Gm Global Technolgoy Operations, Inc.

Dongyi Zhou

Xinyu Du

Shengbing Jiang

Mutasim Abdurrahman Salman

Wen-Chiao Lin

Xiaoyu Huang

Kevin A Cansiani

Atul Nagose

Rod Niner

David Gumpert

Aaron D Motyl

A method of identifying a fault in a friction brake actuated by hydraulic brake pressure and configured to decelerate a vehicle road wheel includes detecting, via a first sensor, a vibration at the road wheel and communicating data indicative of the detected vibration to a controller. The method additionally includes detecting, via a second sensor, upon application of the hydraulic brake pressure, a hydraulic brake pressure variation and communicating data indicative of the detected hydraulic brake pressure variation. The method additionally includes comparing, via the controller, the data indicative of the detected vibration with a threshold vibration value and the data indicative of the detected hydraulic brake pressure variation with a threshold hydraulic brake pressure value. Furthermore, the method includes generating a sensory signal indicative of the brake fault when magnitudes of the detected vibration and the detected hydraulic brake pressure variation are greater than their respective threshold values.

Detection of a friction brake fault

A system includes control modules, a low-voltage communications bus, e.g., a CAN bus of a vehicle, a voltage sensor that measures a bus voltage and outputs 2.5-3.5 VDC high-data and 1.5-2.5 VDC low-data, and a host electronic control unit (ECU). The host ECU detects a recoverable fault using a data pattern in the bus voltage data when the data is outside of a calibrated range, and recalibrates the sensor. Recalibration may be by adjustment to a scaling factor and/or a bias value. Non-recoverable 'stuck-at-fault'-type or 'out-of-range'-type faults may be detected using the pattern, as may be a ground offset fault. A method includes measuring the bus voltage using the sensor, comparing the output data to a range to detect the fault, and isolating a sensor fault as a recoverable fault using the data pattern when the data is outside of the range. The sensor is then be recalibrated.

Analog-to-digital fault detection, isolation, and mitigation for a low-voltage communications network

Growing community of inventors

Sterling Heights, MI, United States of America

Dongyi Zhou