Diminution of Smart Grid with Renewable Sources Using Support Vector Machines for Identification of Regression Losses in Large-Scale Systems. (8th August 2022)

• Record Type: Journal Article
• Authors: Teekaraman, Yuvaraja; Kirpichnikova, Irina; Manoharan, Hariprasath; Kuppusamy, Ramya; Angadi, Ravi V.; Thelkar, Amruth Ramesh
• Other Names: K K Aruna Academic Editor.
• Is Part Of: Wireless communications and mobile computing. Volume 2022(2022)
• Publication Date: 2022-08-08
• Subjects: Wireless communication systems -- Periodicals; Mobile communication systems -- Periodicals; 621.38205
• DOI: 10.1155/2022/6942029
• Languages: English
• ISSNs: 1530-8669

Abstract

Abstract : This article examines the effect of smart grid systems by implementing artificial intelligence (AI) technique with application of renewable energy sources (RES). The current state generation smart grid system follows a high demand on supply of equal energy load to all grid states. However, in conventional techniques, high demand is observed as manual operation is preformed and load problems are not solved within the stipulated time period due to lack of technological advancements. However, applications of AI in smart grid process reduces risk of operation as manual adjustments are converted to highly automated procedures. This type of automatic process identifies the fault location at stage 1 and diagnosis of identified faults will be processed at stage 2. The abovementioned two stage processes will be incorporated with two constant parameters as dummy load is produced to overcome high- to low-power flows. Additionally, a scrap model has been designed to reduce the wastage of power as 100 percent effective progress can be achieved for low- to high-power supplies. To detect the corresponding regression losses in the grid systems, support vector machine (SVM) which completely identifies the previous state loss in the system is integrated. Hence, to analyze the effectiveness of the SVM model, four different scenarios are evaluated and compared with heuristic algorithms, long short-term memory (LSTM), autoregressive indicated moving average (ARIMA), adaptive ARIMA, and linear regression models with distinct performance analysis that includes error in percentage values where a total efficiency of 81% is achieved for projected SVM in all power lines including large-scale systems as compared to existing approaches.