Data processing: generic control systems or specific application – Generic control system – apparatus or process – Optimization or adaptive control
Reexamination Certificate
2000-10-05
2003-12-02
Patel, Ramesh (Department: 2121)
Data processing: generic control systems or specific application
Generic control system, apparatus or process
Optimization or adaptive control
C700S041000, C700S042000, C700S043000, C700S044000, C700S052000, C700S067000, C700S068000, C700S069000, C700S070000
Reexamination Certificate
active
06658304
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to a computer based method of controlling an industrial process, including the steps of
measuring the values of at least one process variable, and
predicting future deviations of at least one process variable with regard to the measured value of said at least one process variable, and providing a control signal based on said prediction by means of a first control law.
The invention also relates to a computer based system for controlling an industrial process, comprising
means for providing a control signal based on a prediction of future deviations of a process variable with regard to a measured value of at least one process variable, said means comprising a first control law for executing said provision of the control signal.
The term “measuring” as defined herein should be regarded in a broad sense. Accordingly, in-line, on-line, and lab measurements should be included.
The term “industrial process” includes continuous, semi-continuous, and batch processes.
Furthermore, the invention particularly relates to methods and systems in which predictions of future variable deviations are preceded by corresponding estimations of inner states in the process.
A future deviation of a process variable, as defined herein, refers to the difference between a predicted future value of said process variable and a future set point value for the variable in question.
The invention particularly includes the application of data driven stochastic system models or empirical models for the purpose of control of different continuous or semi-continuous processes.
Even though the invention is applicable to all kinds of continuous and semi-continuous processes, it is particularly useful in connection to the manufacture of chemical, petrochemical, and polymer products, and continuous pulp and paper production, in which there is a need of controlling the stream compositions in order to obtain the required product properties.
Typically, the invention is applicable to processes like the ones in distillation towers. Here, process variables such as the concentration in the product, the pressure, the temperature, etc., are measured. For example, the future deviation of the concentration in the product is the one being predicted with regard to the measured value or values of said process variables. Process parameters, such as the flow in the tower, the supply of heat, and pressure build-up are adjusted with regard to the executed predictions. Similar applications are found in the processes of fractionators and cracker devices.
The invention is also typically applicable to digesters for continuous pulp production. In such an application, process variables, such as Kappa number, liquor phase variables like the effective alkali, dissolved lignin concentration, dissolved solid concentration, and sulphidity are measured, normally for the purpose of predicting a future value of the Kappa number. Process parameters, such as the supply of heat, are then controlled in order to control the process.
PRIOR ART
Systems and methods of prior art for controlling continuous or semi-continuous processes of the kind defined above also take measurable disturbances in the process in consideration. Such disturbances could be of any kind affecting the value of the process variable for which a future deviation is predicted by means of a given control law.
According to prior art, said control law is adapted to deal with a plurality of variables or inputs, including measured disturbances. As a result, the control law in question becomes relatively complicated. Furthermore, such an arrangement necessitates tailored adaptation of the control law for each individual process which is to be controlled in this way with regard to measured values of certain process variables and measured disturbances.
OBJECT OF THE INVENTION
The object of the present invention is to provide a computer based method and system for controlling a continuous or semi-continuous process, said method and system being adapted to the need of controlling the process with regard to a measurement of a measurable disturbance in the process. In particular, the inventive method and system shall devise ways of facilitating the tuning of a specific control system for a given process with regard to contradictory requirements on rapidity and stability.
SUMMARY OF THE INVENTION
The object of the invention is achieved through a method as initially defined, which is characterised in that it comprises the further steps of measuring a measurable disturbance in the process, predicting future deviations of said process variable with regard to said disturbance but without regard to the measured value of said at least one process variable, and providing a control signal based on said prediction by means of a second control law. Accordingly, two separate control laws are used, and two degrees of freedom are obtained. The function of the first control law can be optimised with regard to the predictions based on the measured value of the process variable or variables, while the second control law can be optimised with regard to the prediction of the deviations of said process variable with regard to the measured disturbances.
For most processes, the adjustments of different process parameters based on the prediction of a future deviation of a given process variable have to be cautious due to complex dynamic behaviours of the process and/or to the measured process variable not being easily accessible. However, upon an introduction of a disturbance in the process, a more rapid or extensive adjustment of such a process parameter is often desired. Control laws according to prior art which has to consider these requirements simultaneously tend to get complex and complicated. With the inventive solution using two degrees of freedom and two separate control laws, the alertness of the control of the process with regard to measurable disturbances on one hand and model errors and variable measurements on the other hand can easily and separately be determined.
One or more process parameters in the process are controlled with regard to the control signals provided by means of said first and second control laws. The control signals, based on the predicted deviations and provided by said control laws are added to each other and the control of said parameter or parameters is executed with regard to the sum of the separate control signals. For the implementation of the method, signals corresponding or representative to said sum should be delivered to relevant units from a computer environment in which said control laws define software.
Often, it is also desired that the adjustment of said process parameter or parameters should not be too extensive. Therefore, according to the inventive method, a minimum value and a maximum value are set for the sum of the control signals based on predicted deviations. Furthermore, the control signal values are added sequentially to each other in a predetermined order. Thereby, the value of the first control signal, to which a second control signal is added, will possibly delimit the contribution of the second control signal in combination with the minimum value and maximum value restrictions. By adding the control signals in a predetermined order, the inventive method proposes a way of prioritise the contribution to the sum of the signals from the different control laws.
According to a preferred embodiment, the prediction used by the first control law is based on a measurement of a first and/or a second process variable, the second process variable being more readily accessible and/or more instantly responsive to non-measured disturbances in the process than the first process variable.
Preferably, it is the future deviations of the first process variable which are predicted and then treated by means of said first and second control laws. As the second variable is more readily accessible and/or more instantly responsive to the non-measured or non-measurable disturbances in the process than the first process variable, a mo
Lundh Michael
Molander Mats
ABB AB
Birch & Stewart Kolasch & Birch, LLP
Patel Ramesh
Shute Douglas M.
LandOfFree
Computer based method and a system for controlling an... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Computer based method and a system for controlling an..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Computer based method and a system for controlling an... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3158104