Method for optimizing multi-objective parameters based on fuzzy analysis and visualization technology

Abstract

The invention relates to a method for optimizing multi-objective parameters based on a fuzzy analysis and visualization technology, which provides an effective measure for processing, analysis, optimization and prediction of parameter data in an industrial process. The method comprises the following steps of: classifying the data by an evolution algorithm and the conventional fuzzy C average clustering hybrid algorithm to select an optimized data set; improving the conventional non-linear mapping evaluation function, introducing a distance error weight and a differentiation factor to ensure that a dimension reduction error of a distance between the farthest point and the nearest point in the data set is smallest to the greatest extent, concentrating the dimension reduction error on secondary information, procedurally solving by a fastest reduction method, and finally reducing high-dimensional spatial data to a two-dimensional plane; and finally, performing axis mapping on multi-dimensional data along each dimension, drawing contour lines, observing influences and optimization potentials of all the parameters on operating points to provide abundant intuitive information for people, repeatedly selecting the points, performing inverse mapping, and reducing to an actual operation space to fulfill the aim of parameter optimization.

Description

technical field [0001] The invention relates to a method for screening an optimal operation interval in an industrial process, in particular to a multi-objective parameter optimization method. Background technique [0002] In modern industrial production, due to the complexity of the process flow, many operating variables, high complexity of the system, and numerous influencing factors, the obtained production process parameter data is complicated, high-dimensional, nonlinear, and noisy, and difficult to obtain. Give people a clear and intuitive feeling. Therefore, it is difficult for people to effectively organize, analyze, judge and predict historical data. [0003] In response to the above problems, people often use classification and visualization methods to optimize and visualize the data structure. First, classify the data, remove outliers, select a set of optimal points, and then map the data to a two-dimensional plane. Display images on the plane, such as charts, h...

AI Technical Summary

Problems solved by technology

Its shortcoming is that in the existing technology, it is difficult to screen out nonlinear and noise information, so that the final output data tends to be distorted, which affects the operator's judgment

[0063] The above methods often only appear in data processing. In the prior art, there is no technical solution for using the above-mentioned multiple data processing methods for the target parameter optimization industrial control process.

Images