Evaluation and optimization of the man-machine interface in nuclear power plants using the HEROS procedure in PSA

For the probabilistic evaluation and optimization of the man-machine-system a new procedure is developed. This and the resulting expert system, which is based on the fuzzy set theory, is called HEROS, an acronym for Human Error Rate Assessment and Optimizing System. There are several existing procedures for the probabilistic evaluation of human errors, which have several disadvantages. However, in all of these procedures fuzzy verbal expressions are often used for the evaluation of human factors, also in the best known procedures. This use of verbal expressions for describing performance-shaping factors, enabling the evaluation of human factors is the basic approach for HEROS. A model of a modem man-machine-system is the basis of the procedure. Results from ergonomic studies are used to establish a rule-based expert system. HEROS simplifies the importance analysis for the evaluation of human factors, which is necessary for the optimization of the man-machine-system. It can be used in all areas of probabilistic safety assessment. The application of HEROS in the scope of accident management procedures and the comparison with the results of other procedures as an example for the usefulness and substantially more extensive applicability of this new procedure will be shown. (author)