对于传统的灰色模型,其中的模型参数a和b是通过最小二乘法获得,但a和b可能不是最优的。该文用改进的模拟退火算法代替最小二乘法求模型参数a和b,保证了最后能搜索到全局最优解,列出最优的灰色模型。经过实例检验,改进的模拟退火-灰色模型的相对误差比传统灰色模型小的多,提高了预测精度。 For the traditional gray model, the model parameters a and b are obtained by the least square method, but a and b may not be optimal. In this paper, we use the improved simulated annealing algorithm instead of the least square method to find the model parameters a and b, which ensures that we can search the global optimal solution finally and list the optimal gray model. After an example test, the relative error of the improved simulated annealing-gray model is much smaller than the traditional gray model, which improves the prediction accuracy.