Production systems, also called rule-based systems, are an important artificial intelligence (AI) programming language, but it is hard to be applied in real-time field. Based on the analysis of the two major deficiencies (low execution rate and insensitive to asynchronous external events) in production systems for real-time applications, several possible approaches and methods to realizing real-time production system were investigated.
Rule-based systems are most important knowledge representation techniques and have been widely used to build large expert systems, but their deficiencies make them unsuitable for real-time application. The ideal space production has real-time requirements. This paper discusses the design and implementation of a real-time rule-based diagnosis and maintenance system (RTDAMS), which can be used in the automatic GaAs production line on the space station.<>
Based on traditional production systems and the member system model, the paper puts forward a concurrent production system (CUPS). This CUPS is described by outlining its syntax, basic structure of a member, working principle of the interpreter, and compatibility with OPS5. A test example is presented to illustrate that CUPS is a concurrent production system programming language supporting general and cooperative problem solving.<>
Thinking, conflict avoidance and planning results optimization are the key for co-ordinated robot problem solving. The paper presents a knowledge-based, robot co-ordinated problem solver (KBCOPS). It also illustrates the planning power of the KBCOPS system by solving and analyzing some typical planning problems in the simulated double robots world.<>
This paper presents an architecture of an on-line diagnosis system for manufacturing organization management. This diagnosis system consists of five component parts: a model base, a rule base, a diagnosis unit, and two interfaces. The functions of each part in this diagnosis system are described. An example about quality diagnosis in production management is given at the end of this paper.<>