Instrument surveillance and calibration verification through plant wide monitoring using autoassociative neural networks

The approach to instrument surveillance and calibration verification (ISCV) through plant wide monitoring proposed in this paper uses an autoassociative neural network (AANN) which utilizes digitized data presently available in the Safety Parameter Display computer system fi-om Florida Power Corporations Crystal River #3 nuclear power plant to verify correct sensor calibration during operation. An autoassociative neural network is one in which the outputs are trained to emulate the inputs over an appropriate dynamic range. The relationships between the different variables are embedded in the weights by the training process. As a result, the output can be a correct version of an input pattern that has been distorted by noise. missing data, or faulty sensors. Plant variables that have some degree of coherence with each other constitute the inputs to the network. Once the network has been trained with normal operational data it has been shown to successfully monitor the selected plant variables to detect sensor drift or failure by simply comparing the network inputs with the outputs. The Sequential Probability Ratio Test (SPRT) is used as a decision logic module to detect incipient changes in the sensor values using the residual signal between the network estimate and the actual sensor signal. The entire plant wide AANN ISCV system was designed and implemented on a Personal Computer platform. The system simultaneously monitors 78 plant variables for sensor drift as well as gross sensor failures. The system proved to be very sensitive to sensor drift with detection setpoints on the order of 0.1% to 2.5% of the full scale range of the instrument. The AANN method of monitoring many variables not only detects sensor failures, it clearly indicates the signal channel in which the signal error has occurred.