Abstract:Acoustic Emission (AE) is commonly utilized in equipment condition monitoring. However, traditional Kalman filtering for noise reduction of AE signals suffers from uncertain system mathematical models and unknown statistical characteristics of measurement noise, leading to system errors that cause reduced filtering accuracy or even divergence, and it cannot achieve adaptive filtering of system noise, resulting in poor filtering effects. To address this issue, this paper proposes a Reinforcement Learning Adaptive Kalman Filter (AR-QLKF) algorithm for AE signal’s autoregressive (AR) modeling. Firstly, the AE signal is used to establish the AR model for Kalman filtering, and the Q-learning algorithm is utilized to learn the statistical characteristics of the system noise and measurement noise of the state space model. The difference of the filtering innovation sequence is used as an indicator to evaluate the quality of the estimated value of the system noise and the measurement noise. The noise variance is adaptively adjusted to enhance the adaptive Kalman filtering noise reduction effect. Computer simulation results demonstrate that the filtering effect of the AR-QLKF algorithm is significantly better than that of the Kalman filter based on the AR model, validating the effectiveness of the algorithm. The AR-QLKF algorithm is applied to online filter the AE signal of the grinding wheel grinder’s condition monitoring and to distinguish between different working conditions, confirming the algorithm’s effectiveness and practicality in engineering applications.

Acoustic Echo Cancellation Algorithm Based on Kalman Filtering of Skewed Observation Noise

Low-Complexity Acoustic Echo Cancellation with Neural Kalman Filtering

Joint Acoustic Echo Cancellation and Speech Dereverberation Using Kalman filters

NeuralKalman: A Learnable Kalman Filter for Acoustic Echo Cancellation

Robust Uncertainty Control of the Simplified Kalman Filter for Acoustic Echo Cancelation

Dynamic Neural Network-based Solutions for Acoustic Echo Suppression

Reinforcement Learning Adaptive Kalman Filter for AE Signal’s AR-mode Denoise

A New Multi-Channel Echo Cancellation Algorithm

A Joint Optimized Robust Acoustic Echo Cancellation Algorithm

A Channel Shortening Approach for Nonlinear Acoustic Echo Cancellation

A Novel Subband Acoustic Echo Cancellation System Based on Fast Affine Projection Algorithm

A New Real-Time Acoustic Echo Cancellation Algorithm Using Blind Source Separation and Multi-delay Filter

A Residual Echo Suppression Technique for Systems with Nonlinear Acoustic Echo Paths

An efficient acoustic echo cancellation design for systems with long room impulses and nonlinear loudspeakers

A Speech Enhancement Method Based on Kalman Filtering

Comparative Study of Acoustic Echo Cancellation Algorithms for Speech Recognition System in Noisy Environment

Research on Adaptive Acoustic Echo Cancellation Algorithm in Digital Hearing Aids

A Robust and Cascaded Acoustic Echo Cancellation Based on Deep Learning

Adaptive Beamforming for Acoustic Echo Cancellation Using Measured Array Models and Subband Filtering

On Semi-blind Source Separation-based Approaches to Nonlinear Echo Cancellation Based on Bilinear Alternating Optimization

A Nonparametric Variable Step-Size Subband Adaptive Filtering Algorithm for Acoustic Echo Cancellation