| 摘要 |
<p>Diagnostic monitoring apparatus is described which uses neural net software for use in assessing machines, quality control and reliability. The apparatus is implemented by portable data collecting means which receives signals from the apparatus to be monitored. The data collecting means includes removable storage means preferably in the form of a PCMCIA portable memory card which stores the monitoring signals corresponding to known outcomes or pre-determined results. The PCMCIA card is removable from the portable unit and is then inserted into computer terminal means. The data from the memory card is read by the computer which also has neural network sofware and which generates the neural network pattern corresponding to the input data from the memory card. This creates calibration data in the form of a neural net pattern which is transferred back to the PCMCIA memory card. The memory card is then replaced in the portable data collector so that when the data coupled to tools or components for monitoring the live signal data received from the data collector is processed through the stored neural net pattern producing an output or a set of outputs. Means are provided for comparing the measured outputs against pre-determined outputs and for providing a result of the comparison. An indication is given to the user whether the monitored tool or machine is within the pre-determined criteria for acceptability.</p> |
