Enhancement of tabu search algorithm for optimized rescue routing operation

ABSTRACT: The Tabu Search algorithm while effective for optimization problems, faces challenges with getting stuck in suboptimal solutions, slow processing for large-scale issues, and inconsistent performance across different problem sizes. The TS algorithm specifically struggles in three critical care areas: it often gets trapped in suboptimal solutions and cannot find the best possible solutions, it suffers from high computational complexity leading to excessive runtime, and it encounters scalability issues that produce unexpected results when applied to problems of varying sizes. To address these limitations, researchers established three primary objectives: to enhance the balance of exploration and exploitation by incorporating a modified perturbation technique to diversity the search space, to optimize the high-complexity operations by swapping selected Points of Interest (POIs) to speed up the generation of neighborhood solutions, and to improve scalability by utilizing a modified dynamic tabu tenure mechanism that adjusts parameters based on problem size. The methodology implemented three innovative approaches: first, a wave-resonance based perturbation technique which mimics wave interference patterns observed in transmission frequency analysis to better explore the solution space; second, a local point sampling method for neighborhood generation that strategically examines only the most promising route segments rather than exhaustively checking every possibility; and third, a Wave-Inspired Dynamic Tenure mechanism (WIDT) that dynamically adjusts memory length using principles derived from wave properties. Testing demonstrated significant improvements over previous approaches: solution diversity increased substantially (scoring 0.21 versus 0.03), processing time for 160-location problems became eight times faster (3.33 seconds versus 28.88 seconds), and performance remained consistent across all tested scenarios regardless of size.