模糊控制在平台稳定回路系统中的研究

1、1 本课题研究的的目的，意义稳定回路是其中事关导航精度的关键部分。对平台稳定回路进行了建模，将模糊控制和带多个修正因子的模糊控制方案引入平台稳定回路的双闭环回路系统，并对此控制方案进行了仿真分析，理论上证明了模糊控制方案在平台稳定回路控制中的可行性。本课题将模糊控制引人平台稳定回路控制，理论研究了引入模糊控制器后系平台式惯性导航系统依靠由陀螺稳定的机械平台，为导航系统和姿态稳定系统提供测量基准，平台稳定回统整体性能，为模糊控制在稳定回路中的工程应用奠定理论基础。2 国内外研究现状模糊控制是以模糊集合论作为它的数学基础，1964年美国的LAZadeh教授创立了檬朔集合理论，提出用“隶属函数”概念

2、来定量描述事物模糊性，奠定了模糊数学的基础。自1974年英国的EHMamdani研制出第一个模糊控制器，进行蒸汽机和锅炉控制方面研究获得成功后，模糊控制的研究和应用一直十分活跃。1984年成立了国际模糊系统协会IFSA，模糊集与系统杂志与IEEE（美国电气与电子工程师协会）。近几年模糊控制已经应用于生活的各领域。随着科学的发展对惯导的要求也越来越高，相应的就对稳定回路领域提出了更高的要求，近年来国内外专家对稳定回路做了很多专门研究，实现有很多方法，有锁相环控制，数字控制，模糊PID控制，基于神经网络的PID控制等。3，拟采用的研究路线I，惯导平台的稳定原理与稳定回路的组成三轴液浮积分陀螺稳定平

3、台，具有三条参数不同而基本工作原理相同的伺服回路通道，用以保证平台台体相对于惯性空间稳定。当台体转动时，陀螺转子的主轴相对惯性空间要保持稳定，陀螺传感器输出陀螺主轴相对惯性空间的角差信号，经过放大和校正后馈送到平台力矩电机，力矩电机产生扭转力矩，使平台向减少角差的方向扭转，直至信号器输出为零，平台相应轴完成对陀螺主轴跟踪，平台稳定于惯性坐标系内。II，惯性平台稳定回路双闭环控制分析框图与被控对象数学模型 H为液浮积分陀螺的角动量；为陀螺传感器的放大倍数；K为耦合放大器和前置放大器的总放大倍数；伺服分解器变比系数；KI功率放大器放大倍数；Kg力矩马达放大倍数；Ka校正网络放大倍数；Wa(s)校正

4、网路；J内框组合件绕轴的转动惯量；J2浮筒组件绕进动轴的转动惯量；C：积分陀螺阻尼系数；力矩马达电枢绕组电磁时间常数；K反馈系数。 平台稳定回路单通道双闭环开环传递函数， III，平台稳定回路双闭环系统模糊控制研究IV，对于模糊控制在稳定回路系统中的应用研究软件设计我们可以采用MATLAB版中的Fuzzy Logic 工具箱，在模糊逻辑工具箱中建立模糊推理系统后，可以再Simulink中进行仿真分析。4，进度安排第一周-第三周 查阅资料，社会调研，文献检索，整理材料，完成开题报告及文献综述。第四周第六周 英文资料翻译，初步确定论文的实现方案第七周第九周 方案论证：理论分析论证，确定可行性方案，

5、选择元件，设计原理图，进行计算机仿真和优化设计。第十周-第十三周 完成各单元电路及总体电路硬件及软件编程设计。第十四周-第十六周 撰写论文 准备答辩5文献综述稳定回路是高精度惯导的关键技术之一，它的性能优劣直接影响着惯导系统的精度。随着科学技术的发展，对惯导精度的要求越来越高。相应的就对稳定回路这个控制系统就提出了更高的要求。这些年来，国内外很多专家都对稳定回路做了许多专门研究，实现的方法也是有很多种，但目前投入使用的最多也是PID控制方式。其中的模糊PID控制就是我们选择的一种常用控制方式，这种方式比较成熟，易于实现。对于稳定平台，最根本的任务就是准确的保持加速度计在空间的位置。惯性稳定平台

6、的主体有陀螺仪，加速器计及安装他们的台体与常平架系统构成。稳定回路的构成：对于控制的核心控制器-模糊控制器，模糊控制器主体部分是由计算机或单片机构成，多采用二维模糊控制结构，其控制过程大致为：计算机采集被测参数的精确值，并将该值与给定值作比较，得出误差e 以及误差变化率ec作为模糊控制器的输入语言变量，将e 和ec 经模糊化得到模糊量E 和EC，再由E 、EC 和模糊控制规则，根据推理的合成规则得到模糊控制量U ，U 经模糊决策、清晰化得到精确控制量u 。一个模糊控制系统得性能优劣，主要取决于模糊控制器的结构、所采用的模糊规则、合成推理算法，以及模糊决策的方法等因素。系统输出的观测量y(t)和

7、给定值r(t)比较，经过采样和A/D 转换得到一个确定的数字量，操作人员根据获得的信息和自己已有的操作经验进行比较分析、模糊推理，从而判定应该对被控对象作什么样的调整，因此在模糊自动控制中必须将操作人员的经验总结成若干条用自然语言描述的控制规则，利用模糊数学这一工具进行处理，构成一个模糊关系矩阵R 存放在计算机的存储中，这些规则称为模糊控制规则。仿照人脑的模糊推理过程，在模糊自动控制中，应该确定推理法则，做出模糊决策，模糊决策后经过模糊化处理，以便得到一个确切。这就构成了一个模糊控制器。 模糊控制器通常由模糊化、知识库、模糊推理和非模糊化几部分组成。模糊化的作用是将参考输入、对象输出等精确量转

8、化成模糊量，实际上就是得到模糊推理中的前提；模糊推理是模糊控制器的核心，具有模拟人的基于模糊概念的推理能力；非模糊化的作用是将模糊推理得到的控制量转化为实际可用的精确量；而知识库则包含了用于模糊化和非模糊化的各语言变量的模糊分割、隶属度函数等数据库信息，以及用于模糊推理的一系列控制规则构成的规则库，知识库构造了模糊推理中的前提。 模糊控制器的基本结构图中的点画框内是模糊控制器，它主要根据误差信号的大小和变化情况进行推理并给出正确的控制信号。模糊控制器通过计算机，根据由精确量转化来的模糊输入信息，按照总结人工控制策略而得到的语言控制规则进行模糊推理，并将其转化为精确量，作为反馈控制作用施加到被控

9、对象上。这反映了人在对被控对象进行控制时，不断将观察到的对象输出精确量转化为模糊量，经过人脑的思维与逻辑推理取得模糊判决后，再将判决的模糊量转化为精确量，去实现人工控制的整个过程。模糊控制器的各部分中，需解决以下三个方面的问题：1）精确量的模糊化，采用合适的隶属度函数，把语言变量的语言值转化为某适当论域上的模糊集合。2）知识库的设计，通过一组模糊条件语句构成模糊控制规则，并计算模糊控制规则决定的模糊关系，实际上就是构造合适的模糊控制算法。3）非模糊化，把模糊推理的结论通过适当方法转化成实际可用的精确量。这几个问题目前都没有惟一的求解步骤，具体采用何种方法需要从实际性能要求出发，通过不断调整而来

10、。可供参考的模糊控制器原则性设计步骤有：1定义输入输出变量；2定义所有变量的模糊化条件；3设计控制规则库；4设计模糊推理结构；5选择精确化策略的方法。根据设计原则，我们可以选择采用Mamdani极大极小推理法进行模糊逻辑推理。针对模糊控制的推广应用，MathWorks公司在其MATLAB软件中添加了Fuzzy Logic工具箱，该工具箱以其强大的功能和方便易用的特点得到了广泛应用。它的模糊逻辑工具箱提供了与SIMULINK的无缝连接功能。因此，我们在模糊控制逻辑工具箱中建立了模糊推理系统后，可以在Simulink仿真环境中对其进行仿真分析【1】， 参考文献来源：1，高金源等，计算机控制系统【M

11、】高等教育出版社，2010；2，刘金琨等，先进PID控制及其MATLAB仿真M北京：电子工业出版社，20033，李国勇，智能预测控制及其MATLAB实现【M】北京：电子工业出版社，20104，韦巍等，智能控制技术【M】机械工业出版社，20005，陈磊，惯性平台稳定回路的设计【J】西北工业大学，20066,6外文翻译外文资料Research on Fuzzy Control for Steam Generator Water LevelWei Peng Da-fa ZhangI. INTRODUCTIONThe steam generator is one of the main devices

12、 in PWR nuclear power plant, in order to ensure the safety of nuclear power plant during operation; the steam generators water level must be controlled in a certain range. When the nuclear power plant is running, as the steam flow or the water flow changing, the amount of boiling bubbles in the stea

13、m generator will change due to local pressure or temperature change, the instantaneous water level showed “false water level” phenomenon . The existence of “false water level” made it difficult to control the water level. The introduction of feed-forward control to the traditional single-loop PID co

14、ntrol can, in a certain extent, overcome the "false water level" phenomenon. But the conventional PID control method in the process of steam generator water level control has some shortcomings. To the steam generator that has highly complex, large time-delay and nonlinear time-varying char

15、acteristics, the PID parameters tuning is a tedious job and the control effect is very poor. Furthermore, to achieve good control performance still as conditions changing, it often needs to change the PID controller parameters. But the analog PID controller parameters are difficult to regulate onlin

16、e. Fuzzy control is a kind of nonlinear control strategy based on fuzzy reasoning, which express operating experience of skilled manipulation men and common sense rules of inference through vague language. Fuzzy control do not need to know precise mathematical model of controlled object, is not sens

17、itive to the change of process parameters, is highly robust and can overcome non-linear factors, so, fuzzy control has faster response and smaller ultra- tone, can get better control effect. Based on understanding above, this paper design a steam generator water level fuzzy controller, the simulatio

18、n shows that the controller has good control performance and practical value.II. DYNAMIC CHARACTERISTICS OF STEAM GENERATORThe transfer function of PWR steam generators mathematical model of the general form shows below:y(s)=GW(s)QW(s)+GS(s)QS(s) （1）where y is the steam generator water level; QW for

19、 the water flow; QS for the steam flow; GW (s) for the impact of the water flow to the steam generator water level; GS (s) for the effect of the steam flow (load) to the steam generator water level.The balance of the steam generator water level is maintained through the match between the water flow

20、and steam flow. The process that water level changes with the steam flow or water flow changing can be regarded as a simple integration process, but impact of the water flow and steam flow s change on water level is different.A. Dynamics Characteristics under Water Flow DisturbanceSuppose steam flow

21、 GS remains unchanged, and water flow GW step increases, on the one hand because the temperature of feed water is much lower than the temperature of saturated water in the steam generator, so that , when feed water entering, it will absorb a lot of extra heat, the vapor phase bubble contents will re

22、duce, resulting in water level decreasing; on the other hand, the increase in water flow GW made it greater than steam load, and cause water level increases linearly. Comprehensive two factors, after the step increase of the water flow, the water level rise has a time delay process, showing a down t

23、hen up.B. Dynamic Characteristics under Steam Load DisturbanceSuppose feed water flow GW remains unchanged, and steam load GS step increases, on the one hand the water level will flow down because the steam flow rate is greater than the water flow rate. On the other hand, as the steam load increased

24、, vapor pressure is reduced; the bubble volume on the liquid surface increases, causing the water level increased. Comprehensive two factors, after the step increase of the steam flow rate, the water level down has a time delay process, showing a up then down. The impact on the water level of water

25、flow or steam flow stepping decreased has similar principle as above.As analysis can be seen as above, when the water flow or steam load change, the water level did not follow the change immediately, but there is an opposite process at first. This phenomenon is called "false water level" p

26、henomenon.III. DESIGN OF WATER LEVEL FUZZY CONTROLLERThe conventional PID controller has a poor control performance to the steam generator that exist “false water level” characteristics, showing a greater overshoot in the tracking time. But a well-designed fuzzy controller is able to overcome the &q

27、uot;false water level" phenomenon, and has good control performance. A. Sstructure of Fuzzy ControllerThe structure showed in Figure 1.FuzzycontrollerK1valveGW(s)dx/dtK2GS(s)expected water leverStream flowWater leverwater floweec+-+-+Figure 1. Structure of steam generator water level fuzzy cont

28、rollerChoose the water level error (e) and change rate of error (ec) as input of the fuzzy controller, the output of the fuzzy controller is the added value of the valve opening signal u. Meanwhile, use the steam flow feed-forward to overcome the "false water level" phenomenon, use water f

29、low feedback to overcome fluctuations in water supply side . k1, k2 were water flow and steam flow transmitter conversion factor. To ensure the water flow to match the steam flow, k1 and k2 values should be equal to.B. Fuzzy theory, fuzzy subset and Membership FunctionThe fuzzy Analects of e, ec and

30、 u are -6, 6, both with seven fuzzy sets NB (negative big), NM (negative middle), NS (negative small), ZO (zero), PS (positive small), PM (positive middle) and PB (positive big) to describe. e, ec and, u are all using triangular membership function (see Figure 2).Figure 2. Input and output variable

31、membership functionC. Fuzzy control rule tableThe establishment principle of fuzzy control rules are: when the error is large, the output control volume should give priority to eliminate error as soon as possible; when the error is small, the output control volume should givepriority to prevent over

32、shoot. Where ec is negative ,it shows that water level has a rising trend, if the water level is high at this time, then we should reduce the valve opening signal; whereas, we should open the valve more. Through a comprehensive analysis of expertise, the establishment of rule table shown in Table 1.

33、Table 1. Fuzzy control rule table eecNBNMNSZOPSPMPBNBNBNBNMNMNSNSZONMNBNMNMNSNSNSZONSNMNMNSNSZOPSPSZONMNSNSZOPSPSPMPSNSNSZOPSPSPMPMPMNSNSZOPSPSPMPMPBZOPSPSPMPMPBPBD. Fuzzy Reasoning and SolutionThis fuzzy inference system uses Mamdani. The basic properties of fuzzy inference system set to: "and

34、" operation with a very small operation; "or" operation uses the maximum operation. Using a very small operation fuzzy implication, fuzzy rules integrated with great operations center Defuzzification method used. IV. SIMULATION EXAMPLESA pressurized water reactor steam generator in Ch

35、inese Qinshan nuclear power station has empirical model G1 (s), G2 (s) below:where Ps denote the rated load. When load at 15% 90% Ps, use (6) and (8); when load less than 15% Ps, use (7) and (8).Figure 3. Expected water level step response diagramThe coefficients in Control system are k1=k2=0.5. Wat

36、er control valve is a king of linear valve, its gain is 4. The quantitative coefficients of e and ec are 6 and 60 respectively; the scale factor of u is 0.5. We limit water flow the range of 0 kg / s to the rated flow 258kg / s when simulation. Consider the expected level step from the initial 0m to

37、 10m, water level response is shown use the solid line in Figure 3. For contrasting the increase effect of fuzzy controller, we also carried out using the traditional PID control simulation. We can see, compared with traditional PID control, fuzzy controller has reported significant improvements in

38、overshoot, settling time, steady degrees.V. CONCLUSIONThis paper designed a water level fuzzy control system aimed at steam generators characteristics of large time delay and model uncertainty. We also gave a simulation to the steam generator of Qinshan nuclear power plant, and achieved satisfactory

39、 results. The method can also be used for other large time -delay and time-varying process control model, and has broad application prospects.蒸汽发生器水位模糊控制研究彭威 张达发1.导论蒸汽发生器是压水反应堆式核电厂里的一个重要的设备。为了保证核电厂运行的安全性，蒸汽发生器的水位必须控制在一定的范围内。核电厂的运行中，因为蒸汽流量和给水流量的改变，蒸汽发生器里沸水中的气泡数量会随着局部气压和温度的变化而改变，瞬时水位呈现“虚假液位”现象。正是由于“虚假

40、液位”的存在使得水位控制变得困难。将前馈控制引入到传统的单回路PID控制中，可以在一定程度上克服“虚假液位”的问题。但是蒸汽发生器的传统PID控制仍然存在着一些不足。对于具有高度复杂，大滞后，非线性特征的蒸汽发生系统，不仅PID参数的调整单调乏味，控制效果也很差。并且当条件改变时，为了获得好的控制性能，通常需要改变PID控制器的参数，但是模拟量的PID控制器参数的在线调整是很难的。模糊控制是一种基于模糊推理的非线性的控制方法，它体现了熟练操作人员的实际经验和模糊语言推理的一般规则。模糊控制不需要知道被控对象的精确的数学模型，它对过程参数的变化并不敏感，鲁棒性很强，能够克服非线性因素，因此，模糊

41、控制有更快的响应速度，更小的超调，更好的控制效果。基于以上了解，本文设计了一个蒸汽发生器水位的模糊控制器，仿真结果表明这个控制器有更好的控制效果和实用价值。2.蒸汽发生器的动态特性压水堆蒸汽发生器一般形式的数学模型的传递函数如下所示:y(s)=GW(s)QW(s)+GS(s)QS(s) （1）其中，y代表蒸汽发生器的水位；QW代表给水流量；QS代表蒸汽流量；GW代表给水流量对蒸汽发生器水位的作用；GS代表蒸汽流量对蒸汽发生器的水位的作用。蒸汽发生器水位的平衡是靠蒸汽流量和给水流量的匹配来维持的。可以将水位随蒸汽流量或者给水流量变化而变化看作一个简单的一体化过程，蒸汽流量变化和给水流量变化对水位

42、的影响又是不同的。(1) 给水流量扰动下的动态特性假设蒸汽流量保持不变，而给水流量阶跃增加，一方面，由于新增给水的温度要比蒸汽发生器中的饱和水的温度低很多，因此，当新水进入后就会吸收大量的额外热量，水中的气泡含量大大减少，从而导致水位下降；另一方面，给水流量大于蒸汽负荷，引起水位线性增加。综合以上两点，当给水阶跃增加，水位增长会有一个延迟的过程，表现为先下降后上升。(2) 蒸汽负荷扰动下的动态特性假设给水流量保持不变，蒸汽负荷阶跃增加，一方面，由于蒸汽流速比给水流速大，水位会下降；另一方面，随着蒸汽负荷的增加，内部蒸汽压力降低，液面的气泡容积增加，从而引起水位增加。综合以上两个因素，当蒸汽流量阶跃增加以后，水位下降会有一个延迟的过程，表现为先上升后下降。给水流量或者蒸汽流量