Study Of Project Duration And Cost Monitoring Based On Statistical Process Control

With the rapid development of China’s economic construction, the construction project becomes more and more complex and huge, the schedule and cost goal of the project is more difficult to achieve, once out of control will result in huge losses, however, the problems of schedule delay and cost overrun are ubiquitous in the engineering project, therefore, effective project schedule and cost monitoring will become the key to the success of the project. We adopted the earned value method to monitor the project schedule and cost in the past, the earned value method can reflect the current performance of the project, the schedule and cost deviate from the plan whether or not. Usually we think that the project is under control when the schedule performance index SPI and the cost performance index CPI is1.But in the process of project construction, index of schedule and cost will produce certain fluctuation due to some reasons, the traditional method don’t reflect the acceptable level of the deviation. Statistical process control technique can analyze and control process by using the statistical regularity of process fluctuation, can reflect the level of the project implementation process being in acceptable and stable state in the statistical significance.This paper studies project schedule and cost monitoring from the perspective of statistical process control, the following aspects are studied:First, this paper introduce that two fluctuation types which is certainly ubiquitous in the process and causes of fluctuation, three kinds of main control chart in statistical process control theory, and how to use control chart identify fluctuations in process. From the aspects of calculation method and the applicable range, several kinds of process capability index for judging process capability in the stable controlled process is discussed. Second, the traditional earned value method can’t detect the small variation in the process of monitoring the project schedule and cost, and the small variation may include the early warning information of the project current or future state, project managers need to take corrective measures.Especially,compared with the ordinary project, the small variation(still around1) is more important in major project, can make a huge impact to the project. However, traditional methods of evaluating and monitoring the project can’t distinguish this variation, thus provide early warning information. In this paper, individual moving range (ImR) control chart is used to monitor project schedule and cost performance index, the control limit of ImR control chart is deduced, the method which transforms non-normal distribution data to the normal distribution data is discussed, the example verifies that ImR control chart can provide timely warning when the project process is out of control, and provide decision-making information for project manager.Third, the traditional Shewhart control chart in monitoring the autocorrelated process appear a large number of false alarms, and the time series model can make serially autocorrelated data to be independent, thus the autoregressive moving average model ARMA(2,1) is used to fit autocorrelated process data, the residuals of process data are used to establish control chart. The controlled process and the uncontrolled process is studied by numerical simulation, we analyse the control effect of the conventional individual control chart and the residual control chart based on ARMA (2,1) model in two kinds of simulation process. Finally, a practical autocorrelated process is used to verify the utility of the residual control chart based on ARMA (2,1) model.Fourth, a specific project example is used to illustrate the application of statistical process control technique during the project monitoring.This paper introduces the statistical process control technique into the project schedule and cost monitoring process, and discusses the monitoring method in the autocorrelated process. Research findings make up for the lack of the earned value method and improve the monitoring level of the project schedule and cost, and has certain reference value in the actual project management.