The identification of Error Path is very important in Active Noise Control. I used several technologies in achieving the model of the error path. The result of the identification was very precise.In paper, the theory of the identification was introduced. These theory included Correlation method, Least Squares method and Neural Networks, the advantages and disadvantages of these methods was analyzed. The identification of the error path was carried out by using these methods. The FIR of the Error Path was achieved by using Correlation method; The IIR of the Error Path was achieved by using Least Squares method; the off-line model of the Error Path was got by using Neural Networks method. The testing of the identification's result showed that dynamic characteristic of the model and that of Error Path was coincident.A new off-line method was brought forward which used several Neural Networks. There were different types in these Neural Networks. This method can converge in a short time and got a precise identification. This method can be used not only in identification but also in the other fields.
|