|
Development of a Robust Adaptive, Model-Based Nonlinear Control Strategy for an Automotive Powertrain, 03-9466 Printer Friendly VersionPrincipal Investigator Inclusive Dates: 04/01/04 - Current Background - Emerging markets in Asia and Europe have production lines that are tooled for manual transmissions. In North America, the truck and heavy-vehicle industry has also used manual transmissions in its drive trains. There is a new effort in these markets to preserve the manual transmission and retrofit an actuator-controller package to make them automated manual transmissions (AMTs). The AMT system is inherently nonlinear because of the manufacturing tolerance stack up of the transmission and the friction at the clutch, shift forks, and synchronizers. Clutch engagement during vehicle launch is therefore a challenging control problem. Approach - The purpose of this project is to develop a control strategy to execute automatically the gearshift and clutch disengagement/engagement using actuators and sensors developed specifically for this application. The goal is to minimize the shift time while emulating the smoothness of a professional driver. The developed solution will enable customization efforts to various production transmissions. A basic control package to automate a manual transmission has been developed (Autoshifter) by SwRI under a cooperative R&D program with an industrial sponsor. This internal research effort will enhance this package with improved mechanism and control algorithms. A commercially available light-duty truck will be used as a test platform. The manual transmission has the typical H-pattern to move between one to five gears and reverse gear. The transmission will be retrofitted with suitable clutch and gearshift mechanisms. The speed of clutch mechanism is critical to the controllability of the engine-clutch combination. Hence the clutch mechanism will be optimized for minimum time between engage and disengage positions without enhancing the component cost excessively. Shift and select actuators will move the shift fork to minimize time between extreme gears in each straight shift. For cross shifts that involve multiple shift forks, special algorithms will be used to find the neutral gate. This rapid location of neutral gate will minimize overall shift time. A fuzzy logic based controller will co-ordinate the feathering between engine throttle and clutch to accomplish a smooth vehicle launch. Control techniques such as quantitative feedback theory and H-infinity will be used as needed to achieve a robust and repeatable shift and clutch displacement control. Component wear will be minimized by speed matching synchronizers and clutch plates during gear engagement. Accomplishments - To date, the following accomplishments have been completed:
|