Advanced Discrete-Time Control Methods for Industrial Applications (1504.04423v2)

Abstract: This thesis focuses on developing advanced control methods for two industrial systems in discrete-time aiming to enhance their performance in delivering the control objectives as well as considering the practical aspects. The first part addresses wind power dispatch into the electricity network using a battery energy storage system (BESS). To manage the amount of energy sold to the electricity market, a novel control scheme is developed based on discrete-time model predictive control (MPC) to ensure the optimal operation of the BESS in the presence of practical constraints. The control scheme follows a decision policy to sell more energy at peak demand times and store it at off-peaks in compliance with the Australian National Electricity Market rules. The performance of the control system is assessed under different scenarios using actual wind farm and electricity price data in simulation environment. The second part considers the control of overhead crane systems for automatic operation. To achieve high-speed load transportation with high-precision and minimum load swings, a new modeling approach is developed based on independent joint control strategy which considers actuators as the main plant. The nonlinearities of overhead crane dynamics are treated as disturbances acting on each actuator. The resulting model enables us to estimate the unknown parameters of the system including coulomb friction constants. A novel load swing control is also designed based on passivity-based control to suppress load swings. Two discrete-time controllers are then developed based on MPC and state feedback control to track reference trajectories along with a feedforward control to compensate for disturbances using computed torque control and a novel disturbance observer. The practical results on an experimental overhead crane setup demonstrate the high performance of the designed control systems.