PowerBiMIP V0.1.0

PowerBiMIP V0.1.0

We are pleased to announce the release of PowerBiMIP V0.1.0. This update brings significant functional expansions and usability improvements, marking a step forward in our support for complex decision-making problems in power and energy systems.

New Features

• Two-Stage Robust Optimization (TRO) Support:

  • Introduced a new solver interface solve_TRO specifically designed for Two-Stage Robust Optimization problems.
  • Currently supports problems with Linear Programming (LP) recourse using the Column-and-Constraint Generation (CCG) algorithm.
  • The subproblems in CCG are treated as special BiMIPs, leveraging the core efficiency of PowerBiMIP.

• Optimized BiMIP Solver Interface:

  • Refactored the solve_BiMIP calling signature for better ease of use.
  • Old: Required passing multiple individual variable vectors (var_xu, var_zu, etc.).
  • New: Now accepts a unified model structure containing all variables, constraints, and objectives, along with an ops structure.

• Expanded Benchmark Library:

  • Added new benchmark cases in the examples/ directory.
  • Updated existing toy examples to align with the new API.

Improvements

• Enhanced User Experience: Optimized the installation process and solver output logs for better clarity and debugging.
• Documentation: Comprehensive updates to the documentation to reflect the new interfaces and mathematical formulations.

Installation & Upgrade

You can update to the latest version via GitHub Desktop or git command.
Or download the source code from the assets below.

PowerBiMIP v0.0.1-alpha

This commit marks the first official public release of the PowerBiMIP!

PowerBiMIP is an open-source MATLAB toolbox designed to efficiently solve bilevel mixed-integer programming (BiMIP) problems for power and energy systems. This initial version establishes the core framework and provides fundamental solving capabilities for optimistic BiMIP problems.

Key Features:

• A unified solver interface solve_BiMIP for easy user interaction.
• Support for modeling BiMIP problems using the YALMIP syntax.
• A preprocessing module to automatically handle and reformulate coupled constraints in the upper-level problem.
• A core solver based on the R&D algorithm (optimistic_solver), supporting methods such as KKT-based, strong duality-based, and a quick mode.
• Flexible algorithm and solver options via the BiMIPsettings function.
• Basic convergence plotting and command-line logging.

This release lays the foundation for future development, including support for pessimistic models and the integration of more advanced algorithms.

Project Links:

• GitHub Repository: https://github.com/GreatTM/PowerBiMIP
• Documentation: https://docs.powerbimip.com