| 摘要 |
PROBLEM TO BE SOLVED: To eliminate the need for a large capacity memory for storing a filter coefficient, to eliminate the need for much time for operation processing the filter coefficient and to make possible obtaining a sufficient noise reduction effect regardless of a temp. change in noise propagation space by varying a cutoff frequency of a low-pass filter and a sampling frequency in convolutional operation processing according to the temp. in the noise propagation space. SOLUTION: Given the sampling frequency of convolutional operation in a filter convolutional operation part 13 when the temp. in the noise propagation space SSN isϕ0 , the cutoff frequency of the low-pass filters 11, 15 is Fc0 is defined F0 , the sampling frequency when the temp. isϕ1 is Fs1 , and the cutoff frequency is Fc1 , the control is performed by a sampling frequency revision part 20 and a cutoff frequency revision part 21 so that following relations I, II are satisfied by all values. The equation I: Fc1 =μ.Fc0 , the equation II: Fs1 =μ.Fs0 . Where,μ=((273.15+ϕ1 )/(273.15+ϕ0 ))1/2 .
|
