Research on the relationship between pH and uranium concentration in leachate of sandstone uranium deposit in-situ leaching area based on TCN-transformer hybrid model
摘要
This study investigates the coupling between pH and uranium concentration in leachate during in-situ leaching (ISL) of sandstone-type uranium deposits. To address the limitations of conventional CNNs regarding gradient instability and the challenges of long-term dependency modeling, a hybrid TCN-Transformer architecture is proposed. With comparable parameter counts and computational complexity, the proposed model achieves superior predictive accuracy for pH-uranium dynamics. Benchmark evaluations demonstrate significant performance enhancements over LSTM, RNN, CNN and GRU based methods. The results establish a computationally efficient framework for real-time process control in uranium ISL.