How nature-inspired algorithms enhance mining safety?

Did you know that algorithms inspired by animals could enhance mining safety? A new study combines three optimization algorithms inspired by tunas, whales, and jellyfish with Random Forest to predict the shear bond strength of cable bolts. These bolts are crucial for stabilizing underground tunnels. Scientists created a database of 860 samples by adding random noise to the original data. Three hybrid models (RF-WOA, RF-JSO, RF-TSA) were trained and evaluated using six different metrics. The RF-JSO model, inspired by jellyfish, achieved the best results with an R2 of 0.981 and an RMSE of 11.063. Additionally, tools like SHAP and LIME showed that confining pressure, elastic modulus, and standard cable type are the most important factors in the prediction. This study demonstrates that hybrid swarm intelligence can significantly improve mining safety.

Nature-inspired optimization algorithms, such as the Tunicate Swarm Algorithm (TSA), Whale Optimization Algorithm (WOA), and Jellyfish Search Optimizer (JSO), were combined with Random Forest (RF) to predict the shear bond strength of cable bolts. These algorithms mimic the behavior of marine animals to find optimal solutions efficiently. The RF-JSO model showed the best performance with a coefficient of determination (R2) of 0.981, a root mean squared error (RMSE) of 11.063, a mean absolute error (MAE) of 6.457, a mean absolute percentage error (MAPE) of 9, a variance accounted for (VAF) of 98.168, and an A-20 index of 0.891. Interpretability tools like SHAP, PDP, and LIME revealed that confining pressure, elastic modulus, and standard cable type are the most influential factors in the prediction. This hybrid approach represents a significant step forward in predicting the shear bond strength of cable bolts, enhancing safety in mining operations.