Dissertation/ Thesis
Robust Nonlinear Estimation and Control of Clutch-to-Clutch Shifts
| العنوان: | Robust Nonlinear Estimation and Control of Clutch-to-Clutch Shifts |
|---|---|
| المؤلفون: | Mishra, Kirti D. |
| بيانات النشر: | The Ohio State University / OhioLINK, 2016. |
| سنة النشر: | 2016 |
| المجموعة: | Ohiolink ETDs |
| Original Material: | http://rave.ohiolink.edu/etdc/view?acc_num=osu1452184768 |
| مصطلحات موضوعية: | Mechanical Engineering, Clutch-2-clutch, robust control, nonlinear estimation, transmission, model order reduction, shift hydraulics |
| الوصف: | Shift quality is an integral component of drivability. Thus, with the extension of feedback control to more powertrain functions, it is of paramount importance that systematic and robust control methods be developed for ensuring good shift quality. Increasingly, automatic transmissions employ a combination of two actively controlled clutches to implement shifts, which makes the gearshifts clutch-to-clutch shifts. In the current study, a model-based approach is presented for the development of a robust clutch-to-clutch shift controller. A detailed nonlinear model of the powertrain consisting of engine, torque converter, transmission system, and vehicle dynamics is developed. The transmission system consists of mechanical and hydraulic subsystems. While the mechanical subsystem is easier to model, the hydraulic subsystem, commonly referred to as shift hydraulic system, is usually challenging to model due to high dynamic order and significant nonlinearities. Starting from a high order nonlinear model of the shift hydraulic dynamics, developed from first principles in an earlier study, we apply systematic methods to arrive at a lower order control-oriented model, and use it to develop a closed loop clutch pressure controller. Specifically, state-space averaging and singular perturbation techniques were used to achieve this. The reduced order model was validated against experimental measurements.Automotive systems in general, and transmission systems in particular, lack adequate sensing essential to successful operation of feedback controllers; thus there is a need for observers for online estimation of important operating variables. In particular, the output shaft torque, turbine torque, reaction torque at the offgoing clutch, and online clutch pressures are critical to the performance of the nonlinear controller proposed in the current study. An observer scheme is proposed to estimate these variables, with the emphasis on robustness to modeling error and parametric variations. A combination of robust Luenberger and unknown input sliding mode observers (UI-SMO) is used to solve the estimation problem. The estimates show reasonable correspondence with the actual signals, even in the presence of appreciable parametric uncertainty.A closed-loop torque phase controller is proposed to ensure good coordination between the offgoing and oncoming clutches, which is an open problem in controlling clutch-to-clutch shifts. The controller uses information on load transfer during the torque phase, represented by the estimated reaction torque at the offgoing clutch, to manipulate the offgoing clutch pressure so that it behaves functionally similar to a one-way clutch. A secondary objective of controlling the duration of the torque phase is also served by this controller. During the inertia phase, a reference trajectory for the oncoming clutch slip speed is specified, tracking which would satisfy the conflicting control objectives associated with this phase, namely, smooth synchronization of the input and output shaft speeds of the transmission system and a short phase duration. Specifically, for each phase, a combination of sliding mode and feedback linearization-based controllers are used. The sliding mode controller implements closed-loop clutch pressure control, where the reference clutch pressure trajectories are generated by the feedback linearization controller. Again, as was the case with estimator design, robustness of the proposed nonlinear controller against different sources of uncertainty was deemed essential. Three strategies to achieve robustness were proposed and validated numerically. The third strategy is novel and can serve as a method for incorporating robustness into feedback linearization-based controllers in general. The validation results are encouraging and indicate feasibility of the proposed observer-based controller. |
| Original Identifier: | oai:etd.ohiolink.edu:osu1452184768 |
| نوع الوثيقة: | Text |
| اللغة: | English |
| الاتاحة: | http://rave.ohiolink.edu/etdc/view?acc_num=osu1452184768 |
| Rights: | unrestricted This thesis or dissertation is protected by copyright: all rights reserved. It may not be copied or redistributed beyond the terms of applicable copyright laws. |
| رقم الانضمام: | edsndl.OhioLink.oai.etd.ohiolink.edu.osu1452184768 |
| قاعدة البيانات: | Networked Digital Library of Theses & Dissertations |
| الوصف غير متاح. |