2.1. Description¶
MindOpt is an efficient large-scale mathematical optimization problem solver software, which covers a variety of efficient optimization algorithms for solving multiple types of optimization models, and can help quickly solve the best solution for business problems. In this chapter, we briefly introduce MindOpt’s solving capabilities and interaction methods.
2.1.1. Overview of Solving Capabilities¶
MindOpt currently supports solving the following types of optimization problems.
The ability to solve more optimization problems is under development, please pay attention to our update notice.
2.1.2. Platform Usage and Acquisition of the Solver¶
Users can obtain a free community edition version with no problem-size limits and trial period for non-commercial usage only. MindOpt community edition is the ideal choice for students, researchers, and developers seeking full access without restrictions for non-deployment use. Note the community edition operates with a cloud-based license, requiring an internet connection.
For commercial usage, we provide customized versions to accommodate various usage scenarios.
Single License: Designed for high-performance servers, this local license has no restrictions on CPU cores or computational resources, and it supports unlimited concurrent usage on a single machine
Floating License: Ideal for complex environments, this license is tied to a server’s MAC address and CPU ID, offering a pool of tokens that can be shared across multiple machines within a local network. Tokens are dynamically allocated to client machines as needed, supporting multi-application and multi-machine deployments.
You can consult and obtain them through via Contact Us. Our team will respond promptly with a detailed quote.
2.1.3. Interaction and Invocation of the Solver¶
Users can call the solver or write their optimization programs through the command line or APIs in the following languages:
We also provide some modeling and optimization examples to explain how to establish, modify optimization models and solve, or use the initial basis to warm start optimization algorithms, etc., to help users master the skills of using the API for modeling and solving.
MindOpt also supports direct reading of optimization problems in the following standard formats, as well as their corresponding GZIP and BZIP2 compressed format files:
.mps
format
.lp
format
.dat-s
format
.nl
format
2.1.4. Modeling Tools¶
Considering users’ modeling needs, we also provide several modeling language docking methods and examples. Among them, the Mindopt APL modeling language was independently developed by the MindOpt team:
2.1.5. Advanced Modeling Techniques¶
We provide the following advanced modeling techniques and examples to help users better solve optimization models:
Irreduciable Infeasible Subsystem (IIS): Helps users identify key constraints that lead to feasibility conflicts;
Callback: Helps users practice personalized heuristic solving strategies to optimize solving speed.