Zhang Huizhen

The Study and Application of Gray-Orthogonal Wavelet Network Forecasting Mode on Cementing Quality

The interrelated characteristics and the complexity of cementation quality affecting factors are considered, a new method of forecasting cementing quality based on Gray-Orthogonal wavelet network is set up. The method regards some factors affecting the cementing quality as the input parameters of the model, and regards quantifying cementation quality as the output parameter of the model. Evaluating and forecasting the cementing quality of Bei1 block of 20 wells in Daqing Oilfield. The maximum fractional error between forecasting results and actual results is 2.66%. The associated problem of impacting factors is solved by the new forecasting method，and the prediction accuracy of cementing quality is improved. A reliable theoretical basis of cementing operation parameters design and construction program adjustment for new wells in oil field is provided. Key words : Cement job quality; Gray system; Orthogonal wavelet neural network

Fold Catastrophe Model of Fracture Propagation of Hydraulic Fracturing

According to energy conservation from the destruction of rock catastrophe, a new calculation method of the length of fracture propagation in hydraulic fracturing is proposed, and assuming the crack extends to approximate ellipse, the width calculation model of fracture extension is established. By using this new method in Changsheng well of Jilin oilfield, And the maximum error between the productivity that calculated by using the new method and the actual oil production is anastomosing. the method is an innovation in parameter calculation of crack propagation with the destruction of rock catastrophe, and it supposes some new theoretical foundation for the design of fracture parameter optimization. Key words : Hydraulic fracture; Fold catastrophe; Fracture parameters

Robust fault-tolerant control for uncertain discrete system with time-delay

This paper focuses on the problem of robust fault-tolerant control for uncertain linear discrete systems with time-delay. On the basis of Lyapunov theory, it is found that a controller based on a kind of memory feed-back control law can keep the uncertainty systempsilas integrity against the sensor failures, and make use of the LMI toolbox in MATLAB, we can get the parameter of the controller. The simulation results are given to demonstrate the effectiveness of the proposed method.