University dormitory is not only a place for college students to learn cultural knowledge, but also a place for college students to have leisure and entertainment. Reasonable allocation of dormitory for college students can make college students quickly integrate into college life and allow students to have a happy college life. However, the traditional college student dormitory allocation system has a slow allocation rate, which increases the burden on managers and cannot be comprehensively analyse students according to their personality characteristics. In this study, the data fusion optimisation algorithm was applied to the intelligent distribution system of student dormitories in colleges to effectively extract students' personality characteristics and allocate suitable dormitories for students. By conducting experimental tests on students in different schools in four aspects: allocation rate, recall rate of student characteristics, allocation effect, and student satisfaction, it was found that the allocation rate was improved after using the data fusion optimisation algorithm. The recall rate of student characteristics increased by 5.01%, the distribution effect increased by 6.44%, and the student satisfaction increased by 1.368 points. The data fusion optimisation algorithm has an optimisation effect on the intelligent distribution system of student dormitory in colleges.