电涡流测功机加载控制系统的研制
发布时间：2018-10-21 08:10
【摘要】：在对车辆发动机与变速箱组成的动力传动系统进行性能试验时,大都采用电涡流测功机作为扭矩加载装置。测功机加载装置的扭矩加载精度与动态性能,决定了动力传动装置性能试验结果的好坏。因此,对于动力传动装置试验测试设备中电涡流测功机控制系统的研究具有重要的实际价值。在确定了测功机加载控制系统的总体方案后,本文完成了加载系统的硬件驱动板与控制板设计,编写了系统软件,搭建了加载系统的软硬件平台。论文在分析了电涡流测功机工作原理的基础上,建立了测功机关于电流与转速、扭矩的机理模型,并利用加载系统的现场实验数据加以验证,证明了该机理模型的正确性,并应用于后续的系统仿真研究中。在控制算法方面,以单神经元自适应PID为系统控制器,双隐层动态递归神经网络和状态延迟输入动态递归神经网络为系统辨识器,设计了基于动态递归神经网络的单神经元自适应PID控制算法。通过实际测试数据训练辨识网络,利用恒扭矩与恒转速实验对算法进行仿真验证。仿真实验证明,该算法能够实现对加载系统的恒扭矩与恒转速控制,并达到较高的控制精度和较快的响应速度。两种辨识器相比,以状态延迟输入动态递归神经网络为辨识器的控制单元具有更强的系统逼近能力。本文设计的电涡流测功机加载控制系统已经在实际系统中得到应用,其不仅能够完成对动力传动装置的扭矩加载,而且具有良好加载精度和动态响应速度。
[Abstract]:Eddy current dynamometer is used as torque loading device in the performance test of power transmission system composed of vehicle engine and gearbox. The torque loading precision and dynamic performance of dynamometer load device determine the performance test result of power transmission device. Therefore, it is of great practical value to study the control system of eddy current dynamometer in the test equipment of power transmission. After the overall scheme of load control system of dynamometer is determined, the hardware driving board and control board design of loading system are completed, the system software is written, and the hardware and software platform of loading system is built. Based on the analysis of the working principle of the eddy current dynamometer, the mechanism model of the dynamometer about current, speed and torque is established and verified by the field experiment data of the loading system, which proves the correctness of the mechanism model. And applied to the subsequent system simulation research. In control algorithm, single neuron adaptive PID is used as system controller, double hidden layer dynamic recurrent neural network and state delay input dynamic recurrent neural network are used as system identifiers. A single neuron adaptive PID control algorithm based on dynamic recurrent neural network is designed. The identification network is trained by actual test data, and the algorithm is simulated and verified by constant torque and constant speed experiments. The simulation results show that the algorithm can control the constant torque and speed of the loading system, and achieve higher control accuracy and faster response speed. Compared with the two kinds of identifiers, the control unit with state delay input dynamic recurrent neural network as the identifier has stronger system approximation ability. The load control system of the eddy current dynamometer designed in this paper has been applied in the practical system. It can not only complete the torque loading of the power transmission, but also have good loading accuracy and dynamic response speed.
【学位授予单位】：北京理工大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：U467.4
本文编号：2284500
[Abstract]:Eddy current dynamometer is used as torque loading device in the performance test of power transmission system composed of vehicle engine and gearbox. The torque loading precision and dynamic performance of dynamometer load device determine the performance test result of power transmission device. Therefore, it is of great practical value to study the control system of eddy current dynamometer in the test equipment of power transmission. After the overall scheme of load control system of dynamometer is determined, the hardware driving board and control board design of loading system are completed, the system software is written, and the hardware and software platform of loading system is built. Based on the analysis of the working principle of the eddy current dynamometer, the mechanism model of the dynamometer about current, speed and torque is established and verified by the field experiment data of the loading system, which proves the correctness of the mechanism model. And applied to the subsequent system simulation research. In control algorithm, single neuron adaptive PID is used as system controller, double hidden layer dynamic recurrent neural network and state delay input dynamic recurrent neural network are used as system identifiers. A single neuron adaptive PID control algorithm based on dynamic recurrent neural network is designed. The identification network is trained by actual test data, and the algorithm is simulated and verified by constant torque and constant speed experiments. The simulation results show that the algorithm can control the constant torque and speed of the loading system, and achieve higher control accuracy and faster response speed. Compared with the two kinds of identifiers, the control unit with state delay input dynamic recurrent neural network as the identifier has stronger system approximation ability. The load control system of the eddy current dynamometer designed in this paper has been applied in the practical system. It can not only complete the torque loading of the power transmission, but also have good loading accuracy and dynamic response speed.
【学位授予单位】：北京理工大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：U467.4
【相似文献】
相关期刊论文 前10条
1 王宏伟,刘清华;电涡流测功机扭矩波动原因浅析[J];设备管理与维修;2001年05期
2 黄秀权,秦贵和,沈长妹,武文治,赵宏伟,崔文博;电涡流测功机自动加载控制[J];吉林工业大学学报;1991年04期
3 杨敬源;宋玉杰;韩道权;;电涡流测功机及其作为能量吸收装置的应用[J];机械工程师;2008年04期
4 邓灿明;;汽车工况法电涡流测功机故障诊断和设备维护[J];机电工程技术;2014年04期
5 曾贤波;金振华;周斌;田颖;卢青春;;电涡流测功机—发动机系统低转速不稳定点的分析研究[J];内燃机工程;2006年02期
6 黄明强;电涡流测功机的正确选型与合理使用[J];车用发动机;1994年04期
7 张续凯;张文明;周海勇;;新型电涡流测功机大功率直流线性稳压电源[J];现代电子技术;2009年10期
8 ;CW型电涡流测功机简介[J];车用发动机;1990年05期
9 田颖;曾贤波;卢青春;;电涡流测功机控制器控制算法研究[J];内燃机工程;2005年06期
10 程广伟;周志立;邓楚南;徐立友;;基于遗传算法的电涡流测功机预测控制[J];农业机械学报;2008年07期
相关重要报纸文章 前1条
1 张梦华;国内外部分牵引试验负荷车现状（二）[N];中国农机化导报;2006年
相关硕士学位论文 前1条
1 范爽;电涡流测功机加载控制系统的研制[D];北京理工大学;2016年
，本文编号：2284500
本文链接：https://www.wllwen.com/kejilunwen/qiche/2284500.html
