DIGITAL POWER SYSTEM STABILIZER ( DPSS )

A digital power system stabilizer ( DPSS ) and its design method are presented in this paper. The mathematical model of a single machine infinite bus power system is obtained using off-line parameter identifications, and on the base of the identified model the control system is synthesized by generalized minimum variance control principles. Through feedback combination of several signals this control system has double functions of damping low frequency oscillation and automatic voltage regulating. In order to avoid static errors the very effective way is to combine discrete generalized minimum variance control with continuous integral control. Besides the proposed control strategy is very simple in structures and very easy to realize. Simulation results under several cases have shown the correctness and effectiveness of the proposed control strategy.     

LOAD FREQUENCY CONTROL USING DECOMPOSED DISCRETE MODELS

Decomposed discrete models, can be used to design load frequency' controllers in multi-area power systems.Considering the exchanging real power in tie-lines as a measurable disturbance, the multi-area power system is decomposed into several subsystems.On the basis of generalized minimum variance control principles, an expanded control rule is induced. Using discrete models, the proposed method greatly decreases the number of varialles in be measured and simplifies the control system. The method is illustrated with a two-area power system example and its simulation Jesuits.     

AN AUTOMATIC VOLTAGE REGULATOR USING IDENTIFIED MODELS

This paper obtains the mathematical model of a single machine infinite bus power system using the idea of parameter identifications. Because the noise is considered in parameter identification process, the model reflects the real system more exactly. On the base of the identified model this paper also presents a discrete automatic voltage regulator by minimum variance control principles. In order to avoid the static error, we combine discrete minimum variance control with continuous integral control and get very satisfactory effects. This proposed control strategy is very easy to realize. The simulation results show that in every case the dynamic qualities of the proposed control strategy are incomparably superior to that of the conventional PI regulator, which means the proposed control strategy has great practical values.     

THE OVERLAPPING DECOMPOSITION DECENTRALIZED CONTRAL FOR GENERATOR''S EXCITER/GOVERNOR SYSTEMS

In this paper the generator's exciter/governor system are decomposed by using the overlapping decomposition method, and an overlapping decomposition decentralized control (ODDC) system is designed for the exciter /governor systems. In or.dar to enhance the dynamic qualities of the decentralized control system, the concept of the second feedback system is introduced to modify the first feedback gain matrix on the condition of the structure of the original decentralized control system maintaining unchanged. The practical decentralized control system design and its; simulation results show that this method greatly improves the control effect of the ODDC system and also holds all the advantages of the ODDC system. Besides, this paper indicates that it is a simple and effective method to control excitor/governor system by overlapping decomposition decentralized control strategy.     

THE APPLICABILITY OF OVERLAPPING DECOMPOSITION DECENTRALIZED CONTROL

The overlapping decomposition decentralized control is a kind of static feedback control system. Using overlapping information sets among subsystems, the reliability of the control system is improved. On the other hand the overlapping decomposition decentralized control systom isn't suitable for all systems waiting control. As for some systems, using overlapping decomposition decentralized control doesn't reach the desired control effect. This paper has discussed the applicability of the overlapping decomposition decentralized control offers the necessary conditions for applying this control technique, and provides theoretical basis for the wide applications of the overlapping decomposition decentralized control systems.     

A SENSITIVITY METHOD FOR THE WEIGHTING MATRIX SELECTION OF LINEAR OPTIMAL CONTROL

The paper analyses the relations between closed-loop system dynamic qualities and linear optimal control weighting matrix Q and offers a sensitivity method for selecting weighting matrix Q. As a practical example, a linear optimal supplement excitation control system is designed for a single machine-infinite bus power system using the presented method. Theoretical . analysis and simulation results have shown that the dynamic qulities of the control system are very satisfactory.     

A SELF-TUNING CONTROL FOR A PARALLEL AC/DC POWER SYSTEM

Based on a non-linear system model, an adaptive control methodology is presented for a parallel AC/DC power system by means of on-line recursive parameter identifications. Taking the advantage of rapid power regulatiing ability of the DC power system and combining with the generalized minimum variance self-tuning control principles, the dynamic qualities of the controlled system has been greatly improved. Using SISO form makes the control system very simple to implement. Through the design, simulation and analysis of the self-tuning control system for a practical example, it shows that the self-tuning control strategy is very promising for AC/DC power systems.