Management of software development project has been the focus of attention recently because of the enormous penalty costs incurred due to poor management. Though the traditional software life cycle concept in project management which deals with problems in each phase separately is invaluable, it does not adequately capture evolutionary or iterative system development mechanism.
A dynamic model of software development project management is developed in this thesis, which shows the dynamic behaviors of key factors during the project life cycle. It also demonstrates various aspects of project behaviors due to different management policies. Simulation experiments were conducted to investigate how the characteristics of software system affect the manpower and time schedule. The results show that the project duration and manpower has trade-off relationship and that the duration curve has increasing slope with decreasing rate as project size increases owing to the effect of embryonic stage. Moreover, experiments demonstrate accurate size estimation is vital to guarantee the stability of manpower.