Aiming at the problem of equipment monitoring in stable operation

a method that based on dynamic baseline in real time is proposed for equipment multiple state estimation. In this method

an equipment health matrix is constructed under stable operating conditions through associative learning and training of equipment running in big data. Then

the real-time deviation of the equipment is calculated according to the current operating observation data

and a real-time dynamic baseline is constructed based on the deviation to realize the judgment for the equipment operating status

and then the abnormal measuring points will be located by timely feedback of the contribution rate of the cause of the failure. This method is applied on air and flow gas system of a unit. The test results show that

this dynamic baseline method can detect the equipment anomalies as early as possible and then carry out fault diagnosis

which effectively solves the uncertainty and lag of fixed thresholds

reduces the probability of misdiagnosis of equipment fault. This method can accurately detect the abnormal state of the equipment and diagnose the fault in the early stage

effectively solve the uncertainty and lag of the fixed threshold judgment of the abnormal state of the equipment

and reduce the probability of fault misjudgment of the equipment.