As mini UAVs become increasingly useful in the civilian work domain, the need for a method for them to operate safely in a cluttered environment is growing, especially for fixed-wing UAVs as they are incapable of following the stop-decide-execute methodology. This paper presents preliminary research to design a reactive collision avoidance algorithm based on the improved definition of the repulsive forces used in the Artificial potential field algorithms to allow feasible and safe navigation of fixed-wing UAVs in cluttered, dynamic environments. We present simulation results of the improved definition in multiple scenarios, and we have also discussed possible future studies to improve upon these results.