抽象的

Implementing Behaviour based hybrid control for Intelligent Vehicles

Ms.Gayathiri.J, Mr.Ganesha Perumal.D

PEV (Plug-in Electric Vehicle) are penetrating the vehicle market due to increasing energy costs, and carbon foot-print. PEV that can manoeuvre automatically are referred to be autonomous, and require automation. Four major operations in autonomous PEV are- cruise control, automated steering control, braking, and pathplanning. Each of these operations in-turn is composed of minor tasks and these tasks are called behaviour. As an example, steering control is composed of behaviour such as left-turn, right-turn, lane changing, obstacle avoidance etc. These behaviours are dependent on the lane conditions, and path. Therefore, vehicle needs to switch among these behaviours leading to hybrid behaviour wherein there is discrete switching among this continuous behaviour. Control design for such hybrid behaviour is complex, and implementing such control actions in dedicated embedded platform is challenging. In this paper first a layered architecture for implementing hybrid controller for autonomous vehicle is presented. There are three layers in this architecture, the first layer is the higher-level control, second layer is called the hybrid control layer, and the third layer is called lower level control. Further, this paper discusses the implementation of lower-level control in an embedded platform such as 8051. In order to implement the lower-level controller, first the performance of the controller is verified in MATLAB using simulation, and then the code is hard-coded in embedded C and ported into the microcontroller. This provides a methodology to perform hardwarein- loop simulation for the autonomous PEV. Simulation results in MATLAB are used to illustrate the working of the hybrid controller, while verification is used to ascertain the implementation.