Research On Multi-quadrotor Cooperative Control Technology Based On Model Predictive Control Algorithm

Due to the short flight distance and the limited load of a single quadrotor,its application in some scenarios is limited to a certain extent.Multi-quadrotor cooperative control can well solve the shortcomings of a single quadrotor.How to coordinate the cooperative control of multi-quadrotor has become a research hotspot.Firstly,the modeling process of the quadrotor is introduced.The first step introduces the composition and structure of the quadrotor and then establishes the relevant coordinate system and the mathematical model of the quadrotor briefly introduces the model predictive control and analyzes the model.The model is decoupled and linearized into linear time-varying state-space expressions of translational and rotational dynamics,and the nonlinear mathematical model of the quadrotor is expressed into pseudo-linear form by the State Dependent Coefficient(SDC).The state-space model which can be used for controller design is obtained,which lays a theoretical foundation for the design of the cooperative controller in the following paper.Secondly,to realize the function of the control system of quadrotor,this paper divides cooperative control problems into formation flight,obstacle avoidance,deployment,and reconstruction of regional research coverage:(1)To solve the formation tracking problem of quadrotor,a two-layer distributed model predictive control scheme with an obstacle avoidance function is proposed.The formation strategy of the quadrotor is designed by using the leader-follower formation method.The problem is divided into two parts: 1)the trajectory generation of the quadrotor;2)Find the relative position of the following quadrotor relative to the leader quadrotor and generate the reference trajectory satisfying the formation constraints.At the top of the control architecture,a distributed nonlinear model predictive controller was designed,and three cost functions were considered to keep the quadrotor in an ideal formation and to realize the function of obstacle avoidance and collision avoidance between the quadrotor.The reference trajectories satisfying the formation constraints were generated for the leading quadrotor and the following quadrotor.At the bottom,to stabilize the quadrotor and follow the reference trajectory generated by the top layer,a linear model prediction trajectory tracking controller was designed for the quadrotor dynamics.(2)After the multi-quadrotor system enters the working area,the solution to the problem of distributed coverage control subject to collision avoidance constraints is discussed.The whole working area was divided in real-time by the Voronoi algorithm,and each quadrotor was strictly kept in a Voronoi cell to meet the constraints.In addition,this paper designs the quadrotor to drive to the Chebyshev center of each Voronoi cell,which can satisfy the regional coverage and reduce the calculation amount.The quadrotor in the formation can perform cooperative tasks without a master-slave relationship.After any quadrotor joins,leaves,or fails,multi-quadrotor cooperation can still restore the optimal coverage state of the region and avoid collision with each other.Finally,based on the Matlab simulation platform,the proposed scheme is further verified.The experimental results show that the designed scheme has good control performance,which illustrates the feasibility and effectiveness of the proposed algorithm.