contents are as follows:Firstly, this paper introduces the composition and characteristics of virtual instrument. According to the flight control system needs, the paper designs the hardware and software of the test system.Secondly, this paper researches the methods of model identification, including least square method, recursive least square method, genetic algorithm, particle swarm optimization algorithm, a method which combines genetic algorithm and genetic algorithm(hybrid algorithm). These algorithms identify and simulate a given model. According to the tracking result, the hybrid algorithm works effectively. So it will be used as the identification algorithm.Then this paper studies the video tracking methods of actuator’s movement. The Mean Shift and template matching methods are studied in detail, and then these two methods are tested in three different scenarios to track a moving object. Compared with the Mean Shift, template matching is a better choice to track the moving object.Finally, the video capture and model identification are carried out in experiment of a certain steering gear. To get the rudder angle information by tracking the steering gear and identify the model parameters by using the hybrid algorithm. Thus we can get the optimal parameters of the steering gear.The research results will be helpful to the design of high performance flight control system, which lays foundation for hardware in the loop simulation.