Russo Mark F, Wild Daniel, Hoffman Steve, Paulson James, Neil William, Nirschl David S
1Research Informatics and Automation, Bristol-Myers Squibb, Co., Princeton, NJ, USA.
J Lab Autom. 2013 Oct;18(5):391-403. doi: 10.1177/2211068213486651. Epub 2013 Apr 19.
A systematic method for assembling and solving complex compound distribution problems is presented in detail. The method is based on a model problem that enumerates the mathematical equations and constraints describing a source container, liquid handler, and three types of destination containers involved in a set of compound distributions. One source container and one liquid handler are permitted in any given problem formulation, although any number of compound distributions may be specified. The relative importance of all distributions is expressed by assigning weights, which are factored into the final mathematical problem specification. A computer program was created that automatically assembles and solves a complete compound distribution problem given the parameters that describe the source container, liquid handler, and any number and type of compound distributions. Business rules are accommodated by adjusting weighting factors assigned to each distribution. An example problem, presented and explored in detail, demonstrates complex and nonintuitive solution behavior.
本文详细介绍了一种用于组装和解决复杂化合物分配问题的系统方法。该方法基于一个模型问题，该问题枚举了描述一组化合物分配中涉及的源容器、液体处理器和三种类型的目标容器的数学方程和约束条件。在任何给定的问题表述中，允许有一个源容器和一个液体处理器，尽管可以指定任意数量的化合物分配。通过分配权重来表示所有分配的相对重要性，这些权重被纳入最终的数学问题规范中。创建了一个计算机程序，该程序在给定描述源容器、液体处理器以及任意数量和类型的化合物分配的参数后，能自动组装并解决一个完整的化合物分配问题。通过调整分配给每个分配的加权因子来适应业务规则。文中详细给出并探讨了一个示例问题，展示了复杂且非直观的解决方案行为。
