摘要
中长期动态优化配矿是研究如何制定在一个较长时期内采矿顺序与配矿的问题,是一个高度非线性受限条件下的多目标优化问题。采用多轮粒子群算法(particle swarm optimization algorithm,PSO)来求解矿山企业动态配矿问题。首先,依据开采条件圈定出可开采的矿块,并给出预测的产品价格趋势结果作为算法的输入条件;然后,通过对PSO算法轮的划分以确定每轮的动态配矿方案,经过多轮PSO算法优化后的最终目标为中长期动态配矿的优化结果。在PSO算法中用粒子的一位来代表矿块,并用0或1来代表选择此矿块来开采,然后重新定义了在约束条件下PSO粒子的运算与飞行规则,最后实现了动态配矿优化的多轮PSO算法。该多轮PSO算法对使用者要求简单,适用性强,通过对实际数据的计算表明了该算法的有效性。
Middle and long-term dynamic optimization mine ore blending is to make a policy which the sequence of mine block to be exploit is decision. This problem is a highly dynamic with complex nonlinear problems,which can be classified to multiobjective optimization problem under the conditions of multi-constrains. This paper proposed several rounds of particle swarm optimization algorithm to solve the problem ore mining enterprises dynamic allocation. First of all,gave the mineable ore blocks according to mining conditions,and gave the product price trends results,all of which could be as the algorithm input conditions. Then,through the division of rounds of PSO to determine the dynamics of each round with mine program,optimized after many rounds of PSO algorithm with the ultimate goal of mine for the long-term dynamic optimization results. In the PSO algorithm,using a particle to represent the ore block,and used 0 or 1 to represent the choice to exploit the ore block,and then redefined the constrained PSO particles computing and fly rule and finally realized the dynamic allocation mine several rounds of PSO algorithm for optimization. This multi round of PSO algorithm is simple and user requirements,applicability,by the calculation of the actual data show the effectiveness of this algorithm.
出处
《计算机应用研究》
CSCD
北大核心
2010年第10期3711-3717,共7页
Application Research of Computers
基金
国家自然科学基金资助项目(50774092)
关键词
配矿优化
多层粒子群算法
多目标优化
效益
ore blending optimization
multi round particle swarm optimization（ PSO） algorithm
multi-objective optimization
efficiency
