Fuzzy logic-model predictive control energy management strategy for a dual-mode locomotive. (1st February 2022)

• Record Type: Journal Article
• Authors: Rodriguez, Rusber; F. Trovão, João Pedro; Solano, Javier
• Is Part Of: Energy conversion and management. Volume 253(2022)
• Publication Date: 2022-02-01
• Subjects: Energy management strategy -- Model predictive control -- Fuzzy logic control -- Load prediction algorithm -- Hybrid electrical vehicle -- Fuel cell -- Battery -- Supercapacitor; Direct energy conversion -- Periodicals; Energy storage -- Periodicals; Energy transfer -- Periodicals; Énergie -- Conversion directe -- Périodiques; Direct energy conversion; Periodicals; 621.3105
• DOI: 10.1016/j.enconman.2021.115111
• Languages: English
• ISSNs: 0196-8904

Abstract

Graphical abstract: Highlights: It proposes a modular energy management system (EMS) for hybrid electric vehicles. It uses a simple linear model for energy sources to apply model predictive control. It proposes a load consumption predictor based on fuzzy logic. It applies the proposed EMS to a dual-mode locomotive. The proposed EMS guarantees a low power variation of the fuel cell and battery. Abstract: Fuel cell hybrid electric vehicles (FC–HEV) combine the high energy density of hydrogen with a high-power density energy storage system. This favors the response to sudden changes in load and the vehicle's autonomy. A challenge on FC–HEVs is to develop an adequate energy management strategy (EMS) to determine the power distribution among the available sources. This paper proposes a modular EMS for a dual-mode locomotive FC–HEV. The EMS seeks to minimize a cost function, including the energy cost and embedded sources degradation. This EMS uses i) fuzzy logic control to guarantee the state of charge of the energy storage system (ESS) at the desired range, ii) model predictive control (MPC) to define the operation of the ESS and iii) rule-based control to manage additional sources and ensure power balance. The MPC strategy uses a fuzzy logic predictor algorithm for load prediction based on previous power and speed values. The proposed EMS is validated in a hybrid locomotive equipped with a fuel cell, batteries, and supercapacitors. Results show 95% global electrical efficiency and 0.20 and 1.38 power variation coefficient for fuel cell and battery, respectively.