Diagnostics of analogue sensors by means of an intelligent measuring channel

Many break-downs of technological units of power stations (nuclear and fossil-fuel), chemical plants, food factories and similar complex technologies, have often an origin in a wrong interpretation of the data from the analogue sensors as failures of sensors are not recognised in time. An intelligent driver for the analogue sensor fault detection (called Intelligent Measuring Channel - IMC) has been developed. The first aim of the paper is the comparison of redundancy models TMR, TMR/S, TMR/S/S with Intelligent Measuring Channels and theoretical evaluation of their safety. There are worked out equations for computation of survival probabilities for the IMC with n sensors. These equations were formed to select an optimum number of redundancy sensors. The results show an enhancement of a safety level, using the IMC in comparison with classic sensor majority voting methods. The second aim of this paper is to present the newly implemented test for a recovery of a measurement after passing of transient failures based on the sensor-input estimation.