METHOD FOR PROBABILISTIC ESTIMATION OF NOISY MEASUREMENTS

摘要

The method involves dividing a measurement range into discrete values and forming a model of the process producing the measurement signal with discrete states corresponding to discrete measurement region values. Probabilities of remaining in the current state and changing to another state by the next sampling time are allocated. The probabilities are used to re-compute the probability for the present sampling time. The method involves dividing a measurement range into discrete values and forming a model of the process producing the measurement signal with discrete states corresponding to discrete measurement region values. At each sampling time a probability value is associated with each state and at least one state is determined using the probability value. Probabilities of remaining in this state and for changing to another state by the next sampling time are allocated. The probabilities for the current and previous sampling times are used to re-compute the probability for the present sampling time.