Analysis of Real-Time Simulation Model for Determining the Drop Point from Unmanned Aerial Vehicles

Abstract

With the increasing importance of initial response in emergency maritime rescue systems, there has been a rise in applications involving aerial delivery or dropping of objects from unmanned vehicles. Training plays a vital role in supporting mission execution in military and emergency situations. However, real-world training encounters limitations in terms of cost and safety, making virtual training a viable alternative. This paper analyzes and proposes approaches for realtime simulation of dropping object from unmanned vehicles for educational training purposes. The educational training simulation models can be classified into three categories: physics-based simulation mathematical models, data-driven search models, and probability-based simulation estimation models. Physics-based models ensure accuracy, but real-time processing is challenging. Data-driven models, on the other hand, cannot adapt to new input conditions. Therefore, a probability-based simulation estimation model, considering uncertainties, is deemed suitable for educational training simulations. The probability-based model provides estimation outputs based on probability distributions, accommodating diverse variables. To implement a specific probability-based estimation model, diverse environmental input conditions should be utilized, and simulation results must be compared and validated against mathematical models. The model