| In the noise control of ship cabin,the active control technology has attracted lots of attention due to excellent low-frequency noise control performance.Early research has focused on active noise control in the global space of cabins.The active noise control in global space is very difficult to achieve because of the limitations of complex cabin noise sources and narrow room.Thus,the local space active noise control technology has become a new research direction.The noise reduction effect of the local space active noise control system is not only affected by the sound field environment in the cabin,but also more dependent on the local space active noise control algorithm.However,these algorithms are limited by factors such as timevarying secondary paths,arrangement of error microphones,noise increase abnormally and listening comfort when used in local space active noise control systems,which affect the practical application of local space active noise control technology.Although these problems can be solved through local space active noise control algorithms such as secondary path modeling,virtual sensing,impulsive noise and sound quality,these control algorithms still have limitations such as slow convergence speed,high steady-state error and insufficient stability.Therefore,this paper first conducts a theoretical analysis of local space active noise control systems under different sound field environments and carry out research on local space active noise control algorithms to address the shortcomings of the aforementioned algorithms.The details are as follows:To evaluate and analyze the noise reduction effects of local space active noise control systems under different acoustic environments,unified models of local space active noise control systems are established to explore the quiet zone distributions of local space active noise control systems under different acoustic environments.Under two ideal sound field conditions,free sound field and diffuse sound field,the quiet zones generated by local space active noise control systems are compared and analyzed for different noise frequencies,control system configurations,and positions between primary or secondary sound sources and error sensors.On the basis,aiming at the acoustic environment where the primary or secondary sound field is a non-diffuse sound field,the influence of reverberation components on the generation of the quiet zone in a local space active noise control system is simulated and analyzed.The simulation results show that the acoustic environment affects the distribution of the quiet zone in the local space active noise control system.Furthermore,the reverberant components in the sound field can improve the noise reduction effect of the local space active noise control system to a certain extent and make the distribution of the quiet zone more uniform.For time-varying secondary path in a local space active noise control system,an improved secondary path online modeling method is proposed based on a noise perturbation strategy and the influence mechanism of perturbed noise on algorithm stability was revealed through algorithmic analysis.Further,a sign algorithm is used to update the weight coefficients of the overall modeling filter to extend the algorithm.Some simulations are carried out to verify the effectiveness of the proposed algorithm.In order to reduce the computational complexity,an improved multi-channel simultaneous modeling and control algorithm is proposed by combining with noise perturbation strategy.Then,theory and simulation analysis are performed.The simulation results show that the proposed algorithm has better control performance compared with simultaneous modeling and control algorithms.On the basis,a simultaneous perturbation stochastic approximation algorithm with a lower computational complexity is proposed based on the estimated cross-correlation function.The advantages of the algorithm in gradient estimation are analyzed and simulation analysis is carried out under different types of input signals.The results show that the proposed algorithm has better convergence performance and can achieve effectively broadband noise control.The noise reduction in the target position can be reduced because of change in sphere of personnel activity in a local space active noise control system.To solve this issue,the robustness analysis is first conducted based on two single-channel virtual sensing algorithms: a remote microphone method and an additional filter method.From the perspective of multi-channel coupling and decoupling algorithms,the computational complexity is analyzed.The simulation and experimental research under different sound field environments are also conducted.Simulation and experimental studies are also performed in different sound field environments.The convergence performance and noise reduction performance of two virtual sensing algorithms are verified through comparative analysis.Moreover,the influences of sound field environment and change in sphere of personnel activity on the performance of virtual sensing algorithms are summarized,providing theoretical reference and experimental support for the practical application of active noise control systems in local cabin rooms in the future.In order to improve the convergence performance of local space active impulsive noise control algorithms,an improved filtered-x affine projection sign algorithm is proposed by introducing a post-adaptive filter based on the filtered-x affine projection sign algorithm.Algorithm analysis was carried out to reveal the influence of post filter parameters on the convergence speed and steady-state error of the algorithm.On this basis,the filtered-x affine projection sign algorithm with a variable step size is proposed by utilizing a convex combination approach and a variable step size strategy.The effectiveness of the proposed algorithm is verified through algorithmic simulations.The results show that the filtered-x affine projection sign algorithm with a variable step size has a faster convergence speed than the filtered-x affine projection sign algorithm based on a post-adaptive filter.In response to the issue of reduced listening comfort in a local space active noise control system,the hybrid feedforward and feedback control structure is analyzed,which reveals the influence mechanism of the hybrid control structure on the stability of the hybrid active sound quality control system.On this basis,two improved hybrid active sound quality control algorithms are proposed by considering different synthesis methods of reference signals in the feedback control structure,which improves the stability of the hybrid active sound quality control system.Moreover,a novel leakage filtered-x least mean square algorithm is proposed based on the anti-Hebbian constraint idea,which solves the problem that the stability and noise reduction of the feedback control structure cannot be considered simultaneously in a hybrid active sound quality control algorithm. |
PDF Full Text Request