Application of functional modelling in the design of industrial control systems

Industrial control systems are gradually becoming more and more complex putting still stronger demands on their design. Difficulties in meeting these requirements have led to increased costs in production of control systems, and to systems that can be hard to understand and maintain and even may fail in satisfying desired reliability figures. One step towards a minimisation of these problems is to look for design methods which can address more effectively the special problems involved in control systems design. The paper presents a new design method called Goal Function Modelling (GFM), developed particularly for (reuse-based) design of complex industrial control systems. Founded upon ideas from the domain of functional modelling, GFM has a strong concern for the concept of function, and for identification and organisation of different kinds of knowledge. Combined with a careful attention to practical issues, these capabilities have been embedded in a design model providing an effective framework for organisation of design descriptions for complex systems. The features include support for separation of knowledge associated to different system properties, and for building links between related pieces of information. Both are essential for production of design descriptions which are relatively easy to understand, verify, and maintain, and therefore should have the potential for bringing about control systems with reduced life-cycle costs and increased quality.