# US 2010/0161289 A1 · Framework for interoperability of solvers

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2010/0161289 A1
- Document ID: US20100161289A1
- Status: Published application
- Kind: A1
- Application number: 12/338,159
- Filed: 2008-12-18
- Published: 2010-06-24
- Earliest claimed: 2008-12-18
- Assignees: Microsoft Corporation
- Inventors: Alexander Sasha Stojanovic; Min Wei
- Web page: https://patentgrep.com/patent/US20100161289A1

## CPC

- G06F 17/10: Complex mathematical operations {(function generation by table look-up G06F1/03; evaluation of elementary functions by calculation G06F7/544)}
- G06F 8/51: Source to source
- G06N 5/01: Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

A framework may be provided for facilitating adoption of mathematical and logical solvers. A model, representing a problem, may be defined. The model may be defined: via a series of statements in a type safe common-language-specification compliant modeling language; via an intermediate format; or via a series of calls to a standardized application program interface. The framework may translate the defined model to an intermediate format, which many solvers may understand. The framework may provide a number of solver-related services including, but not limited to, services for creating a model, analyzing a model, selecting a well-suited solver based on analysis of a model, decomposing a model into multiple sub-models and providing the multiple sub-models to multiple solvers as multiple parallel threads, data binding, simulation, and post optimality analysis.

## Claims

### Claim 1

A system to be executed within one or more processing devices to provide a framework for facilitating adoption of a plurality of solvers, the system comprising:
a modeling language translator embedded in the framework for receiving a model defined in a particular modeling language and outputting the model defined in a second modeling language;
a plurality of solvers for solving a problem as defined by a model;
an application program interface for interfacing with any of the plurality of solvers; and
a data binder for binding either modeling input to a data source or modeling output to a data sink, or for binding both the modeling input to the data source and the modeling output to the data sink.

### Claim 2 (depends on 1)

The system of claim 1, wherein the particular modeling language is common-language-specification compliant.

### Claim 3 (depends on 2)

The system of claim 2, wherein the particular modeling language is type safe.

### Claim 4 (depends on 1)

The system of claim 1, wherein the second modeling language includes an intermediate format.

### Claim 5 (depends on 1)

The system of claim 1, wherein
the second modeling language includes Mathematical Programming System.

### Claim 6 (depends on 5)

The system of claim 5, further comprising:
a solver selector for automatically analyzing a model and determining a well-suited type of one of the plurality of solvers to select to solve a problem defined by the analyzed model.

### Claim 7 (depends on 6)

The system of claim 6, further comprising:
a solver scheduler for providing the model to the one of the plurality of selected solvers of the well-suited type, launching the one of the selected solvers, and collecting a produced solution.

### Claim 8 (depends on 7)

The system of claim 7, further comprising
a parallel solver scheduler for analyzing the model, decomposing the model into a plurality of sub-models of the model, providing respective ones of the sub-models to respective ones of the plurality of solvers well-suited suited for solving respective problems represented by the respective ones of the sub-models, launching the respective ones of the plurality of solvers as multiple parallel threads, and collecting respective produced solutions from the respective ones of the plurality of solvers.

### Claim 9

A tangible machine-readable medium having instructions recorded thereon for at least one processor, the instructions comprising:
instructions for providing a facility for specifying a model associated with a problem, the facility providing for the specifying of the model via a common-language-specification-compliant modeling language, an intermediate format, and a standardized application program interface;
instructions for providing a facility for translating the specified model in the common-language-specification-compliant modeling language to the intermediate format;
instructions for providing the model to one of a plurality of solvers for solving the problem, the model being provided to the one of the plurality of solvers as input in the intermediate format; and
instructions for collecting a solution produced by the one of the plurality of solvers and reporting a result.

### Claim 10 (depends on 9)

The tangible machine-readable medium of claim 9, wherein the instructions further comprise:
instructions for providing a facility for binding either modeling input to a data source or modeling output to a data sink, or for binding both the modeling input to the data source and the modeling output to the data sink.

### Claim 11 (depends on 9)

The tangible machine-readable medium of claim 9, wherein the instructions further comprise:
instructions for binding a parameter for a model to a data source; and
instructions for binding a decision from a model to a data sink.

### Claim 12 (depends on 9)

The tangible machine-readable medium of claim 9, wherein the instructions for providing the model to one of a plurality of solvers further comprise:
instructions for providing the one of the plurality of solvers with configuration information according to one or more specified directives.

### Claim 13 (depends on 9)

The tangible machine-readable medium of claim 9, wherein the instructions further comprise:
instructions for simulating the solution to determine a sensitivity of the model with respect to specified constraints and decisions from the one of the plurality of solvers.

### Claim 14 (depends on 9)

The tangible machine-readable medium of claim 9, wherein the instructions further comprise:
instructions for performing a post-optimality analysis and producing a sensitivity report.

### Claim 15

A processing device comprising:
at least one processor; and
a memory connected to the processor, the memory comprising:
instructions for transforming common-language compliant modeling language statements, for defining a model, to an intermediate format,
instructions for providing the intermediate format, defining a model, to a solver,
instructions for providing a standardized application program interface for creating and defining a model,
instructions for binding modeling input to a data source,
instructions for binding modeling output to a data sink, and
instructions for reporting a solution produced by a solver.

### Claim 16 (depends on 15)

The processing device of claim 15, wherein the memory further comprises:
instructions for analyzing a model and decomposing the model into a plurality of sub-models,
instructions for providing respective ones of the sub-models to respective ones of a plurality of solvers well-suited suited for solving respective problems represented by the respective ones of the sub-models, and
instructions for launching the respective ones of the plurality of solvers as multiple parallel threads over a plurality of processors.

### Claim 17 (depends on 16)

The processing device of claim 16, wherein the memory further comprises:
instructions for collecting results from the respective ones of the plurality of solvers, and
instructions for presenting a report including information regarding the collected results.

### Claim 18 (depends on 15)

The processing device of claim 15, wherein the memory further comprises:
instructions for providing a model from a spreadsheet application, and
wherein the instructions for binding modeling output to a data sink include instructions for binding the modeling output to one or more cells of a spreadsheet of the spreadsheet application.

### Claim 19 (depends on 15)

The processing device of claim 15, wherein the memory further comprises:
instructions for automatically configuring individual solvers based on one or more provided directives.

### Claim 20 (depends on 15)

The processing device of claim 15, wherein the instructions for binding modeling output to a data sink further comprise:
instructions for automatically and declaratively binding the modeling output to the data sink.
