Development, testing, and implementation of an optimized sliding control scheme for unstable under-actuated mechanical systems

Abstract

Many prominent mechanical systems used in today's world, such as robotic manipulators, drones, and unmanned vehicles can be accurately described with nonlinear dynamic models. Sliding Mode Control (SMC) is a powerful tool that can drive uncertain systems towards the desired trajectory while also rejecting disturbances. This is a valuable property when dealing with fast systems that require precise operating points. This thesis details the design, validation, and implementation of a novel sliding control scheme. The sliding surface is made up of proportional, integral, and derivative terms which generate superior performance when compared to other sliding schemes and is easier to design. The controller is then tuned using a custom genetic algorithm in order to achieve optimal performance. A detailed overview of the controller design as well as stability proofs are provided in the following thesis. The new control scheme was implemented on several under-actuated systems to test its performance. The results collected showed that the new Proportional-Integral-Derivative (PID) sliding surface controller produces superior performance when compared to standard sliding and linear schemes.