【摘要】 为对比不同充填材料的成本，指导金属矿山生产，以金川镍矿为研究背景，基于一元非线性回归、曲线估计和约束优化理论，结合掺石灰石粉和粉煤灰充填材料28d强度、塌落度和分层度正交试验结果，建立最低充填成本约束优化问题，约束优化问题中各约束条件数学模型R~2均高于90%.最后，利用OFOA和PSO算法分别求解平均最低充填成本，OFOA计算结果分别为91.615元/kg和97.092元/kg(掺石灰石粉充填成本比粉煤灰低5.477元/kg),而PSO算法未搜索到充填成本最优解.结果表明：(1)利用一元非线性回归和曲线估计理论可针对不同约束条件建立高精度多元非线性数学模型；(2)改良后的OFOA算法可对不同充填材料的成本约束优化问题求解；(3)改良算法较FOA和PSO算法收敛速度更快，计算过程更稳定，利用上述研究方法可对不同金属矿山、充填材料和约束条件进行成本对比，为充填材料比选提供理论指导.更多还原

【Abstract】 The cost of different filling materials is analysed to guide metal mine production.Taking Jinchuan nickel mine as the research background, the minimum filling cost constraint optimization problem was established based on the unary nonlinear regression, curve estimation and constraint optimization theories, combined with the 28d strength, collapse and stratification orthogonal test results of limestone powder and fly ash filling material.The mathematical model R~2 of each constraint condition in the constraint optimization problem was higher than 90%.Finally, the OFOA algorithm and PSO algorithm were used to solve the average minimum filling cost.The calculated results of the OFOA were 91.615yuan/kg and 97.092yuan/kg(the filling cost with limestone powder was 5.477yuan/kg lower than that with fly ash),but the PSO algorithm did not find the optimal filling cost solution.The results show that:(1)Using the theory of unitary nonlinear regression and curve estimation, a high-precision multivariate nonlinear mathematical model can be established according to different constraints;(2)The improved OFOA algorithm can solve the cost constraint optimization problem of different filling materials;(3)Compared with the FOA and PSO,the improved algorithm has a faster convergence rate and a more stable calculation process.The above research methods can be used to carry out cost-comparative analyses on different metal mines, filling materials and constraint conditions, providing theoretical guidance for filling material selection.更多还原