The intelligent control of cleaning of rice–wheat combined harvester is a complex problem, which includes the initial setting of cleaning control, judgment of cleaning loss state, cause analysis and selection of corresponding control strategies and many other sub-problems. The knowledge contained in these sub-problems, including knowledge representation methods and reasoning strategies, is different. Therefore, this paper decomposes the complex problem of cleaning control into a sub-problem of hierarchical structure, and constructs a knowledge model of cleaning control based on binary tree structure. In this way, the cleaning control problem can be decomposed into a small set of sub-problems by the judgment of the nodes of the binary tree, until the sub-problems are small enough to be solved directly so as to get the solution of the original problem. It is proved by examples that this method is of great significance to improve the efficiency of knowledge acquisition, management and maintenance of the expert system of rice–wheat combine harvester, and to enhance the knowledge service ability of the expert system of rice–wheat combine harvester. This method can also be used for reference in other fields.

中文翻译：

---

基于复杂问题求解的清洗控制知识系统

稻麦联合收割机清扫智能控制是一个复杂的问题，包括清扫控制的初始设置、清扫损失状态的判断、原因分析和相应控制策略的选择等多个子问题。这些子问题所包含的知识，包括知识表示方法和推理策略，都是不同的。因此，本文将复杂的清洗控制问题分解为层次结构的子问题，构建了基于二叉树结构的清洗控制知识模型。这样，清洗控制问题可以通过对二叉树节点的判断分解为一小组子问题，直到子问题小到可以直接求解，从而得到原问题的解。实例证明，该方法对于提高稻麦联合收割机专家系统的知识获取、管理和维护效率，增强稻麦联合收割机专家系统的知识服务能力具有重要意义。收割机。此方法也可用于其他领域的参考。