Time-range reasoning for microprocessor systems diagnosis: a probabilistic extension

The concept of time range has been recently proposed to capture the notion of time impreciseness in event occurrence, which is found to be of fundamental importance for temporal reasoning in the domain of microprocessor systems design and diagnosis. A possibility measure is implicitly implied in the time-range representation, although it is not properly quantified for effective temporal reasoning. To overcome this shortcoming, time-range probability models based on truncated Gaussian distribution are proposed as viable possibility measures. The use of simplistic assumptions on these probability models has been shown to enhance the deductive capability of existing time-range constraint models. Difficulties in constraint satisfaction and propagation of the probability models are also discussed.