Design of Sliding Mode PID Controller with Improved reaching laws for Nonlinear Systems

TL;DR

This paper presents an advanced sliding mode PID control method with improved reaching laws for nonlinear systems, utilizing Modified Particle Swarm Optimization for parameter tuning and addressing chattering issues through boundary and second-order sliding mode techniques.

Contribution

It introduces a novel exponential power rate reaching law and a PID-type sliding surface for nonlinear control, optimized via MPSO, and demonstrates robustness and hardware implementation.

Findings

Enhanced finite convergence time with improved reaching law

Effective chattering reduction using boundary and second-order methods

Successful hardware implementation on a conical tank system

Abstract

In this thesis, advanced design technique in sliding mode control (SMC) is presented with focus on PID (Proportional-Integral-Derivative) type Sliding surfaces based Sliding mode control with improved power rate exponential reaching law for Non-linear systems using Modified Particle Swarm Optimization (MPSO). To handle large non-linearities directly, sliding mode controller based on PID-type sliding surface has been designed in this work, where Integral term ensures fast finite convergence time. The controller parameter for various modified structures can be estimated using Modified PSO, which is used as an offline optimization technique. Various reaching law were implemented leading to the proposed improved exponential power rate reaching law, which also improves the finite convergence time. To implement the proposed algorithm, nonlinear mathematical model has to be decrypted without linearizing, and used for the simulation purposes. Their performance is studied using simulations to prove the proposed behavior. The problem of chattering has been overcome by using boundary method and also second order sliding mode method. PI-type sliding surface based second order sliding mode controller with PD surface based SMC compensation is also proposed and implemented. The proposed algorithms have been analyzed using Lyapunov stability criteria. The robustness of the method is provided using simulation results including disturbance and 10% variation in system parameters. Finally process control based hardware is implemented (conical tank system).