Design And Implementation Of Motion Control System For Low-speed Autonomous Vehicles For Park Applications

As an important form of implementation of intelligent vehicle,low-speed autonomous vehicles for park applications have great potential for commercial application in terminal logistics,unmanned buses,closed park logistics,and unmanned sanitation,etc.,and have received continuous attention from all walks of life.It uses sensor sensing technology to understand the environment,decision-making and planning control technology to generate behavior commands,and the vehicle chassis X-by-wire control technology to execute actions and then realize autonomous navigation.Motion control system,as the core part of the system,is responsible for the smooth and accurate driving of the vehicle during the autonomous navigation process.This paper conducts preliminary research on the design and implementation of low-speed autonomous vehicle motion control system for park applications,and elaborates on the three aspects of vehicle hardware architecture and software architecture,positioning perception,and trajectory following control.In terms of hardware architecture and software architecture of the vehicle,a hierarchical automatic driving architecture based on function decomposition is designed on the modular hardware platform,which makes each module work in a clear and coordinated way.In the aspect of positioning perception,GNSS/INS integrated navigation is used to realize the perception of self position and pose,and the robust estimation of vehicle speed and acceleration is realized based on the current statistical model.In this way,Gaussian projection is used to convert longitude and latitude information to plane Cartesian coordinate system and establish the relationship between road and vehicle relative position and pose.In the aspect of trajectory following control,a hierarchical motion control algorithm framework is designed,in which a horizontal and longitudinal decoupled road error tracking model is established in the upper layer,and constraints are introduced to transform it into a quadratic form,and the model prediction algorithm is used to solve the problem online.The lower layer establishes the wheel hub motor execution model and steering response model to realize the upper layer control commands.Finally,the typical campus scenes of the road section around the island and the teaching area were designed for tracking experiments to verify the effectiveness of the whole system architecture and motion control algorithm.