Quiet Zone Enhancement for a Target Location Using an Improved Virtual Sensing Algorithm

Abstract

Generating a quiet zone at a target location is the ultimate goal of active noise control (ANC). Generally, the best noise cancellation is achieved at the locations of error sensors. However, the distribution of physical error sensors at a given location is not always convenient or feasible. To overcome this challenge, a number of virtual sensing algorithms for ANC have been researched. Using the physical error signals and knowledge of the system, the adaptive least mean square (LMS) virtual sensing algorithm estimates the error signal at a location that is remote from the physical error sensor, referred to as the virtual location. It achieves excellent performance under the assumption that the unknown primary path is fixed, but its performance decreases significantly if the path is constantly changing. This paper presents a real-time ANC system with an improved virtual sensing algorithm to solve this problem. The proposed system was tested to enhance the quiet zone in a constantly changing environment; an enhanced quiet zone was created, and the noise cancellation, especially at the target location, obviously improved.Electronics 2017, 6, 76 2 of 13 sensing algorithm cannot maintain excellent performance. This paper presents a modified adaptive LMS algorithm to solve this problem. The proposed algorithm achieves the microphone weights on the basis of the secondary paths, the secondary paths are fixed during the system operation. Thus, the proposed algorithm effectively eliminates the influence of a constantly changing primary path. A multichannel ANC with an improved virtual sensing algorithm was then designed to enhance the noise cancellation at the target location.The organization of this paper is as follows: Section 2 presents the design of real-time multichannel ANC system. In Section 3, the adaptive LMS virtual sensing algorithm is introduced and the improved adaptive LMS virtual sensing algorithm is proposed. Experimental results are presented in Section 4, and the conclusions are summarized in Section 5. Multichannel Real-Time ANC SystemThe filtered-x least mean square (FXLMS) algorithm has been widely used in practical ANC due to its robustness and simplicity. In our proposed system, the multichannel delay-compensated FXLMS algorithm is applied [22], it achieves a faster convergence rate and better performance with only double computation load, and this guarantees the possibility of real-time implementation. The multichannel delay-compensated FXLMS is illustrated in Figure 1. It employs I reference sensors to form the reference signal vector, generates J canceling signals to drive the corresponding secondary sources, and distributes K error sensors over desired locations to measure the residual noise components.Electronics 2017, 6, 76 2 of 12 achieves the microphone weights on the basis of the secondary paths, the secondary paths are fixed during the system operation. Thus, the proposed algorithm effectively eliminates the influence of a constantly changing primary path. A multichannel ANC with an improv...

Handheld Device-Based Indoor Localization with Zero Infrastructure (HDIZI)

“…This approach estimates the position and orientation of a robot as it moves and senses the surrounding environment. Active noise control [61] is a method to reduce noise, generating a quiet zone at a target location. This requires simulation and specific radio signals.…”