/MPC4STL-TID

Model Predictive Control for Signal Temporal Logic with Time Interval Decomposition

Primary LanguagePythonMIT LicenseMIT

MPC4STL-TID

We propose an effective STL synthesis framework - Model Predictive Control for Signal Temporal Logic with Time Interval Decomposition. Here is the paper link. This project is collaborated with Chuwei Wang (main contributor).

Instructions

This project consists of two parts, online part and offline part, which are conducted by Python 3 and Julia respectively.

  • We compared computational efficiency by two ways, with Time Interval Decomposition (TID) and without TID. Both codes can be found in folders online codes - with TID and online codes - without TID respectively.
  • For the offline part, you can check the final terminal set by figure offline codes/terminalSetFigure.png and see some of "intermediate product" by animation offline codes/animation.gif. Note that the offline results shown in the figure and animation is the projection to 2 dimensions and you can also plot them in other dimensions by your own.

Dependencies

In the online part, we use the following packages

matplotlib    v3.6.1
numpy     v1.23.4
Gurobi   v9.5.2

In the offline part, we use the following packages

GR v0.69.5
IntervalArithmetic v0.20.7
LaTeXStrings v1.3.0
LazySets v2.3.0
Plots v1.36.0
Polyhedra v0.7.5

Acknowledgement

We would like to thank our colleague Xiaoduo for discussion and his great help in this work.