研究目的
Investigating the systematic data-driven process for detecting and diagnosing faults in the regenerative braking system of hybrid electric vehicles.
研究成果
The systematic data-driven fault detection and diagnosis approach presented in this paper demonstrated good diagnostic accuracy and can be used for fault analysis in any vehicle system. It has the potential for real-time implementation in automotive and aerospace systems due to the significant reduction in the data size without compromising fault classification accuracy at a reduced computational load/time.
研究不足
The approach requires significant amount of data from monitored variables under nominal and faulty scenarios for data-driven analysis. It is difficult to apply the model-based approach to large-scale systems because it requires detailed analytical models in order to be effective.
1：Experimental Design and Method Selection
The diagnostic process involves signal processing and statistical techniques for feature extraction, data reduction for implementation in memory-constrained electronic control units, and variety of fault classification methodologies to isolate faults in the regenerative braking system.
2：Sample Selection and Data Sources
The RBS model from PSAT is segregated into multiple ECUs and hardware components. The ensuing model consists of a driver model, a component (physical system) model and six ECUs, namely battery control unit, engine ECU, motor1 control unit, motor2 control unit, mechanical brake control unit, and powertrain controller.
3：List of Experimental Equipment and Materials
The RBS is modeled using Powertrain System Analysis Toolkit (PSAT), a vehicle simulation software tool. Although the model is designed using PSAT software, it is converted to run independently in MATLAB/Simulink. Communication among the controllers (ECUs) occurs via a controller area network (CAN) bus and the communication aspects of the model are simulated using the Vector CANoe software.
4：Experimental Procedures and Operational Workflow
The MATLAB/Simulink and Vector CANoe co-simulation environment allowed us to inject faults, evaluate the operational and faulty behavior of the system by monitoring the residuals, and infer/diagnose the failed components.
5：Data Analysis Methods
The fault diagnostic process for RBS primarily involves the following steps: firstly, feature extraction/data reduction techniques are employed to extract the most informative features from the simulation data; secondly, trending and/or threshold-based fault detection tests are designed to detect the occurrence of fault(s); and finally, pattern recognition-based classification techniques, distance-based measures and hidden Markov model (HMM) based inference algorithms are applied to isolate faults in the RBS.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
