APPLIED MATHEMATICS VIA PARAMETERIZATION AND ARTIFICIAL NEURAL NETWORKS FOR STUDY AND IMPROVEMENT OF ELECTRICAL POWER SYSTEMS

• Grant number: 23/03114-9
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: September 01, 2023
• End date: August 31, 2024
• Field of knowledge: Engineering - Electrical Engineering - Power Systems
• Principal Investigator: Alfredo Bonini Neto
• Grantee: Wesley Prado Leão dos Santos
• Host Institution: Faculdade de Ciências e Engenharia. Universidade Estadual Paulista (UNESP). Campus de Tupã. Tupã , SP, Brazil

Abstract

Currently, the continuous trend of increasing load demands, together with economic and environmental constraints for the construction of new power plants and transmission lines, has led power systems to operate closer to their limits, increasing the probability of stability problems. The studies related to static voltage stability require the qualitative evaluation of several operating conditions of the system under different loading conditions and contingencies. The continuation method is one of the main tools used in these studies due to their robustness and versatility, and has been used among others: in the evaluation of the effects of variations of parameters of transmission lines on the power system, in the observation of the behavior of the tensions of the bus system and in the comparison of planning strategies aiming at the adequate proposition of enlargements and reinforcements of the network with the intention of avoiding the cut of load. The publications show an increasing interest on the part of the companies of the electrical sector even in small improvements of this method aiming at the improvement of its performance in the several studies. In the continuation method, the Jacobian matrix singularity is removed by adding parameterized equations in the load flow equations. From the previous studies, created a great expectation that it is possible to be used several plans for the complete design of the P-V curve, an example is the plane formed by the magnitudes versus the angles of the node voltages, resulting in a trajectory of solutions with a linear aspect and thus, enabling the removal of singularity during the tracing of all P-V curve. Another important factor to be investigated is the use of artificial neural networks (ANNs) to analyze the load margin and consequently to obtain the point of maximum loading of the electrical systems. Thus, this project aims to continue the analysis of these techniques of geometric parameterization and also, through the RNAs improve their efficiency in obtaining the load margin and in the reduction of computational time spent by the geometric parameterization techniques.