Design On Autonomous Flight Control System Of Unmanned Aerial Vehicle

UAV gets rid of the bound of human tolerance, can implent high-risk missions in the uncentian object and environment, and has the advantages that manned aircraft cann't compared with, so the research of UAV has been put great attentation by various countries. This article studies autonomous flight control technology of UAV in the following areasIn the first part, a six freedom degrees nonlinear mathematical model of UAV is established, according to the motion characteristics of X-series demonstrators. Furthermore, according to the object-oriented modular design thinking, UAV dynamic models of the movement are established respectively, and these parts are organically combined in the simulink. Eventually, the simulink model of UAV is got, which can reflect the flight dynamic characteristics of UAV.In the second part, the autonomous flight control law algorithms of UAV is researched in both longitudinal and lateral aspects. The modified dynamic inversion and based on functional link neural network are proposed, against the lack of dynamic inversion and trajectory linearization control algorithm, and the optimal generalized predictive control algorithm is applied to the lateral control of UAV. The simulation shows that the controllers, which integrate these control algorithms, have good performance.In the third part, in order to maximize the autonomy of UAV control, an autonomous flight multi-Agents hierarchical control structure is proposed, according to Professor Saridis'hierarchical structure thinking and the idea of humanoid intelligent control, and the structure and function of the task management Agent, decision-making coordination Agent and implementation Agent are further designed. Then the work logic of UAV autonomous flight control is simulinked, using graphics software stateflow based on finite state machine, and the communication problem between Agents is studied.The fourth part, the three diemsional visual simulation software is developed, using the virtual reality simulation software Ceator and Vega, and binding VC++ 6.0. The sofoware can vividly demonstrate the whole process of UAV autonomous flight.