The decision making algorithm based on inverse-design method and its application in the UAV autonomous flight control system design

Abstract

For Unmanned Air Vehicles (UAV) to flight autonomously, the decision making system must be developed to replace the remote control by the pilot. To shorten the learning time and increase the real-time response, the inverse-design (ID) method is presented to calculate the control input decision inversely by using the desired output and the predictor-corrector algorithm which uses the proportional-integral predicted initial value to compute the control inputs. In the mission description process, the quasi-uniform B-spline is used to generate the desired output by a succession of control points. The genetic algorithm (GA) is also implemented to compare the computing time and the development complexity with the ID method. Simulation shows the ID method based on the predictor-corrector algorithm can be validly and efficiency used in the autonomous flight control system decision making of the UAV, and the ID method is preponderant in online computing, development and computing time reducing than the GA. The analysis and ID method presented in this paper will produce a direct benefit in relation to the decision making method and the design of the UAV autonomous flight control system.