| Summary: | 碩士 === 國立臺灣科技大學 === 營建工程系 === 99 === Forecasting project cash-flow greatly impacts on project cost management. A financial plan according to the S-Curve should be executed to modify progress schedules and increase project success. This study used an inference engine, Evolutionary Gaussian Process Inference Model (EGPIM), which fuses three Artificial Intelligence (AI) approaches together, that are Gaussian Process (GP) Bayesian inference and Particle Swarm Optimization (PSO). The EGPIM was the core to forecast the desired cash-flow quantity with three front cash-flow quantities basing on clustered historical data. With the developed project cash-flow forecasting model, the trend of cash-flow could be predicted. Comparing the results with the S-Curve, the project cash-flow can be managed and strategies can be yield. The strategies will be planed according to real-world factors such as float, progress execution time, construction rates, resource demands, and so on, to optimize the current financial plan with the concept of further cash-flow management. Managers can use the predicted cash-flow to handle project progress.
With cases studies, this study demonstrated the developed project cash-flow forecasting model which can offer accurate prediction results immediately. For project cash-flow management, the financial demands can be resulted as references for project managers. That assists managers to modify progress schedules and projects success under original budgets.
|
|---|
