Yuan-Yih Hsu

Short term load forecasting using a multilayer neural network with an adaptive learning algorithm

A multilayer feedforward neural network is proposed for short-term load forecasting. To speed up the training process, a learning algorithm for the adaptive training of neural networks is presented. The effectiveness of the neural network with the proposed adaptive learning algorithm is demonstrated by short-term load forecasting of the Taiwan power system. It is found that, once trained by the proposed learning algorithm, the neural network can yield the desired hourly load forecast efficiently and accurately. The proposed adaptive learning algorithm converges much faster than the conventional backpropagation-momentum learning method. >

Short term load forecasting of Taiwan power system using a knowledge-based expert system

A knowledge-based expert system was developed for the short-term load forecasting of the Taiwan power system. The developed expert system, which was implemented on a personal computer, was written in PROLOG using a 5-year database. To benefit from the expert knowledge and experience of the system operator, eleven different shapes, each with different means of load calculations, were established. With these load shapes at hand, some peculiar load characteristics pertaining to the Taiwan Power Company can be taken into account. The special load types considered by the expert system include the extremely low load levels during the week of the Chinese New Year, the special load characteristics of the days following a tropical storm or a typhoon, the partial shutdown of certain factories on Saturdays, the special event caused by a holiday on Friday or on Tuesday, etc. A characteristic feature of the knowledge-based expert system is that it is easy to add new information and new rules to the knowledge base. To illustrate the effectiveness of the system, short-term load forecasting is performed on the Taiwan power system by using both the developed algorithm and the conventional Box-Jenkins statistical method. It is found that a mean absolute error of 2.52% for one year is achieved by the expert system approach as compared to an error of 3.86% by the statistical method. >

Fuzzy dynamic programming: an application to unit commitment

A new approach using fuzzy dynamic programming is proposed for the unit commitment of a power system. A characteristic feature of the approach is that the errors in the forecast hourly loads can be taken into account by using fuzzy set notations, making the approach superior to the conventional dynamic programming method which assumes that the hourly loads are exactly known and there exist no errors in the forecast loads. To reach an optimal commitment strategy under the fuzzy environment, a fuzzy dynamic programming model in which the hourly loads, the cost, and system security are all expressed in fuzzy set notations is developed. The effectiveness of the approach is demonstrated by the unit commitment of the Taiwan power system, which contains six nuclear units, 48 thermal units, and 44 hydro units. >

Dispatch of direct load control using dynamic programming

An approach based on dynamic programming is presented for the dispatch of direct load control (DLC). The objective is to coordinate DLC strategies with system unit commitment such that the system production cost is minimized. To achieve this goal, the DLC dispatch is first integrated into the unit commitment problem. An optimization technique based on dynamic programming is then developed to reach the optimal DLC dispatch strategy and system generation schedule. To demonstrate the effectiveness of the approach, results from a sample study performed on the Taiwan power system are described. >

Coordinated Synthesis of Multimachine Power System Stabilizer Using an Efficient Decentralized Modal Control (DMC) Algorithm

This paper presents a new method for simultaneously selecting the power system stabilizer (PSS) parameters in multimachine power systems. Design of local output feedback power system stabilizer using local signal of the generator (e.g. speed or power) is formulated as a decentralized modal control (DMC) problem. Exact model reduction based on modal control theory is proposed to make the determination of PSS parameters highly efficient. Results obtained from the coordinated tuning of the parameters of three power system stabilizers equipped on a multimachine power system show that exact assignment of the eigenvalues associated with the poorly damped electromechanical modes can be achieved in a very efficient manner.

Low Frequency Oscillations in Longitudinal Power Systems: Experience with Dynamic Stability of Taiwan Power System

Sustained low frequency oscillations have been observed in Taiwan power system which is of longitudinal structure. It is the purpose of this paper to examine the various factors affecting the damping characteristics of these oscillations which caused dynamic instability problem in the operation of Taiwan power system. It is observed that the amount of power flow on the EHV transmission line and the characteristics of load have a significant effect on the damping of the system while the speed-governing system and the gain of automatic voltage regulator have only a minor one. Detailed investigation using both the frequency domain and time domain approaches also reveals that power system stabilizers can be employed as an effective means for improving dynamic stability of Taiwan power system.

A combined artificial neural network-fuzzy dynamic programming approach to reactive power/voltage control in a distribution substation

Reactive power/voltage control in a distribution substation is investigated in this work. The purpose is to determine proper capacitor on/off status and suitable load tap changer (LTC) positions for the 24 hours in the next day. To reach this goal, an artificial neural network (ANN) is designed to reach a preliminary dispatch schedule for the capacitor and LTC. The inputs to the ANN are main transformer real power and reactive power and primary and secondary bus voltages and the outputs are the desired capacitor on/off status and LTC tap positions. The preliminary dispatch schedule is further refined by fuzzy dynamic programming in order to reach the final schedule. To demonstrate the effectiveness of the proposed method, reactive power/voltage control is performed on a distribution substation in Taipei, Taiwan. Results from the example show that a proper dispatch schedule for capacitor and LTC can be reached by the proposed method in a very short period.

Damping of subsynchronous resonance using excitation controllers and static VAr compensations: a comparative study

The results of a comparative study on the application of two countermeasures, i.e. the excitation controller and the static VAr compensator (SVC), for damping of subsynchronous resonance (SSR) are presented. To stabilize all the SSR modes, a unified approach based on modal control theory is proposed for the design of the excitation controller and the SVC, which are essentially dynamic output feedback compensators. The two damping schemes differ in the way they modulate the reactive power flow in the system to damp out the subsynchronous oscillations. To demonstrate the effectiveness of the proposed damping schemes under disturbance conditions, time-domain simulations based on a nonlinear system model are also performed. The relative merits of the two countermeasures are compared with respect to their validities under various loading conditions and different degrees of series compensations and their capabilities to expand the stable region on the real-capacitive reactance plane. >

Fuzzy-set based contingency ranking (power system security)

An approach based on fuzzy set theory is developed for contingency ranking of the Taiwan power system. In the proposed approach, the post-contingent quantities are first expressed in a fuzzy set notation. The heuristic rules used by system operators in contingency ranking are then compiled and coded in the form of fuzzy reasoning rules. These reasoning rules form the basis for combining the evidence from each post-contingent quantity and reaching the overall system severity index. Since the set of severe contingency cases have been derived based on operators past experience, it is expected that they will match the human experts judgments on what contingency cases are severe. The proposed fuzzy set approach has been applied to contingency ranking of the Taiwan power system which consists of 170 buses, 121 power lines, and 86 transformers. >

Design of PID static VAr controllers for the damping of subsynchronous oscillations

Damping of subsynchronous resonance using static VAr compensator (SVC) is investigated. To damp the various unstable torsional modes, a proportional-integral-derivative (PID) controller is used to modulate the reactive power output feedback controller with only one feedback signal (generator speed deviation), the proposed controller is relatively simple for practical implementation. Eigenvalues analysis of the closed-loop system reveals that the PID controller is quite effective over a wide range of operating conditions. Digital simulations using a nonlinear system model are also performed to demonstrate the effectiveness of the proposed controller under disturbance conditions. >