A particle filter based intruder state estimation method for UAV collision avoidance

A critical safety concern with the rapid increase in UAVs is developing algorithms that can solve the pressing airborne collision avoidance problem. A fundamental necessity in solving the airborne collision avoidance problem is the need to estimate the state of the intruder using the onboard sensor measurement which is usually noisy. Advanced filtering algorithm, such as Particle Filters can provide very accurate estimates of the target state. However, unlike the way that sensors located off-board, the measurements made by the onboard sensor are relative measurements between the UAV aircraft and the intruder. The accuracy of the UAV's state will also affect the location of the Intruder. In this paper, we will take the accuracy of the UAV's state into consideration and use the SIR particle filter to estimate the state of the Intruder. The feasibility and performance of the proposed approach is not only mathematically analyzed, but also verified through simulation. Through the simulation results, we can see that our approach is more accurate in estimating the state of the intruder.