基于CAN总线的神经元智能控制器的研究
详细信息    本馆镜像全文|  推荐本文 |  |   获取CNKI官网全文
摘要
CAN总线是一种目前比较流行的现场总线,以其高可靠性和独特设计被公认为最有前途的现场总线之一。本文分析现场总线和CAN总线的特点和优点,讨论了CAN总线的通信协议,CAN控制器的结构和功能,分析其应用范围和前景。
     单神经元控制算法不依赖于控制对象的数学模型,具有自学习能力,算法简单,控制精度高,自适应性和鲁棒性都比较强,其算法可由单片机系统来实现,特别适合于参数不易辨识的黑箱系统和在过程中参数会改变的系统,已经在很多领域得到了很好的应用并且取得了很好的效果。本文分析单神经元控制几种不同学习规则下的学习算法,从理论上对其自适应性和鲁棒性进行讨论和分析。
     基于以上两种先进的技术,设计基于CAN总线的智能控制器,即控制器采用单神经元的控制算法,并具有CAN通讯的功能,可与其他智能仪表一起组成现场总线控制系统。
     以Matlab/Simulink为工具,建立算法模型和控制对象的数学模型进行仿真,通过改变不同的参数观察输出曲线,从而确定算法中参数对控制效果的影响。利用Simulink建立了状态变量模型、单神经算法模型。然后通过仿真分析了算法的阶跃特性,动态跟随特性以及抗干扰性。并通过与常规PID控制算法的比较分析。同时借助九点控制算法的思想提出了九点在线修改参数的改进算法。
     设计了一个控制器,从硬件和软件两方面探讨了控制器的实现过程。设计了通信接口部分,人机接口部分和存储器部分的硬件电路。用C51进行软件设计,包括CAN通讯部分和单神经元算法,给出了主要部分程序框图。
     本课题的研究成果,将进一步促进现场总线技术和单神经元控制算法这两种先进技术的融合与发展,对于促进现场总线控制系统的进一步推广,乃至国内的自动化水平的提高都有一定的意义。
CAN (Controller Area Network) bus is a kind of popular field bus, which is recognized as one of the most promising field bus with its high reliability and particular design style. The characteristics and the merits of the field bus and CAN bus are analyzed, and then the communication protocol of CAN bus and the structure and function of CAN controller are discussed in this dissertation. Furthermore, its application scope and the perspective are proposed.
    The control algorithm of single neuron possesses many characteristics such as self-study ability, simplicity and high precision, which do not depend on the mathematics model of the plant. In addition, its self-adaptability and robustness behave well too. This algorithm can be implemented by MCU (Micro-programmed Control Unit), especially suits for the black-box system whose parameters identification is difficult and for the system in which parameters may change in process. It has been applied in many fields and achieved good effectiveness. Some study algorithms of single neuron control under different study rules have been introduced in this dissertation. The self-adaptability and robustness of the single neuron are discussed and analyzed in theory.
    Based on the two kinds of advanced technology mentioned above,an intelligent controller is designed, which adopts the algorithm of single neuron and possesses CAN communication ability. It can be used to form a field bus control system in combination with other intelligent instruments.
    The algorithm model and the mathematics model of the plant are set up and simulated by the software Matlab/Simulink. The influence of the parameters upon the control result can be assessed with changing parameters and observing the output curve. The state variable model and the algorithm model of single neuron can be set up with Simulink. Consequently the step response, dynamic
    
    
    
    track and anti-disturbance performance of the algorithm can be analyzed, whose results are compared with those of the traditional FED algorithm. Moreover, the improved algorithm is advanced, in which parameters can be modified online based on the nine-point control algorithm.
    At last, a controller is designed in which the hardware circuits and software are introduced. The controller mainly consists of communication interface, human-machine interface and the memorizer. The program including CAN communication and the control algorithm of single neuron is programmed with C51 language. The main program diagrams are depicted.
    The research result of this dissertation will contribute to the combination and the development of the field bus technology and the control algorithm of single neuron. It will have great signification on the development of field bus control system and the advancement of the national automation level.
引文
[1]阳宪惠.现场总线技术及其应用.清华大学出版社,2001,2
    [2]俞文光.分布式控制系统的现状及趋势.自动化与仪表,1996,11(3):20~23
    [3]宋祖涛、毛静涛.自动化仪表的重大变革—论现场总线智能仪表.上海电力学院学报,2000,16(4):35~41
    [4]邬宽明.CAN总线原理和应用系统设计.北京航空航天大学出版社,1996,11
    [5]刘建昌.现场总线概述.基础自动化,2000,7(6):1~4.
    [6]孙立红,崔敬花.现场总线及其应用.微处理机,2000,(4):4~8.
    [7]付兴建,申东日、陈义俊.智能传感器与现场总线技术,2000,20(2):63~68
    [8]张培仁,王洪波.独立CAN总线控制器SJA1000.国外电子元器件,2001:20~23
    [9]黄飞,赵秀菊.基于DS80C390芯片的CAN总现在智能楼宇中的应用.微型机与应用,2001:22~23
    [10]蒋建文,林勇,韩江洪.智能型CAN总线测控节点的研究.仪器仪表学报,2001,22(3):282~283
    [11]陈杨,刘曙生,龙志强.基于CAN总线的数据通信系统研究.测控技术,2000,19(10):53~55
    [12]程军.用87C196CA实现CAN总线通信的方法.微型机与应用,21~23,2000,4
    [13]朱坚民.基于单神经元的智能温度控制仪的研制.洛阳工学院学报,2000,21(1):32~37
    [14]李光明.单神经元自适应PID智能控制器及其应用.工业仪表与自动化装置,1997,16(2):43~47
    [15]姜平.神经元自适应PID控制器的研究.工业仪表与自动化装置,1997,39(5):668~671
    [16]田杰.纯滞后系统的单神经元Smith预测控制.计算机仿真,2001,18(1):40~44
    
    
    [17]纪宝,王勇.神经元自适应PSD控制器在直流调速系统中的应用.江苏理工大学学报,2000,21(4):32~36
    [18]杜立强,韩兵欣.啤酒发酵过程的单神经元自适应PID控制.河北建筑科技学院学报,2001,18(2):40~44
    [19]王顺晃,谢四江.单神经元自适应PSD智能控制器及其应用.河北理工学院学报,1997,19(2):39~44
    [20]王以骥,涂健.神经元网络控制.机械工业出版社,1999,7
    [21]段军.基于单神经元智能PID控制的柴油机数字式电子调速系统.大连铁道学院学报,1999,3
    [22]刘卫国,马瑞卿.无刷直流电动机控制系统的单神经元控制器.工业大学学报,1999,17(1):124~129
    [23]许伟明,居滋培,张凤登.单神经元自适应控制器在液位控制中的应用.上海理工大学学报,2001,23(2):180~182
    [24]杨智,高靖.神经元自适应预测PID控制器及实现.信息与控制,1999,28(5):345~349
    [25]郭创新,梁年生,叶鲁卿.基于神经网络实现水轮机自学习PID调节.华中理工大学学报,1999,(3):79~86
    [26]李巍,郎力,马树青.一种改进的单个自适应神经元控制学习算法.哈尔滨工业大学学报,1997,29(2),50~53
    [27]宋宜斌.一种神经元自适应伺服控制器的研究.计算机工程与应用,1999,(3):35~37
    [28]翟守忠.用于现场总线的新型控制器.矿业研究与开发,1999,19(3):19~22
    [29]徐颍莉.一种适用的工业过程测控网络CAN.交通与计算机,2000,2
    [30]蔡立虹,陶煜.符合现场总线的智能执行器.工业仪表与自动化装置,2000,(3):61~64
    [31]肖海荣,周风余.基于SJA1000的CAN总线系统智能节点设计.计算机自动测量与控制,2000,9(2):48~51
    [32]张义和.Protel PCB 99 设计与应用技巧.科学出版社2000,6
    [33]闻新,周露.MATLAB神经网络应用设计.科学出版社,2000,9
    [34]陈在平.基于MATLAB(Simulink)语言的智能控制系统仿真.天津理工学院,1998,4
    
    
    [35]孙涵芳.INTEL16位单片机.北京航空航天大学出版社,1995,6
    [36]马忠梅.单片机的C语言应用程序设计.北京航空航天大学出版社,1998,10
    [37]刘乐善.微型计算机接口技术原理及应用.华中理工大学出版社.1996,3
    [38]李人厚.智能控制理论和方法.西安电子科技大学出版社,1999,10
    [39]陈静,范宣志,童名文.一种基于泛布尔代数的空调控制方法.武汉理工大学学报,2002,24(1):50~52
    [40]陈静,范宣志,童名文.AT89C51单片机在超发电计量仪中的应用.自动化与仪器仪表,2002,102(4):28~33
    [41]陈静,童名文,范宣志.ICL7135与单片机AT89C51的接口设计方法.中国科技论坛,(2001增刊):202
    [42]Nobes, T. SMART instruments and Fieldbus. Nuclear Engineer, 42(2) Mar/Apr 2001
    [43]Piggin, R. A fieldbus for machine safety, IEE Review; 46 (4) Jul 2000
    [44]CAN specification 2.0, part A. Can in automation, Am Weichselgarten26, D~91058 Erlangen
    [45]Jonas Berge. Fieidbus Enables Innovative Measuirments. Advances in Instrumentation and Control, Vol.51, 1996
    [46]PHILIPS Semiconductors. CAN controller interface-PCA82C250, 2000, 3
    [47]PHILIPS Semiconductors. SJA1000 Stand- alone controller application note. Philips Electronics, 1997
    [48]DALLAS Semiconductor. Data Book and CD ROM, 2000
    [49]Davis Chapman(骆长乐译).学用Visual C++6.0,清华大学出版社,1999,9
    [50]lliam Stallings(张娟,王海,林东,张兴元译).Data and Computer Communications (5th Eidition),《数据与计算机通信(第5版)》.PRENTICE HALL PUBLICATION,电子工业出版社,1996,3

© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号

地址:北京市海淀区学院路29号 邮编:100083

电话:办公室:(+86 10)66554848;文献借阅、咨询服务、科技查新:66554700