A health state real-time monitoring method and system of a vacuum drying device

Abstract

The present application relates to the technical field of data processing, in particular to a health state real-time monitoring method and system of a vacuum drying equipment, comprising: obtaining standard data sequences and real-time data sequences of subsystems in the vacuum drying equipment and each subsystem; taking the subsystems as nodes and the relationships between the subsystems as edges to construct a directed topological graph of the vacuum drying equipment; inputting the real-time data sequences into an improved machine learning model to obtain a working state of the vacuum drying equipment, and monitoring the vacuum drying equipment based on the working state. The present application solves the problem of low abnormality recognition sensitivity.

Claims

1. A health condition real-time monitoring method of a vacuum drying apparatus, characterized by, include: Acquire the subsystems in the vacuum drying equipment, as well as the standard data sequences and real-time data sequences of each subsystem; A directed topology graph of the vacuum drying equipment is constructed by treating subsystems as nodes and the relationships between subsystems as edges. Obtain the target data sequence at any set time step from the standard data sequence and the real-time data sequence; calculate the cross-correlation function values ​​between the standard data sequence and the real-time data sequence and the target data sequence respectively, and take the set time step corresponding to the maximum value of the cross-correlation function value as the target time step; obtain the cross-correlation deviation degree of any side and the propagation degree of any node, wherein the cross-correlation deviation degree is positively correlated with the absolute value of the difference between the maximum value of the cross-correlation function between the standard data sequence, the real-time data sequence and the target data sequence respectively; Obtain a correction factor, which is positively correlated with the degree of propagation and the degree of cross-correlation deviation; The real-time data sequence is input into an improved machine learning model to obtain the working status of the vacuum drying equipment, and the vacuum drying equipment is monitored based on the working status; wherein, the improved machine learning model includes a Gaussian kernel bandwidth, which is inversely correlated with the correction factor.

2. The method for real-time monitoring of the health status of a vacuum drying device according to claim 1, characterized in that, No. Cross-correlation deviation of the strip for: , , These represent the maximum value of the cross-correlation function between the target data sequence and the standard data sequence at any set time step, and the target time step corresponding to the maximum value of the cross-correlation function, respectively. , These represent the maximum value of the cross-correlation function between the target data sequence and the real-time data sequence at any set time step, and the target time step corresponding to the maximum value of the cross-correlation function, respectively. This is the preset time step.

3. The method for real-time monitoring of the health status of a vacuum drying device according to claim 1, characterized in that, No. The degree of propagation of each node for: , To find the minimum value function, To find the maximum value function, For the first The cross-correlation deviation of the first edge corresponding to each node. For the first The cross-correlation deviation of the second edge corresponding to each node It is a non-zero constant.

4. The method for real-time monitoring of the health status of a vacuum drying equipment according to claim 1, characterized in that, The correction factor for: , For the first Cross-correlation deviation of the edges For the first The degree of propagation of each node, Let be the total number of edges. The total number of nodes. These are the preset hyperparameters.

5. The method for real-time monitoring of the health status of a vacuum drying device according to claim 1, characterized in that, Gaussian kernel bandwidth for: , The initial Gaussian kernel bandwidth, This is a correction factor.

6. The method for real-time monitoring of the health status of a vacuum drying equipment according to claim 1, characterized in that, The standard data sequence and the real-time data sequence are subjected to denoising and standardization processing.

7. The method for real-time monitoring of the health status of a vacuum drying equipment according to claim 1, characterized in that, The subsystem includes a vacuum pump system, a cavity system, a heating system, and a temperature control system.

8. The method for real-time monitoring of the health status of a vacuum drying equipment according to claim 1, characterized in that, The machine learning model is the Anomaly Transformer model.

9. The method for real-time monitoring of the health status of a vacuum drying equipment according to claim 1, characterized in that, This also includes training machine learning models, specifically: The training set is fed into a pre-built machine learning model for training. During the training process, the loss between the output prediction value and the label is calculated. Use gradient descent to adjust model parameters to minimize prediction error; The parameters of the machine learning model are iteratively adjusted until the loss is less than a set value or the set number of training iterations are reached, and finally a well-trained machine learning model is obtained.

10. A real-time health status monitoring system for a vacuum drying equipment, characterized in that, It includes a memory and a processor, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, it implements the real-time health status monitoring method of any one of claims 1 to 9.

Images