Controller Design Of Multi-propeller Airship By Using Nonlinear PID And Polytopic LPV Robust Approaches

Traditional airship has a streamlined shape and use thrust andaerodynamic control surfaces for operation, thus it has small forward dragcoefficient and simple control configuration. However, the lateralanti-wind ability is weak, and the control ability of rudder is decreasedobviously with the decreasing of airspeed. A multi-propeller airship withspecial shape is introduced in this paper, it has higher lateral anti-windcapability, at the same time, the whole vectored thrust configuration canimprove the control ability in low speed condition. The different controlstrategies are proposed for position tracking of the multi-propeller airship,thus different control systems are developed in this paper. The paper isorganized as follows:The nonlinear mathematical model of airship is established based onthe Newton-Euler method, then the nonlinear model is linearized under thecondition of small perturbation theory, and then the linearized model isdecoupled into the longitudinal and lateral equations respectively. All themodels above are used in the controller design process.The PID controller is designed based on the nonlinear model withinner and outer loop separated. The redundant configuration of actuators istaken into consideration, the dynamic inverse is used to achieve theallocation of actuators. The direct and indirect position control methods areproposed, and the corresponding simulation results are given. Simulationresults show that both of the two methods can achieve good positiontracking, but the control accuracy, response time and energy consumptionare different. The polytopic LPV model is established based on the longitudinalLPV model by using the HOSVD method. Then the state-feedbackcontroller satisfying H∞performance is designed in each vertex and arobust variable gain H∞controller is synthesized. This method hasadvantages over the traditional gain schedule method from the respects ofstability theory and switch process. The effectiveness of designedcontroller is illustrated in velocity tracking simulation.Some results of the designed controller have been applied into the realcontrol system, and validated by the flight test.