嵌入式可編程邏輯控制器算法中英文翻譯

1、<p><b>　　外文文獻翻譯</b></p><p>　　Advanced control algorithms embedded in a programmable</p><p><b>　　Abstract</b></p><p>　　This paper presents an innovative

2、 self-tuning nonlinear controller ASPECT (advanced control algorithms for programmable logic controllers). It is intended for the control of highly nonlinear processes whose properties change radically over its range of

3、operation, and includes three advanced control algorithms. It is designed using the concepts of agent-based systems, applied with the aim of automating some of the configuration tasks. The process is represented by a set

4、 of low-order local linear </p><p>　　Keywords: Control engineering; Fuzzy modelling; Industrial control; Model-based control; Nonlinear control; Programmable logic controllers; Self tuning regulators</p&g

5、t;<p>　　1. Introduction</p><p>　　Modern control theory offers many control methods to achieve more efficient control of nonlinear processes than provided by conventional linear methods, taking advanta

6、ge of more accurate process models (Bequette, 1991; Henson & Seborg, 1997; Murray-Smith & Johansen, 1997). Surveys (Takatsu, Itoh, & Araki,1998; Seborg, 1999) indicate that while there is a considerable and g

7、rowing market for advanced controllers, relatively few vendors offer turn-key products. Excellent results of advanced control</p><p>　?。?）Because of the diversity of real-life problems, a single nonlinear co

8、ntrol method has a relatively narrow field of application. Therefore, more flexible methods or a toolbox of methods are required in industry.</p><p>　?。?）New methods are usually not available in a ready-to u

9、se industrial form. Custom design requires considerable effort, time and money.</p><p>　　（3）The hardware requirements are relatively high, due to the complexity of implementation and computational demands.&l

10、t;/p><p>　　（4）The complexity of tuning (Babusˇ ka et al., 2002) and maintenance makes the methods unattractive to nonspecialised engineers.</p><p>　?。?）The reliability of nonlinear modelling is oft

11、en in question.</p><p>　　（6）Many nonlinear processes can be controlled using the well-known and industrially proven PID controller. A considerable direct performance increase (financial gain) is demanded whe

12、n replacing a conventional control system with an advanced one. The maintenance costs of an inadequate conventional control solution may be less obvious.</p><p>　　The aim of this work is to design an advance

13、d controller that addresses some of the aforementioned problems by using the concepts of agent-based systems (ABS) (Wooldridge & Jennings, 1995). The main purpose is to simplify controller configuration by partial au

14、tomation of the commissioning procedure, which is typically performed by the control engineer. ABS solve difficult problems by assigning tasks to networked software agents. The software agents are characterized by proper

15、ties such as autonom</p><p>　　This work does not address issues of ABS theory, but rather the application of the basic concepts of ABS to the field of process systems engineering. In this context, a number o

16、f limits have to be considered. For example: initiative is restricted, a high degree of reliability and predictability is demanded, insight into the problem domain is limited to the sensor readings, specific hardware pla

17、tforms are used, etc. The ASPECT controller is an efficient and user-friendly engineering tool for impl</p><p>　　The controller parameters are automatically tuned from a nonlinear process model. The model is

18、 obtained from operating process signals by experimental modelling,using a novel online learning procedure. This procedure is based on model identification using the local learning approach (Murray-Smith & Johansen,1

19、997, p. 188). The measurement data are processed batch-wise. Additional steps are performed before and after identification in order to improve the reliability of modelling, compared to adap</p><p>　　The con

20、troller comprises the run-time module (RTM) and the configuration tool (CT). The RTM runs on a PLC, performing all the main functionality of real-time control, online learning and control performance monitoring. The CT,

21、used on a personal computer (PC) during the initial configuration phase, simplifies the configuration procedure by providing guidance and default parameter values. The outline of the paper is as follows: Section 2 presen

22、ts an overview of the RTM structure and describes its</p><p>　　2. Run-Time Module</p><p>　　The RTM of the ASPECT controller comprises a set of modules, linked in the form of a multi-agent system

23、. Fig. 1 shows an overview of the RTM and its main modules: the signal pre-processing agent (SPA), the online learning agent (OLA), the model information agent (MIA), the control algorithm agent (CAA), the control perfor

24、mance monitor (CPM), and the operation supervisor (OS).</p><p>　　2.1. Multi-faceted model (MFM)</p><p>　　The ASPECT controller is based on the multi-faceted model concept proposed by Stephanopou

25、lus, Henning, and Leone (1990) and incorporates several model forms required by the CAA and the OLA. Specifically, the MFM includes a set of local first- and second-order discrete-time linear models with time delay and o

26、ffset, </p><p>　　which are specified by a given scheduling variable s(k).The model equation of first order local models is </p><p><b>　　(1)</b></p><p>　　while the model

27、equation of second order models is</p><p><b>　?。?）</b></p><p>　　where k is the discrete time index, j is the number of the local model, y(k) is the process output signal, u(k) is the

28、 process input signal, v(k) is the optional measured disturbance signal (MD), du is the delay in the model branch from u to y, dv is the delay in the model branch from v to y, and ai,j, bi,j, ci,j and rj are the paramete

29、rs of the jut local model. The set of local models can be interpreted as a Takagi–Surgeon fuzzy model, which in the case of a second order model can be expressed </p><p><b>　?。?）</b></p>&

30、lt;p>　　Where bj( k) is the value of the membership function of the jut local model at the current value of the scheduling variable s(k). Normalized triangular membership functions are used, as illustrated in Fig. 2.&l

31、t;/p><p>　　The scheduling variable s(k) is calculated using coefficients kr, ky, ku, and kv, using the weighted sum</p><p><b>　　(4)</b></p><p>　　The coefficients are config

32、ured by the engineer according to the nature of the process nonlinearity.</p><p>　　2.2. Online Learning Agent (OLA)</p><p>　　The OLA scans the buffer of recent real-time signals, prepared by the

33、 SPA, and estimates the parameters of the local models that are excited by the signals. The most recently derived parameters are submitted to the MIA only when they pass the verification test and are proved to be better

34、than the existing set. The OLA is invoked upon demand from the OS or autonomously, when an interval of the process signals with sufficient excitation is available for processing. It processes the signals batch-w</p>

35、;<p>　　2.2.1. Copy signal buffer</p><p>　　The buffer of the real-time signals is maintained by the SPA. When the OLA is invoked, the relevant section of the buffer is copied for further processing.<

36、;/p><p>　　2.2.2. Excitation check</p><p>　　A quick excitation check is performed at the start, so that processing of the signals is performed only when they contain excitation. If the standard devi

37、ations of the signals r(k), y(k), u(k), and v(k) in the active buffer are below their thresholds, the execution is cancelled.</p><p>　　2.2.3. Copy active MFM from MIA</p><p>　　The online learnin

38、g procedure always compares the newly identified local models with the previous set of parameters. Therefore, the active MFM is copied from the MIA where it is stored. A default set of model parameters is used for the lo

39、cal models that have not yet been identified (see Section 2.3).</p><p>　　2.2.4. Select local models</p><p>　　A local model is selected if the sum of its membership functions bj(k) over the activ

40、e buffer normalized by the active buffer length exceeds a given threshold. Only the selected local models are included in further processing.</p><p>　　2.2.5. Identification</p><p>　　The local mo

41、del parameters are identified using the fuzzy instrumental variables (FIV) identification method developed by Blazˇ icˇ et al. (2003). It is an extension of the linear instrumental variables identification procedure (Lju

42、ng, 1987) for the specified MFM, based on the local learning approach (Murray-Smith & Johansen, 1997). The local learning approach is based on the assumption that the parameters of all local models will not be estima

43、ted in a single regression operation. Compared to th</p><p>　　Estimation. The vector of parameters and the covariance matrix are updated only if the absolute weighted difference between the process output an

44、d its prediction is above the configured noise threshold.</p><p>　　In case of lack of excitation in the branch from u to y or in the model branch from v to y (or when measured disturbance is not present at a

45、ll), variants of the method with reduced parameter estimate vectors are used.</p><p>　　2.2.6. Verification/validation</p><p>　　This step is performed by comparing the simulation output of a sele

46、cted local model with the actual process output in the proximity of the local model position. The normalized sum of mean square errors (MSEj) is calculated. The proximity is defined by the membership functions bj. For ea

47、ch of the selected local models, this step is carried out with three sets of model parameters: and the set with the lowest MSEj is selected. Then, global verification is performed by comparing the simulation out</p&

48、gt;<p>　　2.2.7. Model structure estimation</p><p>　　Two model structure estimation units are also included in the OLA. The dead-time unit (DTU) estimates the process time delay. The membership functio

49、n unit (MFU) suggests whether a new local model should be inserted. It estimates an additional local model in the middle of the interval between the two neighboring local models that are the most excited. The model is su

50、bmitted to the MIA if the global validation of the resulting fuzzy model is sufficiently improved, compared to the original fuzzy mo</p><p>　　2.3. Model Information Agent (MIA)</p><p>　　The task

51、 of the MIA is to maintain the active MFM and its status information. Its primary activity is processing the online learning results. When a new local model is received from the OLA, it is accepted if it passes the stabi

52、lity test and its confidence index is sufficient. If it is accepted, a “ready for tuning” flag is set for the CAA. Another flag indicates whether the local model has been tuned since start-up or not. If the model confide

53、nce index is very low, the automatic mode may be dis</p><p>　　2.4. Control Algorithm Agent (CAA)</p><p>　　The CAA comprises an industrial nonlinear control algorithm and a procedure for automati

54、c tuning of its parameters from the MFM. Several different CAAs may be used in the controller and may be interchanged in the initial configuration phase. The following modes of operation are supported:</p><p&g

55、t;　　_ Manual mode: open-loop operation (actuator constraints are enforced).</p><p>　　_ Safe mode: a fixed PI controller with conservatively tuned parameters.</p><p>　　_ Auto mode (or several aut

56、o modes with different tuning parameters): a nonlinear controller. The CAAs share a common interface of interaction with the OS and a common modular internal structure, consisting of three layers:</p><p>　　1

57、. The control layer offers the functionality of a local linear controller (or several local linear controllers simultaneously), including everything required for industrial control, such as handling of constraints with a

58、nti-windup protection, bump-less mode switching, etc.</p><p>　　2. The scheduling layer performs real-time switching or</p><p>　　Scheduling (blending) of tuned local linear controllers, so that i

59、n conjunction with the control layer a fixed-parameter nonlinear controller is realized.</p><p>　　3. The tuning layer executes the automatic tuning procedure of the controller parameters from the MFM when th

60、e MIA reports that a new local model is generated if auto-tuning is enabled. The parameters of the control layer and the scheduling layer are replaced in such a manner that real-time control is not disturbed. Three CAAs

61、have been developed and each has been proved effective in specific applications: the Fuzzy parameter-scheduling controller (FPSC), the dead-time compensation controller (D</p><p>　　2.4.1. Fuzzy parameter-sch

62、eduling controller</p><p>　　An overview of the FPSC is shown in Fig. 4.</p><p>　　ARTICLE IN PRESS</p><p>　　The control layer of the FPSC includes a single PID controller in the form

63、 suitable for controller blending using velocity-based linearization. It is equipped with anti-windup protection and bump-less transfer. The scheduling layer of the FPSC performs fuzzy blending of the controller paramete

64、rs (in the case of Ti, its inverse value) according to the scheduling variable s(k) and the membership functions bj(k) of the local models. The instrument of velocity-based linearization enables the dynamic</p>&l

65、t;p>　　2.5. Control performance monitor (CPM)</p><p>　　The CPM scans the buffer of recent real-time signals for recognizable events. When events are detected, it estimates the features of closed-loop contr

66、ol response and an overall performance index. Like the OLA, it is invoked autonomously or upon demand from the OS and runs as a low-priority task. It consists of three modules: the buffer pre-processor (BP), the situatio

67、n classifier (SC), and the performance estimator (PE), as shown in Fig. 5. When the CPM is invoked, the BP copies the relevant sec</p><p>　　Regime options are listed below:</p><p>　　_ The OLA an

68、d/or the CPM may be invoked autonomously (during regular operation) or upon OS demand (following scheduled experiments), or both.</p><p>　　_ The OLA may estimate the process dead-time continuously or not.<

69、;/p><p>　　_ The OLA may attempt to insert additional local models when appropriate, or estimate the local models at the fixed pre-selected positions only.</p><p>　　_ Controller retuning may be trig

70、gered automatically immediately after each change of the model in MIA(“adaptive” operation), or following an engineer’s confirmation (“self-tuning” operation).</p><p>　　_ Online learning may also be used for

71、 monitoring of the process dynamics without the intention of controller tuning, either by adaptation of the model or by cross-validation of a fixed model. At start up, the system is initialized from a configuration file,

72、 which may put it into any phase of the</p><p>　　Configuration procedure. Afterwards, the configuration procedure may be continued or repeated using the HMI dialogue windows.</p><p>　　3. Configu

73、ration tool</p><p>　　The CT is used on a PC, connected to the PLC running the RTM. The CT contains a configuration“wizard” that guides the engineer through the typical scheduling controller commissioning pro

74、cedure. It is intended for less experienced users. Experienced engineers may find it more efficient to perform configuration only using the RTM, where other sequences of the tasks are possible. The procedure is decompose

75、d into small steps (25 dialogue windows). In each step, instructions are displayed and default</p><p>　　The main phases of the configuration procedure are:</p><p>　　_ Basic settings: selection o

76、f the control signals, the signal limits, the sampling time, the control algorithm, the scheduling variable, the model order._ Safe mode configuration: estimation of the process dynamics (experimentation and identificati

77、on using the RTM may be used), self-tuning of the ‘‘safe’’controller parameters (using the RTM), optional performance verification._ Fuzzy model initialization: initialization of the model positions, initialization of t

78、he local model parameters, disp</p><p>　　_ CAA settings: initialization of the default values, advanced auto-tuning parameters.</p><p>　　_ OLA settings: initialization of the default values adva

79、nced OLA settings.</p><p>　　_ CPM settings: initialization of the default values advanced CPM settings.</p><p>　　_ Experiment settings: initialization of the default values advanced automatic ex

80、perimentation settings.</p><p>　　_ Local controller tuning: sequence of automatic (open- or closed-loop) experimentation, online learning and tuning using the RTM around each local model position.</p>

81、<p>　　_ Performance verification: sequence of automatic experimentation and performance evaluation using the RTM around each local model position. The CT relies on the functionality of the RTM wherever possible. Ho

82、wever, the PC development environment is more convenient for graphical user interface design and enables better visualization of results, compared to typical PLC systems.</p><p>　　4. Field test</p>&l

83、t;p>　　The ASPECT controller has been tested in several pilot applications, for example on a pH control benchmark (Blazˇ icˇ et al., 2003) and a gas. Liquid separator (Kocijan et al., 2003). This section presents a pil

84、ot application on an apparatus for testing hydraulic valves, located in a hydraulic equipment production plant. A simplied scheme of the apparatus is shown in Fig. 6. It comprises a boiler with local temperature control,

85、 three pumps, a pressure sensor, a valve test stand with a pressure d</p><p>　　(a) The steady-state relation between the pressure difference on the valve pv and the mass .own through the valve Qm (related t

86、o the pump rotation speed o) is quadratic,</p><p>　　(b) The openness of the valve Sv can be changed during a test, while the signal Sv is generally not available (manual valves), and</p><p>　　(

87、c) Different pumps (or combinations of pumps) can be used, according to the size of the valve. These factors severely affect the process dynamics, which results in the unsatisfactory performance of a previously existing

88、 control system based on a .axed PI controller.</p><p>　　Scheduling variable selection is a crucial step when applying a parameter-scheduling controller. While the nonlinearity (a) alone may be easily solved

89、 using scheduling from pv, the condition (b) makes the problem considerably more difficult. Process modelling was used to and a suitable scheduling variable1.</p><p>　　5. Conclusion</p><p>　　An

90、advanced self-tuning nonlinear controller has been successfully implemented on an industrial PLC platform. Several pilot applications, including the one presented in this paper, have also been successfully completed. Com

91、pared to the industry standard PI controller, an expected considerable improvement in the control performance was achieved using the FPSC control algorithm. Moreover, this performance was easily achieved in practice with

92、 self-tuning using the online learning procedure, by perf</p><p>　　to the reliability of adaptive control.</p><p>　　Acknowledgment</p><p>　　The authors would like to acknowledge the

93、 contributions of all other project team members. The ASPECT</p><p>　　Project was financially supported by EC under contract IST-1999-56407. ASPECT r2002 software is the property of INEA d.o.o., Indelec Euro

94、pe S.A., and Start Engineering JSCo. Patent pending PCT/SI02/00029.</p><p>　　嵌入于可編程邏輯控制器的先進算法</p><p><b>　　摘要</b></p><p>　　本文主要介紹一種新型的非線性自我調(diào)節(jié)器的可編程控制器的可編程邏輯控制器規(guī)則算法，被用于高階非

95、線性控制的處理，這種處理特性能從根本上啟用了代理服務(wù)系統(tǒng)，使用的目標(biāo)是使結(jié)構(gòu)工程實現(xiàn)自動化，這種處理描述一系列的低階的局部線性模型，這種模型的參數(shù)可以通過在現(xiàn)學(xué)習(xí)程序來識別，該程序?qū)⒛Ｐ妥R別與識別前后步驟連接起來以便提高可靠的這些操作，控制器可以監(jiān)視和評價閉環(huán)系統(tǒng)的控制性能，該控制器通過PLC實現(xiàn)其功能。</p><p>　　關(guān)鍵字：控制工程，模糊控制，工業(yè)控制，基于模型的控制，非線性控制，可編程邏輯控制，自我調(diào)

96、節(jié)器。</p><p><b>　　1、 引言</b></p><p>　　現(xiàn)代控制理論提供了很多控制方法，這些控制方法可用于達到比連續(xù)線性方法更有效的非線性控制處理，可以采用更有效的處理模型的優(yōu)點。調(diào)查表明如果有一個相當(dāng)可觀的并且正在逐步壯大的先進控制器市場，那么相對的就不會有很多小販倒賣盜版產(chǎn)品。</p><p>　　基于模糊參數(shù)規(guī)則的先進

97、控制觀念及到可喜的成績，復(fù)合控制模型或自適應(yīng)控制已經(jīng)在演講報告中提到過，然而，也有很多的關(guān)于這些方法在工業(yè)應(yīng)用中受到了限制，總結(jié)如下：</p><p>　　（1）現(xiàn)實生活的差異，一個單獨的非線性控制方法環(huán)只是在一個比較狹小的范圍內(nèi)應(yīng)用，因此我們需要更加行之有效的方法用于工業(yè)技術(shù)領(lǐng)域。</p><p>　?。? ）新的方法通常在實際工業(yè)中不能取得很好的效果，傳統(tǒng)的設(shè)計又要花費大量的人力、物力

98、和財力。</p><p>　?。?） 硬件要求相對要高，由于有復(fù)雜的執(zhí)行和估算要求。</p><p>　?。?） 復(fù)雜的調(diào)整和維修使得這些方法不會引起非專業(yè)工程師的注意。</p><p>　?。? ）非線性模型控制也經(jīng)常成為問題。</p><p>　?。? ）許多非線性處理可用于已知的工業(yè)PID控制器得到控制，當(dāng)我們要用一個高級控制器取代一個

99、連續(xù)控制系統(tǒng)時需要大量的直接的參數(shù)，并且不完全連續(xù)控制系統(tǒng)，維護費也會明顯的減少。</p><p>　　本文的目標(biāo)就是設(shè)計一個規(guī)則控制器用于解決一些前面提到的問題，而且是通過代理服務(wù)器來解決的，最主要的目的就是通過代理程序的部分自動化來簡化控制器結(jié)構(gòu)，這些代理程序主要是控制工程師來操作運行的，代理服務(wù)系統(tǒng)通過網(wǎng)絡(luò)軟件代理的分配工作來解決難題，軟件代理通過其性質(zhì)特征化，其中這些性質(zhì)有自控（無人直接參與的操作），社會