Abstarct: The stuck pipe is one of most important problems that it is costly and time consuming in drilling. If problem not fix, cause financial damage. There isn’t an exact method to detect the stuck pipe. Studies show that the parameters and structure of the ground during drill operation is effective in the stuck pipe. One of these parameters is torque that generated by the rotation of the drill string. Engineers monitor drilling parameters to control the operation and diagnostics deals moment stuck pipe. This approach has many weaknesses with regard to the possibility to human mistake and lack sufficient precision. Experience has shown that the torque is suddenly rises, in the moment of stuck and drill bit rotates (rotation per minute RPM) get more slowly. Thus, by detecting torque and rotation per minute spikes, the stuck moment can be detected. Various methods have been introduced to spike detection .in the proposed method, we use smooth nonlinear energy operator. In this method, signal is detected positive and negative changes. In this study , introduced a method for the detection of positive spike for torque and negative spike for RPM. CA and ER measures is used to present and simulation results indicated that proposed method, compared to existing methods. In addition, it has been shown that the proposed method needs little time to analyze a signal, So it can be used in real time drilling systems. In the second part of this study for prediction stuck pipe, a new method introduce to identify various ground boundary laxyer, as high risk areas in the stuck pipe. In this method, torque signal are segmented as an effective parameters on this topic. There are two methods for segmentation, fixed and adaptive. Segmentation methods have proved poor results. In this study , a new method for segmentation , adaptive wavelet transform and fractal dimension is used. It has been shown that the boundaries between the signal segmentation, with good accuracy stone boundary laxyer defines the well path. This research can be as a step to optimizing intelligent drilling.