中文摘要：

采用神经网络的CA模型对古建筑物理性质变化进行预测．应用X光探测采集的古建筑木质构件的灰度图，使古建筑木质构件的灰度图中的每个像素点与神经网络的CA模型中的每个元胞一一对应，运用CA模型中“灰度”的概念通过神经网络的训练学习计算出每个元胞的灰度值的变化，从而得出古建筑木质构件随时间推移而受损的情况．通过实例得到了经过预测后木质材料随时间推移而受损情况的图片．

英文摘要：

Using the cellular automata （CA） model of neural network to forecast the changes of physical property of historic buildings makes the result of forecasting physical property of historic buildings conform with changes of historic buildings when the time goes. Using the X ray to scan the wooden unit of historic buildings and collect the gray intensity image of wooden unit of historic buildings make the every single pixel of gray intensity image corresponding with the cell of CA model. Based on the CA model and neural network training system,it will calculate the change of every ingle pixel of gray intensity image corresponding with the cell of CA model; thus get damaged circumstance of wooden unit of historic buildings with the passage of time. According to an actual example, it will acquire the damaged circumstance of picture from the predicted wooden uint with the passage of time.