摘要: Neutron multiplicity counting is a non-destructive, passive technique for monitoring plutonium invento-ry. However, when extending the application from plutonium metal/plutonium oxides to large-volume plutonium solution systems, the original “point model” reveals significant shortcomings. Currently, while two types of improvements for original “point model” has been proposed:(a) improved “point model” suitable for small-volume solution systems and (b) volume-weighted “point model” correcting for spatial dependence of solid systems, neither is suitable for large-volume solution systems. Based on the im-proved “point model”, we firstly employ the volume-weighted approach to derive a volume-weighted model suitable for solution systems, which however neglects the disparity between the induced fission source distribution and the initial source distribution in our opinion. Furthermore, by additionally incor-porating the distribution of induced fission reactions as a weighting coefficient for the spatial dependence correction factor, we propose the composite-weighted model. Comparative analysis of simulation results from the improved “point model”, volume-weighted model, and composite-weighted model demonstrates that the composite-weighted model has the best performance, offering superior universality and accuracy, thereby confirming the necessity and validity of the improvements. Theoretically, the methodology for addressing spatial dependence in large-volume solution systems introduced in this study can also be ex-tended to other large-volume plutonium-containing material systems.