双发动机动力系统的控制研究 (1).pdf

南京航空航天大学 博士学位论文 双发动机动力系统的控制研究 姓名：潘松 申请学位级别：博士 专业：车辆工程 指导教师：魏民祥 20081101 南京航空航天大学博士学位论文 I 摘 要 多动力源动力装置不仅可以提高动力系统的动力性能，还增加了动力系统的可 靠性，即其中一个动力源发生故障失去动力以后不致造成整个动力系统瘫痪，但是 这却增加了动力系统的复杂程度，给动力系统的设计分析和控制研究带来了很大的 难度本文根据某国防创新项目要求，完成了某型双发动机动力装置的转速控制 该装置配备两台同型号的专用活塞式二冲程汽油发动机，在工作中需要对两台发动 机的转速进行控制 为了对双发动机动力系统进行理论研究，本文对发动机及动力传动系统进行了 数学分析建模根据双发动机系统的动力传递特性和负载特性，在理论分析和实验 验证的基础上得到了动力系统的模型针对所采用的二冲程活塞式发动机，采用系 统辨识的方法建立了发动机扭矩模型，采用模块化的建模方法建立了各个模块的数 学模型，并用仿真软件 Matlab/Simulink 进行了仿真模拟 根据课题要求，该动力装置工作中要保持发动机转速恒定，要求控制器具有较 强的鲁棒性。

首先，设计了鲁棒 PI 控制算法，采用线性矩阵不等式的方法得到了使 得闭环控制系统稳定的控制参数然后，根据被控对象的特性，采用真值模型的方 法构造了模糊 T-S 模型；基于模糊 T-S 模型，采用并行分配补偿策略设计了非线性 鲁棒控制器，采用 Lyapunov 稳定性理论对闭环系统的控制稳定性进行了分析，得 到了具有全局稳定性的状态反馈控制器 针对双发动机动力系统研究了双动力控制策略，搭建了双动力半物理仿真平 台通过在半物理仿真平台上对几种控制策略的仿真，研究了适合于双发动机的控 制策略，降低了实验成本，缩短了控制器的开发周期 以嵌入式 16 位微控制器为核心，设计了数据采集和控制硬件电路，并且根据 所研究的控制算法和控制策略完成了控制软件设计，实现了发动机的远程控制和测 试为实现发动机节气门的自动控制，在原发动机节气门的基础上设计了发动机节 气门远程操纵系统，根据操纵系统的特性，采用自适应脉动 PID 控制算法设计了节 气门控制器，实现了节气门的精确位置控制基于 Visual Studio /C++平台设计了地 面实验监控系统，可以完成对实验数据的实时观测、保存及历史数据的回放 为了验证所设计的控制算法和控制策略，采用所设计的发动机控制器，在实际 动力装置上进行了闭环发动机控制实验。

在实验中，在稳定工况和改变发动机的设 定转速时，控制器可以控制发动机的转速误差在规定的范围内，改变动力装置的负 载时，节气门在闭环控制下可以迅速响应，最终的控制结果达到了课题所提出的控 双发动机动力系统的控制研究 II 制要求和目标 关键词关键词：双发动机动力系统，发动机建模，鲁棒 PI，模糊 T-S 模型，线性矩阵不等 式，双发动机控制策略，嵌入式控制器 南京航空航天大学博士学位论文 III ABSTRACT With multi-engine the power system can enhance its performance and has more reliability, i.e. if one engine of the power system does not work the whole system will not lost its power. But multi-engine power system makes the system more sophisticated and makes its analysis, design and control more complex. According to the project’s requirement, this paper studied a control problem of a power system, which equipped with two special two-stroke piston gasoline engines, i.e., twin-engine power system. Twin-engine speed control and regulation is the main objective of this project. The twin-engine power system model was constructed for theoretical research. According to the properties of the power transmission system, a power model was presented through theoretical analysis and experiments. Based on the experimental data of a two-stroke gasoline engine, a steady-state torque model was established by mathematical tool of stepwise regression. The established engine model provided a basis for research of drive-train system and automatic control of engine. The whole model of the objective system was verified by Matlab/Simulink. It’s needed to keep the engine speed constant and the controller must be robust. A method for designing robust proportional-integral (PI) controller was presented. The problem of parameter designing of the robust PI controller was solved with linear matrix inequalities approach. According to the properties of researched system, the engine Takagi-Sugeno models were approached using sector nonlinearity. A nonlinear speed controller was developed, and a system augmented method was used for design of engine speed control algorithm without steady-state error. The stability conditions were derived using Lyapunov approach, and the problem was formulated as linear matrix inequality problems. Simulation results attested to the properties of these methods. Twin-engine power control strategies were researched through a self-constructed hardware-in-the-loop simulating platform. By simulating on the simulating platform, a suitable control strategy was confirmed to reduce the experimental cost, and shorten the development cycle. Based on 16bits embedded microcontroller, the circuit of control and data sampling was designed. The control strategies were realized through software, and the designed 双发动机动力系统的控制研究 IV controller can control the engine remotely. In order to control engine automatically, an engine throttle control system was researched. An adaptive impulse PID controller for throttle was presented, by which an exact position control was achieved. Based on Visual Studio/C++ platform, a ground monitoring system was developed, and on which the obtained data could be displayed and saved in real time. In order to testify the efficiency of researched control strategies, experiments on real power system were delivered through designed controller. During experimenting, the designed controller could track the reference speed and keep the engine speed fluctuating under permitted range. Under the close-loop control the throttle responds quickly while the load demand was changed. The satisfactory control results were achieved. Key Words: twin-engine power system, engine simulating, robust PI, T-S fuzzy model, linear matrix inequality, twin-engine control strategy, embedded controller 南京航空航天大学博士学位论文 IX 图表清单 图 1.1 动力装置的结构框图2 图 1.2 双发动机汽车动力系统的示意图3 图 1.3 双发动机单桨动力装置模型3 图 1.4 直升机动力系统模型4 图 1.5 旋翼和推进系统结构框图4 图 1.6 Cook 和 Powell 建立的模型结构框图5 图 1.7 四冲程发动机转速控制模型6 图 1.8 混合动力汽车结构图8 图 1.9 并列调速器并车控制方式9 图 1.10 主从调速器并车控制方式9 图 1.11 单调速器并车控制方式10 图 1.12 发动机 ECU 开发 HIL 结构. 11 图 2.1 动力装置总体结构框图13 图 2.2 传动系统结构14 图 2.3 模型辨识过程框图19 图 2.4 试验数据与辨识模型输出数据对比21 图 2.5 发动机模型22 图 2.6 节气门阶跃22 图 2.。