Scholarlink

In this study, a control strategy has been devised to reduce the diesel consumption in conventional trains while using solar photovoltaic (PV) power integrated with the battery storage system known as a tri-source hybrid system. Since management of three sources requires robust control, therefore, the controller is designed using two different techniques to ensure its robustness and validate its performance. A Fuzzy logic controller is compared with the Adaptive Neuro Interference System (ANFIS). Membership functions and their performance are validated using Root Means Square Method Error (RSME) comparison between the controllers Both the controllers have been tested for triangular, trapezoidal, and Gaussian membership functions. The results of ANFIS are compared with FLC to determine the most optimized control strategy. It is concluded that the root mean square error (RMSE) for ANFIS (0.261) is much lesser than the FLC (11.788).

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Performance Assessment Between Adaptive Neuro Fuzzy Interference System (ANFIS) and Fuzzy Logic Membership Functions to Reduce the Diesel Consumption for Electrical Load of Train in a Hybrid Power Generation System

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Zulfiqar Memon,
  • Abdullah Tariq Sipra,
  • Fawad Azeem

摘要

In this study, a control strategy has been devised to reduce the diesel consumption in conventional trains while using solar photovoltaic (PV) power integrated with the battery storage system known as a tri-source hybrid system. Since management of three sources requires robust control, therefore, the controller is designed using two different techniques to ensure its robustness and validate its performance. A Fuzzy logic controller is compared with the Adaptive Neuro Interference System (ANFIS). Membership functions and their performance are validated using Root Means Square Method Error (RSME) comparison between the controllers Both the controllers have been tested for triangular, trapezoidal, and Gaussian membership functions. The results of ANFIS are compared with FLC to determine the most optimized control strategy. It is concluded that the root mean square error (RMSE) for ANFIS (0.261) is much lesser than the FLC (11.788).