The purpose of this study is to provide a correct and timely diagnosis mechanism of shearer failures by knowledge acquisition through a fuzzy inference system which could approximate expert experience. Concerning a question of uncertain knowledge expression and reasoning in shearer malfunction, the fuzzy inference theory is used in shearer malfunction fault diagnosis. The fuzzy relation matrix of faults and signs is deduced based on deep research of failure mechanism and expert experience, which agrees with fault and fault symptoms non one-to-one relationship and human thinking. Fault characteristic parameter is calculated to corresponding subordinate degree, then is operated with fuzzy relation matrix and get fault fuzzy vector. Finally, the shearer malfunction fault is diagnosed according to certain diagnosis principle. The example proves that the method has less calculation, explicit conclusion and other merits.

Numerical simulation of the three-dimensional steady and unsteady turbulent flow in the whole flow field of a multi-blade centrifugal fan is performed. Unstructured grids is used to discrete the computational domain. Pressure boundary conditions are specified to the inlet and the outlet. The SIMPLE algorithm in conjunction with the RNG k-ε turbulent model is used to solve the three-dimensional Navier-Stokes equations. The moving reference frame is adopted to transfer data between the interfaces of the rotating field and the stationary field. Based on the calculation of the inner-flow in the fan, the pressure pulsation of some important monitoring points and the aerodynamic noise distribution, banding together experiment data were farther analyzed The simulation results are of important significance to the optimal design and noise control of the fan.

The full flow field model of a widely used multi-blade centrifugal was built, using the CFD method, the steady and unsteady numerical simulation of the inner flow in the fan at different working conditions are presented. The numerical simulation results were validated by contrasting to the experiment results. The results displayed the characteristics of the velocity field, pressure field and pressure fluctuate in the centrifugal fan. The results can provide basis for optimizing the fan design and the internal flow, and have important value of engineering applications in the increase of the overall performance in operation.

Air-assist Boom Spraying is a crucial technique and spray method of plant protection which has a significant effect on improving the efficiency of pesticide application and reducing droplet drift. Spraying drift is the main reason for the low utilization rate of pesticides, at the same time, the interaction of natural wind and the air bag wind is the main cause of spray drift. But at present, the regularity of spray droplet deposition and motion trajectory under the common influence of natural wind and air bag wind is not yet clear. The kinematics characteristics of the droplets deposition of air-assist boom spraying in the plants is investigated with numerical simulation method. The Phase Doppler Particle Analyzer (PDPA) is used to validate the numerical simulation results and it indicates that the deposition process can be well reproduced. The results show that the bigger droplets (>150 μm) get more kinetic energy from the vertical airflow and locate in windward side. These droplets are not sensitive to the horizontal airflow and are easier to deposit within a short time. The droplets with smaller size (<150 μm) are minimally influenced by the vertical airflow and suspend in the upside of the flow field. The vertical airflow not only prevents the horizontal airflow drift the droplets, but also introduces a vortex region near the canopy of the plants, which rotates the plant canopy and increases the droplet deposition on backside of the leaves. This study improves the understanding of the droplets deposition under the multi-wind fields and provides the theoretical basis for the innovation of the air-assist boom spraying system.

Expert system of full-mechanized face equipment fitting based on artificial neural network is researched and processing method is presented in this paper. On the basis of this, BP neural network models for the forecast of production capability of mining face and the parameters of equipment fitting of full-mechanized mining face are built up. Employing these models forecasts the output, work efficiency and main technical parameters of full-mechanized equipment of the test mining face in Ji'er Colliery.