In spite of many efforts made a complete model of machine spinning processes, due to its complexity, multidimensionality of the decision space and the present state of knowledge, is unachievable. The paper addresses the issues of constructing a local process model to enable the search for a locally optimal course of the process, within a short time and with the cost as low as possible. Comparison was made between the theoretically well-grounded response surface designs method with a few approaches to the model construction based on intuitively understood heuristic bases justified by their successful practical applications. In order to determine a set of Pareto-optimal solutions for a discrete decision space, the durations of process execution were generated through a virtual simulation. In order to outline and justify the adopted solutions a comprehensive example of the practical construction of the machine spinning process model was presented, including its various versions. The results obtained were validated and evaluated. The main utilitarian conclusion is the indication whereby basing on a partial experiment plan it is possible, thanks to simple heuristic methods, to obtain Pareto-optimal solutions which are close to those obtained when the full experiment plan is carried out.

尽管进行了许多努力，但由于机器旋转过程的复杂性，决策空间的多维性和当前的知识状态，仍无法实现完整的模型。本文探讨了构建本地过程模型的问题，以便能够在短时间内以尽可能低的成本搜索过程的局部最优过程。在理论上有良好基础的响应面设计方法与几种模型构建方法之间进行了比较，这些方法基于直觉理解的启发式基础，并以其成功的实际应用为依据。为了确定离散决策空间的一组Pareto最优解，通过虚拟仿真生成了流程执行的持续时间。为了概述并证明所采用的解决方案的合理性，给出了机器纺丝工艺模型实际构建的综合示例，包括其各种版本。获得的结果经过验证和评估。功利主义的主要结论是这样的指示：借助简单的启发式方法，可以根据部分实验计划获得帕累托最优解，该解与进行完整实验计划时获得的解接近。