Life cycle cost analysis of maintenance costs and budgets for university buildings is essential. This paper briefly describes the meaning and application of the standard of maintenance costs and budgets, and addresses the meaning of maintenance cost and budgeting with regard to engineering. To examine this issue, a case study on the operations maintenance phase of four university buildings on the campus of National Taiwan University is presented. Using historical data on maintenance and repair over a 42-year period, a cost prediction model using the life-cycle cost (LCC) was determined using three different methods: (1) simple linear regression (SLR); (2) multiple regression (MR); and finally (3) a back propagation artificial neural network (BPN). The research results showed that the BPN model had good estimation ability. This paper implemented the BPN model in a case study to analyze the problems of maintenance costs and budgeting for university buildings. The study helps to set a legitimate standard for predicting repair maintenance costs, and proposes a plan and standard for the repair maintenance strategy of structures. The results of this study are provided to show how to establish a cost prediction model of maintenance and how these university buildings can be used to obtain the optimal life cycle maintenance scenario.

No abstract