An in - depth study on lattice distortion of a chemical sensor composed of calcium doped lead titanate (CaxPb1 -xTiO3 ) thin film is present.The micro- structural lattice distortion (defined as tetragonality,δ) is crucial for sensor's sensitivity and response time.To evaluation and predict the tetragonality,a three -layer artificial neural network (ANN) model is applied,based on experimental results related to the tetragonality values of the thin films determined via X - ray powder diffraction.The fabrication parameters,including heat- treatment temperature,dopant content,and heating rate have been considered as the input parameters,whereas the tetragonality as the output parameter.Function approximation was employed and simulation was implemented on the MatlabTM.The predicted results were compared with experimental results and it was found out that the results obtained from ANN model were accurate in predicting the nano - structural distortion of the thin film.The results showed that ANN is an effective tool in the simulation and prediction of thin - film tetragonality and highly useful compared with traditional trial - error experimental processes.Since predicted data by ANN model were essentially identical to the experimental results,this model can be used to estimate the tetragonality of different thin films for humidity sensors.%利用人工神经网络对化学传感器的晶格畸变进行了深入研究.实验结果表明:传感原件钙掺杂钛酸铅薄膜材料随合成参数的变化而呈现不同的晶格畸变度,该参数通过X射线粉末衍射精细结构确定,并作为人工神经网络的输出变量；合成参数作为输入变量,以MatlabTM作为操作平台,利用三层识别方法对传感元件的纳米结构进行评估与预测.实验结果与预测结果十分吻合,因此神经网络与传统实验方法结合,可以对传感原器件的性能进行准确快捷评估.
展开▼
