Sliding window time difference-Gaussian process regression modeling method based on local time lag reconstruction

Abstract

The invention relates to a sliding window time difference-Gaussian process regression modeling method based on local time lag reconstruction. The method is suitable for chemical processes with time lag, nonlinear and time-variant characteristics. By means of the method, the latest process time-variant dynamic state can be tracked step by step through a sliding window strategy; meanwhile, parameters are extracted from process periodical time lag characteristics through a fuzzy curve analysis method in a sliding window and used for time lag reconstruction of local model training samples and testing samples; the variable drift characteristics on the locally-reconstructed sliding window are depicted through a time difference-Gaussian process regression (TDGPR) model. An effective real-time prediction and control technical support means is provided for the industrial process, product quality can be easily improved, production cost is controlled, and potential safety hazards are avoided.

Description

technical field [0001] The invention relates to a soft sensor modeling method of a sliding window time-difference Gaussian process regression (LTR-MWTDGPR) based on local time-delay reconstruction, and belongs to the field of complex industrial process modeling and soft sensor. Background technique [0002] With the continuous strengthening of modern industrial processes for product quality control and optimization requirements, the online measurement technology of process variables also puts forward higher requirements accordingly. In many practical applications, quality-related variables (such as gas flow concentration, the content of certain chemical components in products, etc.) are often difficult to measure online, but they need to be obtained in a timely manner. In this context, soft-sensing technology emerged as the times require. It infers the real-time value of difficult-to-measure leading variables by constructing a mathematical model of auxiliary variable sets an...

AI Technical Summary

Problems solved by technology

[0009] In order to solve the above-mentioned technical problems, the object of the present invention is to provide a TDGPR model that does not need to wait for the arrival of new input data, but only needs to use the corresponding time-delay reconstruction samples as model input, so as to obtain the real-time predicted value of the leading variable in advance , to a certain extent solve the problem of leading variables with time lag Sliding window time difference-Gaussian process regression modeling method based on local time-delay reconstruction

Images