Method for synchronizing data acquired in hot continuous rolling process

Abstract

The invention discloses a high-dimension time sequence data synchronization method which is applied to a hot continuous rolling process, aims to solve the problems that the obtaining of the variables of different frames from a hot continuous rolling manufacturing process and other time sampling has hysteresis, time-varying and other phenomena so that the difficulty in quality detect cause analysis or control system debugging and optimization analysis is increased, builds spatio-temporal transformation relationship with the relationship among invariants in the hot rolling process, converts high-dimensional data which is sampled in equal time into synchronous sampling data with the same volume in a space, and establishes the spatial synchronous corresponding relationship of the high-dimensional data on strip steel. During a time-space conversion process, an undetermined coefficient in equal volume sample is first removed, then equal volume sampling is carried out with the spatio-temporal transformation relationship, and the sampling values of non-aligned points are estimated through an interpolation method during an equal volume acquisition process. A hot continuous rolling process is taken for an example to verify the effectiveness of the method, and the example shows that the synchronized data benefits to the quality defect cause analysis of some types of products.

Description

technical field [0001] The invention belongs to the product quality analysis technology of the metallurgical process, and is mainly used for the analysis and pretreatment of product quality defects in the hot continuous rolling product manufacturing process. Background technique [0002] Multi-stage time-varying complex time-delay systems are common systems in the process industry, such as hot rolling process, cold rolling process and so on. In these complex systems, isochronous sampling methods are generally used to record various control variables, process variables, and product quality parameters in the rolling process for the purpose of control system debugging or quality improvement analysis. Due to the multi-stage nature of the rolling process and the lag of quality observation data, the high-dimensional time-series data collected and recorded has the following characteristics: [0003] 1) The control variables of the same rack correspond to the same position of the p...

AI Technical Summary

Problems solved by technology

[0005] 3) Due to the hysteresis in the quality monitoring stage, the collected and recorded product quality parameters are not in a point-to-point relationship with the control variables and process parameters of each rack;

[0006] 4) Due to the reasons of speed-up and speed-down in the rolling process, there is no fixed time-delay influence relationship between data points between different racks;

[0007] The above characteristics increase the difficulty of debugging analysis or quality analysis of a class of multi-stage time-varying complex time-delay systems such as hot rolling

When quality defects occur, it is difficult to quantitatively establish the corresponding relationship between process variables and quality defects in the control process, making many effective analysis methods for synchronous systems difficult to apply

Images