Noise Time-Domain Signal Reconstruction of Passenger Head Position Considering Compressed Sensing and Multi-source Data Fusion

Sound field reconstruction technology is used to provide accurate primary reference signals for active noise control systems by reconstructing the interior sound field. Traditionally, time-domain noise signal-based reconstruction modeling has certain deficiencies, such as large data volume, noise reconstruction model complexity and considerable time consumption. Hence, a novel Signal compression optimisation-based BP network for passenger head position signal reconstruction (CBHSR) algorithm is proposed. Based on compressed sensing, the proposed algorithm converts raw multi-source signals into the compressed domain to implement compressed sampling. The signal reconstruction model is created by regarding the optimal fitness value as the initial weight and the threshold of the signal reconstruction BP network, and training with the compressed multi-source data. The recovery compression signal method realizes the time-domain signal reconstruction of the passenger head position. The effectiveness of the proposed CBHSR algorithm is validated using noise signal sources collected from a vehicle. Compared with the reconstruction model of the BP algorithm, the proposed algorithm is superior in reconstruction accuracy and time consumption.