Integrated system for intelligent control (Q685755)

This book describes the concept of intelligent control as an integrated system encompassing automatic control facilities and the theory and applications of artificial intelligence (AI). The author defines two objectives of intelligent control as follows: ``1. enhancing product quality through automation, and 2. improving efficiency through AI techniques.'' Development of an expert system component of an automatic control system is seen not only as a collection of the existing experiences but as a process of exploration of the problem domain in which new expertise can be acquired. There are three types of architectures proposed for the intelligent control system depending on the number of expert systems they use and the level of their integration. Modern approaches used in AI for expert systems development in relation to control systems are discussed. An implementation of an integrated intelligent system for real time control technology (IDSCA) due to the author, \textit{T. Jiang} and \textit{J. J.-P. Tsai} [Int. J. Intell. Syst. 3, No. 4, 361-379 (1988; Zbl 0657.93022)] is proposed and described. The author explains details related to the IDSCA development process, software and architectural requirements, as well as a background research methodology for the project that resulted in the system. The book has 8 main parts introduction and conclusion. The introduction presents a general idea of intelligent control for industrial processes and brings a number of issues related to the development and implementation of such systems. This section concludes with the proposed architectures for the IDSCA system. Part two describes several components of the IDSCA system and illustrates the control system and reasoning process in IDSCA. Part three reviews several expert systems research projects for intelligent support in decision making for optimal control problems that have been investigated recently. The author has been involved in some of these projects. Problems of knowledge modelling and integration are discussed. A hierarchy of expert systems linked by one meta-system is proposed as an intelligent control architecture. Each sub- system from this hierarchy can contain specific knowledge and use different suitable software platforms in order to provide integration of such components by means of meta-knowledge. Several functions and features of meta-systems are listed and discussed. Implementation methods for meta-systems in several software environments are presented in section four. Parts five and six describe several expert systems that were implemented based on the proposed architecture. Section five gives a theoretical description of the problem of intelligent optimal control based on AI approach. Knowledge representation and maintenance issues are discussed in this context. Section six presents examples of pulp and paper process control as an illustration of such applications. Section seven describes architecture and application of intelligent maintenance support systems for air-traffic control. Integrated intelligent system for manufacturing is presented in section eight. The configuration of integrated intelligent manufacturing include intelligent CAD/CAM sub-components for design, planning and programming the manufacturing process. In section nine (`New knowledge generation') the author presents some results from both control engineering and AI techniques that have been derived from the intelligent control systems research. In conclusion, the author gives a brief summary of the principles of intelligent control systems development and puts forward the idea that domain experts after training in AI technology and being provided with appropriate software environment can act as intelligence engineers and develop better expert systems for their needs then professional knowledge engineers. The writing style of the book is easy to follow and the content should be understandable, even to relative novices in the field of expert system. The book provides rich material for discussion and further research.