There are many software tools available to solve optimization problems ranging from free and open-source to proprietary commercial packages. In general, students want to chose a software platform that will be both state-of-the-art and accessible long-term. Solving optimization problems requires some familiarity with a computer programming language. Some of the popular programming languages (see TIOBE index) also include capabilities for scientific computing.
Information on APMonitor, MATLAB, OptdesX, and Python is listed below. Assignments and projects can be completed with any of these software platforms.
The APMonitor Modeling Language is optimization software for differential and algebraic equations. It is coupled with large-scale nonlinear programming solvers for data reconciliation, real-time optimization, dynamic simulation, and nonlinear predictive control. It is available as a web service through an internet browser or as an add-in package to MATLAB or Python programming languages.
- APMonitor Documentation
- APM Model File Editor
- APMonitor MATLAB Toolbox
- APMonitor Python Toolbox
- Wikipedia: APMonitor
MATLAB allows mathematical operations, plotting, creation of user interfaces, and interfacing with programs written in other languages, including C, C++, Fortran, and Java. A number of optimization tools are available in the Optimization Toolbox. An additional package, Simulink, adds graphical simulation and design for dynamic systems. MATLAB users come from engineering, science, and economics in both academic research and industrial applications. MATLAB is available through the following methods:
- Download MATLAB from software.byu.edu for installation on a personal computer
- BYU Citrix logon with a BYU CAEDM account
- BYU RGS servers with a BYU CAEDM account
Please note that Citrix and RGS perform computing on shared servers with performance that can be less desirable at times of high loading. The most reliable route is to download MATLAB on a personal computer.
The OptdesX software was developed at BYU. It contains some of the best algorithms still in use for constrained nonlinear optimization, including the algorithms in Excel and Matlab. We use OptdesX as a learning tool and we can get inside the code and see what the algorithms are doing, which we cannot do with other software. We can also look at gradients, which are very useful in understanding scaling and in developing robust designs. OptdesX is available through CAEDM Linux workstations or through remote logon to shared servers.
Python is an open-source, versatile, interpreted high-level programming language that is supported on Windows, Linux, MacOS, etc. Python is extended by a number of add-on packages for optimization, numeric analysis, and graphical analysis. The following packages are recommended for Windows users and include Python 2.7, Numpy 1.6.2, Matplotlib 1.2.0, Scipy 0.11.0, and Pyserial 2.5.
- Python 2.7 for Windows
- Numpy 1.6.2 for Windows
- Matplotlib 1.2.0 for Windows
- Scipy 0.11.0 for Windows
- Pyserial 2.5 for Windows
- Pywin32 218 for Windows
Any additional packages must be compatible with Python 2.7 (32-bit version). The 32-bit version is recommended because official releases of Numpy and Scipy packages are only available in that format. There are many optimization packages that have been developed for Python with a few of them listed below:
This course uses the APM Python package for tutorials and class instruction. The APMonitor Modeling Language is integrated with Python through a set of functions that allow optimization problems to be solved as a web-service.
- Comparison of Python Integrated Development Environments (IDEs)
- Python IDE with Eclipse
- Python IDE with Wingware
- MIT Open Course Ware Tutorial on Python and IDLE
- Wikipedia: Python
comments powered by Disqus