Pulse tube refrigerator working condition prediction method and system based on machine learning

Abstract

The invention relates to a pulse tube refrigerator working condition prediction method and system based on machine learning, and the method comprises the following steps: collecting working condition parameter data during the operation of a pulse tube refrigerator, and dividing the working condition parameter data into two parts: training data and inspection data; building a working condition prediction model by adopting an LM optimized back propagation algorithm based on the training data, and continuously improving the precision of the working condition prediction model through adoption of an iteration method; and inputting the test data into the working condition prediction model, comparing a predicted value calculated by the working condition prediction model with actually measured data, and verifying the precision of the working condition prediction model. The system comprises a data acquisition module, a learning training module and an inspection module. According to the invention, the training learning model is set up by collecting the working condition parameter data during operation of the pulse tube refrigerator, the piston stroke and the pressure amplitude, which are difficult to measure, of the given PTC are accurately predicted, the measurement cost of the method is lower than that of an additionally-installed sensor, and the deviation between a predicted value and an actual value is small.

Description

technical field [0001] The invention relates to the technical field of intelligent working condition monitoring of a pulse tube refrigerator, in particular to a method and system for predicting the working condition of a pulse tube refrigerator based on machine learning. Background technique [0002] Pulse tube refrigerator (PTC) has a wide range of application prospects, but its mechanical structure is compact and its cost is high. In order to prevent unnecessary maintenance costs from errors during its operation, a precise and efficient monitoring method is needed to promote its business. progress. [0003] The current methods applied to PTC operating condition monitoring are generally to install complex thermal sensors for direct monitoring and construct numerical simulation models for indirect prediction. Directly measuring the piston stroke and pressure amplitude of the PTC has high requirements on the sensor and the installation of the sensor will affect the operation...

AI Technical Summary

Problems solved by technology

[0004] For this reason, the technical problem to be solved by the present invention is to overcome the problem of insufficient monitoring of PTC operating conditions in the prior art, and to provide a method and system for predicting the operating conditions of pulse tube refrigerators based on machine learning. The working condition parameter data during operation is used to build a training and learning model to accurately predict the piston stroke and pressure amplitude of a given PTC that is difficult to measure. The measurement cost of this method is lower than that of installing sensors and the deviation between the predicted value and the actual value is small.

Images