Iranian Researcher Offers New Method for Optimal Switch Placement in Power Distribution Systems

“Optimal Placement of Switching Devices in Electric Power Distribution Systems Using Machine Learning Methods" is the title of Ph.D. dissertation of Mehrdad Ebrahimi at Iran’s Shiraz University.

“Supplying high-quality and uninterrupted electrical power is one of the most important goals of power system planners. Also, the wideness of the electric power distribution system and the variety of equipment have made the distribution systems to be exposed to various damages, such as the collision of trees and birds with electrical equipment.

Therefore, it is important to create suitable infrastructure and equip the electrical power distribution system with switching equipment such as circuit breakers, separators to facilitate and accelerate the power outage management process,” the Shiraz University researcher said.

Machine learning (ML) is a field of study in artificial intelligence concerned with the development and study of statistical algorithms that can learn from data and generalize to unseen data and thus perform tasks without explicit instructions.

“Today, the high accuracy and efficiency of machine learning methods have made them very useful in solving many problems such as foreign currency prices prediction, image processing, and translation from one language to another,” he said, adding that the ML is a growing trend that has been used for controlling, planning and developing power systems in recent years.

According to the researcher, the use of machine learning methods to solve power system planning problems is in two ways: In the first case, the ML algorithm is used as an independent processor to solve the problem. The second form is the use of ML algorithms along with mathematical methods. In the second case, the problem solving algorithm is a combination of ML and mathematical methods, and ML algorithms are used as a helper to reduce the volume of problem solving processing and increase the accuracy and efficiency of mathematical methods.

“One of the achievements of this research project is to provide accurate and scalable models that make problem solving possible for large real systems. This is while using the existing mathematical optimization methods, solving the problem for large systems faces a challenge due to the complicated calculations and the limits to computer processors,” he continued.

“The development of knowledge in relation to the use of ML methods for the optimal placement of switching equipment, the feasibility and design of internal software based on artificial intelligence for the design and planning of electric power distribution systems and the drawing a roadmap for the optimal development of electric power distribution systems are among the goals of this project,” Ebrahimi concluded.

4155/i