Research On Modeling And Control Of Mini Unmanned Helicopter Based On Frequency Identification

Miniature unmanned helicopter (MUH) has lots of advantages, such as small size, small weight, strong ability to remain in concealment, good flexibility, easy to hover, easy to achieve aggressive flight and so on. It is widely used in both military and civil aviation and being studied by many institutions and organizations all over the world in recent years.The key technologies for MUH include research for MUH modeling and autonomous flight control. The main work of this dissertation is giving a systematic and complete design for MUH,including the structure of theoretical prototype helicopter, methods for solving difficulties in engineering practice, selection and design for different kinds of experiment setups, a general method for MUH modeling based on frequency identification, design and realization of algorithms for autonomous flight control, and further improvement for these algorithms. The key parts of this dissertation are research on MUH modeling and design for autonomous flight controller. We believe the outcome of this dissertation is very important and meaningful for further development of MUH and design for autonomous flight control machine.The main work of this dissertation is as follow:In Chapter 1, we give the background and purposes of this dissertation, also including a general introduction of current progress in MUH. Then we give a brief introduction to technologies of MUH modeling and flight control. After that we end Chapter l by using a frame chart to show the whole contents of this dissertation.In Chapter 2, we mainly introduce the MUH System and some related key technologies. First we introduce different modules with different functions in this system, then we discuss some technical difficulties arising in engineering practice, such as sensor data fusion and so on.At the same time, we also give solutions to each of these difficulties. Finally in this Chapter, we give a very specific explanation to the testbed of ground flight simulation.In Chapter 3, we are aiming to construct the reduced model of MUH in hover. For better understanding of this chapter, we give a very specific analysis for the dynamic characteristics of MUH. Based on the analysed results, we carry on a very thorough trim computing process under the condition that the MUH stays in hover. And later, the outdoor flight experiment confirms the correctness of our theoretically computing results. By now, we can use the 6-DOF rigid body equations of motion to establish a full state nonlinear equation for MUH. Further, by taking a small disturbance linearization method, we can deduce a small disturbance linear differential equation of motion for the MUH of constant motion. At last, according to these characters held by the MUH in hover, we propose some reduced conditions and get a linear model for the MUH in hover. We believe this model means a lot to autonomous hover control.In Chapter 4, we proposed a modeling method based on parameter identification for a miniature unmanned helicopter. This method combines the advantages of mechanism modeling and system identification. By using this method and strict mechanism derivation, we can construct a parameterized dynamical angle model of roll and pitch channel for a miniature unmanned helicopter. Then, we get the dynamical parameterized transfer function of MUH by taking a frequency identification technique, which is based on partial coherence analysis. Till now, we can take both mechanism model and identification results together into consideration, so as to obtain the key parameters of a certain MUH. At last, this model is verified by comparing the model output with the output collected during flight test. This method is very feasible, and can be benefit for modeling research of other miniature unmanned helicopters.In Chapter 5, we are aiming to design a autonomous flight control system based on the model of roll and pitch angle dynamic we've got. This can be also regarded as an applicability validation for this model in real flight control system. First of all, by taking a PID control strategy based on cascade multi-loop, we achieve the autonomous hover control of MUH in a condition that there isn't any disturbance from wind. Further, using a PID hybrid control strategy based on single nerve cell, we improve velocity closed loop and achieve the autonomous hover control of MUH with disturbance from 6m/s gust. At last, we briefly discuss the situation when flying at a small speed and carry out an experiment, which also proves that the reduced hover model built in chapter 3 and chapter 4 is also applicable when flying at a small speed.In Chapter 6, for unmanned helicopter which exhibits a complex and nonlinear dynamic behavior, we propose an attitude control method based on adaptive output feedback. First, we assume that the controlled system satisfies the output feedback linearization conditions, then we design the controller by considering the quasi-linear model of the system as the diffeomorphism of the system. As for the errors produced by nonlinear, uncertainty and distanbance, they can be compensated by adaptive neural networks. What'more, we can prove the boundedness of tracking errors and weight errors by Lyapunov stability theorm. Also, we apply this method in tracking and control simulation for unmanned helicopter, and the simulation results show that the system designed in this method can can possess strong robustness and excellent tracking performance. In Chapter 7, we summarize all the research work of this dissertation and propose some further work for deep research.