Safe Optimization

Project Overview

The Safe Optimization project focuses on developing algorithms that enable efficient optimization under safety constraints. Unlike traditional optimization methods, Safe Optimization ensures that the exploration process avoids unsafe regions, making it highly suitable for applications where safety is critical, such as robotics, autonomous systems, and medical treatments.

Demonstration

SafeOpt

SafeOpt is an algorithm for optimizing black-box functions that converges to the optimal reachable target point under safety constraints.

1-d Example

StageOpt

StageOpt is an algorithm for optimizing black-box functions based on SafeOpt. It divides the optimization process into two stages: safe region expansion and utility function maximization, thereby satisfying safety constraints while approximating the optimal point.

1-d Example

HdSafeBO

HDSAFEBO is the first high-dimensional safe Bayesian optimization algorithm that combines optimistic safety identification with isometric embedding to efficiently optimize complex embodied systems while guaranteeing probabilistic safety.

For details, please refer to our project page: HdSafeBO.