Evaluation of Neural Networks Performance in Active Cancellation of Acoustic Noise
نویسندگان
چکیده
Active Noise Control (ANC) works on the principle of destructive interference between the primary disturbance field heard as undesired noise and the secondary field which is generated from control actuators. In the simplest system, the disturbance field can be a simple sine wave, and the secondary field is the same sine wave but 180 degrees out of phase. This research presents an investigation on the use of different types of neural networks in active noise control. Performance of the multilayer perceptron (MLP), Elman and generalized regression neural networks (GRNN) in active cancellation of acoustic noise signals is investigated and compared in this paper. Acoustic noise signals are selected from a Signal Processing Information Base (SPIB ) database. In order to compare the networks appropriately, similar structures and similar training and test samples are deduced for neural networks. The simulation results show that MLP, GRNN, and Elman neural networks present proper performance in active cancellation of acoustic noise. It is concluded that Elman and MLP neural networks have better performance than GRNN in noise attenuation. It is demonstrated that designed ANC system achieve good noise reduction in low frequencies.
منابع مشابه
Active Noise Cancellation using Online Wavelet Based Control System: Numerical and Experimental Study
Reaction wheels (RWs) used for attitude control of space vehicle systems usually encounter with undesired wide band noises. These noises which significantly affect the performance of regulator controller must tune the review or review rate of RWs. According to wide frequency band of noises in RWs the common approaches of noise cancellation cannot conveniently reduce the effects of the noise. Th...
متن کاملRecursive filter estimation for feedforward noise cancellation with acoustic coupling
The basic principle of feedforward noise cancellation for broadband active noise reduction is based on the availability of a reliable measurement of the noise source. Reliability is compromised in case the noise measurement is acoustically coupled to the actual active noise cancellation (ANC), as stability and performance of the feedforward compensation becomes ambiguous. This paper presents a ...
متن کاملPersian Phone Recognition Using Acoustic Landmarks and Neural Network-based variability compensation methods
Speech recognition is a subfield of artificial intelligence that develops technologies to convert speech utterance into transcription. So far, various methods such as hidden Markov models and artificial neural networks have been used to develop speech recognition systems. In most of these systems, the speech signal frames are processed uniformly, while the information is not evenly distributed ...
متن کاملPerformance evaluation of artificial neural networks in statistical downscaling of monthly precipitation (Case study: Minab watershed)
متن کامل
Active Noise Control in Pardis Coach using Different Fuzzy Controllers
In recent years, need to increase the convenience of trips in railway vehicles causes that train operators and manufacturers focus on reducing the noise level which is sensed by passengers. In this paper, first the state of modeling acoustic noise in cab train is discussed and natural frequencies and acoustic mode shapes are derived and then formulation of acoustic pressure in the cab will be o...
متن کامل