The library builds on the PuLP linear programming library to provide trajectory modeling and non-convex constraints.
The latter are expressed as "union" constraints in the form C:={C1,C2,C3...} where at least one of the sets of constraints Ci must all be satisfied. This handles a variety of avoidance constraints in a trajectory setting.
Problems are solved by converting to Mixed-Integer Linear Programming (MILP) form and solving using PuLP's existing solver capabilities. An experimental direct branch-and-bound solver is under development, using only LP solvers.
Notebook pytro_example.ipynb walks through a simple shortest-path example with obstacle avoidance.
The aircraft_example.py file implements the method from Richards A. G. and How J. P. "Aircraft Trajectory Planning With Collision Avoidance Using Mixed Integer Linear Programming" American Control Conference 2002
with the additional corner-cutting prevention from Richards A. G. and Turnbull O. D. N. "Inter-sample avoidance in trajectory optimizers using mixed-integer linear programming" IJRNC 2013.
See also the reproduction at (https://github.com/auralius/richards-how) using CVXPY
carplan.py is a version of the ideas covered in Bali C. and Richards A. G. "Merging Vehicles at Junctions using Mixed-Integer Model Predictive Control" ECC 2018 on using MILP to separate vehicles moving along fixed tracks with junctions.