当确定各种各样的过程变量对生产操作结果的影响时,通常应用二因次评估法。根据任一控制或操作规程的建议而得到的判据远非最佳值。一个变量变化时,将影响其他许多变量。虽然多变量系统的方法学已有较多的文献资料证明,但高效率的现代化仪器,如快速分析仪和大功率计算机没有被采用,理论和实践之间还存在着差距。析因设计工作的专门目的是改善由实验结果引出的结论的可靠性,并且定量地表示不同变量相应的重要性。本文涉及作者提交到国际自动控制联盟第九次大会(1984年7月2日至6日在布达佩斯举行)上的一篇论文。这里以一个铜矿浮选过程为实例,介绍析因设计在工业中的应用,并表示出各种数据的关系。 When determining the impact of a wide range of process variables on the outcome of production operations, a two-way assessment is usually applied. Criteria based on the advice of either control or operating rules are far from optimal. When a variable changes, it affects many other variables. Although the methodology of multivariable systems has been documented in many documents, highly efficient and modern instruments such as rapid analyzers and high-power computers have not been adopted and there is still a gap between theory and practice. The special purpose of factoring design work is to improve the reliability of the conclusions drawn from the experimental results and to quantify the corresponding importance of the different variables. This article deals with a paper presented by the author to the Ninth Congress of the International Union of Automatic Controls (Budapest, Hungary, 2-6 July 1984). Here a copper flotation process as an example, introduces the design of factoring in the industrial application, and showed a variety of data.