The fuel consumption cost of open-pit mine trucks accounts for up to 22% of the total cost. However, research on the fuel consumption of mine trucks has been hindered by low monitoring accuracy and unclear fuel consumption patterns. Based on the fuel consumption per cycle of mine trucks, this study analysed the fuel consumption of 5179 transportation cycles with three types of mine trucks. First, a regression analysis was applied to reveal the non-linear pattern of the fuel consumption caused by multi-dimensional features, such as haulage distance, lifting height, and operation time per cycle. Then, based on multi-dimensional features and the XGBoost algorithm, a prediction model for the fuel consumption of mine trucks was proposed, and the R-squared and mean absolute percent error index values were determined to be 0.93 and 8.78%, respectively. The experimental results showed that the lifting height, load, and haulage distance were the key factors influencing the fuel consumption of mine trucks. The fuel consumption of mine trucks during queuing cannot be ignored. These results can be applied to formulate reasonable fuel consumption assessment indicators, truck dispatch strategies, and cost budget plans.

露天矿卡车的油耗成本占总成本的22％。然而，由于监测精度低和燃料消耗模式不明确，阻碍了矿用卡车燃料消耗的研究。基于矿用卡车每周期的油耗，本研究分析了三种矿用卡车在5179个运输周期中的油耗。首先，进行回归分析以揭示由多维特征（例如牵引距离，提升高度和每个周期的操作时间）引起的燃料消耗的非线性模式。然后，基于多维特征和XGBoost算法，提出了矿用卡车油耗的预测模型，确定R平方和平均绝对百分比误差指数值分别为0.93和8.78％。实验结果表明，起升高度，载荷和牵引距离是影响矿用卡车油耗的关键因素。排队期间矿用卡车的油耗不容忽视。这些结果可用于制定合理的油耗评估指标，卡车调度策略和成本预算计划。