The wavelet transform is developed to identify the differentphases in a fermentation process. In this method, the wavelettransform modulus maxima are used to estimate the local maximumpoints of the second derivative o...The wavelet transform is developed to identify the differentphases in a fermentation process. In this method, the wavelettransform modulus maxima are used to estimate the local maximumpoints of the second derivative of the growth curve in order toclassify the different phases of fermentation process. Moreover, themethod can effectively get rid of noise from the signal, making useof the different characters showed by signal and noise in the wavelettransform modulus maxima. Compared with neural network modeling, thepresented method needs less quantity of information and calculation.The results of experiments show that this method is effective.展开更多
基金Supported by the Natural Science Foundation of Shandong Province(Q99B01).
文摘The wavelet transform is developed to identify the differentphases in a fermentation process. In this method, the wavelettransform modulus maxima are used to estimate the local maximumpoints of the second derivative of the growth curve in order toclassify the different phases of fermentation process. Moreover, themethod can effectively get rid of noise from the signal, making useof the different characters showed by signal and noise in the wavelettransform modulus maxima. Compared with neural network modeling, thepresented method needs less quantity of information and calculation.The results of experiments show that this method is effective.
