This paper presents a follow-up and improvement of our previous work on conflict detection and resolution (CDR) for unmanned aircraft “sense-and-avoid” (SA) applications. More specifically, we propose an extension of our previous model predictive control formulation and algorithm that takes into account costs incurred by possible deviation from the desired destination so that the optimized solution produced by the algorithm naturally drives the aircraft towards its destination. Furthermore, we work out a solution to the SA problem in aircraft-weather conflicts within the multiple model framework of our approach. The performance of the proposed method is evaluated via simulation of a large variety of SA encounter scenarios, and compared with an existing method and with the optimal solution.