In this paper a new method of neural filtering using artificial neural network systems is presented for the filtering problems of linear and nonlinear, stationary and nonstationary stochastic signals. The neural filter (denoted neurofilter) developed in this paper has either finite impulse response (FIR) structure or infinite impulse response (IIR) structure. The neurofilter differs from the conventional linear digital FIR and IIR filters because the artificial neural network system used in the neurofilter has a nonlinear structure due to the sigmoid function. Numerical studies for the estimation of a second-order Butterworth process are performed by changing the structures of the neurofilter in order to evaluate the performance indices under changes of the output noises or disturbances. The results obtained from these studies verified the capabilities which are essentially necessary for on-line filtering of various stochastic signals.
[1]
L. Ljung,et al.
Recursive input-output and state-space solutions for continuous-time linear estimation problems
,
1976,
1976 IEEE Conference on Decision and Control including the 15th Symposium on Adaptive Processes.
[2]
Masanori Sugisaka.
The Design of a Chandrasekhar-Type Filter via Imbedding Method
,
1988
.
[3]
M. Morf,et al.
Fast time-invariant implementations of Gaussian signal detectors
,
1978,
IEEE Trans. Inf. Theory.
[4]
Masanori Sugisaka,et al.
Continuous-time Chandrasekhar smoothers in detection and tracking
,
1995
.
[5]
Thomas Kailath,et al.
Some new algorithms for recursive estimation in constant linear systems
,
1973,
IEEE Trans. Inf. Theory.
[6]
Masanori Sugisaka,et al.
On-line Estimation and Learning Via Neural Network Systems Based on Digital Chandrasekhar Filter
,
1991
.