A Rotary Kiln Firing State Recognition Method with Imitation Feedback Adjustment Mechanism

Abstract

The invention discloses a method for identifying the firing state of a rotary kiln with an imitation feedback adjustment mechanism, and its features include: the first stage is to extract the cognitive features of the flame image, and establish a complete cognitive feature space of the sample; the second stage is to Establish a reduced cognitive feature space based on information granularity; the third stage is to design a cognitive system using an integrated RVFL classifier; the fourth stage is to evaluate the reliability of cognitive results, freely adjust the information granularity according to the evaluation results, and reactivate the second To the fourth stage, the cognitive feature space is updated and reduced, and the cognition is fed back multiple times. The invention can imitate manual free adjustment of complete cognitive feature level to carry out multi-level feedback cognition, significantly improve the cognitive accuracy and reliability, so as to truly realize "machine watching fire" instead of "manual fire watching", in order to realize the quality of rotary kiln clinker The closed-loop control of indicators lays the foundation.

Description

technical field [0001] The invention belongs to the technical field of flame image recognition of a rotary kiln, and in particular relates to a method for identifying a firing state of a rotary kiln with an imitation feedback adjustment mechanism. Background technique [0002] Rotary kiln is a large-scale sintering equipment widely used in metallurgy, environmental protection, chemical industry and cement. Due to the difficulties in the sintering process of the rotary kiln, such as the difficulty of online measurement of the clinker quality index and the difficulty of accurately identifying the firing state of key process parameters closely related to the clinker quality index, the existing sintering process of the rotary kiln is still in the "manually watching the fire" Open-loop operation phase. However, the results of manual identification of the firing state are restricted by subjective factors such as operator experience, responsibility, and attention, which can easily...

AI Technical Summary

Problems solved by technology

[0007] The open-loop cognitive system extracts simple features for different flame images at a specific granularity level. Once the feature space is established, it will not be updated, which will easily lead to the fact that the simple feature space is not enough to fully represent the flame image.

In addition, the classifier is not robust enough to establish classification cognition criteria only by comparing peak signal-to-noise ratio values.

The method used does not conform to the characteristics of using multi-level features to repeatedly deliberate and compare when human beings judge the firing state, and the recognition accuracy is not satisfactory

Images