System and method of optimizing blending ratios for producing product

Patent metadata

Field	Value
Publication number	US-9773097-B2
Application number	US-201414453140-A
Country	US
Kind code	B2
Filing date	Aug 6, 2014
Priority date	Aug 6, 2014
Publication date	Sep 26, 2017
Grant date	Sep 26, 2017

Abstract

An optimization system includes a) one or more processors; b) a loop search engine configured to perform a loop search based on a random blending ratio of blending components for at least a product, and to generate a first optimized blending ratio; and c) a local search engine configured to perform a local search based on the first optimized blending ratio, and to generate a second optimized blending ratio. The loop search engine and the local search engine are implemented on the one or more processors.

First claim

What is claimed is:

1. An optimization system comprising:
one or more processors;
a recipe optimizer configured to randomly distribute solutions within predefined upper and lower limits of i) acceptable ranges in volumes of feed tanks and final product tanks; and ii) acceptable ranges in properties of an oil product, to generate random solutions including a random blending ratio of blending components for at least a product;
a loop search engine configured to perform a loop search based on the random solutions, and to generate a first optimized blending ratio; and
a local search engine configured to perform a local search based on the first optimized blending ratio, and to generate a second optimized blending ratio,
wherein the loop search engine and the local search engine are implemented on the one or more processors.

2. An optimization system comprising:
one or more processors;
an initialization module configured to create a random blending ratio of blending components automatically;
a loop search engine configured to perform a loop search based on the random blending ratio of blending components for at least a product, and to generate a first optimized blending ratio; and
a local search engine configured to perform a local search based on the first optimized blending ratio, and to generate a second optimized blending ratio; and
wherein the initialization module, the loop search engine and the local search engine are implemented on the one or more processors.

3. The optimization system of claim 2 , wherein the loop search engine and the local search engine are configured to reiterate the loop search and the local search cyclically.

4. The optimization system of claim 3 , wherein the loop search comprises a genetic algorithm loop search, and the local search comprises a Nelder-Mead simplex method search.

5. The optimization system of claim 2 , further comprising:
a generation counter module configured to count a generation number every time the loop search engine and the local search engine have performed the loop search and the local search, respectively,
wherein the generation counter module is implemented on the one or more processors,
wherein the loop search engine and the local search engine are configured to reiterate the loop search and the local search cyclically until the generation number counted by the generation counter module reaches a predetermined number, and
wherein the loop search comprises a genetic algorithm loop search, and the local search comprises a Nelder-Mead simplex method search.

6. The optimization system of claim 2 , wherein the initialization module is configured to create the random blending ratio, based on at least:
an optimization target;
configuration parameters of the loop search and the local search; and
a set of boundary conditions.

7. The optimization system of claim 6 , wherein the set of boundary conditions comprises:
a specification of a product, the product being a blend of a plurality of blending components;
an available amount of each of the blending components; and
relevant properties of each of the blending components.

8. The optimization system of claim 7 , wherein the specification of the product includes at least one of: sulfur content, research octane number, motor octane number, Reid vapour pressure, flash point, viscosity, olefin, benzene, oxygen, aromatics, boiling point, true specific gravity, and price of each product.

9. The optimization system of claim 8 , wherein the relevant properties include at least one of: sulfur content, research octane number, motor octane number, Reid vapour pressure, flash point, viscosity, olefin, benzene, oxygen, aromatics, boiling point, true specific gravity, and cost of each of the blending components.

10. The optimization system of claim 6 , wherein the optimization target includes an optimization of an objective function configured to:
reduce a total cost of all the blending components to be used;
increase a total profit of all the products together;
reduce a total remaining mass balance of each blending component; and
increase throughput of at least one of the products.

11. The optimization system of claim 2 , further comprising:
a user interface configured to display the second optimized blending ratio on a display screen.

12. The optimization system of claim 2 , wherein the loop search engine comprises a genetic algorithm loop search engine, and wherein the local search engine comprises a Nelder-Mead simplex method search engine.

13. The optimization system of claim 2 , further comprising:
a memory device accessible by the one or more processor, the memory device stores a set of program components that, when executed by the one or more processor, cause the processor to act as the loop search and the local search engine.

14. The optimization system of claim 13 , wherein the memory device stores: a set of information, the set of information comprising:
a plurality of predefined boundary conditions;
a plurality of minimized objective functions;
a plurality of maximized objective functions, which are the reciprocals of the minimized objective functions; and
a plurality of configuration parameters for a genetic algorithm loop search and a Nelder-Mead simplex method for the loop search and the local search engine, respectively.

15. The optimization system of claim 2 , wherein the one or more processor is programmed to implement the loop search engine and the local search engine.

16. An optimization system comprising:
one or more processors;
a loop search engine configured to perform a loop search based on a random blending ratio of blending components for at least a product, and to generate a first optimized blending ratio;
a local search engine configured to perform a local search based on the first optimized blending ratio, and to generate a second optimized blending ratio; and
a user interface configured to display the second optimized blending ratio on a display screen,
wherein the user interface is configured to receive to entry of at least:
an optimization target;
configuration parameters of the loop search and the local search; and
a set of boundary conditions.

17. The optimization system of claim 16 , further comprising:
an initialization module configured to create the random blending ratio based on at least:
the optimization target,
the configuration parameters of the loop search and the local search, and
the set of boundary conditions.

18. The optimization system of claim 17 , wherein the set of boundary conditions comprises:
a specification of a product which is a blend of a plurality of blending components;
an available amount of each of the blending components; and
relevant properties of each of the blending components.

19. The optimization system of claim 18 , wherein the optimization target includes an optimization of an objective function configured to:
reduce a total cost of all the blending components to be used;
increase a total profit of all the products together;
reduce a total remaining mass balance of each blending component; and
increase throughput of at least one of the products.

20. The optimization system of claim 16 , wherein the user interface is configured to receive an entry of information of an existing volume of each product in its product tank.

21. A computer-implemented method of optimizing a blending ratio of blending components for at least a product, the method comprising:
randomly distrib

Assignees

• Yokogawa Electric Corp

Inventors

• Mu Sheng Jing
• Zeng Dan
• Wang Wen Hua
• Gu Jin Sheng

Classifications

• G06F19/70 Primary

  Physics · mapped topic

• G16C20/10 Primary

  Analysis or design of chemical reactions, syntheses or processes · CPC title

• G16C20/40

  Searching chemical structures or physicochemical data · CPC title

• G16C99/00 Primary

  Subject matter not provided for in other groups of this subclass · CPC title