Process quality homogenization level evaluation method and system based on Birch clustering algorithm

Abstract

The invention relates to a process quality homogenization level evaluation method and system based on a Birch clustering algorithm. The process quality homogenization level evaluation method comprisesthe following steps: collecting key parameter data of tobacco primary processing; extracting historical batch data in the production process; preprocessing the data, and solving a characteristic value of the dimension of the batch by utilizing a Birch clustering algorithm and a Kmeans algorithm; determining dynamic quality stability characterization of steady-state and non-steady-state processes;constructing a quality stability evaluation model from the minimum research unit-process parameter scale to the process scale and then to the batch scale step by step, so as to provide a systematic and effective method for analyzing and evaluating the cut tobacco processing quality from different scales. The clustered data sets record characteristic parameters and time labels of the data, normaldata sets and outlier data sets are distinguished according to the characteristic parameters of the data sets and treated differently in a targeted mode, then the abnormal time periods of the data aretraced according to the time labels of the data, and prerequisite conditions are provided for batch quality abnormity tracing and equipment intelligent maintenance.

Description

technical field [0001] The invention relates to an evaluation technology of tobacco shredded technology, in particular to a multi-scale correlation evaluation method and system for tobacco shredded technology based on Birch clustering algorithm. Background technique [0002] "Cigarette Process Specifications" emphasizes the transformation of process control from control indicators to control parameters, and highlights the important role of process control in cigarette production. The inventors found that the current management and control evaluation methods of cigarette production process mainly include process capability index Cpk and Sigma level, but the above methods have certain limitations (1) The production process objects covered by the evaluation are not comprehensive enough, and the current management and control evaluation only focuses on the steady-state processing process , but no attention has been paid to the non-steady-state production process that significant...

AI Technical Summary

Problems solved by technology

The inventor found that the current management and control evaluation methods for the cigarette production process mainly include process capability index Cpk and Sigma level, but the above methods have certain limitations (1) The production process objects covered by the evaluation are not comprehensive enough, and the current management and control evaluation only focuses on the steady-state processing process , but no attention has been paid to the non-steady-state production process that significantly affects the quality of the process, such as material head, material tail, and severe material breakage; (2) The existing evaluation methods have certain limitations, whether it is Cpk or Sigma level and other characterizations , its statistical calculations are based on the assumption that the data conform to the normal distribution, however, the data of the cigarette production process is difficult to fully satisfy the normal distribution

Images