Automated and risk-based Box Jenkins forecasting model for infrastructure system capital planning (ISCP).

摘要

Infrastructure system capital planning (ISCP) involves complex decision-making processes over long time. This characteristic urges engineers to properly employ forecasting methods for scientific and reliable information about risk & uncertainty of the future. The information, subsequently, supports decision-making. However, due to limitations of current forecasting methods, easy-to-use methods cannot handle complex situations by providing satisfactory forecasts and desired risk analysis function; while such advanced and capable methods as Box Jenkins methodologies are prohibited from large-scale application due to difficult methodology comprehension and implementation.;The author presents a forecasting model based on Box Jenkins methodologies. It simplifies and automates implementation of Box Jenkins modeling and is also capable of risk-based forecasting for risk & uncertainty analysis. Following validation tests using public data verify the model's functionality. A case study on cash flow forecasting for a subway extension project in City of Edmonton further proves the model's usability in industry cases.