Design and implementation of guidance control system for autonomous light electric vehicles

In this work; a light AGV is presented. Light AGVs are usually designed to move small payloads and their distinctive feature is their high flexibility and re-configurability in the load handling. However; they have usually limited load weight capacity. In fact; large load weight values have a high impact on the guidance performances; even affecting the vehicle mechanical stability and strong speed limitations are usually applied. We will focus our attention to the vehicle guidance control problem and aim to design controllers that guarantees desirable specifications for a wide range of possible load mass values (from 50kg to 1000kg). Hence; we assume that the payload mass is an uncertainty in the model and resort to robust control design methods. Moreover; we'd like not only to find a single suitable controller but to analyze the stability domain in the controller parameter space; so that we can evaluate the effects on the guidance performances produced by changes in the controller parameters; preserving control system stability. In order to obtain such results we will pay special attention to randomized techniques. Randomized techniques are oriented to deal with basic notions for any engineering characteristics – gain or phase margin; overshoot or other time-response characteristics; robustness margin – as well as mathematical objectives such as H2 or Hinf norm. Specifically; randomized method are applied to the tuning of PI controllers; taking into account control action limitations; due to the limited current supply values; system speed specifications; defined by desired values for settling time; Hinf performances.