Wavelet Neural Network (WNN) is a new form of neural network combined with the wavelet theory and artificial neural network. The wavelet neural network model based on Morlet wavelet and the corresponding learning algorithm were studied in this paper. And through learning the wavelet neural network model is applied to all kinds of engineering examples, it proved that the wavelet neural network prediction model which has a more flexible and efficient function approximation ability and strong fault tolerance, and with high predicting precision.

The realization of strain measurement based on digital image processing technique is discussed. Combined with the strain measurement principle and the optical imaging feature, considered the axisymmetrical and non-axisymmetrical distortion, the automatic surface-fitting calibration method based on region growing is adopted. Cubic polynomial is applied to establish the relationship between the object space and pixel coordinate system. The expectations based on wavelet coefficients are used for sub-pixel edge detection, measurement accuracy can reach 0.02 pixels. Accuracy of displacement for strain measurement achieves micrometer-level. The theoretic validity and practical feasibility of the system are proved by the strain measurement of a cantilever steel beam.

Under the control of the tow-Yun fracture, the geological conditions are complicated and the rock strata is broken in Shiyan area ,in the northwest of Hubei province, and the rock mass is weathered into fragment. When Highway is built in this area, a lot of tunnel portal slope and landslide problems will be met. To properly solve these problems, it is needed to analyze the slope stability and to propose appropriate management measures. In the paper, taken a highway tunnel portal slope in Shiyan area for example, a detailed research was made on the mechanism of landslide formation, and using laboratory soil test results and the simple Janbu method, the local and global slope stability is calculated. The results show that the slope is stable in natural state, and in heavy rains the stability coefficient is reduced quickly and the slope will be sliding. So the slope needs to manage and the rain water needs to control.