Applying the PCR6 Rule of combination in real time classification systems

Abstract

Modern navies face the introduction of new ships with reduced and less trained manning and new, more complex, sensor systems. These trends could result in operator overload which might be solved by introducing a higher level of automation. In case of classification, a sub process of picture compilation, the problem of conflicting sensor data can be solved by proportional redistribution of conflict according to the PCR6 algorithm. This method is computationally demanding to such an extend that the applicability is limited in practice. The methodology in this paper adapts the PCR6 algorithm by simplification of the solution space of the algorithm by preprocessing the input data. Comparison tests show that this simplification reduces computing time by at least a factor hundred while the effects on the output are limited. We conclude that by applying preprocessing the pcr6 algorithm can be applied in real time classifications systems using common computer hardware.