Detection of Abnormal Ship Behavior Based on Grid Partitioning and KDE Method

Abstract

Ensuring the security and efficiency of maritime transportation is critical for the global commercialization of biotechnological products, particularly in supply chain logistics involving temperature-sensitive and high-value goods. Traditional ship behavior monitoring systems face challenges in detecting anomalies that may disrupt the timely delivery of biotechnology-related cargo. This study proposes a novel approach for detecting abnormal ship behavior using a combination of grid partitioning and the Kernel Density Estimation (KDE) method. The grid partitioning technique divides maritime routes into manageable segments, allowing for precise spatial analysis, while the KDE method models the probability distribution of normal ship behaviors, facilitating the identification of deviations indicative of abnormal activities. Through extensive simulations and real-world data validation, the proposed model demonstrates high accuracy in detecting abnormal ship movements, including unauthorized deviations, suspicious slowdowns, and unusual docking patterns. The findings underscore the model's ability to enhance maritime supply chain resilience by enabling early detection of potential disruptions. For the commercial biotechnology industry, this approach ensures the secure and efficient transportation of bioproducts, mitigating risks associated with shipment delays, contamination, and regulatory non-compliance. This research contributes a robust analytical framework for integrating advanced maritime monitoring systems into biotechnology logistics, supporting the reliable and secure commercialization of biotechnological innovations worldwide.