Ma, Zedong (China North Vehicle Research Institute), Lin, Shibo (Dalian University of Technology), Chai, Hua (China Northern Engine Research Institute), Kang, Mingxin (Northeastern University), Wu, Yuhu (Dalian University of Technology)

Mixed Fault Diagnosis Algorithm for Thermal Management System of Heavy-Duty HEV Based on Logical Threshold and Fuzzy Logic Methods

Scheduled for presentation during the Regular session "Thermal Management" (ThB4), Thursday, October 31, 2024, 13:50−14:10, Room T4

7th IFAC Conference on Engine and Powertrain Control, Simulation and Modeling, Oct 30 - Nov 1, 2024, Dalian, China

This information is tentative and subject to change. Compiled on May 16, 2025

Abstract

The system complexity of hybrid electric vehicles (HEVs) brings more challenges to the maintenance of the powertrain system. Especially to the heavy-duty HEV vehicle, the health conditions of the thermal management system (TMS) determines the engine, and motor's normal operation. There are many potential feature states to influences the system health conditions, including the temperature difference and pressure difference between inlet and outlet of the radiator, coolant flow rate, etc. Since the feature states often indicate the inconsistent fault status and interact with each other, therefore the exact fault diagnosis results is difficult to obtain. To address this problem, this paper proposes a mixed fault diagnosis algorithm for TMS of heavy-duty HEV based on logical threshold and fuzzy logic methods. Firstly, the heat dissipation model of TMS is established consisting of three independent cooling cycles for engine, motors, and generators, respectively. And then the fault simulation can be realized by adjusting the key model parameters. Secondly, the logical threshold method is introduced for judge a preliminary diagnosis state. Thirdly, a fuzzy logic scheme is designed to deduce the final diagnosis results based on logical threshold method, to improve the diagnosis precision. Finally, some typical fault simulation experiments have been conducted and the results demonstrate the effectiveness of the proposed mixed fault diagnosis algorithm.