Froth Depth Soft Sensor Of Sulfur Froth Flotation And Its Application

Flotation is a widely used method for separating mineral particles. Sulfur flotation is part of Zinc direct leaching process. The adaption of froth flotation recovers sulfur from the acid leaching residue. This not only can avoid the pollution on the environment caused by the emissions of the waste residue but also will get a high grade of sulfur concentrate products and enhance productivity. Froth depth is a key parameter in the flotation. It directly affects the sulfur flotation recovery and concentrate grade. Accurate detection of froth depth is the prerequisite for controlling it. In the industrial field, due to the harsh and extreme environment, the float-ultrasonic froth depth detection meter can not work properly shortly after its adoption. Therefore, this paper looks at how to model froth depth soft sensor so as to realize froth depth online measurement.The main work are as follows:Based on the deep analysis of the mechanism of sulfur flotation technology, the effects of flotation cell pulp change on the froth depth are discussed. Through the ananylsis of foam layer famation mechanism, the relationship between froth surface imagine features and froth depth is highlighted, which gives a theoretical guidance for secondary variables selection of froth depth soft measurement modeling.The methods of extracting the flotation cell surface foam’s characteristics are also studied. After that, the froth image features data and industrial field data is preprocessed through deleting outline data and denoising with wavelet.A Bayesian framework based Relevance Vector Machine learning method is used to develop the froth level soft sense model. As the regression performance of RVM is influenced by the kernel parameter selection, Particle Swarm Optimization method is introduced to the RVM model training process. In the adaptive kernel RVM, different kernel parameters are set for different input features, and PSO is utilized to optimize the kernel parameters at an interval of several iterations so as to improve the regression accuracy of RVM. Based on the froth depth modeling method, sulfur flotation froth imagines monitoring system is developed and has been applied to zinc hydrometallurgy sulfur flotation process. The practical data of a flotation plant validated the effectiveness of the model. The regression accuracy of the proposed model can meet the requirement of industrial application. So, the work of this paper lays a foundation for the following overall control and optimization of sulfur flotation.