A Research On Schedule Plan And Control Of Software Development Project

To meet the need of the larger and larger scale in the software development and more and more people joining in this career, which results from the rapid development of info-tech, software project managing has widely been introduced into software companies. Thus, it is necessary to work out an effective way of managing software project. It is known that the vital elements in managing software project must be planning and controlling of the project. The present dissertation is an attempt of a systematic study on the approaches adopted in the process of controlling the plan of a software project, which might offer a practical reference to software companies.The prediction of IT project scheduling as well as the scheduling in parallel is carried out by adopting the theory of Dynamic Programming and the method of network planning and making use of the model of the Markov Chain in this thesis. We have studied the application of schedule prediction of IT project with the actual research data on software development projects, establishing Markov's dynamic predict model of parallel IT project schedule.In this paper, firstly, I introduce the background and basis, purpose and significance, content and framework of this paper. Secondly, based on the achievements made in the fields of project managing, software engineering and software project managing, managing of plan and controlling in the software development project is dealt with in this dissertation, and especially the estimating of workload, assignment of the task, and three techniques most taken (Gantt, PERT and CPM) in plan-making. Then I introduce four common techniques, milestone progress, activities progress decided by people, work modules progress, earned value method.Based on the knowledge mentioned above, this paper introduce the Basic principles of Markov, the stage experience and method of IT project development. Then we studied the concurrent engineering of IT project development and predict the schedule of IT project with the methods of the Markov Chain and DELPHI. Based on the above analyses, a project of OA system is taken to testify the feasibility and validity of this frame base on Markov Chain.Finally, I give the conclusion of the whole paper. Markov Chain prediction is based on analysis of previous statistics. With the help of detailed and accurate historical data series analysis, the present model can give satisfactory forecasting. However, as a probability forecasting model of scheduling, the accuracy and reliability of the present model should be verified through a huge amount of data forecasting and testing. The Markov Chain in this paper is used to predict within a certain period; however, it can be expanded into an unlimited one, which has more feasible value in life cycle.Its innovation process lies in the combination of the Markov Chain and technology of the network planning, giving a concise analysis and prediction of schedule.