Enhancing Ultrasonic Time-of-Flight Estimation Using Adaptive Differential Evolution and Levenberg–Marquardt Algorithm

A parameter estimation algorithm based on an asymmetric Gaussian model (ACGM) is proposed to address the problems of low accuracy and susceptibility to noise in the traditional ultrasonic time-of-flight (TOF) estimation method, which improves the accuracy of TOF estimation and is less affected by noise. The algorithm is a combination of adaptive differential evolution (DE) and Levenberg–Marquardt (LM), using the DE algorithm as a global search strategy and the LM algorithm as a local search strategy. First, the received echo signal is processed using the Hilbert transform to extract the envelope, reducing dimensionality, and computational burden. Then, the adaptive DE algorithm converges quickly to a better solution, and then, the result of the adaptive DE algorithm is used as the initial value of the LM algorithm to further improve the accuracy, which solves the problem that the LM algorithm needs a good initial value to provide an accurate output. Finally, the model parameters are used to derive an accurate echo leading edge arrival time. The proposed algorithm is compared with several well-known DE variants as well as the latest algorithm in numerical simulations and ranging experiments. The results demonstrate the excellent performance of the adaptive differential evolution (ADE)–LM algorithm in the estimation of ultrasonic TOF.