• The Journal of Korea Robotics Society
• /
• v.17 no.1
• /
• pp.76-85
• /
• 2022

One of the most fundamental challenges when designing controllers for dynamic systems is the adjustment of controller parameters. Usually the system model is used to get the initial controller, but eventually the controller parameters must be manually adjusted in the real system to achieve the best performance. To avoid this manual tuning step, data-driven methods such as machine learning were used. Recently, reinforcement learning became one alternative of this problem to be considered as an agent learns policies in large state space with trial-and-error Markov Decision Process (MDP) which is widely used in the field of robotics. However, on initial training step, as an agent tries to explore to the new state space with random action and acts directly on the controller parameters in real systems, MDP can lead the system safety-critical system failures. Therefore, the issue of 'safe exploration' became important. In this paper we meet 'safe exploration' condition with Control Barrier Function (CBF) which converts direct constraints on the state space to the implicit constraint of the control inputs. Given an initial low-performance controller, it automatically optimizes the parameters of the control law while ensuring safety by the CBF so that the agent can learn how to predict and control unknown and often stochastic environments. Simulation results on a quadrotor UAV indicate that the proposed method can safely optimize controller parameters quickly and automatically.

• Journal of Drive and Control
• /
• v.15 no.1
• /
• pp.28-37
• /
• 2018

When the wind speed rises above the rated wind speed, the produced power of the wind turbines exceeds the rated power. Even more, the excessive power results in the undesirable mechanical load and fatigue. A solution to this problem is pitch control of the wind turbines. This paper presents a systematic design method of a collective pitch controller for the wind turbines using a discrete fuzzy Proportional-Integral (PI) controller. Unlike conventional PI controllers, the fuzzy PI controller has variable gains according to its input variables. Generally, tuning the parameters of fuzzy PI controller is complex due to the presence of too many parameters strongly coupled. In this paper, a systematic method for the fuzzy PI controller is presented. First, we show the fact that the fuzzy PI controller is a superset of the PI controller in the discrete-time domain and the initial parameters of the fuzzy PI controller is selected by using this relationship. Second, for simplicity of the design, we use only four rules to construct nonlinear fuzzy control surface. The tuning parameters of the proposed fuzzy PI controller are also obtained by the aforementioned relationship between the PI controller and the fuzzy PI controller. As a result, unlike the PI controller, the proposed fuzzy PI controller has variable gains which allow the pitch control system to operate in broader operating regions. The effectiveness of the proposed controller is verified with computer simulations using FAST, a NREL's primary computer-aided engineering tool for horizontal axis wind turbines.

• 제어로봇시스템학회:학술대회논문집
• /
• 2005.06a
• /
• pp.2557-2561
• /
• 2005

The PID controller design and choosing PID parameters according to system response are proposed in this paper. Here PID controller is employed to control DC motor speed and Matlab program is used for calculation and simulation. Choosing PID parameters are demonstrated by several contrast experiments and a way for setting PID parameters values is discussed.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.5 no.2
• /
• pp.3-12
• /
• 1995

A controller utilizing fuzzy logic is developed to control the speed of a motor in a washing machine by choosing an appropriate phase. Due to the hardship imposed on obtaining a result from a relation established for inputs, present speed and present rate of speed, and ouput, a phase, of the system that can be tested against an experimental result, it is impossible to apply a genetic algorithm to fine-tune the fuzzy logic controller. To avoid this difficulty, a proper assumption that the parameters of an if-part of a primary fuzzy logic controller have a functional relationship with an error between computed values and experimental ones in made. Setting up of a fuzzy relationship between the parameters and the errors is then achieved through experimentally obtained data. Genetic Algorithm is then applied to this secondary fuzzy logic controller to verify the fuzzy logic. In the verification process, the primary fuzzy logic controller is used in obtaining experimental results. In this way the kind of difficulty in obtaining enough experimental values used to verify the fuzzy logic with genetic algorithm is gotten around. Selection of the parameters that would produce the least error when using the secondary fuzzy logic controller is done with applying genetic algorithm to the then-part of the controller. In doing so the optimal values for the parameters of the if-part of the primary fuzzy logic controller are assumed to be contained. The experimental result presented in the paper validates the assumption.

• Proceedings of the KIEE Conference
• /
• 2005.10b
• /
• pp.158-160
• /
• 2005

This paper presents a new approach to the problem of designing a cascaded three-parameters controller for a given linear time invariant (LTD plant in unity feedback system. We consider a proportional-integral-derivative (PID) and a first-order controller with specified overshoot and settling time. This problem is difficult to solve because there may be no analytical solution due to the use of low-order controller and furthermore. the zeros of controller just appear in the zeros of feedback system. The key idea of our method is to impose a constraint on the controller parameters so that the zeros of resulting controller are distant from the dominant pole of closed-loop system to the left as far as the given interval. Two methods realizing the idea are suggested. We have employed the characteristic ratio assignment (CRA) in order to deal with the time response specifications. It is noted that the proposed methods are accomplished only in parameter space. Several illustrative examples are given.

• Journal of Institute of Control, Robotics and Systems
• /
• v.3 no.6
• /
• pp.555-562
• /
• 1997

With only the classical PID controller applied to control of a DC motor, good (target) performance characteristic of the controller can be obtained if all the model parameters of DC motor and operating conditions such as external load torque, disturbance, etc. are known exactly. However, in case when some of system parameters or operating conditions are uncertain or unknown, the fixed PID controller does not guarantee good performance, which is assumed with precisely known system parameters and operating conditions. In view of this and the robustness enhancement of DC motor control system, we propose a PID learning controller which consists of a set of learning rules for PID gain tuning and learning of an auxiliary input. The proposed PID learning controller is shown to drive the state of uncertain DC motor system with unknown system parameters and external load torque to the desired one world wide asymptotically. Computer simulation and experimental results are given to demonstrate the effectiveness of the proposed PID learning controller, thereby showing its superiority to the conventional fixed PID controller.

• The Transactions of the Korean Institute of Electrical Engineers D
• /
• v.55 no.1
• /
• pp.20-23
• /
• 2006

This paper presents a new approach to the problem of designing a cascaded three-parameters controller for a given linear time invariant (LTI) plant in unity feedback system. We consider a proportional-integral-derivative (PID) and a first-order controller with the specified overshoot and settling time. This problem is difficult to solve because there may be no analytical solution due to the use of low-order controller. Furthermore, the zeros of controller just appear in the zeros of feedback system. The key idea of our method is to impose a constraint on the controller parameters so that the zeros of resulting controller are distant from the dominant pole of closed-loop system to the left as far as the given interval. Two methods realizing the idea are suggested. We have employed the characteristic ratio assignment (CRA) in order to deal with the time response specifications. It is noted that the proposed methods are accomplished only in parameter space. Several illustrative examples are given.

• Journal of Advanced Marine Engineering and Technology
• /
• v.24 no.5
• /
• pp.97-102
• /
• 2000

In this paper, a methodology for designing a controller based on inverse dynamics for speed control of DC motors is presented. The proposed controller consists of a prefilter, the inverse dynamic model of a system and the PI controller. The prefilter prevents high frequency effects from the inverse dynamic model. The model of the system in characterized by a nonlinear equation with coulomb friction. The PI controller regulates the error between the set-point and the system output which may be caused by modeling error, variations of parameters and disturbances. The output which may be caused by modeling error, variations of parameters and disturbances. The parameters of the model and the PI controller are adjusted offlinely by a genetic algorithm. An experimental work on a DC motor system is carried out to illustrate the performance of the proposed controller.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2002.05a
• /
• pp.225-228
• /
• 2002

When the fuzzy logic controller (FLC), which is designed based on the plant model, is applied to the real control system, satisfactory control performance may not be attained due to modeling errors from the plant model. In such cases, the control parameters of the controller must be adjusted to enhance control performance. Until now, the trial and error method has been used, consuming much time and effort. To resolve such problem, response surface methodology (RSM), a new method of adjusting the control parameters of the controller, is suggested. This method is more systematic than the previous trial and error method, and thus optimal solutions can be provided with less tuning. First, the initial values of the control parameters were determined through the plant model and the optimization algorithm. Then, designed experiments were performed in the region around the initial values, determining the optimal values of the control parameters which satisfy both the rise time and overshoot simultaneously.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.15 no.2
• /
• pp.127-134
• /
• 2005

This paper proposes a nonlinear fuzzy PID controller with variable parameters to improve slow rising time and divergence occurred by limited input spaces and a resultant limited control input during fuzzification in a fuzzy PID controller with fixed parameters, and describes the design principle and tracking performance of a proposed fuzzy PID controller. The parameters of a proposed controller are adjusted by the stability conditions derived from 'small gain theorem' and satisfy the BIBO stability of overall control system.