The problem of improving the methodological and algorithmic support of the complexing process by developing models of adaptive redundant complexing of weighted interval data has been solved. The object of this study is the process of complexing interval data obtained from several independent sources; the subject is the algebraic methods of excessive complexing of weighted interval data. The relevance of the task is due to the severity of the problem of consolidating homogeneous data in order to obtain more accurate and relevant information about the object or process under study. Models have been developed that, unlike the existing ones, make it possible a posteriori to take into account the accuracy of experts at the preliminary stage of expert evaluation. A single analytical form of the model for processing weighted interval and point estimates with the possibility of structural and parametric tuning is proposed. It allows one to increase the degree of automation of processing expert assessments under conditions of interval uncertainty. Recommendations for the practical application of the proposed models have been formulated. Options for parametric configuration of preference functions were indicated depending on the characteristics of weighted interval estimates. The commonality of the limiting cases of the proposed models with previously known ones is proved. The example shows the shift of the integrated assessment to the side of more accurate assessments at the previous stage of source assessment. The adaptability of the proposed models is illustrated. At the same time, a slight, on average, about 10 %, expansion of the complexed interval relative to the primary ones was registered. The built models and algorithms can be used in automated expert systems, as well as in cascade models of information processing and compression.