The KulmanJilter is a very eficient optimal$ltel; however it has the precondition that the noises of the process and of the measurement are Gaussian. In this paper we introduce 'The General Distribution Filter' which is an optimal jilter that can be used even where the distributions are not Gaussian. An eficient practical implementation of theJilter is possible where the distributions are discrete and compact or can be approximated as such. The problem is that when the measurement is not a

more »

... ian distribution, the Kalman filter cannot be used. This was the motivation to develop the general distribution filter which can be used for any distribution function. As we will see, although the filter is defined for any probability distribution of the measurement, an efficient computer implementation of this filter is possible when the probability distribution of the measurement is discrete or can be approximated as such.