An approach to reduce computational effort in multiobjective path planning and task allocation for multi robot systems

Abstract

In today s modern automated and semi-automated sectors, multi-objective path planning algorithms for mobile robot applications are in high demand. Multi-objective optimization algorithms require a more significant amount of computational effort to produce an optimal path. The proposed grid-based multi-objective global path planning algorithm [Quadrant selection algorithm (QSA)] plans the path from the initial location to the target location by considering the direction of movements from the starting position to the target position while requiring the least amount of computation. The direction of movements is divided into quadrants as the primary classification in this method. The created feasible paths are evaluated based on the cumulative path distance travelled in obstacle avoidance. The cumulative angle turned to achieve an ideal path to avoid obstacles. Finally, to make navigation easier for the robot, the ideal path is further smoothed to prevent sharp curves and reduce the distance travelled. The suggested QSA, taken as a whole, decreases the need to seek paths in other quadrants that are unnecessary. The newly designed algorithm is evaluated in various contexts and compared to current algorithms based on the number of cells examined to find the shortest path to the goal. In contrast to existing algorithms, the suggested QSA provides an optimal path by drastically lowering the number of cells examined throughout the computation. The experimental verification of the suggested QSA demonstrates that the approach is feasible in terms of cost and time. newlineIn deploying multiple robots, allocating tasks with the lowest path cost is one of the most frequently encountered difficulties. Our research tackles the multi-robot task allocation problem with the lowest path costs, the minimum computing time, and the most equitable task distribution. newline