# Search model based on Kalman Filter and Monte Carlo simulation

## TL;DR

This paper proposes a search model for submersibles using Kalman Filter and Monte Carlo simulation to improve search and rescue operations.

## Contribution

The novelty lies in combining Kalman Filter and Monte Carlo simulation for submersible search prediction and introducing a nearest neighbor correlation algorithm for multi-target tracking.

## Key findings

- The proposed model effectively predicts submersible locations using HYCOM marine environment data.
- Monte Carlo simulation helps quantify the probability of submersible locations after a fault.
- The model outperforms previous methods in predicting and tracking multiple submersibles.

## Abstract

In the search and rescue operation of the submersible, to better search for the missing or faulty submersible, taking the marine environment simulated by the HYCOM model as the sample, it is necessary to use the Kalman filter model to predict the time location of the submersible and provide information support for the follow-up search and rescue operations according to the position information transmitted to the main ship when the submersible is running normally. Monte Carlo simulation is used to quantitatively analyze the probability of the possible area of the submersible in four possible cases after the fault, to obtain the location of the initial search deployment point, that is, the minimum plane projection area of the covering sample. Python software was used to quantitatively analyze the probability of finding the submersible with the passage of time and cumulative search results. Moreover, we conducted a comparative analysis of the method proposed in this paper with previous methods to illustrate the advancement of the method proposed in this paper. By introducing the nearest neighbor correlation algorithm into the multi-target tracking algorithm, the motion position of multiple submersibles in the same area can be predicted.