The Use of Simulated Environments in Developing Critical Systems

摘要

A major problem arises when developing high integrity and safety critical systems: it is usually undesirable or impractical to test them within their operating environment, which means that the simulation assumes increased importance. In any application, the accuracy with which a simulated environment models its real counterpart determines its effectiveness in testing. For high integrity systems, the quality of a simulated environment becomes a key factor because the results of testing with the simulator are often used as the basis for acceptance or certification The guidelines and standards that regulate the development of high integrity systems stipulate very little about how the simulators that are used to test them should be produced or validated. Similarly, most of the life cycle models currently in widespread use place little emphasis of this aspect of a project. This paper argues that increased importance should be placed upon the development and validation of simulators, and in particular that the choice of - life cycle model used must encompass environment simulation. It describes life cycle models that emphasise the development of the environment simulators that are used in the testing of safety critical systems and the effect of ihese models upon the management of projects involving the development of high integrity systems. It is believed that this will encourage a more systematic approach to the production of simulators and allow recommendations to be made concerning the various phases of their development in line with their required integrity level. The work described in this paper forms part of CONTESSE, a collaborative project undertaken within the United Kingdom by a consortium of universities and major companies representing several industrial areas including defence, nuclear, aerospace, automotive, and process control.