Abstract

Optimal Control of Two-Wheeled Mobile Robots for Patrolling Operations
Walaaeldin Ahmed Ghadiry,
Concordia University, 2015
This work studies the use of the two-wheeled mobile robots in patrolling operations, and provides the most distance-e�cient as well as time-e�cient trajectories to patrol a given area. Novel formulations in the context of constrained optimization are introduced which can be solved using existing software. The main concept of the
problem is directly related to the well-known Traveling Salesman Problem (TSP) and its variants, where a salesman starts from a base city and visits a number of
other cities with minimum travel distance while satisfying the constraint that each city has to be visited only once. Finally, the salesman returns back to the starting base city after completing the mission. Two di�erent patrolling con�gurations that are related to the TSP and its variants, namely the Single Depot multiple Traveling Salesman Problem (mTSP) and the Multidepot multiple Traveling Salesman Problem (MmTSP) are investigated. Novel algorithms are introduced for the trajectory planning of multiple two-wheeled mobile robots, either with two di�erential motors (which can turn on the spot) or with Dubins-like vehicles. The output trajectories for both types of wheeled robots are investigated by using a model predictive control scheme to ensure their kinematic feasibility for the best monitoring performance. The proposed formulations and algorithms are veri�ed by a series of simulations using e�cient programming and optimization software as well as experimental tests in the lab environment.