Process trend analysis and prediction method based on intelligent instrument

Abstract

The invention discloses a process trend analysis and prediction method based on an intelligent instrument, and the method comprises the steps: enabling the intelligent instrument to record historical perception data; training a data prediction model according to the historical perception data; adding perceptual data predicted by the data prediction model into a sliding window; calculating a kurtosis value in the sliding window; and inputting the kurtosis value into the state transition model to complete updating and prediction of the state. On the basis of historical perception data in the intelligent instrument, the data prediction model is trained through the operational capability of the intelligent instrument, the sliding window is added, and the kurtosis value is input into the state transition model, so that state updating and prediction of the intelligent instrument are finally realized. The kurtosis value, multiple linear regression and finite-state machine methods are adopted to accurately predict the process state of the instrument, and compared with complex machine learning methods such as an artificial neural network, the method is very small in calculation amount and can be completely completed at the instrument end.

Description

technical field [0001] The invention relates to an instrument process trend prediction method, in particular to an intelligent instrument-based process trend analysis and prediction method. Background technique [0002] At present, traditional mainstream instruments often have single system functions, limited computing resources, and weak self-processing capabilities. In the traditional mainstream instrument and control system, the instrument is responsible for measurement, and the control system is responsible for calculation and control. In this mode, the instrument cannot know its current state, so it is often difficult to meet the higher measurement and control of smart factories due to delay and asynchrony. Accuracy requirements. [0003] In order to reduce the asynchronous problem caused by the delay, a lightweight process trend analysis method on the instrument side is needed. At the same time, in order to ensure the long-term reliable measurement of the instrument e...

AI Technical Summary

Problems solved by technology

The gray forecasting model has the problem of short forecasting period and is not suitable for long-term forecasting. The support vector machine has problems such as difficulty in parameter selection and slow training speed. The artificial neural network has problems such as complex calculations and difficulty in ensuring real-time performance.

Images