A Robust Leak Localization Method Based on Energy-Weight TDOA Using Accoustic Sensor Networks

This paper introduces an improved Time Difference of Arrival (TDOA) method for pipeline leak localization using acoustic signals. Unlike the classical TDOA, which assumes uniform sensor reliability, the proposed method incorporates energy-based weights derived from acoustic signal strength. The resulting weighted cost function enhances robustness against noisy or poorly positioned sensors. Monte Carlo simulations in both 2D and 3D settings show that the weighted TDOA reduces the root-mean-square error (RMSE) by more than 50% compared to the classical approach. This demonstrates its effectiveness in accurately locating leaks under realistic noisy conditions.