Numerical simulation of backward flow of methane roof layer in an airway

We study methane backward flow in the laneways at various wind velocities and angles based on computational fluid dynamics theory; analyze the backward flow area length, the concentration distribution and the methane roof layer thickness. The results show that the methane roof layer would move against the airflow in a declensional ventilated airway if the downward air velocity is not large enough. The faster the airflow speed, the smaller the methane backward flow area becomes, and the methane roof layer becomes thicker and shorter. In a declensional ventilated airway, the methane/air mixing capacity increases with the increase of the inclination of the airway, and the thickness of methane roof layer decreases with the increase of the inclination of the airway.     

Settlement Analysis of a Laboratory Model Test of Dewatering of a Metro Station in Shanghai

With the exploitation of urban underground spaces, deep foundation pit constructions increasingly are evaluated. Pits in soft soil areas adjacent to operating subway are especially difficult to excavate. After indoor testing of water drainage for a station on the Shanghai Metro line 9, and on the basis of treatment and analysis of water level drain hole and flux monitoring data, pore water pressure, layered settlement and surface soil settlement, we summarize the relationship between factors influencing settlement during water drainage. This research offers design experience and blueprints for water drainage and ensures the safety of metro pit construction and enclosure structure stability.     

Application of project cost forecasting with data mining and neural network technologies for power engineering

A model based on data mining technology and neural network theory was put forward to forecast and censor complicated engineering problems. Data mining technology was applied to normalize the real data, feature selection and clustering, which generates fuzzy rules. A forecasting model of engineering cost was constructed by using an improved back propagation(BP) neural network fuzzy system. A good calculation result was obtained by analyzing the history project training and example samples of a city. The result demonstrates the accuracy and stringency of this method and the validity and practicability of this model for forecasting and censoring complicated engineering problems.     

A control strategy for braking energy distribution and regenerative braking in a 4WD hybrid electric vehicle

Based on the braking theory of conventional vehicles and high efficient energy recovery and security braking, a control strategy for braking energy distribution and regenerative braking of a 4WD hybrid electric vehicle was proposed. Considering limiting factors, such as motor and battery efficiency, a simulation model of a regenerative braking system was built and a simulation of typical braking conditions was conducted. The simulation result shows that the highest energy recycle ratio can reach 87.5%, 47.8％, and 28.6％ when z is 0.1, 0.3, and 0.5, respectively, at 30 km/h. The control strategy for braking energy distribution and regenerative braking can satisfy demand for vehicle braking and obtain high regenerative energy.