[go: up one dir, main page]

WO1996009573A2 - Arrangement for the adaptive control of a section - Google Patents

Arrangement for the adaptive control of a section Download PDF

Info

Publication number
WO1996009573A2
WO1996009573A2 PCT/DE1995/001223 DE9501223W WO9609573A2 WO 1996009573 A2 WO1996009573 A2 WO 1996009573A2 DE 9501223 W DE9501223 W DE 9501223W WO 9609573 A2 WO9609573 A2 WO 9609573A2
Authority
WO
WIPO (PCT)
Prior art keywords
controller
parameters
neural network
route
control
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/DE1995/001223
Other languages
German (de)
French (fr)
Other versions
WO1996009573A3 (en
Inventor
Hans-Peter Preuss
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Siemens AG
Siemens Corp
Original Assignee
Siemens AG
Siemens Corp
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Siemens AG, Siemens Corp filed Critical Siemens AG
Publication of WO1996009573A2 publication Critical patent/WO1996009573A2/en
Publication of WO1996009573A3 publication Critical patent/WO1996009573A3/en
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05BCONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B13/00Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
    • G05B13/02Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
    • G05B13/0265Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric the criterion being a learning criterion
    • G05B13/027Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric the criterion being a learning criterion using neural networks only

Definitions

  • the invention relates to a device for adaptive control of a route according to the preamble of claim 1.
  • a control loop there consists of a comparator which forms a control difference from a comparison of a control variable measured on the route with a control variable, and a controller which determines a control variable for the route as a function of the control difference.
  • the basis of the PID controller parameterization is a step response of the controlled system, from which sampled values are obtained and stored at an appropriate time interval.
  • the measured curve is approximated by a special model approach, a PT n model, in order to be able to use a controller setting method tailored to it, the optimum amount.
  • the invention has for its object to provide a device for adaptive control of a route in which a model of the route can be dispensed with and which nevertheless ensures good control behavior.
  • the new device has the Merl ⁇ nal mentioned in the characterizing part of claim 1.
  • Advantageous further developments are given in the subclaims.
  • the invention has the advantage that it is possible with the neural network, directly from the discrete samples of the
  • Step response of the route as input variables is a suitable parameter, for example a PID controller, as an output large to calculate.
  • the neural network can be implemented on comparatively simple hardware, since its use only requires the creation of the sample values of the step change response serving as input variables and the one-off calculation of the neural network. Because of its generalization properties, the network immediately supplies suitable parameters of the controller for a largely arbitrary step response.
  • FIG. 1 shows a block diagram of a device according to the invention for adaptive control
  • FIG. 2 shows a basic illustration of a signal preprocessing and a neural network.
  • an adaptive control loop consists of a comparator 1, a controller 2, a system 3, a signal preprocessing 4 and a neural network 5.
  • a control variable x measured on the control system 3 is compared with a Reference variable w compared and thus a control difference xd is formed.
  • the course of the controlled variable x and the manipulated variable y is sampled and stored in the signal preprocessing 4.
  • the neural network 5 calculates an assignment of the parameters Kp, Tn and Tv of the PID controller 2 that is suitable for the route 3.
  • a step in the manipulated variable is advantageously applied.
  • any time course of the manipulated variable y is possible as an excitation of the route 3, which leads to a course of the controlled variable x which is characteristic of the route behavior. If a calculation of the controller parameters by the neural network 5 is to be possible for different courses of the manipulated variable y, in addition to samples of the course of the controlled variable x, samples of the manipulated variable y must also be supplied to the neural network 5 by the signal preprocessing 4.
  • the sampling time is fixed in the signal preprocessing 4. If, on the other hand, it is to be kept variable, it is required for the controller parameter calculation and must be passed on to the neural network 5 as an additional input variable.
  • a signal PI / PID is present at an input of a neural network 6 and is used to select the controller type. Further input signals are the discrete sample values of the controlled variable x obtained in signal preprocessing, which are obtained from the course of the step response, which is shown in a time diagram. With these input variables, the neural network 6 directly calculates the parameters Kp, Tn and Tv according to the selected controller type.
  • Multi-dimensional, non-linear relationships can, as is known for example from WO 94/06095, be modeled by artificial neural networks.
  • Such multi-dimensional, non-linear relationships the knowledge of which is the solution to the controller design problem, also exist between the sample values of the step response and the optimal ones Controller parameters of an associated PID controller.
  • These relationships can only be stated in analytical form in very special cases.
  • the setting rules for the optimum amount are such a special case.
  • PT n _ systems it provides rational functions with the parameters gain, K, time constant T and order n of the system as setting instructions for the parameters Kp, Tn and Tv of a PID controller, which are specified in the article mentioned at the beginning. In general spelling they read:
  • Kp f x (n, K, T)
  • Tn f2 (n, T)
  • Tv f 3 (n, T).
  • Kp g 1 (x 1 , ...,, ⁇ t),
  • An MLP (Multi Layer Perceptron) network is a neural network with ten inputs for the base values of the step response, one input for the time increment ⁇ t, possibly one input for the switching signal PI / PID and one output each for Kp, Tn, Tv suitable.
  • the sample values of the step responses of various analytical route models are first calculated and the controller parameters based on the known setting rules (e.g. optimum amount for PT n - Routes). Further learning data are obtained by control loop simulation and numerical optimization of the controller parameters.
  • the neural network is then trained with the learning data obtained in this way. The neural network therefore delivers optimal results for this data. It is then recommended to test the generalization properties of the network on lines for which no controller parameters were available as training data.
  • the result is a trained neural network, the internal parameters of which are fully known. It can therefore be implemented on comparatively simple hardware, since high computing power is only required for the learning process.
  • the use of the neural network in the control loop merely requires the creation of the sample values of the step change response serving as input variables and the one-off calculation of the network. Without any further computation-intensive optimization, the network immediately delivers good controller parameters that are optimal at the learned base values for any step response. A model of the route is no longer required for the actual adaptation.

Landscapes

  • Engineering & Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Computation (AREA)
  • Health & Medical Sciences (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Medical Informatics (AREA)
  • Software Systems (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Automation & Control Theory (AREA)
  • Feedback Control In General (AREA)

Abstract

The invention concerns an arrangement for the adaptive control of a section (3) with a signal pre-processing system (4) by means of which sampled values (a) can be generated from the time characteristic of the controlled variable (x), and with a neural network (5) which calculates the parameters (Kp, Tn, Tv) of the controller (2) in dependence on the sampled values (a). The neural network (5) is trained with reference to learned data of the parameters (Kp, Tn, Tv) and sampled values (a) of the section step response, which have been generated by analytically calculated parameter adjusting rules or control circuit simulation and numerical optimization for one or a plurality of section models. The neural network (5) thus supplies the learned data with optimum control parameters which are well-suited to intermediate values owing to its generalization properties. The invention is used in automation technology.

Description

Beschreibungdescription

Einrichtung zur adaptiven Regelung einer StreckeDevice for adaptive control of a route

Die Erfindung betrifft eine Einrichtung zur adaptiven Rege¬ lung einer Strecke nach dem Oberbegriff des Anspruchs 1.The invention relates to a device for adaptive control of a route according to the preamble of claim 1.

Eine derartige Anordnung ist bereits aus dem Aufsatz "Robuste Adaption in Prozeßreglern" von Hans-Peter Preuß, verδffent- licht in der Zeitschrift "Automatisierungstechnische Praxis" 33 (1991), Seiten 178 bis 187, bekannt. Ein Regelkreis be¬ steht dort aus einem Vergleicher, der aus einem Vergleich ei¬ ner an der Strecke gemessenen Regelgröße mit einer Pührungs- grδße eine Regeldifferenz bildet, und einem Regler, der in Abhängigkeit von der Regeldifferenz eine Stellgröße für die Strecke bestimmt. Grundlage der PID-Reglerparametrierung ist eine Sprungantwort der Regelstrecke, von der in angemessenem Zeitabstand Abtastwerte gewonnen und abgespeichert werden. Der gemessene Verlauf wird durch einen speziellen Modell- ansatz, ein PTn-Modell, approximiert, um ein darauf zuge¬ schnittenes Reglereinstellverfahren, das Betragsoptimum, an¬ wenden zu können.Such an arrangement is already known from the article "Robust Adaptation in Process Controllers" by Hans-Peter Preuss, published in the journal "Automatisierungstechnik Praxis" 33 (1991), pages 178 to 187. A control loop there consists of a comparator which forms a control difference from a comparison of a control variable measured on the route with a control variable, and a controller which determines a control variable for the route as a function of the control difference. The basis of the PID controller parameterization is a step response of the controlled system, from which sampled values are obtained and stored at an appropriate time interval. The measured curve is approximated by a special model approach, a PT n model, in order to be able to use a controller setting method tailored to it, the optimum amount.

Der Erfindung liegt die Aufgabe zugrunde, eine Einrichtung zur adaptiven Regelung einer Strecke zu schaffen, bei der auf ein Modell der Strecke verzichtet werden kann und die dennoch ein gutes Regelverhalten gewährleistet.The invention has for its object to provide a device for adaptive control of a route in which a model of the route can be dispensed with and which nevertheless ensures good control behavior.

Zur Lösung dieser Aufgabe weist die neue Einrichtung das im kennzeichnenden Teil des Anspruchs 1 genannte Merlαnal auf. In den Unteransprüchen sind vorteilhafte Weiterbildungen angege¬ ben.To solve this problem, the new device has the Merlαnal mentioned in the characterizing part of claim 1. Advantageous further developments are given in the subclaims.

Die Erfindung hat den Vorteil, daß es mit dem neuronalen Netz möglich ist, direkt aus den diskreten Abtastwerten derThe invention has the advantage that it is possible with the neural network, directly from the discrete samples of the

Sprungantwort der Strecke als Eingangsgrößen gut geeignete Parameter, beispielsweise eines PID-Reglers, als Ausgangs- großen zu berechnen. Dabei kann das neuronale Netz auf einer vergleichsweise einfachen Hardware implementiert werden, da seine Anwendung lediglich das Anlegen der als Eingangsgrößen dienenden Abtastwerte der Streckensprungantwort und die ein- malige Durchrechnung des neuronalen Netzes erfordert. Auf¬ grund seiner Generalisierungseigenschaften liefert das Netz zu einer weitgehend beliebigen Sprungantwort sofort geeignete Parameter des Reglers.Step response of the route as input variables is a suitable parameter, for example a PID controller, as an output large to calculate. The neural network can be implemented on comparatively simple hardware, since its use only requires the creation of the sample values of the step change response serving as input variables and the one-off calculation of the neural network. Because of its generalization properties, the network immediately supplies suitable parameters of the controller for a largely arbitrary step response.

Anhand der Figuren, in denen ein Ausführungsbeispiel der Er¬ findung dargestellt ist, werden im folgenden die Erfindung sowie Ausgestaltungen und Vorteile näher erläutert.Based on the figures, in which an embodiment of the invention is shown, the invention as well as embodiments and advantages are explained in more detail below.

Es zeigen: Figur 1 ein Blockschaltbild einer erfindungsgemäßen Einrich¬ tung zur adaptiven Regelung und Figur 2 eine Prinzipdarstellung einer Signalvorverarbeitung und eines neuronalen Netzes.FIG. 1 shows a block diagram of a device according to the invention for adaptive control and FIG. 2 shows a basic illustration of a signal preprocessing and a neural network.

Ein adaptiver Regelkreis besteht nach Figur 1 aus einem Ver¬ gleicher 1, einem Regler 2, einer Strecke 3, einer Signalvor¬ verarbeitung 4 und einem neuronalen Netz 5. In dem Verglei¬ cher 1 wird eine an der Regelstrecke 3 gemessene Regelgröße x mit einer Führungsgröße w verglichen und so eine Regeldiffe- renz xd gebildet. Der Regler 2, der als PID-Regler ausgeführt ist und seine Parameter Kp, Tn und Tv von dem neuronalen Netz erhält, bestimmt aus der Regeldifferenz xd eine Stellgröße y, die auf die Strecke 3 wirkt. Der Verlauf der Regelgröße x und der Stellgröße y wird in der Signalvorverarbeitung 4 abgeta- stet und gespeichert. Aufgrund der durch die Signalvorverar¬ beitung 4 gewonnenen Abtastwerte a wird durch das neuronale Netz 5 eine für die Strecke 3 geeignete Belegung der Para¬ meter Kp, Tn und Tv des PID-Reglers 2 berechnet. Vorteilhaft wird für die Erzeugung der Abtastwerte a in der Signalvorver- arbeitung 4 der Strecke 3 durch den Regler 2 ein Stellgrδßen- sprung aufgeschaltet. Prinzipiell ist aber jeder beliebige Zeitverlauf der Stell¬ größe y als Anregung der Strecke 3 möglich, der zu einem für das Streckenverhalten charakteristischen Verlauf der Regel¬ größe x führt. Wenn eine Berechnung der Reglerparameter durch das neuronale Netz 5 für verschiedene Verläufe der Stellgröße y möglich sein soll, müssen neben Abtastwerten des Verlaufs der Regelgröße x auch Abtastwerte der Stellgröße y durch die Signalvorverarbeitung 4 dem neuronalen Netz 5 zugeführt wer¬ den. Da aber nicht für jeden Verlauf der Stellgröße y eine sinnvolle Berechnung der Reglerparameter möglich ist, bei¬ spielsweise im stationären Zustand der Regelung, wird in die¬ sem Fall ein Kriterium benötigt, mit dessen Hilfe in Abhän¬ gigkeit vom Stellgrößenverlauf die Parameterberechnung frei¬ gegeben wird.According to FIG. 1, an adaptive control loop consists of a comparator 1, a controller 2, a system 3, a signal preprocessing 4 and a neural network 5. In the comparator 1, a control variable x measured on the control system 3 is compared with a Reference variable w compared and thus a control difference xd is formed. The controller 2, which is designed as a PID controller and receives its parameters Kp, Tn and Tv from the neural network, determines a manipulated variable y, which acts on the system 3, from the control difference xd. The course of the controlled variable x and the manipulated variable y is sampled and stored in the signal preprocessing 4. Based on the sample values a obtained by the signal preprocessing 4, the neural network 5 calculates an assignment of the parameters Kp, Tn and Tv of the PID controller 2 that is suitable for the route 3. For the generation of the sample values a in the signal preprocessing 4 of the section 3, a step in the manipulated variable is advantageously applied. In principle, however, any time course of the manipulated variable y is possible as an excitation of the route 3, which leads to a course of the controlled variable x which is characteristic of the route behavior. If a calculation of the controller parameters by the neural network 5 is to be possible for different courses of the manipulated variable y, in addition to samples of the course of the controlled variable x, samples of the manipulated variable y must also be supplied to the neural network 5 by the signal preprocessing 4. However, since a sensible calculation of the controller parameters is not possible for every course of the manipulated variable y, for example in the steady state of the control, a criterion is required in this case, with the aid of which the parameter calculation is enabled depending on the course of the manipulated variable becomes.

Die Abtastzeit ist in der Signalvorverarbeitung 4 fest einge¬ stellt. Wenn sie dagegen variabel gehalten werden soll, wird sie für die Reglerparameterberechnung benötigt und muß als weitere Eingangsgröße auf das neuronale Netz 5 gegeben wer- den.The sampling time is fixed in the signal preprocessing 4. If, on the other hand, it is to be kept variable, it is required for the controller parameter calculation and must be passed on to the neural network 5 as an additional input variable.

In Figur 2 ist das Adaptionsprinzip anschaulich dargestellt. An einem Eingang eines neuronalen Netzes 6 liegt ein Signal PI/PID an, das zur Auswahl des Reglertyps dient. Weitere Ein- gangssignale sind die in einer Signalvorverarbeitung gewonne¬ nen diskreten Abtastwerte der Regelgröße x, die aus dem Ver¬ lauf der Sprungantwort, die in einem Zeitdiagramm dargestellt ist, gewonnen werden. Mit diesen Eingangsgrößen berechnet das neuronale Netz 6 dem gewählten Reglertyp entsprechend direkt die Parameter Kp, Tn und Tv.The adaptation principle is clearly illustrated in FIG. A signal PI / PID is present at an input of a neural network 6 and is used to select the controller type. Further input signals are the discrete sample values of the controlled variable x obtained in signal preprocessing, which are obtained from the course of the step response, which is shown in a time diagram. With these input variables, the neural network 6 directly calculates the parameters Kp, Tn and Tv according to the selected controller type.

Mehrdimensionale, nichtlineare Zusammenhänge können, wie bei¬ spielsweise aus der WO 94/06095 bekannt ist, durch künstliche neuronale Netze modelliert werden. Solche mehrdimensionalen, nichtlinearen Zusammenhänge, deren Kenntnis die Lösung des Reglerentwurfsproblems darstellt, bestehen auch zwischen den Abtastwerten der Streckensprungantwort und den optimalen Reglerparametern eines zugehörigen PID-Reglers. Allerdings lassen sich diese Zusammenhänge nur in sehr speziellen Fällen in analytischer Form angeben. Die Einstellregeln des Betrags¬ optimums sind ein solcher Spezialfall. Es liefert für PTn_ Strecken als Einstellvorschrift für die Parameter Kp, Tn und Tv eines PID-Reglers rationale Funktionen mit den Kennwerten Verstärkung K, Zeitkonstante T und Ordnung n der Strecke, die in dem eingangs genannten Aufsatz angegeben sind. In allge¬ meiner Schreibweise lauten sie:Multi-dimensional, non-linear relationships can, as is known for example from WO 94/06095, be modeled by artificial neural networks. Such multi-dimensional, non-linear relationships, the knowledge of which is the solution to the controller design problem, also exist between the sample values of the step response and the optimal ones Controller parameters of an associated PID controller. However, these relationships can only be stated in analytical form in very special cases. The setting rules for the optimum amount are such a special case. For PT n _ systems, it provides rational functions with the parameters gain, K, time constant T and order n of the system as setting instructions for the parameters Kp, Tn and Tv of a PID controller, which are specified in the article mentioned at the beginning. In general spelling they read:

Kp = fx(n, K, T), Tn = f2 (n, T) und Tv = f3(n, T) .Kp = f x (n, K, T), Tn = f2 (n, T) and Tv = f 3 (n, T).

Da die Kennwerte n, K, T der Strecke die Sprungantwort x(t) eindeutig beschreiben, gibt es ebenfalls einen Zusammenhang zwischen m äquidistanten, diskreten Werten x^, i - l, ... , m, der Streckensprungantwort x(t) und den betragsoptimalen Reg¬ lerparametern:Since the characteristic values n, K, T of the route clearly describe the step response x (t), there is also a relationship between m equidistant, discrete values x ^, i - l, ..., m, the route step response x (t) and the amount-optimal controller parameters:

Kp = g1(x1, ... , , Δt) ,Kp = g 1 (x 1 , ...,, Δt),

Tn = 92 <xl» •••» Xm' Δ > und Tv = g3 (__l t ..., %, Δt) .Tn = 92 < x l »•••» X m ' Δ > and Tv = g 3 (__ lt ...,%, Δt).

Auch für andere Streckentypen existieren entsprechende Bezie¬ hungen, wobei die Funktionen g_ vom jeweils verwendeten Ent¬ wurfskriterium (Betragsoptimum, quadratische Regelfläche, ... ) abhängen.Corresponding relationships also exist for other route types, the functions g_ depending on the design criterion used in each case (optimum amount, square control area, ...).

Als neuronales Netz ist ein MLP (Multi Layer Perceptron) -Netz mit zehn Eingängen für die Stützwerte der Sprungantwort, ei¬ nem Eingang für das Zeitinkrement Δt, eventuell einem Eingang für das UmsehaltSignal PI/PID und je einem Ausgang für Kp, Tn, Tv geeignet. Zur Lerndatengenerierung werden zunächst die Abtastwerte der Sprungantworten verschiedener analytischer Streckenmodelle berechnet und die Reglerparameter anhand der bekannten Einstellregeln (z. B. Betragsoptimum für PTn- Strecken) vorgegeben. Weitere Lerndaten werden durch Regel- kreissimulation und numerische Optimierung der Regler¬ parameter gewonnen. Mit den so erhaltenen Lerndaten wird schließlich das neuronale Netz trainiert. Für diese Daten liefert also das neuronale Netz optimale Ergebnisse. An¬ schließend empfiehlt sich ein Test der Generalisierungseige - Schäften des Netzes an Strecken, für die keine Reglerpara¬ meter als Trainingsdaten zur Verfügung standen.An MLP (Multi Layer Perceptron) network is a neural network with ten inputs for the base values of the step response, one input for the time increment Δt, possibly one input for the switching signal PI / PID and one output each for Kp, Tn, Tv suitable. To generate the learning data, the sample values of the step responses of various analytical route models are first calculated and the controller parameters based on the known setting rules (e.g. optimum amount for PT n - Routes). Further learning data are obtained by control loop simulation and numerical optimization of the controller parameters. The neural network is then trained with the learning data obtained in this way. The neural network therefore delivers optimal results for this data. It is then recommended to test the generalization properties of the network on lines for which no controller parameters were available as training data.

Als Ergebnis erhält man ein trainiertes neuronales Netz, des¬ sen interne Parameter vollständig bekannt sind. Es kann damit auf einer vergleichsweise einfachen Hardware implementiert werden, da eine hohe Rechenleistung nur für den Lernvorgang erforderlich ist. Die Anwendung des neuronalen Netzes im Re- gelkreis erfordert lediglich- das Anlegen der als Eingangs¬ größen dienenden Abtastwerte der Streckensprungantwort und die einmalige Durchrechnung des Netzes. Ohne jede weitere rechenintensive Optimierung liefert das Netz zu einer belie¬ bigen Sprungantwort sofort gute, an den gelernten Stützwerten optimale Reglerparameter. Ein Modell der Strecke wird für die eigentliche Adaption nicht mehr benötigt. The result is a trained neural network, the internal parameters of which are fully known. It can therefore be implemented on comparatively simple hardware, since high computing power is only required for the learning process. The use of the neural network in the control loop merely requires the creation of the sample values of the step change response serving as input variables and the one-off calculation of the network. Without any further computation-intensive optimization, the network immediately delivers good controller parameters that are optimal at the learned base values for any step response. A model of the route is no longer required for the actual adaptation.

Claims

Patentansprüche claims 1. Einrichtung zur adaptiven Regelung einer Strecke (3)1. Device for adaptive control of a route (3) - mit einem Vergleicher (1) , der aus einem Vergleich einer an der Strecke (3) gemessenen Regelgröße (x) mit einer- With a comparator (1), the comparison of a controlled variable (x) measured on the route (3) with a Führungsgröße (w) eine Regeldifferenz (xd) bildet,Command variable (w) forms a control difference (xd), - mit einem Regler (2) , der in Abhängigkeit von der Regel- differenz (xd) eine Stellgröße (y) für die Strecke (3) be¬ stimmt, und - mit einer Signalvorverarbeitung (4) , mit der aus dem zeit¬ lichen Verlauf der Regelgröße (x) Abtastwerte (a) erzeug¬ bar sind, gekennzeichnet durch ein neuronales Netz (5) , das in Abhängigkeit von den Abtast- werten (a) die Parameter (Kp, Tn, Tv) des Reglers berechnet.- With a controller (2), which determines a manipulated variable (y) for the system (3) as a function of the control difference (xd), and - With a signal preprocessing (4), with which from the temporal Course of the controlled variable (x), sample values (a) can be generated, characterized by a neural network (5) which calculates the parameters (Kp, Tn, Tv) of the controller as a function of the sample values (a). 2. Einrichtung nach Anspruch 1, dadurch gekenn¬ zeichnet, daß der Regler (2) ein PI- oder PID-Regler ist.2. Device according to claim 1, characterized gekenn¬ characterized in that the controller (2) is a PI or PID controller. 3. Einrichtung nach Anspruch 1 oder 2, dadurch gekenn¬ zeichnet , daß für die Erzeugung der Abtastwerte (a) in der Signalvorverarbeitung (4) der Strecke (3) durch den Reg¬ ler (2) ein Stellgrδßensprung aufschaltbar ist.3. Device according to claim 1 or 2, characterized gekenn¬ characterized in that for the generation of the samples (a) in the signal preprocessing (4) of the route (3) by the controller (2) a manipulated variable jump can be switched on. 4. Einrichtung nach einem der vorhergehenden Ansprüche, da¬ durch gekennzeichnet , daß das neuronale Netz (5) anhand von Lerndaten der Parameter (Kp, Tn, Tv) und der Ab¬ tastwerte (a) der Streckensprungantwort trainiert ist, die durch Parameter-Einstellregeln für ein oder mehrere Strecken¬ modelle generiert wurden.4. Device according to one of the preceding claims, characterized in that the neural network (5) is trained on the basis of learning data of the parameters (Kp, Tn, Tv) and the sampling values (a) of the step response, which is parameterized Setting rules for one or more line models were generated. 5. Einrichtung nach einem der vorhergehenden Ansprüche, da¬ durch gekennzeichnet , daß das neuronale Netz (5) anhand von Lerndaten der Parameter (Kp, Tn, Tv) und der Ab¬ tastwerte (a) der Streckensprungantwort trainiert ist, die durch Regelkreissimulation und numerische Optimierung der Reglerparameter (Kp, Tn, Tv) für ein oder mehrere Strecken¬ modelle generiert wurden. 5. Device according to one of the preceding claims, characterized in that the neural network (5) is trained on the basis of learning data of the parameters (Kp, Tn, Tv) and the sampling values (a) of the step response, which by means of control loop simulation and numerical optimization of the Controller parameters (Kp, Tn, Tv) were generated for one or more line models.
PCT/DE1995/001223 1994-09-19 1995-09-07 Arrangement for the adaptive control of a section Ceased WO1996009573A2 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
DE19944433332 DE4433332A1 (en) 1994-09-19 1994-09-19 Device for adaptive control of a route
DEP4433332.3 1994-09-19

Publications (2)

Publication Number Publication Date
WO1996009573A2 true WO1996009573A2 (en) 1996-03-28
WO1996009573A3 WO1996009573A3 (en) 1996-05-30

Family

ID=6528585

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/DE1995/001223 Ceased WO1996009573A2 (en) 1994-09-19 1995-09-07 Arrangement for the adaptive control of a section

Country Status (2)

Country Link
DE (1) DE4433332A1 (en)
WO (1) WO1996009573A2 (en)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5847952A (en) * 1996-06-28 1998-12-08 Honeywell Inc. Nonlinear-approximator-based automatic tuner
DE10302585B4 (en) * 2003-01-22 2004-12-30 Endress + Hauser Wetzer Gmbh + Co Kg Procedure for setting a controller
EP3825788B1 (en) * 2019-11-19 2022-11-09 Asco Numatics GmbH Control device, control system and control method for regulating a physical quantity of a fluid

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH03201008A (en) * 1989-12-28 1991-09-02 Toshiba Corp Gain scheduling controller
EP0617806B1 (en) * 1991-12-18 1998-05-20 Honeywell Inc. A closed loop neural network automatic tuner

Also Published As

Publication number Publication date
WO1996009573A3 (en) 1996-05-30
DE4433332A1 (en) 1996-03-21

Similar Documents

Publication Publication Date Title
DE102007001025B4 (en) Method for computer-aided control and / or regulation of a technical system
EP2519861B1 (en) Method for the computer-aided control of a technical system
AT512251B1 (en) Method of designing a nonlinear controller for non-linear processes
DE10341573A1 (en) Integrated model-based predictive control and optimization within a process control system
DE69719058T2 (en) IMPROVED NEURONAL NETWORK
DE19606480C2 (en) Method and arrangement for adapting a fuzzy controller
DE4035099A1 (en) OVEN TEMPERATURE CONTROL DEVICE WITH ONE ADJUSTING INPUT
DE3839675C2 (en) Optimizer for a parameter-dependent control system
EP0520233A2 (en) Arrangement for identification of a controlled system
EP3542229B1 (en) Device and method for determining the parameters of a control device
EP3349078A1 (en) Diagnosis device and method for monitoring and/or optimizing a control device
WO1996009573A2 (en) Arrangement for the adaptive control of a section
DE19525907A1 (en) Self-adjustable control device and method for self-adjustment of this controller
DE3418501C2 (en)
EP4058666A1 (en) Method and system for calibrating a controller of a machine
EP1217472B1 (en) Method for controlling a process having delay with compensation as well as control device for carrying out the method
EP0563719A2 (en) Digital modulating method and digital modulating apparatus
DE102019206541A1 (en) Method for performing computer-aided XiL simulations
EP0700536B1 (en) Regulating device
DE20117191U1 (en) Device for operating a technical system
DE29610789U1 (en) Device for identifying a transmission path, in particular a control path
DE102021116388A1 (en) Method for optimizing a controller and control system with a controller optimized in this way
EP0826166B1 (en) Process for determining a feedforward control
DE19932794A1 (en) Method of adapting a process controller to the characteristics of a controlled system, requires initially identifying the process using complete sets of measured values
EP0724748B1 (en) Control device

Legal Events

Date Code Title Description
AK Designated states

Kind code of ref document: A2

Designated state(s): JP US

AL Designated countries for regional patents

Kind code of ref document: A2

Designated state(s): AT BE CH DE DK ES FR GB GR IE IT LU MC NL PT SE

AK Designated states

Kind code of ref document: A3

Designated state(s): JP US

AL Designated countries for regional patents

Kind code of ref document: A3

Designated state(s): AT BE CH DE DK ES FR GB GR IE IT LU MC NL PT SE

121 Ep: the epo has been informed by wipo that ep was designated in this application
122 Ep: pct application non-entry in european phase