Research On Control Allocation With Constraint For Multi Effectors Aircrafts

To improve the flight performance and the safety of the aircraft, the advanced aerodynamiclayout aircraft has been widely developed in several military power countries. By configuringmulti-control-effectors, the ability of fault reconstruction of multi-effectors aircraft is improved. Theradar reflection area and key factors of some effectors are reduced at the same time. On the other hand,the aircraft has control redundancy. However, the flight control system has become much morecomplicated with the increasing of the configuration of the aircraft control surface than the traditionalaircraft. Thus, how to achieve the goal of redundant configuration of the multi-control-effector aircraftand obtain the optimal control with the restriction of the deflection combination is the most importantproblem. In this thesis, the constrained control allocation methods are studied for the multi-effectorredundancy aircraft. The main contents and results are as follows:Firstly, the control diagram of the multi-effectors aircraft control system with distributor isestablished. By introducing the control distributor, a multi-control-effector aircraft with" main loop+servo loop" cascade flight control structures is given. On this basis, the mathematical models ofaircraft and actuators are established, respectively. And then, the sliding mode control scheme ispresented based on nonlinear disturbance observer for multi-effectors aircraft. The obtained controlinput is considered as the desired virtual control signal which will be used in the followingconstrained control allocation.Secondly, the restricted space of the control surface is divided into multiple grids according tothe position constraint and rate constraint of actuators. The global temporary optimal control solutionis achieved through the ant colony algorithm. On the basis of the global temporary optimal solution,the global optimal solution is obtained via using the pattern search technique. The simulation resultsshow that the virtual control input can be assigned to the control surface properly by using thedeveloped control allocation method.Later, the piecewise linear control allocation scheme based on ant colony algorithm anddifferential evolution algorithm is developed for the allocation problem with nonlinear controleffectiveness of effectors. The segmentation sections are obtained by employing the ant colonyalgorithm to search the best combination cells of piecewise linear control effectiveness. The optimalsection coefficients are given through the differential evolution. Control surface deflection commandis calculated by the results of the ant colony algorithm and differential evolution. Simulation results show that the developed control allocation method is effective and the control requirement can beachieved.Following, the unstructured robust least-squares control allocation strategy is applied directly forcontrol allocation problem with interval matrix uncertain of the efficiency matrix. After that, theinterval control efficiency matrix is transferred into polytopic uncertain matrix. The robustleast-squares control allocation method using the linear matrix inequality (LMI) is developed on thebasis of polytopic uncertain control efficiency matrix. The position constraint and rate constraint ofactuators are taken into account at the same time. Simulation results show that the uncertain ofeffectiveness matrix can be better dealt with by applying the designed method.Finally, the robust model predictive constrained control allocation strategy is studied for thecontrol allocation of overactuated systems with polytopic uncertain dynamic of actuators. In thismethod, the discrete augmented system model of the actuator is established. Then, the robustpredictive control method is introduced for the control allocation scheme design. And the LMItechnique is used to solve the allocation problem. The simulation results show that the dynamiccontrol allocation object with uncertain model of actuator is achieved by considering physicalconstraint under the developed control allocation method.