Minimum time, optimizing and stabilizing multivariable control method and system using a constraint associated control code

A method for integrating on-line optimization without destabilizing for a refinery, chemical process or the like associated with a plurality of independently controlled manipulated plant inputs X and Y, and one or more controlled plant outputs Z dependent upon said manipulated plant inputs X, Y in which closed loop, proportional control of the manipulated plant inputs X, Y can occur in accordance with a series of pre-tested optimizing control components, Uo, whereby optimization of the X plant inputs can be accelerated, deaccelerated, zeroed, or reversed without use of optimal set points in association therewith, said Y plant inputs being reserved for stabilization in connection with set points established for themselves as well as the Z plant outputs, incremental changes of said Y plant inputs being essentially defined by a series of stabilizing control components, Us, reflecting constraints associated with the X plant inputs but bringing about stabilization employing only changes in the Y plant inputs, whereby said X plant inputs do not require optimal set points to define optimal conditions, said method comprising

• (i) establishing mapping logic for relating changes in the X and Y plant inputs plus measured disturbances to changes in the Z plant outputs in matrix format over a series of stepped changes 1 . . . N;
• (ii) after establishing measurements in said Z plant outputs and in prior incremental changes in connection with X and Y plant inputs and measured disturbances, generating a stabilizing and optimizing control (OSC) vector normalized to present time t by means of a series of matrix-vector multiplications, said OSC vector including elements in rows 1 . . . Y and rows 1 . . . X corresponding to incremental changes in plant inputs Y and X, respectively, normalized to present time t as well as incremental future changes in at least Y plant inputs based on said mapping logic of step (i), and being of a form:
• ##EQU12## where M is a limit in future time for controlling said process, (iii) deriving a present time t control code from said OSC vector and causing incremental changes in said inputs X and Y in accordance therewith, whereby said inputs X can be driven toward optimization without use of optimal set points while inputs Y can maintain stability using only incremental changes therein to control outputs Z, in linear proportion thereto,
• (iv) without waiting for steady-state conditions, repeating step (ii) and (iii) so as to pursue optimization of said inputs X without the need for steady-state conditions between control cycles.

A method for integrating on-line optimization without destabilizing of a large-scale refinery, chemical plant or the like associated with a plurality of independently controlled manipulated plant inputs X, Y, and one or more controlled plant outputs Z dependent upon said manipulated plant inputs X, Y in which proportional control of the manipulated plant inputs X, Y can occur in a manner that allows rapid optimization of the X plant inputs wherein said X plant inputs are equal to at least two (2) in number undergoing simultaneous rapid optimization without the need of optimal set points, said method comprising

• (i) after establishing measurements in said Z plant outputs and in prior incremental changes in connection with X and Y plant inputs and measured disturbances, generating a stabilizing and optimizing control (OSC) vector normalized to present time t by means of a series of matrix-vector multiplications, said OSC vector including elements in rows 1 . . . Y and rows 1 . . . X corresponding to incremental changes in plant inputs Y and X, respectively, normalized to present time t as well as incremental future changes in at least Y plant inputs based on mapping logic relating changes in plant inputs to changes in Z plant outputs, and being of a form:
• ##EQU13## where M is a limit in future time for controlling said process, (ii) deriving a present time t control code from said OSC vector and causing incremental changes in said inputs X and Y in accordance therewith, whereby said inputs X can be driven toward optimization without use of optimal set points while inputs Y can maintain stability using only incremental changes therein to control outputs Z in linear proportion thereto,
• (iii) without waiting for steady-state conditions, repeating step (i) and (ii) so as to pursue optimization of said inputs X without the need for steady-state conditions between control cycles.

A method for controlling a two-stage hydrocracking process including a first-stage reactor for reacting a feed and a hydrogen charge under hydrocracking conditions to produce a hydrocracked product and an unreacted hydrocracking effluent, a distillation column fractionator for producing a jet fuel fraction and a jet fuel fraction of said hydrocracked product as well as separating the hydrocracked product from said hydrocracking effluent, into recycle and reboil effluent, and a second stage reactor for reacting the recycled hydrocracking effluent also under hydrocracking conditions with addition hydrogen to produce additional hydrocracked product, by a computer-dominated, constrained-dependent control system integrating on-line optimization without destabilizing the process by controlling controlled outputs Z including particular set points associated with gasoline end point and jet fuel freeze point properties, recycle effluent nitrogen content and hydrocracking effluent level within said distillation column fractionator, via appropriate, proportional manipulation of a series of independently controlled manipulated plant inputs X, Y, said plant inputs Y being associated with temperatures within the first-stage and second-stage reactors, and withdrawal rate of the one of the jet fuel and gasoline fractions from the fractionator, while said plant inputs X being associated with at least the rapid optimization the withdrawal rate of the other of the jet fuel and gasoline fractions whereby said X plant inputs are no longer limited to slow steady-state pursuit toward optimization nor require optimal set points to define said optimal conditions, said method comprising

• (i) establishing mapping logic for relating changes in the X and Y plant inputs plus measured disturbances in connection with feed and recycle flow rates, to changes in the Z plant outputs in matrix format over a series of stepped changes 1 . . . N;
• (ii) after establishing measurements in said Z plant outputs and measurements in connection with prior incremental changes in X and Y plant inputs and in measured disturbances, generating a stabilizing and optimizing control (OSC) vector normalized to present time t by means of a series of matrix-vector multiplications, said OSC vector including elements in rows 1 . . . Y and rows 1 . . . X corresponding to incremental changes in plant inputs Y and X, respectively, normalized to present time t as well as incremental future changes in at least Y plant inputs based on said mapping logic of step (i), and being of a form:
• ##EQU14## where M is a limit in future time for controlling said hydrocracking process,
• (iii) deriving a present time t control code from rows 1 . . . X and 1 . . . Y of said OSC vector and causing incremental changes in said inputs X and Y in accordance therewith, whereby said inputs X can be driven toward optimization without use of optimal set points while inputs Y can maintain stability using only incremental changes therein to control outputs Z in linear proportion thereto.

Method for controlling a two-stage hydrocracking process including a first-stage reactor for reacting a feed and a hydrogen charge under hydrocracking conditions to produce a hydrocracked product and an unreacted effluent, a distillation column fractionator for providing gasoline and jet fuel fractions from the hydrocracked product as well as separating said said hydrocracked product from said unreacted effluent into recycle and reboil effluent, and a second stage reactor for reacting the recycle effluent also under hydrocracking conditions with addition hydrogen to produce additional hydrocracked product, said system integrating on-line optimization without destabilizing the process by controlling controlled outputs Z including particular set points associated with gasoline end point temperature, jet fuel freeze point temperature, recycle nitrogen content and hydrocracking effluent level within said distillation column fractionator, via appropriate, proportional manipulation of a series of independently controlled manipulated plant inputs Y associated with temperatures within the first-stage and second-stage reactors, and withdrawal rate of one of the gasoline and jet fuel fractions from the fractionator, while optimizing other plant inputs X in accordance with a series of pre-tested optimizing control components, Uo, in a manner that allows them to be driven toward optimization under constrained conditions, whereby optimization of said X plant inputs can be accelerated, deaccelerated, zeroed, or reversed without use of optimal set points, said Y plant inputs being reserved for stabilization in connection with maximum set points established for themselves as well as the Z plant outputs, incremental changes of said Y plant inputs being essentially defined by a series of stabilizing control components, Us, reflecting constraints associated with the X plant inputs but bringing about stabilization employing only changes in the Y plant inputs, irrespective of disturbance changes in the process associated with changes in feed and recycle flow rates, said method comprising

• (i) establishing mapping logic for relating changes in the X and Y plant inputs plus measured disturbances to changes in the Z plant outputs in matrix format over a series of stepped changes 1 . . . N;
• (ii) after establishing measurements in said Z plant outputs and in prior incremental changes in connection with X and Y plant inputs and in measured disturbances, generating a stabilizing and optimizing control (OSC) vector normalized to present time t by means of a series of matrix-vector multiplications, said OSC vector including elements in rows 1 . . . Y and rows 1 . . . X corresponding to incremental changes in plant inputs Y and X, respectively, normalized to present time t as well as incremental future changes in at least Y plant inputs based on said mapping logic of step (i), and being of a form:
• ##EQU16## where M is a limit in future time for controlling said hydrocracking process,
• (iii) deriving a present time t control code from rows 1 . . . X and 1 . . . Y of said OSC vector and causing incremental changes in said inputs X and Y in accordance therewith, whereby said inputs X can be driven toward optimization without use of optimal set points while inputs Y can maintain stability using only incremental changes therein to control outputs Z in linear proportion thereto,
• (iv) without waiting for steady-state conditions, repeating step (ii) and (iii) so as to pursue optimization of said inputs X without the need for steady-state conditions between control cycles.

A method for controlling a large-scale hydrocracking process that includes first-stage and second stage hydrocracking reactor means for producing a hydrocracked product and means for providing a gasoline fraction and a jet fuel fraction of said hydrocracked product as well as separating said hydrocracked product from a hydrocracking effluent so as to produce a recyle portion and a reboil portion, in connection with a computer-dominated constraint-dependent control system that integrates on-line optimization of at least two plant inputs X without using optimal set points in connection therewith, with stabilization of Z plant outputs, said X plant inputs including at least one input related to a characteristic associated with incremental change in one of said gasoline and jet fuel fractions and at least another input related to incremental change in a characteristic associated with said reboil portion of said hydrocracking effluent, said plant outputs Z of the process including gasoline end point temperature, jet freeze point temperature, recycle hydrocracking effluent level and nitrogen content of the recycled hydrocracking effluent, said plant inputs Y being reserved for stabilization and including incremental changes in connection with temperatures within said first-stage and second-stage reactor means, and a characteristic associated with other of said gasoline and jet fuel fractions, said system manipulating said inputs using a present time t control code generated from a series of matrix-vector multiplications, whereby said plant inputs X can be driven toward optimization without use of optimal set points while inputs Y can maintain stability using only incremental changes therein to control outputs Z in linear proportion thereto, said method comprising:

• (i) establishing mapping logic for relating changes in the X and Y plant inputs plus measured disturbances to changes in the Z plant outputs in matrix format over a series of stepped changes 1 . . . N;
• (ii) after establishing measurements in said Z plant outputs and in connection with in prior incremental changes in X and Y plant inputs and in measured disturbances, generating a stabilizing and optimizing control (OSC) vector normalized to present time t by means of a series of matrix-vector multiplications, said OSC vector including elements in rows 1 . . . Y and rows 1 . . . X corresponding to incremental changes in plant inputs Y and X, respectively, normalized to present time t as well as incremental future changes in at least Y plant inputs based on said mapping logic of step (i), and being of a form:
• ##EQU18## M is a limit in future time for controlling said hydrocracking process, (iii) deriving a present time t control code from rows 1 . . . X and 1 . . . Y of said OSC vector and causing incremental changes in said inputs X and Y in accordance therewith, whereby said inputs X can be driven toward optimization without use of optimal set points while inputs Y can maintain stability using only incremental changes therein, in linear proportion to deviations from preselected conditions, to control outputs Z,
• (iv) without waiting for steady-state conditions, repeating step (ii) and (iii) so as to pursue optimization of said inputs X without the need for steady-state conditions between control cycles.

A computer-dominated constraint-dependent control system for integrating on-line optimization without destabilizing a refinery, chemical process or the like associated with a plurality of independently controlled manipulated plant inputs X, Y, and one or more controlled plant outputs Z dependent upon said manipulated plant inputs X, Y in which proportional control of the manipulated plant inputs X, Y can occur using control codes generated from a plurality of matrix-vector multiplications, in a manner that allows rapid optimization of the X plant inputs but wherein said X plant inputs do not require optimal set points to define optimal conditions, said system comprising

• a series of flow control means within said refinery, chemical process or the like, associated with said X and Y inputs for effecting incremental changes therein, and a series of transducer means for measuring said Z plant outputs and measured disturbances associated with inputs X, Y,
• controller means including mechanical linkage means operationally connected to said flow control means associated with said X and Y plant inputs for independently effecting incremental change therein, said controller means also including separate transducer means for measuring incremental changes in said X and Y plant inputs,
• computer means operationally connected to said series of transducer means and said controller means for generating a present time t control code as well as for predicting future conditions associated therewith, thereby providing for the systematic control of said Z plant outputs by manipulating said X and Y plant inputs, wherein on an incremental proportional basis alone, said X inputs pursuing optimization, said Y inputs pursuing stabilization in connection with themselves as well as said Z plant outputs
• said computer means including a central processing unit (CPU), a program memory for controlling said CPU, and general memory means for storing general data related to the refinery, chemical process or the like to be controlled, including mapping logic for relating changes in the X and Y plant inputs plus measured disturbances to changes in the Z plant outputs in matrix format over a series of stepped changes 1 . . . N;
• said general memory means also storing a series of limiting set points for each of said Y plant inputs and for one or more of said plant outputs,
• said CPU in association with said program memory and said general memory, generating an optimizing and stabilizing control (OSC) vector normalized to time t as a result of a series of matrix-vector multiplications, said OSC vector including elements corresponding to incremental changes in inputs X and Y normalized to present time t as well as including incremental future changes in at least Y plant inputs, and being of a form ##EQU20## where M is a limit in future time for controlling said process, whereby said inputs X can be driven toward optimization without use of optimal set points while inputs Y can maintain stability as a function of incremental changes therein to thereby maintain plant outputs Z, in linear proportion thereto.

A computer-dominated constraint-dependent control system for integrating on-line optimization without destabilizing a refinery, chemical process or the like associated with a plurality of independently controlled manipulated plant inputs X and Y, and one or more controlled plant outputs Z dependent upon said manipulated plant inputs X, Y in which closed loop, proportional control of the manipulated plant inputs X, Y can occur in accordance with a series of pre-tested optimizing control components, Uo, whereby optimization of the X plant inputs can be accelerated, deaccelerated, zeroed, or reversed without use of optimal set points in association therewith, said Y plant inputs being reserved for stabilization in connection with set points established for themselves as well as the Z plant outputs, incremental changes of said Y plant inputs being essentially defined by a series of stabilizing control components, Us, reflecting constraints associated with the X plant inputs but bringing about stabilization employing only changes in the Y plant inputs, whereby said X plant inputs do not require optimal set points to define optimal conditions, said system comprising

• a series of flow control means for effecting incremental changes in said X and Y inputs within said refinery, chemical process or the like, and a series of transducer means for measuring changes in said Z plant outputs and measured disturbances associated with said plant inputs X and Y,
• universal controller means including mechanical linkage means operationally connected to said control means for independently effecting change in each of said X and Y plant inputs over a time t control cycle, said controller means having separate transducer means for measuring incremental changes in said X and Y plant inputs,
• computer means including input/output means operationally connected to said series of transducer means and said controller means, a central processing unit (CPU) that includes a control unit and general and program memories, for generating a present time t control code therefor as well as for predicting future conditions, based on measurements from both said series and said separate transducer means storable within said general memory in correct format and subsequent manipulations employing a series of instructions stored within said program memory, thereby providing for the systematic control of said Z plant outputs by manipulating said X and Y plant inputs, wherein said X inputs pursue optimization while Y inputs pursue stabilization in connection with themselves as well as said Z plant outputs
• said general memory also being capable of storing mapping logic for relating actual changes in the X and Y plant inputs plus measured disturbances to changes in the Z plant outputs in matrix format over a series of stepped changes 1 . . . N;
• said general memory also being capable of storing a series of limiting set points for each of said Y plant inputs and for one or more of said plant outputs, as well as storing a series of gain matrices Bx, said gain matrices Bx consisting of a proportional gain B matrix associated with constrained stabilization, and a proportional gain Bo/r matrix associated with unconstrained stabilization, for use in synthesizing control of said refinery, chemical process or the like, on a proportional basis alone without need for optimal set points in connection with said X plant inputs,
• said CPU under control of said control unit in association with instructions from said program memory, generating an optimizing and stabilizing control (OSC) vector normalized to time t through a series of matrix-vector multiplications, said OSC vector including elements related to present time t elements from which said time t control code can be derived as well as future times associated with at least said Y plant inputs, and being of the form ##EQU22## whereby said inputs X can be driven toward optimization without use of optimal set points while inputs Y can maintain stability as a function of incremental changes therein to thereby control plant outputs Z in linear proportion thereto.

A computer-dominated constraint-dependent control system for integrating on-line optimization without destabilizing of a large-scale refinery, chemical plant or the like associated with a plurality of independently controlled manipulated plant inputs X, Y, and one or more controlled plant outputs Z dependent upon said manipulated plant inputs X, Y in which proportional control of the manipulated plant inputs X, Y can occur in a manner that allows rapid optimization of the X plant inputs wherein said X plant inputs are equal to at least two (2) in number undergoing simultaneous rapid optimization without the need of optimal set points, said system comprising

• a series of transducer means for measuring said Z plant outputs and changes in known disturbances associated with inputs X, Y,
• controller means including flow control means positioned within said refinery, chemical plant or the like for independently effecting incremental changes in said X and Y plant inputs, said controller means also including separate transducer means for measuring incremental changes in said X and Y plant inputs,
• computer means operationally connected to said series of transducer means and said controller means for generating a present time t control code for effecting incremental changes in said X and Y plant inputs on a systematic basis as well as for predicting future conditions associated therewith, thereby providing for the systematic control of said Z plant outputs, in linear proportion thereto, said at least two X inputs pursuing optimization without use of optimal set points within said controller means while Y inputs pursue stabilization in connection with themselves as well as said Z plant outputs using maximum set point limits,
• said computer means including a central processing unit (CPU), a program memory for controlling said CPU, and general memory means for storing general data related to the refinery, chemical plant or the like to be controlled, including mapping logic for relating changes in the X and Y plant inputs plus measured disturbances to changes in the Z plant outputs in matrix format over a series of stepped changes 1 . . . N;
• said CPU in association with said program memory and said general memory, generating an optimizing and stabilizing control (OSC) incremental vector normalized to time t as a result of a series of matrix-vector multiplications, said OSC vector including present time t elements as well as future time elements and being of a form from which said present t control code can be derived:
• ##EQU23## whereby said inputs X can be driven toward optimization without use of optimal set points while inputs Y can maintain stability as a function of incremental changes therein to control plant outputs Z in linear proportion thereto.

A computer-dominated constraint-dependent control system for controlling a two-stage hydrocracking process including a first-stage reactor for reacting a feed and a hydrogen charge under hydrocracking conditions to produce a hydrocracked product and an unreacted hydrocracking effluent, a distillation column fractionator for producing a jet fuel fraction and a jet fuel fraction of said hydrocracked product as well as separating the hydrocracked product from said hydrocracking effluent, into recycle and reboil effluent, and a second stage reactor for reacting the recycled hydrocracking effluent also under hydrocracking conditions with addition hydrogen to produce additional hydrocracked product, said system integrating on-line optimization without destabilizing the process by controlling controlled outputs Z including particular set points associated with gasoline end point and jet fuel freeze point characteristics, recycle effluent nitrogen content and hydrocracking effluent level within said distillation column fractionator, via appropriate, proportional manipulation of a series of independently controlled manipulated plant inputs X, Y, said plant inputs Y being associated with temperatures within the first-stage and second-stage reactors, and withdrawal rate of the one of the jet fuel and gasoline fractions from the fractionator, while said plant inputs X being associated with at least the rapid optimization the withdrawal rate of the other of the jet fuel and gasoline fractions whereby optimization of said X plant inputs do not use optimal set points, said system comprising

• a series of flow control means within said two-stage hydrocracking process associated with said X and Y inputs for effecting incremental change, and a series of transducer means for measuring changes in said Z plant outputs and measured disturbance changes in connection with flow rates of said feed and recycle effluent,
• controller means including mechanical linkage means operationally connected to said flow control means associated with said X and Y plant inputs for independently effecting incremental change thereto, said controller means also including separate transducer means for measuring changes in said X and Y plant inputs and for establishing maximum set points for said X plant inputs only,
• computer means operationally connected to said series of transducer means and said controller means for generating a present time t control code as well as for predicting future conditions associated therewith, as a result of a series of matrix-vector multiplications, said control code providing for the systematic control of said Z plant outputs by manipulating said X and Y plant inputs on a proportional basis alone, said X inputs pursuing optimization, said Y inputs pursuing stabilization in connection with themselves as well as said Z plant outputs
• said computer means including a central processing unit (CPU), a program memory for controlling said CPU, and general memory means for storing general data related to said two-stage hydrocracking process, including mapping logic for relating changes in the X and Y plant inputs plus measured changes in the feed and recycle rates to changes in the Z plant outputs in matrix format over a series of stepped changes 1 . . . N;
• said general memory means also storing a series of limiting set points for each of said Y plant inputs and for one or more of said plant outputs Z,
• said CPU in association with said program memory and said general memory, generating an optimizing and stabilizing control (OSC) vector normalized to time t for control purpose, but including incremental future changes in at least Y plant inputs, said OSC vector including present t elements from which said time t control code can be derived and being of a form ##EQU24## where M is a limit in future time for controlling said hydrocracking process, whereby said inputs X can be driven toward optimization without use of optimal set points while inputs Y can maintain stability as a function of incremental changes therein to control said plant outputs Z in linear proportion thereto.

A computer-dominated constraint-dependent control system for for controlling a two-stage hydrocracking process including a first-stage reactor for reacting a feed and a hydrogen charge under hydrocracking conditions to produce a hydrocracked product and an unreacted effluent, a distillation column fractionator for providing gasoline and jet fuel fractions from the hydrocracked product as well as separating said hydrocracked product from said unreacted effluent into recycle and reboil effluent, and a second stage reactor for reacting the recycle effluent also under hydrocracking conditions with addition hydrogen to produce additional hydrocracked product, said system integrating on-line optimization without destabilizing the process by controlling controlled outputs Z including particular set points associated with gasoline end point temperature, jet fuel freeze point temperature, recycle nitrogen content and hydrocracking effluent level within said distillation column fractionator, via appropriate, proportional manipulation of a series of independently controlled manipulated plant inputs Y associated with temperatures within the first-stage and second-stage reactors, and withdrawal rate of one of the gasoline and jet fuel fractions from the fractionator, while optimizing other plant inputs X in accordance with a series of pre-tested optimizing control components, Uo, in a manner that allows them to be driven toward optimization under constrained conditions, whereby optimization of said X plant inputs can be accelerated, deaccelerated, zeroed, or reversed without use of optimal set points, said Y plant inputs being reserved for stabilization in connection with maximum set points established for themselves as well as the Z plant outputs, incremental changes of said Y plant inputs being essentially defined by a series of stabilizing control components, Us, reflecting constraints associated with the X plant inputs but bringing about stabilization employing only changes in the Y plant inputs, irrespective of disturbance changes in the process associated with changes in feed and recycle flow rates, said system comprising

• a series of flow control means within said two-stage hydrocracking process associated with said X and Y inputs for effecting incremental change, and a series of transducer means for measuring changes in said Z plant outputs and measured disturbance changes in connection with said feed and recycle flow rates,
• controller means including mechanical linkage means operationally connected to said flow control means associated with said X and Y plant inputs for independently effecting incremental change thereto, said controller means also including separate transducer means for measuring changes in said X and Y plant inputs and for establishing maximum set points for said X plant inputs only,
• computer means including input/output means operationally connected to said series of transducer means and said controller means and a central processing unit (CPU) that includes a control unit and general and program memories for generating a present time t control code therefor as well as for predicting future conditions, based on measurements from both said series and said separate transducer means storable within said general memory in correct format and operatable on employing a series of instructions stored within said program memory, thereby providing for the systematic control of said Z plant outputs by manipulating said X and Y plant inputs, wherein said X inputs pursue optimization while Y inputs pursue stabilization in connection with themselves as well as said Z plant outputs, said general memory also being capable of storing mapping logic for relating actual changes in the X and Y plant inputs plus changes in the measured disturbances in connection with feed and recycle flow rates, to changes in the Z plant outputs in matrix format over a series of stepped changes 1 . . . N;
• said general memory also being capable of storing a series of limiting set points for each of said Y plant inputs and for one or more of said plant outputs, as well as storing a series of gain matrices Bx, said gain matrices Bx consisting of a proportional gain B matrix associated with constrained stabilization, and a proportional gain Bo/r matrix associated with unconstrained stabilization, for use in synthesizing control of said process, on a proportional basis alone without need for optimal set points in connection with said X plant inputs,
• said CPU under control of said control unit in association with instructions from said program memory, generating an optimizing and stabilizing control (OSC) vector normalized to time t but including incremental future changes in at least Y plant inputs, said OSC vector including present t elements from which said time t control code can be derived and being of the form ##EQU25## where M is a limit in future time for controlling said process, whereby said inputs X can be driven toward optimization without use of optimal set points while inputs Y can maintain stability as a function of incremental changes therein to control said plant outputs Z in linear proportion thereto.

A computer-dominated constraint-dependent control system for a large-scale hydrocracking process that includes first-stage and second stage hydrocracking reactor means for producing a hydrocracked product and means for providing a gasoline fraction and a jet fuel fraction of said hydrocracked product as well as separating said hydrocracked product from a hydrocracking effluent so as to produce a recycle portion and a reboil portion, said system integrating simultaneous on-line optimization of at least two plant inputs X without using optimal set points in connection therewith, one of said X plant inputs being related to a characteristic associated with incremental change in one of said gasoline and jet fuel fractions and the other being related to incremental change in a characteristic associated with said reboil portion of said hydrocracking effluent, said system also stabilizing controlled plant outputs Z of the process including gasoline end point temperature, jet freeze point temperature, recycle hydrocracking effluent level and nitrogen content of the recycled hydrocracking effluent, via manipulating remaining plant inputs Y using a present time t control code generated from a series of matrix-vector multiplications, said plant inputs Y being reserved for stabilization, and including incremental changes in connection with temperatures within said first-stage and second-stage reactor means, and a characteristic associated with other of said gasoline and jet fuel fractions,

• controller means including flow control means positioned within said process for independently effecting incremental changes in said X and Y plant inputs, said controller means also including separate transducer means for measuring incremental changes in said X and Y plant inputs,
• computer means operationally connected to said series of transducer means and said controller means for generating a present time t control code for effecting incremental changes in said X and Y plant inputs on a systematic basis as well as for predicting future conditions associated therewith, thereby providing for improved control of said Z plant outputs, said at least two X inputs pursuing optimization without use of optimal set points within said controller means, said Y inputs pursuing stabilization in connection with themselves as well as said Z plant outputs using maximum set point limits integral therewith,
• said computer means including a central processing unit (CPU), a program memory for controlling said CPU, and general memory means for storing general data related to said process including mapping logic for relating changes in the X and Y plant inputs plus measured disturbances to changes in the Z plant outputs in matrix format over a series of stepped changes 1 . . . N;
• said CPU in association with said program memory and said general memory, generating an optimizing and stabilizing control (OSC) incremental vector normalized to time t, as a product of a series of matrix-vector multiplications with said CPU, said OSC vector also including both incremental future changes in at least Y plant inputs, as well as present t elements from which said time t control code can be derived, and being of a form ##EQU26## where M is a limit in future time for controlling said process, whereby said inputs X can be driven toward optimization without use of optimal set points while inputs Y can maintain stability as a function of incremental changes therein to control said plant outputs Z in linear proportion thereto.