选矿企业决策部门下达的月度综合生产指标是以原矿金属回收率及选矿比的试验理想值为基础编制的,选矿厂通常无法将该指标直接用于指导生产,需根据选矿生产实际达到的金属回收率及选矿比对其进行调整,但目前这一调整过程完全凭选矿厂工程师人工经验来完成,不能保证调整后月度综合生产指标的合理性与优化性.以某选矿厂的实际情况为背景,根据月度综合生产指标的优先程度不同,建立了以精矿品住、精矿产量及精矿成本等指标偏差量最小为优化目标的目标规划模型,并应用遗传算法对其进行求解,利用现场数据的实验研究验证了模型和算法的有效性.
Monthly comprehensive production indices of an ore - dressing plant was made by decision department based on test results of metal recovery rate and concentration ratio of raw ores. But the indices can not directly guide production process and must be adjusted based on actual metal recovery rate and concentration ratio. Currently the adjusting process is done completely by the engineer of ore - dressing plant based on his experience, so the indices are not guaranteed to be rational and optimal. Taking a certain ore - dressing plant as a background, according to the difference of the priority degrees of the indices this paper establishes a goal programming model to minimize the amount of deviation from optimization objective of grade of concentrate, output of concentrate and cost of concentrate, and applies genetic algorithm to solve the model. Experimental research demonstrates the effectiveness of the model and the algorithm.