Device system health degree online monitoring method based on v-gap metric

Abstract

The invention discloses a device system health degree on-line monitoring method based on v-gap metric, which comprises the following steps that a sensor collects input and output data on line, a system model is identified and updated by adopting instant learning, and the data is mapped to an operator space; the distance between the online updating model and the system reference model is measured in the operator space, and v-gap metric is selected as the measurement mode; and finally, judging whether the system breaks down or not according to whether the distance between the models exceeds a set threshold value or not, wherein the threshold value is calculated according to the confidence interval of the reference model obtained by off-line identification. According to the method, the difference between an updating model and a reference model is calculated on line, different from other data measurement methods, the change of dynamic characteristics of a measurement system from the model level is innovatively proposed, and the health degree of the system in all operation stages including a transient transition stage and a steady-state operation stage is effectively monitored. The invention provides a brand new thought for the dynamic monitoring of the health degree of the device system, and provides a powerful guarantee for the safe operation of the device system.

Description

technical field [0001] The invention belongs to the field of process monitoring, and in particular relates to an online monitoring method for device system health based on a ν-gap metric. Background technique [0002] In recent years, with the development of my country's modern chemical industry, metallurgy, machinery, logistics and other industries and the progress of related technologies, the investment and scale of the device system are increasing day by day, the complexity is getting higher and higher, and there are also couplings between systems. For such a complex process, safety and reliability play a crucial role, so health monitoring and fault early warning technologies emerge as the times require. Due to the convenience of process data acquisition, data-based methods are widely used, which can be mainly divided into machine learning and multivariate statistical analysis methods. The two methods are used to monitor the state of the device system according to the ex...

AI Technical Summary

Problems solved by technology

[0004] However, the above examples only monitor the health of the device system in the steady state stage. In the actual production process, due to changes in working conditions or set points, there will often be a dynamic transition stage, that is, the operation of the system is determined by the transient state and the steady state. Composed of state processes, the full cycle work of the system should be completely monitored

In addition, feature extraction at the data and parameter level cannot fully obtain the dynamic information of the system, so the above methods are not effective in monitoring the transient phase.

Images