Jong Shik Kim

Nonlinear quadratic Gaussian control with loop transfer recovery

Abstract A nonlinear quadratic Gaussian with loop transfer recovery control method called NQG/LTR method is presented, which is the integration of statistical linearization, loop-shaping and loop transfer recovery techniques. The proposed NQG/LTR method is a powerful one for designing controllers of nonlinear systems with hard nonlinearities such as Coulomb friction, backlash and dead-zone. In order to show the effectiveness of the proposed method, the LQG/LTR and NQG/LTR methods are applied to a timing-belt driving cart system with Coulomb friction and dead-zone. The results of simulation and experiment for a cart system show that the performance and stability robustness of the NQG/LTR control system are insensitive to the effects of the nonlinear elements of a cart system.

Pre-sliding friction control using the sliding mode controller with hysteresis friction compensator

Friction phenomenon can be described as two parts, which are the pre-sliding and sliding regions. In the motion of the sliding region, the friction force depends on the velocity of the system and consists of the Coulomb, stick-slip, Streibeck effect and viscous frictions. The friction force in the pre-sliding region, which occurs before the breakaway, depends on the position of the system. In the case of the motion of the friction in the sliding region, the LuGre model describes well the friction phenomenon and is used widely to identify the friction model, but the motion of the friction in the pre-sliding such as hysteresis phenomenon cannot be expressed well. In this paper, a modified friction model for the motion of the friction in the pre-sliding region is suggested which can consider the hysteresis phenomenon as the Preisach model. In order to show the effectiveness of the proposed friction model, the sliding mode controller (SMC) with hysteresis friction compensator is synthesized for a ball-screw servo system.

Robust control for rotational inverted pendulums using output feedback sliding mode controller and disturbance observer

This paper presents a system modeling, controller design and implementation for a rotational inverted pendulum system (RIPS), which is an under-actuated system and has the problem of unattainable velocity state. Two control strategies are applied to theRIPS. One is a sliding mode control method using the parameterization of both the hyperplane and the compensator for output feedback. The other is the disturbance observer which estimates disturbance and some modeling errors of RIPS with less computational effort. Some simulations and various kinds of experiments are performed in order to verify that the proposed controller has the ability to control RIPS whose velocity is assumed to be unavailable. The results of the simulations and experiments show that the proposed control system has superior performance for disturbance rejection and regulation at certain initial conditions as well as the robustness to model uncertainties.

Robust Control of Maglev Vehicles with Multimagnets Using Separate Control Techniques

A robust control design scheme using well-developed SISO techniques is proposed for maglev vehicles that are inherently unstable MIMO systems. The proposed seperate control method has basically two control loops: a stabilizing loop by a pole-placement technique, and a performance loop using a novel optimal LQ loop-shaping technique. This paper shows that the coupling terms involved in maglev vehicles with multimagnets should not be neglected but compensated for their stability and performance robustness. The robustness properties of the proposed control system are then evaluated under variations of vehicle masses and air gaps through a computer simulation. This paper also describes the reason why the proposed control technique can be suggested as a tool using only SISO techniques in controlling unstable MIMO systems such as maglev vehicles.

A Self-Tuning PI control system design for the flatness of hot strip in finishing mill processes

A novel flatness sensing system which is called the Flatness Sensing Inter-stand Looper (FlatSIL) system is suggested and a self-tuning PI control system using the FlatSIL is designed for improving the flatness of hot strip in finishing mill processes. The FlatSIL system measures the tension along the direction of the strip width by using segmented rolls, and the tension profile is approximated through the tension of each segmented roll. The flatness control system is operated by using the tension profile. The proposed flatness control system as far as the tension profile-measuring device works for the full strip length during the strip rolling in finishing mills. The generalized minimum variance self-tuning (GMV S-T) PI control method is applied to control the flatness of hot strip which has a design parameter as weighting factor for updating the PI gains. Optimizing the design parameter in the GMV S-T PI controller, the Robbins-Monro algorithm is used. It is shown by the computer simulation and experiment that the proposed GMV S-T PI flatness control system has better performance than the fixed PI flatness control system.

Dynamic System Synthesis in Terms of Bond Graph Prototypes

This paper deals with two synthesis methods for dynamic systems as an application of the bond graph prototypes which are originally developed for dynamic system analysis in the frequency domain. The first method called the analytical synthesis uses the procedures similar to the network synthesis in the electrical field and an already-existing one, but yet demonstrates its own strengths inherited from the bond graph prototypes such as the freedom from the causality assignment and determination of junction types. The second method called the direct synthesis is introduced in this paper to provide physical realization for a given specification of impedance and admittance forms without any mathematical manipulations except a simple division. The two synthesis methods are shown through examples in each case to be a concise method in their usage and a scientific method in managing the bond graph prototypes systematically.

H∞-constrained quasi-linear quadratic Gaussian control with loop transfer recovery

In this paper we propose a new nonlinear controller design method, called quasi-linear quadratic Gaussian/H-infinity/loop transfer recovery (QLQG/H∞/LTR), for nonlinear multivariable, systems with hard nonlinearities such as Coulomb friction and dead-zones. We consider H∞-constraints for the optimization of statistically linearized systems, by replacing the covariance Lyapunov equation by a modified Riccati equation, whose solution leads to an upper bound QLQG performance. As a result, the nonlinear correction term is included in the Riccati equation which, in general, is excessively difficult to solve numerically. To solve this problem, we use the modified loop shaping technique and derive analytic proofs of the LTR condition. Finally, the H∞-constrained, nonlinear controller is synthesized by an inverse random input describing function technique (IRIDF). The proposed design method for a hard nonlinear multivariable systems has better robustness to unstructured uncertainty and hard nonlinearities than the QLQG/LTR method. A flexible link system with Coulomb frictions serves as a design example for our methodology.

Nonlinear Position Servo Design Using the QLQG / LTR Method

A new nonlinear controller design method called the QLQG/LTR method is proposed. This method is related to the use of statistical linearization, LQG/LTR, and IRIDF methods. It is applied to a position servo with Coulomb friction. The computer simulation results show that the QLQG/LTR control system can adapt automatically to changes in input magnitude.

The QLQG/LTR control for nonlinear systems with a non-Gaussian nature

The QLQG/LTR control method is applied to a nonlinear system with Coulomb friction which has a non-Gaussian nature. It is shown that the non-Gaussian nature degrades the effectiveness of the QLQG/LTR design method. Thus, a method for alleviating this problem is proposed. It is the QLQG/LTR control method with a modified model based compensator which can consider the non-Gaussian nature of nonlinear systems. The computer simulation results show that the responses of this nonlinear control system are relatively insensitive to the imput magnitude even if there exist a hard nonlinearity and a non-Gaussian nature in the plant.