Co-DMPC-based chassis multi-agent system (MAS) cooperative control method for autonomous vehicles, controller, and storage medium

A cooperative distributed model predictive control (Co-DMPC)-based chassis multi-agent system (MAS) cooperative control method for autonomous vehicles, comprising:

• S1:
• building a two-degree-of-freedom vehicle model with four-wheel steering and direct yaw moment control (DYC);
• S2:
• establishing a distributed state-space equation, wherein a centralized state-space equation of a vehicle is divided into two agents, namely, a steering agent and a DYC agent according to different control variables;
• as for the steering agent, the distributed state-space equation has control variables of a front wheel steering angle δf and a rear wheel steering angle δr and state variables of a lateral velocity vy and a vehicle lateral displacement Y;
• as for the DYC agent, the distributed state-space equation has a control variable of a direct yaw moment Mz and tracked state variables of a yaw rate γ and a yaw angle φ S3:
• building a cooperative control framework, comprising:
• an input layer, wherein the two-degree-of-freedom vehicle model computes a reference yaw rate γref and a reference lateral velocity vref according to deviation between a current state of the vehicle and a target path;
• a cooperative control layer, wherein a centralized chassis control system is decomposed into a distributed control system consisting of a plurality of controllers, to achieve tracking of expected targets;
• through the distributed state-space equation established in the S2, a Co-DMPC algorithm is employed to select a cost function considering global performance;
• an iterative update method is used in a unit sampling time, enabling agents to independently compute local optimization and exchange computation results multiple times, and an entire chassis multi-agent system obtains a global optimal solution;
• a moment distribution layer, wherein driving/braking torques of four wheels are optimally distributed based on a total driving force computed by a proportional-integral-derivative (PID) speed controller and a direct yaw moment computed by the cooperative control layer as constraints;
• and an actuation layer, wherein the control variables are applied to the four wheels through a steering actuator and a hub motor to achieve accurate control of the vehicle;
• S4:
• establishing an information transmission protocol, wherein a same prediction time domain Np and a same control time domain Nc are used in all the agents;
• as for an agent i, in a control time domain [t, t+Nc], a control variable at a time point t+k is predicted at a time point t as ui(k|t), and following three sets of control trajectories are defined:
• uip(k|t):
• predicted control trajectory;
• ui*(k|t):
• optimal control trajectory;
• uia(k|t):
• assumed control trajectory;
• wherein uip(k|t) represents a control variable at a future time point predicted based on the current state and a control variable at a previous time point;
• ui*(k|t) represents an optimal solution obtained after solving local optimization problems;
• uia(k|t) represents an assumed control variable transmitted by the agent i to a neighbor agent and is obtained by shifting a predicted control variable by one sampling period;
• in a prediction time domain [t, t+Np], a state variable at the time point t+k is predicted at the time point t as xi(k|t), and following three sets of state trajectories are defined:
• xip(k|t):
• predicted state trajectory;
• xi*(k|t):
• optimal state trajectory;
• xia(k|t):
• assumed state trajectory;
• the agent i uses, at the time point t, a predicted state at the future time point and an assumed state xjp(k|t) and an assumed control uja(k|t) from the neighbor agent to solve a distributed optimization problem;
• when k=0, xip(0|t)=xi(t) and xjp(0|t)=xj(t), wherein xi(t) and xj(t) are directly measured current states of the vehicle;
• after obtaining an optimal state xi*(k|t) and an optimal control ui*(k|t), a first ui*(0|t) of a control trajectory is applied to actual control;
• a single-step recursion is performed on the optimal state trajectory and the optimal control trajectory to obtain assumed trajectories as follows:
• uia(k−1|t+1)=ui*(k|t), k∈[1, Nc−1]
• xia(k−1|t+1)=xi*(k|t), k∈[1,Np] (1)
• wherein a length of the control tr to complete a last term, uia(Nc−1|t+1)=0;
• and a state of the vehicle not under control at a next time point is xia(Np|t+1)=Axia(Np−1|t+1)=Axi*(Np|t);
• and S5:
• controlling the four wheels through the steering actuator and the hub motor using the control variables.