| Budget Amount *help |
¥2,100,000 (Direct Cost: ¥2,100,000)
Fiscal Year 1994: ¥600,000 (Direct Cost: ¥600,000)
Fiscal Year 1993: ¥1,500,000 (Direct Cost: ¥1,500,000)
|
|---|
| Research Abstract |
1. We investigated and summarized the properties of Wavelet transforms. Particularly, we characterized the class of Wavelet functions.
2. We applied the Wavelet trnasform to the problem of identification, and showed that it is possible to extract the signal from the noisy data.
3. From the experience of 2, we are motivated to derive the relationship among scalogram and other distributions of energy.
4. To treat the noisy data via Wavelet transform it is necessary to study stochastic features such that weak stationarity, correlations, power spectrum, and innner product. We have shown that these properties hold in a similar manner to Fourier transform.
5. To determine the unique mother wavelet function, we formulate an optimal problem under exogenous noisy inputs, and derived optimal solution. It is shown that the noise effect can be completely reduced for some class of noises.
6. From the idear of time-frequency decomposition, we formulate the problem of designing the controllers each of which processes the signals individually.
|
