The principal research objective of this GOALI project is to study a class of hybrid systems with a view towards developing novel, reliable, on-board diagnostic and monitoring algorithms for air brakes in large vehicles such as trucks, buses and trains. An air brake system in a typical truck consists of ten brake chambers whose input (mass flow rate of air) is governed by the brake pedal input from the driver. The pressure evolves in the air brake chamber depending on the whether the preload of the spring resisting the motion of the piston and pushrod has been overcome or whether the piston and pushrod are in motion or whether the pushrod has moved sufficiently that the brake pads have come in contact with the drum. In this project, the following problems will be dealt with: Given an experimentally corroborated mathematical model of the evolution of pressure in each brake chamber and relevant valves, the brake pedal input and the measurement of pressure in each brake chamber, is it possible (1) to reliably detect and diagnose leaks in the air brake system, and (2) to reliably estimate the stroke of the pushrod? Motivated by this application, sequential hybrid systems will be considered. Such systems consist of finite number of discrete states subject to one possible forward transition and one backward transition. The problem of parameter identification in the transition conditions of general sequential hybrid systems will then be investigated.
A healthy brake system in trucks, buses and trains is essential to guaranteeing safety of vehicles and their occupants on the highway. Given their sheer weight and given that air brakes are sensitive to maintenance, regular brake inspections are necessary for trucks and buses. Not only are such inspections subjective, time and infrastructure intensive, they are at times inconvenient for inspectors to get underneath the vehicle, especially for vehicles with low clearance. The two main brake system defects are leaks and the pushrod stroke exceeding a certain limit. Air leaks degrade the performance of the brake system by increasing the time lag in the pressure response and by limiting the maximum possible brake pressure. The larger the stroke of the pushrod, the longer is the dead-time and the lag in the pressure response. For this reason, this project aims to develop leak detection and pushrod stroke estimation algorithms in support of developing a portable diagnostic system, where data concerning the air brake system from the sensors is processed and stored on a chip which is readily accessible to inspectors via hand-held devices. From an educational viewpoint, this project will enable the undergraduate and graduate students to work on a real-world problem, interact with industrial researchers at Meritor-WABCO and implement the developed algorithms on the experimental test-beds at Texas A&M University and at Meritor-WABCO's test facilities. The economic, social and health implications of the project cannot be overemphasized in view of the safety implications and the optimal method for detecting faults in the component of a vehicle most prone to malfunctioning.
