Discontinuous optimization procedure modelling the run-idle status of plural process components

• US 4,607,325 A
• Filed: 03/30/1984
• Issued: 08/19/1986
• Est. Priority Date: 10/21/1981
• Status: Expired due to Term

• Current Assignee: Honeywell Incorporated (Honeywell International Inc.)
• Original Assignee: Honeywell Incorporated (Honeywell International Inc.)
• Inventors: Horn, Brian C.
• Primary Examiner(s): Smith, Jerry
• Assistant Examiner(s): Lastova, John R.
• Application Number:
  US06/595,117
• Time in Patent Office: 872 Days
• Field of Search: 364/149-153, 364/156, 364/166, 364/180, 364/493-495, 318/561
• US Class Current: 700/31
• CPC Class Codes: G05B 13/04 involving the use of models or simulators

Abstract

A method of optimizing the operation of a process so that desired products are produced at minimum cost. The process has a plurality of process components, with each component having a run status and an idle status. The process has available more than one input and produces more than one output. The process is controlled by a computer which is provided with a mathematical model of the process, which model includes a model of each of its components. The model for each process component includes a logic variable, which logic variable can have only two values, 0 and 1, and which represent the run or idle status of each process component. The computer, when predetermined conditions arise, solves a mixed integer equation to determine the optimum state of the process at a given time to produce the desired outputs, at desired rates and at minimum cost; and, in doing so, determines the value of the logic variable for each component. When a change of status of a process component from run to idle or idle to run is required to place the process in its optimum state, control signals to cause such change of status of process components are applied to such components.

Claims

1. A method of operating a process with the aid of a digital computer to minimize the cost of producing desired outputs of the processor, said process having a plurality of process components with each component having a mathematical model, at least some of said mathematical models being discontinuous, said process components being capable of being placed in a run or idle status, said process being capable of having applied to it one or more inputs and capable of producing one or more outputs, comprising the steps of:

constantly providing said computer with data concerning the status of operation of each component, each output and each input of the process;

providing the computer with a mathematical model of the process which includes mathematical models of each component;

providing the computer with a real time discontinuous optimization procedure for producing an optimum state of operation of the process, said discontinuous optimization procedure determining the run or idle status of each component to minimize cost of the process;

providing the computer with a procedure to update the model;

providing the computer with a scheduling procedure;

said scheduling procedure substantially continuously monitoring the state of operation of the process, comparing the current state of operation of the process with the last optimum state produced by the optimization procedure;

said scheduling procedure initiating the operation of the optimization procedure to produce an updated optimum state if the current state of the process and the most recently calculated optimum state differ significantly or after the passage of a predetermined period of time; and

said scheduling procedure changing the status of process components from run to idle or idle to run to be in accord with the updated optimum state of the process, as well as optimizing the operation of each process component to produce the desired outputs at minimum cost.

2. The method of claim 1 in which the discontinuous optimization procedure solves a mixed integer linear program with a logic variable for each process component.

3. The method of claim 2 in which the mathematical model is a set of linear equations.

4. The method of claim 3 in which the discontinuous optimization procedure includes a cost factor associated with a change of status of each component from run to idle and from idle to run.

5. A method of operating with the aid of a digital computer a process having a plurality of process components with each component capable of being placed in a run status or idle status, each of said process components having a mathematical model, the mathematical model of at least one of the components having a discontinuity between its run status and its idle status, said process being capable of having applied to it a plurality of inputs and capable of producing a plurality of products, comprising:

constantly providing said computer with data concerning the status and operation of each component, of each product and of each input to the process;

providing the computer with a mathematical model of the process, such model including logic variable parameters, and continuous variables and parameters, for each process component;

providing the component with a discontinuous real time optimization procedure for producing the values for the logic variable of each process component and for the operation of each process component to produce an optimum state of operation of the process at minimum cost at a given time;

providing the computer with a procedure to update the parameters of the model;

providing the computer with a scheduling procedure, said scheduling procedure monitoring the state of operation of the process, comparing the current state of operation of the process with the optimum state at said given time as produced by the optimization procedure;

said scheduling procedure initiating the operation of the optimization procedure to produce an updated optimum state if the current state and the optimum state differ significantly or after the passage of a predetermined period of time; and

said optimization procedure producing an updated optimum state and adjusting the process to bring it into substantial accord with the updated optimum state as calculated by the optimization procedure, including changing the status of process components from run to idle or idle to run, as well as optimizing the operation of each process component to produce the desired outputs at desired rates at minimum cost.

6. The method of claim 5 in which the discontinuous optimization procedure solves a set of mixed integer linear equations.

7. The method of claim 6 in which the discontinuous optimization procedure also produces values for the independent variables of the mixed integer linear equations of the process.

8. The method of claim 7 in which the discontinuous optimization procedure includes a cost factor associated with a change of status of each component from run to idle or from idle to run.

9. A method of operating with the aid of a digital computer a utility plant having a plurality of components with each component capable of being placed in a run status or idle status, each component having a mathematical model, the mathematical model of a component of the utility plant being discontinuous between its respective run state and idle state, said plant being capable of having applied to it a plurality of energy inputs and capable of producing a pluraity of energy outputs, comprising the steps of:

constantly providing said computer with data concerning the status and operation of each component, each product of and each input of the plant;

providing the computer with a mathematical model of the plant, such model including logic variables, continuous variables and parameters for each component;

providing the computer with a real time discontinuous optimization procedure for producing the values for the logic variables and continuous variables to produce an optimum state of operation of the plant at a given time;

providing the computer with a procedure to update the parameters of the model;

providing the computer with a scheduling procedure, said scheduling procedure monitoring the state of operation of the plant, comparing the current state of operation of the plant with the optimum state at said given time as produced by the optimization procedure and initiating the operation of the optimization procedure if the current state and the optimum state differ significantly or after the passage of a predetermined period of time; and

said optimization procedure, when initialized, producing an updated optimum state for the process and adjusting the operation of the plant to bring it into substantial accord with the updated optimum state as determined by the optimization procedure, including changing the status of the process components from run to idle or idle to run, as well as optimizing the operation of each process component to produce the desired outputs at desired rates at minimum cost.

10. The method of claim 9 in which outputs of the plant include electrical power and steam.

11. The method of claim 10 in which the steam is produced at a plurality of differing pressures.

12. The method of claim 11 in which the inputs include at least two fuels and electrical power from a utility.

13. The method of claim 12 in which the plant includes a plurality of turbogenerator sets for producing electric power.

14. The method of claim 13 in which the discontinuous procedure solves a set of mixed integer linear equations to determine the optimum state of the plant at any given time.

15. The method of claim 14 in which the discontinuous optimization procedure also produces values for the continuous variables of the mixed integer linear equation.

16. The method of claim 15 in which the discontinuous optimization procedure includes a cost factor associated with a change of status of each plant component from run to idle or from idle to run.

17. The method of claim 16 in which one of the components of the plant is a boiler, the mathematical model of which is discontinuous between its run and idle states.