Zinc flotation dosing prediction control method based on texture degree optimization

Abstract

The invention discloses a zinc flotation dosing prediction control method based on texture degree optimization. The method comprises the following steps: firstly, extracting foam image features including entropy, energy, inverse difference moment, foam size and foam color features of a roughing foam image and texture features of a concentration foam image, and defining the texture complexity according to a relation between texture parameters of the concentration foam image and a concentrate grade; secondly, constructing a selected foam texture complexity prediction model by utilizing the extracted characteristic parameters and adopting a neural network, and finally, calculating a medicament dosage adjustment value by adopting an optimization method by taking a difference square of a selected foam texture complexity prediction value and an expected optimal value as a target function, so as to finish medicament adding control. According to the method, the foam state characteristics of concentration and roughing are combined, so that the control result is more excellent, the mineral recovery efficiency is improved, and the reagent consumption is reduced.

Description

technical field [0001] The invention belongs to the technical field of froth flotation, and in particular relates to a method for predicting and controlling the dosing amount in the zinc flotation process. Background technique [0002] Froth flotation is a process of separating minerals from useless gangue by utilizing the difference in hydrophilicity and hydrophobicity of minerals. The dosage of zinc flotation is mainly observed by experienced operators with naked eyes to observe the color, size, and The state of texture and so on judges the production state and adjusts it. However, manual observation has strong subjectivity and randomness of operation, which leads to large errors and low efficiency in the dosing operation during the flotation process, which may lead to serious waste of chemicals and mineral resources due to inaccurate judgment, and the foam in the selection tank Texture features are closely related to the concentrate grade of foam, so by controlling the t...

AI Technical Summary

Problems solved by technology

However, the flotation process is a complex multi-process process, which is completed through roughing, beneficiation and sweeping. This method only analyzes and judges the foam state in the roughing cell, and does not combine the foam state of multiple flotation cells. , in actual industrial applications, there are problems that the production status cannot be fully monitored and the adjustment is not timely

Images