Junzo Kawakami

Original Contribution: Solving inequality constrained combinatorial optimization problems by the hopfield neural networks

The Hopfield neural networks are extended to handle inequality constraints where linear combinations of variables are lower- or upper-bounded. Then by eigenvalue analysis, the effects of the inequality constraints are analyzed and the following results are obtained: (a) if a combinatorial solution obtained by the networks satisfies the inequality constraints, the eigenvalues corresponding to the solution are the same as those without the inequality constraints; and (b) a combinatorial solution which satisfies the inequality constraints is stable if the energy, without the inequality constraints, of the solution is the smallest among those of the adjacent combinatorial solutions. From these results, the weights in the energy function are determined so that a combinatorial solution which satisfies the equality constraints, but does not satisfy the inequality constraints, is unstable. The results are verified for the knapsack problem and the transportation problem. For the latter problem, convergence to the optimal solution is improved by the introduction of the inequality constraints.

Experiences of Expert Systems for Power System Operation and Planning

Abstract In the stage of system planning and operation, Power system has three aspects; economic, logical and physical analytical aspects. Each aspects is carried out with conventional numerical calculation and also heuristic methodology such as expert systems. This paper presents some experiences of expert systems applied to power system planning and operation.

Research and development on expert systems applications to power systems in Japan

Abstract This paper describes the present status and future prospects for expert system development in the electric power industry in Japan. The paper first reviews the principal activities in expert system application to power systems: a research committee in the Institute of Electrical Engineers of Japan, a special committee in the Society of Electrical Cooperative Research, and an R&D group of utility members. Electric utilities have strong expectations for applying expert system technology to various tasks for power supply services. The paper examines such expectations and also summarizes the present situation of system development within the industry, based on an analysis of extensive questionnaires sent both to power companies and manufacturers. The principal techniques currently used in developing expert systems in this field are also reported. Finally, a future scenario for using expert systems in many activities for advanced power supply services at the beginning of the 21st century is presented. In Japan, there are some organizations where people exchange their information other than collaboration by universities, users and manufacturers in development. There are places for both competition and cooperation.

State Estimation for a Power System with Dense and Short Transmission Lines

Abstract This paper presents experimental results of applying several state estimation methods to a power system. This power system is located in a small island and has high load density. The fast decoupled weighted least squares method turned out to be not suitable to this system, because the assumption of the decoupling between P and Q is not appropriate, and the magnitude of impedances vary widely. As for the variety of the impedance magnitude, the ratio of the maximum to minimum values is larger than 1000. The usual bad data detection method using the statistical approach also turned out to be not applicable to this system, because of the trouble in maintaining accuracy in numerical calculation. To cope with these troubles, we adopted the weighted least squares method without decoupling between P and Q, and developed a new bad data detection method without matrix computation. The method was validated by numerical experiments.