C. C. Huang

System output feedback controller design via O.S.A. method

The paper addresses the design of a system stabilizer using an optimal reduced order model by only the output state variables. System damping can be improved by using eigenvalues assignment and the co-ordination of stabilizers can be achieved through eigenvector assignment by maintaining system mode shape, this is defined as the optimal subeigenstructure assignment (O.S.A) method. The proportional-integral controller is derived via the optimal reduced order model instead of via the whole system model. The effectiveness of this stabilizer is evaluated, and this study reveals that the result of eigenstructure assignment is more stable and much better than in the assignment method based on the whole system model. A system is given as examples to illustrate the advantages and effectiveness of the proposed approach. Results based on the whole system model are included for comparison.

Controllable performance system regulator

The concept of time-varying state feedback control of a system under the desired closed loop scate transition matrix is introduced. By setting the parameters or the preset state transition matrix, the time-varying state feedback can be designed. The theorem and design method are derived and an example of motor control withany desired performance can be achieved under the design of time-varying state feedback regular. By the optimal reduced order method, any system with desired performance can be designed in the same way.

S.T.M. based predictable time-varying controller and estimator design

The concepts of time-varying state feedback under the desired closed-loop state transition matrix S.T.M. and estimator design are introduced in this paper. By introducing the estimation error concept (i.e. e(t) = x−◯) under the assumption that e(t)≡(i.e. x = bx ), the time-varying state feedback F(t) and the estimator matrix L can be designed separately

Optimal pair-layer stabilizer design

The purpose of this paper is to address the design of decoupled-level large system controllers using an optimal reduced order model whose state variables are all output states. The reduced-order model retains their physical meaning and is used to design a decoupled-level linear feedback controller that takes into account the realities and constraints of the large systems. The decoupled-level control strategy is used and a global control signal is generated from the output variables tominimize the eff ect of interactions. As a result, the eff ectiveness of this controller is evaluated and considerable savings in computer memory are achieved. In conclusion, the controllers determined at the subsystem level depend only on local information operating to the particular machine. Example is given to illustrat the advantages of the proposed method. Responses of the system with decoupled-level scheme and optimal reduced order scheme are included for comparative analyses. It is recommended that the proposed method can be applied to supper large system.

Output-regulated optimal grey controller

In this paper, we propose a new method to design a grey prediction system controller using output feedback. The optimal reduced order model will be used to retain the physical meaning of the output states. By using only the output states feedback, the control strategy can be implemented easily. The grey prediction method adopts the forecasting information from the output state variables to control system behavior. As a result, it reduces the oscillation and enhances the dynamic stability of the system. The advantages of the proposed method are verified through a detailed simulation of a multi-machine power system.

Optimal system harmonic filters design

In this paper, we formulate the design harmonic filter problem taking by considering investment cost and operational constraints. System quality can generally be achieved by installing harmonic filters in industry systems, thereby reducing the harmonics. Users prioritize the performance and cost of fi lters as primary concerns. This new formulation uses a combinatorial optimization problem with a non-diff erentiable objective function. A solution methodology based on an optimization technique-simulated annealing is also proposed to determine the size of filters with minimum cost. As a result, the proposed solution methodology can off er the optimal size and minimum cost of design filters. In conclusion, a solution algorithm is developed to derive the global optimal solution and test on a 69kV industrial power system with highly promising results.