Method for forecasting execution time of software process

Abstract

The invention belongs to the field of simulating and modeling software process, and in particular discloses a method forecasting execution time of a software process by building a software process simulation model for considering random factors and simulating the software process. The method comprises the following steps: transforming a general software process workflow graph into an s-TRISO / ML random process model; determining the value of a random variable in the process model through historical data of the productivity of assigned project personnel, the complexity coefficient of activity type and the estimated range of activity scale; inputting the process model assigned through the random variable into a software process simulation tool for repeated simulation; analyzing simulation results; and obtaining the execution time of the same activity in the software process, as well as derivative indexes thereof. The method has the advantages that the method transforms the flow graph into the random process model so as to simplify the complexity of modeling the process, can support the simulation of the software process, is more convenient and efficient than the prior process-model static analysis method, and improves the accuracy and practicability of software process analysis.

Description

technical field [0001] The invention belongs to the field of software process simulation and modeling, in particular to a method for predicting the execution time of a software process by establishing a software process simulation model and software process simulation considering random factors. Background technique [0002] "Software crisis", a concept that began in the 1960s, has always been accompanied by the entire historical development process of the software industry. Over the past few decades, software organizations have often faced software project overruns and delays, and the quality of software products has often failed to meet customer expectations. Many studies from industry and academia have attempted to address this issue. In these attempts, research based on software process technology, such as: software process modeling, software process simulation, etc., are very effective methods. [0003] At present, people have generally recognized that the quality of ...

AI Technical Summary

Problems solved by technology

The weakness of this method lies in: first, this method is relatively complicated, and the calibration process of parameters requires a large amount of historical data, which are generally not easy to collect and analyze; to use this method for progress prediction, the software organization needs to prepare early. Constantly calibrate model parameters

For general small and medium-sized software companies, the cost of accumulating process historical data is very high, so their accumulation of software process historical data is often not enough to support the calibration of the parameters in the model, so this method is not applicable in such software organizations

Second, this method only estimates the process progress based on the process workload and personnel work efficiency, without considering the structure of the software process. If the process structure is unreasonable, it will often affect the work efficiency of the process executives and cause problems such as delays in work. greatly affects the accuracy of the method in predicting the progress of the process

Third, this method does not consider the impact of software reuse, COTS software, management strategies, etc.

Based on the above points, COCOMO's prediction of the progress of the software process is not practical and accurate enough

Images