Computational Intelligence approach to the problem of adaptive noise canceling

A Computational Intelligence (CI) based intelligent noise filtering approach is proposed for the problem of adaptive noise canceling (ANC). A self-constructing neuro-fuzzy system (SCNFS) is used as an adaptive filter to deal with the complex nonlinearity of noise. In the SCNFS, the system learning is composed of parameter learning and structure learning. In the parameter learning, a hybrid machine learning algorithm with the methods of both random optimization algorithm (RO) and least square estimation (LSE) is introduced to enable the SCNFS with learning capability. In the RO-LSE learning, the premises and the consequents of the SCNFS are updated by RO and LSE, respectively. In the structure learning, the system structure can be generated or rearranged using the proposed self-constructing mechanism of rule-splitting and/or rule-expanding. To demonstrate the feasibility and capability of the proposed approach, an example of adaptive speech noise cancellation with 2 cases is illustrated. With the experimental results, the SCNFS adaptive filter shows excellent filtering performance for signal recovering in the sense of statistics.