Logic, Based On The Knowledge Of Time And Action Planning - Intent

BDI (belief-desire-intention) is an essential structure which explains such mental factors affecting subjects' rational behaviors as belief, desire, and intention. BDI logic is a general designation for the logic investigating the influence of BDI on subjects' rational behaviors. It has been widely used in artificial intelligence and kept a focus of the computer science and artificial intelligence research since 1980s. Consequently, logic research on BDI arose to meet the artificial intelligence requirement for related logic theories.The present study aimed to establish a logic with general BDI features—knowledge-intension logic based on time and action planning. This logic was thought to integrate temporal logic, dynamic logic, and cognitive logic and describe the logic features of knowledge and intension in BDI and their dianoetic relations.For this purpose, the following work was carried out in the current research:To start with, Chapter 2 set up PTL, a temporal logic in which natural numbers were used to describe temporal sequence, by improving reflexivity, seriality, transitivity, and discreteness of the natural number temporal logic by [1] and [2]. Also, PTL was proved to be sound and complete.By combining the labeled transition systems [3] based on the logic proposed in Chapter 2, Chapter 3 developed another logic, APTL, which was able to express actions, and proved it was sound and complete. This logic made it possible to depict the features of intention.Then, Chapter 4 created a logic to express the features of intention—I_aptl, an intention logic based on time and action planning. This creation was an important innovation of the current research. In terms of this logic, the semantic definition of intension involved time and action ingredients in addition to general mode semantics, demonstrating the complexity of intension itself. Unlike the logic of other BDI structures, intention in the current research demanded reflexivity, with which the desire for the actions induced by intension could be expressed. I_aptl successfully solved the problem of intention consequences: K-axiom of intention was invalid in I_aptl.Further, Chapter 5 suggested a logic describing the dianoetic relations between knowledge and intension. That was LI_aptl, a knowledge-intension logic based on time and action planning. It directly revealed that the subject's intension was always evoked by his cognition and hence it was rational. To realize this, the present study established the relationship between knowledge and intension in the logic pattern. This logic was another innovation. LI_aptl solved the side effects of the BDI logic.The last chapter put forward the ultimate achievement of the present study: L_aI_ptl, a general awareness knowledge-intention logic based on time and action planning. This logic could figure out logic omniscience and intension consequences. The solution derived from the general awareness logic by [4]. Making awareness operators act on knowledge operators, the current research accounted for the logic omniscience problem by and large. Simultaneously, there is a axiom that express intention is an angent's awared planning action. Then agent's intention is more rational. Thus we solved the five basic problems in BDI logics.In brief, this research involved a series of five logics, among which those presented in Chapter 2 and 3 served as preparation for the coming logics. Those in Chapter 4, 5, and 6 were logics with general BDI features.