EVALUATION OF GENERATION SYSTEM RELIABILITY INDICES USING RBFNN METHOD

• Main Authors: Ravindra M. Moharil, Prakash S. Kulkarni
• Format: Article
• Language: English
• Published: Yeshwantrao Chavan College of Engineering, India 2021-02-01
• Series: Journal of Research in Engineering and Applied Sciences
• Subjects: expected energy not served (eens); forced outage rate (for); loss of load expectation (lole); orthogonal least square (ols); radial basis function neural network (rbfnn).
• Online Access: http://www.mgijournal.com/Data/Issues_AdminPdf/231/EVALUATION%20OF%20GENERATION%20SYSTEM%20.pdf
• ISSN: 2456-6403; 2456-6403

Electrical power system is changing from utility centric to customer centric after introduction of Electricity Act 2003 in India and unbundling of power sector in rest of the world. After unbundling, customer satisfaction has prime importance.Customer expectspower supply 24x7, which can be assuredonly after performingthe reliability analysis of generation system. The purpose of this paper is to present Neural Network (NN) approach, which can overcome limitationsof the conventionalreliability evaluation method such as poor accuracy, complicated models,and large timefor execution. This paper presents anorganized approach forestablishing the learning model with supervision using radial basis function neural network (RBFNN). This model is used for evaluation of various reliabilityindicesused for generation planning. Markov process and basic probabilistic approach is used to develop theinput-output training patterns for neural network. These patterns are normalized and presented to RBFNN. The validation of the proposed technique is confirmed by analyzingRoy Billinton Test System (RBTS) and IEEE-Reliability Test system (IEEE-RTS).