Experimental data – driven algorithm to predict muckpile characteristics in jointed overburden bench using unmanned aerial vehicle and AI tools

ABSTRACT

In the current investigation, an intriguing method called Firefly-XGBoost was put up to predict and integrate blast muck-pile results notably drop, throw, and lateral spread, which are instrumental in regulating and remedying the loading problems by loader as well as deceiving factors in type of loader selection. As a result, the particle swarm optimisation (PSO) algorithm and the firefly algorithm were relied on to shore up the efficacy of the XG Boost conventional model. A total of 164 blast experiments were executed in two different mines and the data collected from these trials data were used to train the model. This data includes specific features such as joints spanning height (JSH) as well as other blast characteristics such as the number of joint sets, decking length, total quantity of explosives, stemming length, decking length and firing pattern. The Firefly-XG Boost algorithm yielded better outcomes in terms of RMSE and R² values when compared to the XG Boost, and PSO-XG Boost algorithms. The developed model was found to be suitable in predicting muck-pile parameters for practical application.

KEYWORDS:

• AI–Muckpile
• blast fragmentation
• muck-pile parameters
• unmanned aerial vehicle
• artificial intelligence application
• firefly-XG boost algorithm

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Data will be available on reasonable request from the corresponding authors.

Ethical statement

Authors state that the research was conducted according to ethical standards.

Additional information

Funding

This research received no external funding.