LPV Flight Control Of Multi-Rotor UAV

In recent years, multi-rotor UAV(Unmanned Aerial Vehicle) has become a popu-lar platform in the UAV flight,due to the simple mechanical structure, low mainte-nance cost and its ability of hovering,vertical take-off and landing. At present, most multi-rotor UAV platforms are quadrotors, but they cannot meet the requirement of altitude, load and anti-interference performance for outdoor flight. In addition, in the research on multi-rotor UAV flight control, most of the established mathematical models based on the hover operating point and ignore air resistance, the center of gravity position, made them more simplified, but they can’t reflect the change of the model when the forward speed increases. Therefore, this article design an octocopter UAV platform, and established a nonlinear mathematical model considering the actual flight status, which changes as forward speed increases. Linearize the model in each working point and get a LPV model, design controller, complete gain scheduling and simulation validation. Design and develop the flight control and task scheduling soft-ware system and complete experimental verification. The main work and contribu-tions are summarized as follows:First, design a complete and reliable platform of octocopter UAV and build a set of motor test platform according to the requirement of actual project, mathematical modeling and engineering test. The octocopter UAV platform consists of core hard-ware and software system, the test platform could detect characteristics of blades and motors and measure the important input and output data for modeling, and send them for remote display and storage in ground control station, the data can be used to estab-lish the input and output model of the motor system.Then, establish the mathematical model of octocopter UAV, based on the relevant theory of aerodynamic and experiment data, build the mathematical model of motors, blades and fuselage, complete the whole nonlinear system model, and linearize the model on different forward speed operating points within an advance ratio range, then build the LPV model. Analyze the differences when the forward speed changes. De-sign the cascade PID controller of changing parameters for the LPV model, to realize the precise control of attitude, speed, complete the gain scheduling and simulation validation and compare to the controller which is not scheduled.In addition, in order to complete the complex flight mission better and reduce the complexity of operator in experimental flight at the same time, several flight patterns and task scheduling strategies are defined and developed. Implement the autonomous flight control in simulation first, then implement the whole software framework from the hardware drive and motion control to the top flight task scheduling, complete a reliable and integrated software system.Finally, as the hardware platform and software system has been established, we can proceed semi-physical simulation on the platform established in DSPACE, which is a real-time simulation system for the calculation of mathematical model, validation of logic and strategy, and the test of control effect. Design several flight experiments on basis of GPS/INS integrated navigation, examine the effect of the various flight mode, flight control and airline flight task, at last, the octocopter UAV was used as an experiment prototype in a project program and realized autonomous flight at the alti-tude of 800 m, and it could come back successfully.