Longitudinal Motion Control For Intelligent Vehicles
Published 2011 · Computer Science
The longitudinal motion control is to control a vehicle according to its relative position with respect to either the lead vehicle or obstacles. There are four dynamical models of the vehicle longitudinal motion which can be described as first order systems, first-order lag systems, second order systems, second-order lag systems. In this chapter, we introduce how to identify the velocity model. And we present an improved Single-Neuron adaptive PID (SN-PID) control module which plays an important role in our system. In the experiment, we use four learning rules: unsupervised Hebb learning rule, supervised Delta learning rule, supervised Hebb learning rule, and improved Hebb learning rule to validate the longitudinal system model. From the experimental results, we can see that the value of K affects the performance of the controller, and the learning quadratic performance index has lower computing burden and clearer physical meaning.