The application discloses a trajectory tracking multi-
algorithm fusion control method considering riding comfort. A vehicle
yaw single-track dynamics model is established; based on an LSTM neural
network model, vehicle motion state parameters are taken as model input information, and prevention of passenger
motion sickness is taken as a target to
train an expected anti-sickness
yaw angle; a MPC
model predictive control method is used, the vehicle
yaw single-track dynamics model is taken as a basis, a prediction model of the MPC is established based on a
state vector, a
control variable and an output variable, a predicted path and a predicted yaw angle are output, a cost function and a constraint are designed based on a deviation of the expected path and the predicted path, a deviation of the expected anti-sickness yaw angle and the predicted yaw angle, and an expected front wheel rotation angle is solved; a PSO
particle swarm optimization algorithm is used to obtain optimal key
control parameters of the MPC, the MPC is updated, and then an updated expected front wheel rotation angle is obtained. The application effectively guarantees safe and smooth driving of the vehicle, and comprehensively improves expected
path tracking accuracy and a passenger
motion sickness discomfort state.