Using Bayesian Formulation of the Equivalent Source Method to Reconstruct the Acoustic Field of a Cylindrical Shell

Accurate reconstruction of the acoustic field is critical for applications such as active sound field control. The equivalent source method (ESM) is a popular reconstruction method. However, traditional ESM is highly sensitive to selection of both equivalent source locations and measuring points. This study proposes a Bayesian-enhanced ESM framework integrating historical data with sparse real-time measurements. The framework first calibrates prior source parameters through pre-acquired numerical/experimental datasets to constrain the placement, followed by Bayesian inference enabling dynamic equivalent source strength updates via sparse measurements. The simulation results demonstrate the effectiveness of the proposed method despite some mismatch of the priors, and experimental data processing further confirms its performance improvement in reconstructing the scattered sound field of a steel cylindrical shell.