Improved artificial potential field algorithm based multi-local minimum solution

Artificial Potential Field (APF) is a path planning algorithm that is considered reliable to produce obstacle-free paths. One of the main drawbacks of the APF algorithm is that it is trapped at a local minimum. Some forms of local minimums are symmetrically aligned robot-obstacle-goal (SAROG) and goal non-reachable due to obstacle nearby (GNRON). Previous research resolved local minimum problems separately. In this study, the local minimum problem in the form of a single local minimum problem and a multi-local minimum problem is solved by the Improved Artificial Potential Field (I-APF) algorithm. The auxiliary function (v) was created to solve the SAROG problem. This function functions to change the resultant value of the force formed by force on each axis. In addition, GNRON is accomplished using adaptive repulsive gain and adaptive distance between the goal and current position. The I-APF algorithm is successfully used to solve local minimum problems in the form of a single local minimum and multi-local minimum. Based on the results obtained at the initial distance = 10, the average of the total trajectory (Dtrav) generated by the I-APF algorithm is 11.93. The average Ergvalue with a tolerance of 0.01 is 0.0077. © 2021, International Association of Engineers. All rights reserved.