DaDRA

Contents:

Introduction

DaDRA (day-druh) is a Python library for Data-Driven Reachability Analysis. The main goal of the package is to accelerate the process of computing estimates of forward reachable sets for a given dynamical system.

Motivation

Currently, there are no prominent Python libraries in use for Data-Driven Reachability Analysis, as most exist strictly in MATLAB. With the ever increasing popularity of the Python language, DaDRA takes advantage of the many open-source Python libraries already well established, such as cvxpy and scipy, and applies them to the domain of control theory.

Limitations

  • As this library is still young, there is currently limited support for certain types of dynamical systems. Further improvements can be made to dadra.DynamicSystem to expand the use and applicability of DaDRA.

  • As of now, there only exists in DaDRA scenario reachability with p-norm balls. Other methods, including non-scenario approach methods such as Christoffel functions, will be implemented in the near future.

  • To take full advantage of Python’s multiprocessing.pool objects, the dadra.DynamicSystem currently makes use of the global keyword upon the instantiation of an inner function within the dadra.DynamicSystem.get_system() method. This allows pooling to be used for parallelization, but prevents multiple instances of dadra.DynamicSystem to draw samples in a non-sequential order, though this is only a problem if each instance is instantiated using the class method dadra.DynamicSystem.get_system(), rather than the default constructor for the dadra.DynamicSystem class.