Abstract:The tailing grades must be predicted in real time for the automatic control and avoidance of the waste of valuable minerals in froth flotation. In most previous prediction models, all input variables are regarded as having equal importance, ignoring their differences in contribution to grade prediction. Additionally, the process data obtained by different sensors at different positions contain abundant information about the temporal changing trend of flotation performance, which can provide useful information for grade prediction. However, few works consider fusing these data for grade prediction. To solve the above problems, variablewise and temporalwise attention is proposed for the prediction of tailing grades in this paper. First, to strengthen the important variables, a variablewise attention module is proposed to automatically learn the importance of variables and assign higher weights to more important variables. Meanwhile, since the froth videos and tailing grades are obtained from different flotation cells and detected by different sensors, they have time lags and inconsistent detection frequencies, such that they cannot be fused directly. Therefore, to facilitate the fusion of both types of data for grade prediction, a temporalwise attention module is proposed to match them in the temporal dimension. Extensive experiments have validated the effectiveness of the proposed model. In addition, compared to the encoder-decoder model, the mean absolute error (MAE) and root mean squared error (RMSE) of the proposed model are decreased by 13.8 % and 13 %, respectively, and the R-squared (R2) score of the proposed model is increased by 9.7 %.

Grade Prediction Modeling for Froth Flotation Based on NARX-NN and Ensemble Learning

Grade prediction of zinc tailings using an encoder-decoder model in froth flotation

Prediction of multi-stage froth flotation efficiency of complex lead–zinc sulfide ore using an integrated ensemble neural network–random forest model

Long short-term memory-based grade monitoring in froth flotation using a froth video sequence

FlotationNet: A Hierarchical Deep Learning Network for Froth Flotation Recovery Prediction

Grade Prediction of Froth Flotation Based on Multistep Fusion Transformer Model

A Novel Semi-Supervised Soft Sensor Modeling Method Based on Deep Dynamic and Semantic Information Extraction for Concentrate Grade Prediction in Froth Flotation

Physics-Informed Machine Learning for Grade Prediction in Froth Flotation

Interpretable Prediction Modeling for Froth Flotation Via Stacked Graph Convolutional Network

Online prediction of concentrate grade in flotation process based on PCA and improved BP neural networks

Towards Tailing Grade Prediction in Zinc Flotation Via Variablewise and Temporalwise Attention

Upstream Process Condition Monitoring for Froth Flotation Based on Feature Performance Evaluation and Parameter-Mapped GRNN

Salinity Time Series Prediction and Forecasting Using Dynamic Neural Networks in the Qiantang River Estuary

ES-net: An Integration Model Based on Encoder–Decoder and Siamese Time Series Difference Network for Grade Monitoring of Zinc Tailings and Concentrate

Grouped Time Series Networks for Grade Monitoring of Zinc Tailings With Multisource Features

Enhanced Short-Term Temperature Prediction of Seasonally Frozen Soil Subgrades Using the NARX Neural Network

Prediction of X-ray Fluorescence Copper Grade Using Regularized Stochastic Configuration Networks

Prediction Model for Ph Value in Flotation Process Based on Hybrid Neural Network

Two-Stream Deep Feature-Based Froth Flotation Monitoring Using Visual Attention Clues

Short-long Temporal Graph Convolution Network for Grade Monitoring in a First Zinc Rougher

Online hybrid modeling method with application for predicting Bauxite production indicators