ABSTRACT
This chapter examines reasoning and control with qualitative knowledge represented by a cloud model rather than through a precise mathematical model, and implementing stable control by qualitative rule controller, reflecting the control experience expressed by human natural language. In the control field, rules such as "perception–action" are usually adopted to express the logic cause-result relation. The cloud-based qualitative rule generator is composed of a precondition cloud generator and a postcondition cloud generator. The multi-condition-single-rule generator can be constructed by expanding the method utilized in the double-condition-single-rule generator. The single-condition-single-rule and multi-condition-single-rule previously generated can be stored in the rule database for further reasoning and control of the qualitative knowledge. In the practice of reasoning and qualitative control, it is difficult to give an accurate mathematical model for the controlled object. The chapter compares the cloud-based qualitative control method, the classical fuzzy control method, and the probability control method to illustrate the qualitative control mechanism of cloud-based method.
