公开(公告)号：WO1994017457A1

公开(公告)日：1994-08-04

申请号：PCT/US1994000753

申请日：1994-01-21

Applicant: HONEYWELL INC.

Inventor： HONEYWELL INC. ,  SAMAD, Tariq ,  FOSLIEN GRABER, Wendy, K. ,  KONAR, Ahmet, F.

IPC: G05B13/02

CPC classification number: G05B13/027

Abstract: Any parameters are selected to be input into a neurocontroller to generate output where that neurocontroller is trained on those parameters. Training systems for accomplishing this, as well as various closed-loop applications, are taught.

Abstract translation: 选择任何参数输入到神经控制器中以产生输出，其中神经控制器被训练在这些参数上。 教授了实现这一目标的培训系统，以及各种闭环应用。

公开(公告)号：WO1993015448A1

公开(公告)日：1993-08-05

申请号：PCT/US1993001089

申请日：1993-01-27

Applicant: HONEYWELL INC.

Inventor： HONEYWELL INC. ,  KONAR, A., Ferit ,  SAMAD, Tariq ,  HARP, Steven, A.

IPC: G05B13/02

CPC classification number: G05B13/027 ,  Y10S706/903 ,  Y10S706/906

Abstract: A PID controller which uses a neural network to introduce non-linear functions into conventional PID paradigms. A neural network is substituted for a linear-equation-based controller processor. Dynamic neural network forms are preferred. One implementation allows for switching between the conventional and a neural network controller. The networks are trained on forms of the error from setpoint (or desired value) and the PID input signals.

Abstract translation: 一个使用神经网络将非线性函数引入常规PID范例的PID控制器。 神经网络代替基于线性方程的控制器处理器。 优选动态神经网络形式。 一个实现允许在常规神经网络控制器之间进行切换。 网络根据设定值（或所需值）和PID输入信号对错误形式进行培训。

公开(公告)号：WO1993012476A1

公开(公告)日：1993-06-24

申请号：PCT/US1992011086

申请日：1992-12-17

Applicant: HONEYWELL INC.

Inventor： HONEYWELL INC. ,  MATHUR, Anoop, K. ,  SAMAD, Tariq

IPC: G05B13/02

CPC classification number: G05B13/027 ,  Y10S706/903

Abstract: A closed loop neural network based autotuner (66) develops optimized proportional, integral and/or derivative parameters (18) based on the outputs (u, y, r, e...) of other elements in the loop. Adjustments are initiated by making a step change in the setpoint which may be done by a user or automatically. A Smith predictor (31) may also be employed.

Abstract translation: 基于闭环神经网络的自动调谐器（66）基于循环中其他元件的输出（u，y，r，e ...）开发出优化的比例，积分和/或微分参数（18）。 通过对用户进行的设定值进行步进更改或自动进行调整。 也可以使用史密斯预测器（31）。

公开(公告)号：WO1998000763A1

公开(公告)日：1998-01-08

申请号：PCT/US1997009473

申请日：1997-06-02

Applicant: HONEYWELL INC.

Inventor： HONEYWELL INC. ,  SAMAD, Tariq

IPC: G05B13/02

CPC classification number: G05B13/027

Abstract: An automatic tuner for control systems that produces as output values for parameters of an arbitrary controller. The controller is in a control loop so that its output effects changes in actuators and regulates a physical process. The controller consists of either linear and nonlinear controller components or a combination of both. The tuner has a nonlinear approximator that has been optimized off-line. The off-line optimization is done without supervised learning so that desired outputs of neural network do not need to be available, and separate optimization to generate the desired outputs is not necessary. The off-line optimization can also rely on optimization criteria that are arbitrary: differentiability, convexity, even continuity of criteria are not required. The off-line optimization ensures robustness of generated controller parameters so that the input process characteristics do not need to be highly accurate. The off-line optimization is performed in such a way that the input parameters that relate to desired closed-loop system behavior include robustness parameters that can be used to effect tradeoffs between robust and nominal performance. The inputs to the nonlinear approximator consist of two sets of input parameters, either of which may be empty. A first set of input parameters can include parameters that relate to process characteristics. A second set of input parameters can include parameters that relate to desired closed-loop system behavior. The output values may be proportional and/or integral and/or derivative gains for PID-like controllers. The output values otherwise may be parameters for delay-compensation controllers, parameters for controllers that consist of lead-lag terms in combination with PID controllers, parameters for higher-order linear controllers, discrete variables that select between different control structures, or parameters for nonlinear controllers of predetermined structure. The nonlinear approximator may be implemented as a compositional sigmoidal mapping, a multilayer perception structure, a fuzzy logic model, a radial basis function network, a polynomial expansion, or other parametrized nonlinear structure.

Abstract translation: 用于控制系统的自动调谐器，其产生用于任意控制器的参数的输出值。 控制器处于控制回路中，使其输出效应在执行器中发生变化并调节物理过程。 控制器包括线性和非线性控制器组件或两者的组合。 调谐器具有已经离线优化的非线性近似器。 离线优化是在没有监督学习的情况下进行的，因此神经网络的期望输出不需要可用，并且分离的优化以产生期望的输出是不必要的。 离线优化还可以依赖于任意的优化标准：可区分性，凸度，甚至连续性的标准都不是必需的。 离线优化确保了生成的控制器参数的鲁棒性，使得输入过程特性不需要高度准确。 离线优化以这样的方式执行，使得与期望的闭环系统行为相关的输入参数包括可用于实现鲁棒性和标称性能之间的权衡的鲁棒性参数。 非线性近似器的输入由两组输入参数组成，其中任一个可能为空。 第一组输入参数可以包括与过程特性相关的参数。 第二组输入参数可以包括与期望的闭环系统行为相关的参数。 PID类控制器的输出值可以是比例和/或积分和/或导数增益。 输出值可以是延迟补偿控制器的参数，控制器的参数包括与PID控制器相结合的超前滞后项，高阶线性控制器的参数，在不同控制结构之间选择的离散变量，或非线性参数 预定结构的控制器。 非线性近似器可以被实现为组合S形映射，多层感知结构，模糊逻辑模型，径向基函数网络，多项式展开或其他参数化的非线性结构。

公开(公告)号：WO1990011568A1

公开(公告)日：1990-10-04

申请号：PCT/US1990000828

申请日：1990-02-21

Applicant: HONEYWELL, INC.

Inventor： HONEYWELL, INC. ,  GUHA, Aloke ,  HARP, Steven, A. ,  SAMAD, Tariq

IPC: G06F15/18

CPC classification number: G06N3/086 ,  G06N3/126

Abstract: The invention relates to a method for using genetic type learning techniques in connection with designing a variety of neural networks that are optimized for specific applications. In the invention herein a general representation of neural network architectures is linked with the genetic learning strategy to create a flexible environment for the design of custom neural networks. A concept upon which the invention is based is the representation of a network design as a ''genetic blueprint'' wherein the recombination or mutation of subsequently generated editions of such blueprints result in different but related network architectures.

公开(公告)号：EP0907909A1

公开(公告)日：1999-04-14

申请号：EP97929758.0

申请日：1997-06-02

Applicant: Honeywell Inc.

Inventor： SAMAD, Tariq

IPC: G06F15 ,  G05B13

CPC classification number: G05B13/027

Abstract: An automatic tuner for control systems that produces as output values for parameters of an arbitrary controller. The controller is in a control loop so that its output effects changes in actuators and regulates a physical process. The controller consists of either linear and nonlinear controller components or a combination of both. The tuner has a nonlinear approximator that has been optimized off-line. The off-line optimization is done without supervised learning so that desired outputs of neural network do not need to be available, and separate optimization to generate the desired outputs is not necessary. The off-line optimization can also rely on optimization criteria that are arbitrary: differentiability, convexity, even continuity of criteria are not required. The off-line optimization ensures robustness of generated controller parameters so that the input process characteristics do not need to be highly accurate. The off-line optimization is performed in such a way that the input parameters that relate to desired closed-loop system behavior include robustness parameters that can be used to effect tradeoffs between robust and nominal performance. The inputs to the nonlinear approximator consist of two sets of input parameters, either of which may be empty. A first set of input parameters can include parameters that relate to process characteristics. A second set of input parameters can include parameters that relate to desired closed-loop system behavior. The output values may be proportional and/or integral and/or derivative gains for PID-like controllers. The output values otherwise may be parameters for delay-compensation controllers, parameters for controllers that consist of lead-lag terms in combination with PID controllers, parameters for higher-order linear controllers, discrete variables that select between different control structures, or parameters for nonlinear controllers of predetermined structure. The nonlinear approximator may be implemented as a compositional sigmoidal mapping, a multilayer perception structure, a fuzzy logic model, a radial basis function network, a polynomial expansion, or other parametrized nonlinear structure.

公开(公告)号：EP0624264A1

公开(公告)日：1994-11-17

申请号：EP93905812.0

申请日：1993-01-27

Applicant: HONEYWELL INC.

Inventor： KONAR, A., Ferit ,  SAMAD, Tariq ,  HARP, Steven, A.

IPC: G05B13

CPC classification number: G05B13/027 ,  Y10S706/903 ,  Y10S706/906

Abstract: Un régulateur PID utilise un réseau de neurones afin d'introduire des fonctions non linéaires dans des paradigmes PID classiques. Un réseau de neurones se substitue à un processeur de régulateur basé sur des équations linéaires. Certaines formes de réseaux de neurones dynamiques sont préférées. Un mode de réalisation permet la commutation entre un régulateur classique et un régulateur de réseaux de neurones. Les réseaux sont formés sur la base de formes de l'erreur par rapport à la valeur de consigne (ou valeur voulue), et des signaux d'entrée PID.

公开(公告)号：EP0465489B1

公开(公告)日：1994-11-30

申请号：EP90904720.1

申请日：1990-02-21

Applicant: HONEYWELL INC.

Inventor： GUHA, Aloke ,  HARP, Steven, A. ,  SAMAD, Tariq

IPC: G06F15/18 ,  G06F15/80

CPC classification number: G06N3/086 ,  G06N3/126

Abstract: The invention relates to a method for using genetic type learning techniques in connection with designing a variety of neural networks that are optimized for specific applications. In the invention herein a general representation of neural network architectures is linked with the genetic learning strategy to create a flexible environment for the design of custom neural networks. A concept upon which the invention is based is the representation of a network design as a ''genetic blueprint'' wherein the recombination or mutation of subsequently generated editions of such blueprints result in different but related network architectures.

公开(公告)号：EP0465489A1

公开(公告)日：1992-01-15

申请号：EP90904720.0

申请日：1990-02-21

Applicant: HONEYWELL INC.

Inventor： GUHA, Aloke ,  HARP, Steven, A. ,  SAMAD, Tariq

IPC: G06G7 ,  G06F15 ,  G06N3

CPC classification number: G06N3/086 ,  G06N3/126

Abstract: L'invention concerne un nouveau procédé d'utilisation de technique d'apprentissage de type génétique, en rapport avec la conception d'une variété de réseaux neuronaux optimisés pour des applications spécifiques. Selon l'invention, une représentation générale d'architectures de réseaux neuronaux est liée à la stratégie d'apprentissage génétique, afin de créer un environnement souple permettant la conception de réseaux neuronaux personnalisés. Un concept sur lequel l'invention est basée est la représentation d'une conception de réseaux sous la forme d'un ''calque génétique'', dans lequel la recombinaison ou la mutation d'éditions générées ultérieurement de tels calques ont pour résultat des architectures de réseaux différentes mais apparentées.

公开(公告)号：EP0624264B1

公开(公告)日：1997-11-26

申请号：EP93905812.9

申请日：1993-01-27

Applicant: HONEYWELL INC.

Inventor： KONAR, A., Ferit ,  SAMAD, Tariq ,  HARP, Steven, A.

IPC: G05B13/02

CPC classification number: G05B13/027 ,  Y10S706/903 ,  Y10S706/906

Abstract: A PID controller which uses a neural network to introduce non-linear functions into conventional PID paradigms. A neural network is substituted for a linear-equation-based controller processor. Dynamic neural network forms are preferred. One implementation allows for switching between the conventional and a neural network controller. The networks are trained on forms of the error from setpoint (or desired value) and the PID input signals.