Industrial process fault diagnosis method based on high order correlation

Abstract

The invention discloses an industrial process fault diagnosis method based on high order correlation and belongs to the field of industrial process monitoring. According to the method provided by theinvention, high order correlation information in measurement data of each sensor is obtained without supervision from an angle of representing learning through utilization of a stack sparse self-coding network, and three monitoring indexes SRE, M2 and C are provided for obtained high order correlation characteristics. In a hierarchical learning mode, according to the method, change of a small fault or an early fault in an industrial process is represented relatively finely, so detection of whether the fault occurs or not is facilitated. According to the provided monitoring indexes, whether process operation is kept in a control domain or not can be monitored and certain guidance for identifying a fault type is provided. In an unsupervised learning mode, influence of insufficient sample labels and imbalanced data in the industrial process is not influenced. The method has important practical significance in solving an industrial process monitoring problem in practice.

Description

technical field

[0001] The invention belongs to the field of industrial process control, and relates to a high-order correlation-based industrial process fault diagnosis method. A reasonable index is designed for the obtained high-order statistics to detect and diagnose faults, thereby realizing real-time monitoring of complex industrial processes. Monitoring, especially for the rapid detection of minor faults and early failures. Background technique

[0002] Data-driven technology is an effective tool in metrology for identifying abnormal processes, and multivariate statistical process monitoring has achieved great success in the field of process control and has become one of the most active areas of research in the past few decades. Multivariate statistical analysis methods and their improved methods are widely used in industrial processes, including chemical processes, microelectronics manufacturing, and pharmaceutical processes.

[0003] The purpose of fault detection i...

Examples

Embodiment Construction

[0040] The method proposed by the present invention includes two parts of off-line modeling and on-line monitoring, and its flow chart is as follows image 3 shown.

[0041] The offline part of the method is as follows:

[0042] Step 1: Use the collection of all monitoring quantities collected under normal operating conditions in industrial processes (such as pressure values, concentration values, feed ratios, etc. in chemical processes) as the training set X train , after standardized preprocessing based on its mean and variance, we get

[0043] Step 2: Set the initialization structure parameters of the stacked sparse autoencoder network, and initialize its connection parameters randomly at the same time;

[0044] Step 3: Use layer-by-layer greedy method to train network parameters, including network structural parameters and weights, and end training until the cost function is minimized;

[0045] Step 4: The final output of the stacked autoencoder network is the high-or...

technical field

[0001] The invention belongs to the field of industrial process control, and relates to a high-order correlation-based industrial process fault diagnosis method. A reasonable index is designed for the obtained high-order statistics to detect and diagnose faults, thereby realizing real-time monitoring of complex industrial processes. Monitoring, especially for the rapid detection of minor faults and early failures. Background technique

[0002] Data-driven technology is an effective tool in metrology for identifying abnormal processes, and multivariate statistical process monitoring has achieved great success in the field of process control and has become one of the most active areas of research in the past few decades. Multivariate statistical analysis methods and their improved methods are widely used in industrial processes, including chemical processes, microelectronics manufacturing, and pharmaceutical processes.

[0003] The purpose of fault detection i...

Examples

Embodiment Construction

[0040] The method proposed by the present invention includes two parts of off-line modeling and on-line monitoring, and its flow chart is as follows image 3 shown.

[0041] The offline part of the method is as follows:

[0042] Step 1: Use the collection of all monitoring quantities collected under normal operating conditions in industrial processes (such as pressure values, concentration values, feed ratios, etc. in chemical processes) as the training set X train , after standardized preprocessing based on its mean and variance, we get

[0043] Step 2: Set the initialization structure parameters of the stacked sparse autoencoder network, and initialize its connection parameters randomly at the same time;

[0044] Step 3: Use layer-by-layer greedy method to train network parameters, including network structural parameters and weights, and end training until the cost function is minimized;

[0045] Step 4: The final output of the stacked autoencoder network is the high-or...

Images