Abstract: It is difficult to use the current methods to make predictions.A novel method based on grey theory is proposed.The design thought，key technique and experiment are given.This method gets the effort sequence changing feature-“changing ratio” from the completed stage effort sequences，and gets the“changing ratio threshold” from historical projects by machine learning methods，then uses grey models to make predictions.The experimental results show that，compared with linear regression method，the prediction accuracy of the proposed method has been improved by 10%~50%.

摘要： 针对软件成本估算困难的问题，提出一种基于灰色理论的估算方法。给出了设计思想、关键技术以及估算的过程，并进行了实验验证。该方法主要借鉴灰色理论可以基于小样本和贫信息的特性，结合GM（1，1）和Verhulst各自的优点，利用成本序列及其变化率动态构造灰色理论模型进行软件成本估算。实验结果表明，该方法平均估算误差比线性回归方法低10%~50%。