×

Waveform evaluating apparatus using neural network

  • US 5,630,019 A
  • Filed: 05/30/1995
  • Issued: 05/13/1997
  • Est. Priority Date: 05/23/1992
  • Status: Expired due to Fees
First Claim
Patent Images

1. A waveform evaluating apparatus using a neural network comprising:

  • a plurality of neural network modules each being made up of at least an input layer, a hidden layer, and an output layer, connected in the neural network having the neural weight ratio;

    first means for providing a first predetermined output value by inputting a first standard waveform to said neural network module, said first means including means for adjusting said neural weight ratio;

    second means for providing a second predetermined output value by inputting a second standard waveform to said neural network module, said second means including means for adjusting said neural weight ratio, said first standard waveform being predeterminedly preferred to said second standard waveform;

    means for producing a third output value in response to inputting an input waveform to said neural network, wherein said third output value'"'"'s closeness to said first and second predetermined output value is ascertainable; and

    a signal extracting portion disposed in said input layer for extracting a signal which is in phase with the first standard waveform, wherein said signal extracting portion supplies the extracted signal to said input layer as an input waveform for said neural network;

    wherein said signal extracting portion is formed of a phase detecting portion for detecting phasic information and a waveform slicing portion for performing waveform slicing using the phasic information detected by said phase detecting portion.

View all claims
  • 1 Assignment
Timeline View
Assignment View
    ×
    ×