Research On Control Algorithm Of UAVs Formation

The UAV is widely used because of its small size,light weight and flexible mobility.However,with the poor airborne capability,it is impossible for a single drone to accomplish tasks which are more difficult.The UAVs formation flight technologies have been developed in this demand.In recent years,UAV technology has improved rapidly with lower costs,which promotes the UAVs formation control technology directly.As a common UAVs formation control algorithm,traditional PID control algorithm is simple and easy to be realized in projects.But parameters of the controller are fixed,when the dynamic changes emerge in the system,the control effect is often not satisfactory.Multi-agent adaptive consensus control is an intelligent control algorithm,which is often used for the UAVs formation control.Neural Network(NN)or Fuzzy Logic System(FLS)can be used in the algorithm to approximate the unknown function in the complex System.However,there are many research on linear systems in the existing algorithms,and when designing controllers,the disturbance term and time delay problem is often ignored,while the actual system is usually nonlinear and time-delay.Besides,for obtaining the desired approximation accuracy,these adaptive consensus methods often require a large number of adaptive parameters.Therefore,if these consensus methods are applied to practical engineering systems,it will result in heavy online computation burden.In formation control,path planning algorithm is also involved,which is usually divided into two types,global and local.Moreover,for the global path planning,the overall environment needs to be considered every step,so the calculation would be a lot and complex.While for the local path planning,only the nearby environment needs to be considered,so the calculation would be less,but it is easy to fall into the local optimal.Motivated by above discussion,this paper addresses the leader-following UAVs formation control algorithm.The main work and contributions of the thesis are as follows:Firstly,the UAVs formation control algorithm based on PID control is designed and implemented,then the algorithm is improved,and the fuzzy controller is adopted to realize the adaptive adjustment of PID control algorithm parameters.According to the simplified single UAV movement model,the relative motion relation of UAVs formation is analyzed,and the mathematical description of the relative movement of UAVs is presented.The system chooses first-order yaw angle retainer,first-order velocity retainer and second order height retainer as the UAV autopilot in formation,then the PID controller of every channel is designed,and engineering methods are obtained for setting the parameters of the PID controller.Finally,the PID controller is improved by a fuzzy controller,and the parameters of fuzzy PID controller are determined according to fuzzy rules.Secondly,this paper proposes an adaptive fuzzy consensus method of UAVs formation control algorithm.There is only one adaptive parameter in the algorithm,which greatly reducing the online computation costs of the control system.Lyapunov-Krasosvskii functional is introduced to compensate the unknown state delay of the system;and FLS is adopted to approximate unknown nonlinear dynamics and external interference,which reduce design difficulty of the nonlinear control system.Over all: graph theory is used to described communication topology relations of the UAVs formation system.Given all functions are known in the nonlinear multiple UAVs system,an UAVs formation controller is designed.Lyapunov-Krasosvskii functional is used to compensate the system delay and analyze the stability of the controller algorithm.Then the reliability of controller is proved.However,not all the system functions are known.So,the FLS is used to approximate the unknown function and design the UAVs formation control algorithm.Finally the stability of the control algorithm is analyzed and the UAVs formation control is achieved.Thirdly,an improved dynamic window path planning algorithm is proposed with a little computation,and it solves the problem that the dynamic window algorithm gets into the local optimal situation easily.The main contribution is that: the speed space(linear velocity and angular velocity)of UAVs formation is analyzed.First,speed space constraints of UAV movement and its equation is established.Then,the dynamic velocity window in the process of UAV movement is analyzed to get the velocity constraint set.What’s more,the evaluation function of dynamic window algorithm is improved and its effectiveness is analyzed.Finally,the algorithm is realized by programming,which verifies the path of the improved algorithm.