HUMAN RELIABILITY WITH PROBABILISTIC LEARNING IN DISCRETE AND CONTINUOUS TASKS - CONCEPTUALIZATION AND MODELING

This paper presents human reliability models in discrete and continuous tasks based on a nonstationary Bernoulli process and a nonhomogeneous Poisson process when probabilistic learning takes place. The learning process is modeled to be compatible with the general learning curve theory. Expressions are given for several human performance measures. Parameter estimation procedures are derived.