A novel energy based design methodology of sliding mode controllers for tracked autonomous guided vehicles

Tracked Autonomous Guided Vehicles (TAGVs) are becoming an important tool for surveillance and security operations so as to mitigate harm to human who otherwise have to conduct such operations. To accomplish steering of such tracked vehicles, skid steering is used. This entails varying the speed of the individual tracks to impose a turning moment on the vehicle. In skid steering, the interaction between the tracks and the terrain is in the form of slippage and soil shearing during the process of turning. These phenomena of slippage and soil shearing are highly nonlinear and difficult to quantify. As such they result in inaccurate kinematics/dynamics of the vehicle. Achieving good path tracking performance of the vehicle in the presence of such inaccuracies requires a robust controller design.; Sliding mode Controllers (SMCs) are insensitive to system parameter uncertainties, modeling errors, and disturbances. Such controllers are considered to be robust. Consequently, they have been proposed as the proper choice for this work. One of the major drawbacks of the published approaches of designing SMCs is the need to assume the values of some of the controller parameters. In this thesis, a novel design methodology of SMCs is developed based on the system total energy. This approach provides a relationship between the SMC parameters and some limits on these parameters. However, the choice of the value of these parameters is still wide enough such that the approach of sensitivity analysis is carried out to tighten these limits. Using SMCs result in undesirable chattering. To reduce or eliminate chattering, the boundary layer technique is employed.; The new design methodology of the SMC is exemplified first on a DC motor model where the model is a Single-Input-Single-Output (SISO) system. Since the TAGV is a special case of a Multi-Input-Multi-Output (MIMO) system, an extension of the new SMC approach is developed. Inverted pendulum systems are popular examples in many control text because of the many practical applications they can model, particularly in the field of robotics. The inverted pendulum is a Single-Input-Multi-Output (SIMO) system similar to the TAGV, and is selected to test the new approach. SMC with hyper sliding surface is also introduced. Simulation results of the motor and pendulum show very good performance of the SMC especially when it is compared to that of proportional and proportional-derivative controllers.; Simulation results of the TAGV using SMCs both decoupled and hyper sliding surfaces based on the proposed design methodology show very good performance when compared to that of proportional controllers.