Optimal iterative learning control design for generator voltage regulation system

Abstract

Abstract:The purpose of this paper is to design a good tracking controller for the generator automatic voltage regulator (AVR) system. It uses an iterative learning control (ILC) algorithm as a control rule and tries to keep voltage error as low as possible, while replying to set point changes as fast as possible. Two models of ILC are discussed: P-type learning and linear quadratic (LQ) optimal design. The results of designing by LQ optimal method are compared with a ZieglerNichols proportional-integral-derivative (PID) controller. The effects of changing different parameters are investigated and illustrated by simulation. In order to provide better robustness for the design, a PID-ILC parallel structure is designed and simulated under system parameter changes.

Control of robot-assisted gait trainer using hybrid proportional integral derivative and iterative learning control

“…The tracking error is modified according to the learning signal to enhance a particular control goal and track the intended trajectory [42], [43]. In brief, because the PID has strong robustness and the ILC has good performance, it is assumed that their combination will have both [44]. Figure 7 depicts a diagram block of hybrid PID-ILC applied to the exoskeleton, where uj is ILC control signal, kp is proportional value, kd is derivative value, ki is integral value, ej is error between desired and actual outputs (yd to yj), and j is number of iterations…”

Application of fractional order iterative learning controller for a type of batch bioreactor

“…Each of the above criteria should have a related equality or non-equality, e.g. for the first criterion |C(jω cg )G(jω cg )| = 0, arg(C(jω cg )G(jω cg )) = −π + ϕ m (16) By solving the set of all these equalities and non-equalities, the optimum settings for FOPID are found. Due to the difficult and time-consuming solution procedure, another module in FOMCON is used for the numerical solution.…”

“…Thus, it can be used when a precise trajectory tracking is expected. ILC has a wide range of application in industry and academic research, for instance, in robotics, glycaemic control in diabetes mellitus, hard disk position control, iterative system identification, human learning behaviour, industrial processes, electro pneumatic servo systems, injection moulding processes, food production facilities, robotic assembly lines, chemical batch reactors, density control of freeway traffic flow [16] and control of liquid slosh transportation movement [17]. Fractional order ILC has the advantage of improved performance, over conventional integer order ILC, for control of fractional order systems.…”

Optimizing Automatic Voltage Regulation: A Deep Reinforcement Learning Approach

“…While new algorithms have been proposed for regulating synchronous generator's terminal voltage, many industrial systems still rely on the conventional Proportional-Integral-Derivative (PID) based AVR design due to its simplicity and ease of implementation [2] . Conventional PID controllers have certain drawbacks.…”