Collision Avoidance for Cooperative UAVs With Optimized Artificial Potential Field Algorithm

Cooperative Control of Multi‐Agent Systems

“…In this case, according to the properties of the APF algorithm there must be some points in the plain in which the repulsive force of the obstacle and attractive force of the target neutralized each other. Consequently, the total force at that point will be zero and if the agent enters that point will atop and will trapped there forever [4]. There is another case that observed when performing MATLAB simulation in the previous section.…”

“…The Artificial Potential Field (APF) is one of the leading algorithms and techniques in the field of Multi Agent Systems (MAS) control. The classic APF algorithm first introduced by O. Khatib in 1986 to make autonomous robots with obstacle avoidance capabilities [4]. This algorithm controls the agents in the environment using the gradients of the potential fields associated with the target and obstacles.…”

“…This algorithm is comprehensively used in the implementation of autonomous robots/UAVs path planning, localization, navigation and mapping. The classic APF algorithm first introduced by O. Khatib in 1986 to make autonomous robots with obstacle avoidance capabilities [4]. Later on, in 1992, Koditschek and Robin introduced navigation functions as a new and powerful approach to address problems within Khatib methodology [5].…”

“…Therefore agent(s) will trap in the neighborhood of the obstacle(s). Agent(s) will oscillate around the obstacle and will be not able to reach their goal [4].…”

“…In this case agents try to find a path that looks like a channel to pass the obstacles. When agents passing the channel, it might be locations that the superposition of the attractive and repulsive forces is in a way that will stop agent [4]. Agents might oscillate inside the channel or just stop forever in the point of local minima.…”