【作者】 陈然；

【导师】 孙冬野；

【作者基本信息】 重庆大学 ， 车辆工程， 2010， 博士

【摘要】 双离合式自动变速器(DCT, dual clutch automatic transmission)通过两离合器的交替切换完成换挡过程,从根本上解决了机械式自动变速器(AMT)换挡动力中断带来的问题。它可以继续使用原有手动变速器的生产设备,生产继承性好,同时其具有很好的换挡品质和车辆动力性、经济性,非常适合我国国情。目前全球范围内的汽车DCT研究尚处于起步阶段,电子控制系统是DCT的关键技术,开展DCT电控系统的研究和产品化必将对提高我国自动变速器的自主开发能力具有积极的意义,为我国在汽车自动变速器领域赶超世界先进水平提供一次重要契机。为了满足整车动力性、经济性以及适应客户需求多样化,变速器电控系统的智能化复杂程度在不断提高,相应的开发难度也在不断增加,通过传统台架试验来对电控单元进行全面综合测试的传统方法,由于耗费大量人力物力,开发周期长,环境不稳定、可重复性差等缺点,已难以适应电控系统智能化、快速升级化的需求。虚拟试验技术作为一种先进的计算机仿真技术,它可以快速完成电控系统研发试验,缩减开发时间和费用,利用虚拟试验技术,可以开发出能够进行DCT电控单元ECU调试、检测及性能评估的虚拟环境,这对于高效、智能、低成本的进行DCT电控单元的开发和研制起到相当重要的作用。本文结合国家863高科技发展项目“经济型轿车高性能自动变速变速器开发”和重庆市科技攻关项目“自动变速器硬件在环仿真平台研究”的资助下重点进行了汽车DCT智能控制及电控单元虚拟试验系统的研究。具体工作内容如下:①系统地分析和研究了DCT车辆动力传动系统的特性,建立了相应的动力学模型。在系统分析DCT的扭矩传递动力学机理基础上,研究了不同工作状态及过渡过程中离合器扭矩传递动态特性,建立了DCT传动系统起步和换挡过程的动力学微分方程。分析了电子节气门的原理和特性,建立了节气门执行机构的非线性动力学模型。研究了基于永磁直流电动机的干式自动离合器执行机构原理和特性,建立了离合器执行机构的动力学模型。依据发动机的试验数据,提出了基于支持向量机的发动机模型辨识方法,研究支持向量机的核函数、损失函数及惩罚参数对系统辨识精度的影响,确定了各参数对模型精度影响的顺序,获得了使支持向量机网络辨识精度达到最优时各参数的取值,并以此建立了发动机转矩及油耗模型。为DCT智能控制策略的制定及电控单元虚拟试验奠定基础。②研究DCT智能控制体系结构。研制DCT电控系统的硬件,完成了电机、传感器等的选型,设计电控单元及接口电路。研究并提出基于分层递阶智能控制的DCT智能控制软件系统,建立基于感知层、中间控制策略层和执行层的DCT三层递阶智能控制结构,详细地分析基于野点剔除、数字滤波、测量算法三步法构成的DCT感知层的传感器信号软处理算法。解决了DCT复杂控制的协调合作问题,为DCT车辆智能控制系统设计提供了参考。③进行执行层控制算法研究。在对DCT中电动干式离合器的原理分析基础上,根据离合器执行机构的性能特点确定采用控制算法简捷且鲁棒性好的滑模控制算法,设计自适应模糊滑模控制器。较好地解决滑模控制的抖振及系统抗干扰问题,证明了控制系统的稳定性,为保证DCT精确控制提供了保障。④开展了DCT智能起步控制策略研究。综合考虑驾驶起步意图、发动机状态、离合器状态、冲击度、滑摩功等多方面因素,基于模糊智能控制方法,提出了双层模糊起步智能控制,制定了起步时离合器的接合模糊控制规则。试验结果证明所设计控制策略的正确性。⑤仿人式智能换挡规律研究。在综合了驾驶员—车辆—环境整个系统基础上,根据量化的驾驶员期望、节气门开度、相对车速及换挡经验模糊推理出驾驶期望,结合动力性因子,实时动态的修正换挡边界,给出了具体的仿人式智能换挡规律的设计方法。仿人式智能换挡规律不仅达到了动力性和经济性的协调,而且还可以避免在特殊行驶环境意外换挡现象,同时可以体现驾驶员意图,实现换挡规律由线到面的跨越。⑥DCT换挡过程控制研究。在分析DCT车辆换挡过程特点基础上,综合冲击度、滑摩功因素,根据离合器状态转移条件,以满足冲击度和滑摩功为边界条件,首次给出了换挡过程两离合器接合时序的定量计算公式方法,为DCT车辆换挡品质的控制提供了新的理论指导。⑦研制了DCT电控单元虚拟试验系统。利用Matlab/xPC Target、MATLAB/Simulink、Matlab/RTW并结合三维虚拟现实技术,将数据采集卡、微机、DCT电控单元、路由器等硬件相结合,研制了具有成本低廉、通用性强、人机交互、虚拟现实的DCT电控单元虚拟试验系统平台。并进行了DCT电控单元的性能试验与分析。试验结果表明:①所建立的DCT虚拟试验系统能够进行DCT电控单元控制系统开发;②所设计的DCT智能控制策略及执行器控制算法能够有效地控制汽车起步、换挡及在各种工况中有效行驶。验证了所设计的DCT智能控制的优劣性及虚拟试验平台的高效性。促进了DCT的产品化进程,为快速开发具有自主知识产权的DCT奠定了基础。更多还原

【Abstract】 Dual-clutch transmission (DCT) which has the intrinsic advantages in high efficiency and low production cost eliminates torque interrupts during gearshift through one clutch separated while another clutch engaged, so that the vehicles can run as smooth as one using an Automated Mechanical Transmission (AMT). Product line of manual transmission (MT) which is widely applied in transmission market of our country can be used to design and manufacture DCT, which is in favor of the upgrade from MT to automatic transmission and coincide with the situation of our country. Nowadays, the researches and application of DCT vehicle are still at an early stage, the study on key technology of DCT is a good chance to pursue the advanced automatic transmission technology of the world is provided to us.To satisfy the dynamic performance, fuel economy of the vehicle and diversification of product demand, the intelligent of electronic control unit has more and more intelligence,and it is more difficult to be developed. Conventionally, electronic control units of vehicle are developed and tested using open-loop testers and prototype vehicles. The final system’s validations are also performed in prototype vehicles at various geographical locations. This is not only expensive but also time consuming and inflexible. Virtual test technology can accomplish tests rapidly, shorten developing time and save study expense. Based on a virtual test bench, ECU’s performance can be examined and evaluated, and the system can play an important role in studying and developing the ECU with higher efficiency and lower cost.In this paper, intelligent control technology of DCT and the virtual test system of DCT electronic control unit are carried out relying on a project of National High-tech R&D Program (863 Program)“Development of high performance automatic transmission for economical car”. The main work as follows:①Dynamic characteristic of DCT during vehicle running are analyzed, and the its dynamic model are built. Theory and characteristic of electronic throttle are analyzed, and the nonlinear model of electronic throttle is described. The BLDC motor-driven dry dual clutch actuator is studied, the mathematical model has been established for the system. Focus on the nonlinearity and time-varying properties of engine, a novel engine identification model is proposed based on support vector machine (SVM). The model is based on engine test data and structure risk minimization (SRM) principle. It has very good generalization ability. Using the MATLAB,many precision influencing factors of engine torque are analyzed, and the precision influencing sequence of these parameters is confirmed. Considering the interaction among the parameters, the optimal parameters of SVM engine model is found out using loop nest program method.②Study on the intelligent control architecture of DCT. The overall electrical control system program and the electronic control system hardware and software structure are determined, the motor and the sensors are selected. A hierarchical intelligent control software system of DCT is proposed。A three-level controlling architecture which consists of perception layer, control strategy layer and execution layer is built. Signal processing algorithm of DCT perception layer based on abnormal data deletion, digital filtering and measure algorithm is described in detail.③A sliding mode controller(SMC) for clutch actuator is designed based on the actuator nonlinear dynamic mode. In order to alleviate the chattering phenomenon of SMC and uncertainty of actuator, an adaptive fuzzy logic based sliding mode control (AFSMC) is proposed. The stability of the control system and convergence of clutch actuator position tracking errors are guaranteed by Lyapunov stability theory. Simulation results show that the proposed control system can achieve favorable tracking performance and robustness with regard to parameter variations and external disturbances.④Research of intelligent control technology during DCT vehicle starting. The engaging process of the clutch is analyzed, clutch starting strategy based on fuzzy control which comprehensively considers the driver intention, engine running state and clutch engaging state is presented. Then, fuzzy control parameters are selected and fuzzy control logic rule is presented. The simulation results prove the correctness of the designed control strategy.⑤Study on human simulating shift schedule. Considering the driver-vehicle-road, the driving expects can be reasoned out by fuzzy control from the driver’s intention, electronic throttle opening and relative speed. Dynamic factor is proposed and the design method of human simulating shift schedule is presented. Simulation results show that the shift point alters dynamically according to driver’s intentions, human simulating shift schedule can eliminate the improper shifting under special situation, and the shift schedule converts from shift curve to shift area.⑥Analysis on gearshift of DCT. Considering factors jerk and slipping work, based on the jump condition of clutch, satisfying the boundary conditions jerk and slipping work, quantitative formula of clutch engagement timing sequence is presented. It provides a design method for developing an intelligent vehicle with DCT.⑦The research of DCT ECU virtual test platform development. According to the scheme of DCT ECU virtual test platform, the ECU virtual test platform was built based on Matlab/xPC Target, MATLAB/Simulink, Matlab/RTW and Virtual Reality (VR). The platform consisted of test bench, sensors, clutch actuators, electronic throttle, real-time platform, signal collecting system and software system. The system feature low cost, more versatility, man-machine interaction, virtual reality, proprietary intellectual property rights. Some ECU virtual tests on some conditions were preceded on the platform based on control algorithm prototype. The test simulations showed:①The DCT ECU virtual test proved that the ECU platform was qualified to the test.②The proposed intelligent control strategy of DCT and control algorithm of clutch actuator can well control DCT vehicle under various conditions; for instance, vehicle starting and shifting. Test results show that the intelligent control of DCT is the superior and the virtual test platform features the high efficiency, thus established basic value purport for the proprietary intellectual property rights of DCT.更多还原