论文部分内容阅读
为了实现对水力旋流器的全面设计,建立了3层BP神经网络模型,该模型可根据分离粒度、生产能力、底流质量浓度等值,选择合适的水力旋流器。经10组数据测试,选型误差为:底流口直径10.43%,溢流口直径7.51%,插入深度17.86%,入料压力20.24%,选型精度高于传统方法。该模型既可用于设备选型,也可用于优化旋流器参数。选择合适的水力旋流器分级加重质,制备得到的粗、细两产品分别满足湿法、干法对加重质要求,对我国选煤业发展有重大意义。
In order to realize the full design of hydrocyclone, a 3-layer BP neural network model is established, which can select the appropriate hydrocyclone based on the separation granularity, production capacity, and the equivalent value of the underflow mass concentration. After 10 sets of data test, the selection error is as follows: the diameter of underflow port is 10.43%, the diameter of overflow port is 7.51%, the insertion depth is 17.86% and the feeding pressure is 20.24%. The selection accuracy is higher than the traditional method. The model can be used both for equipment selection and for optimizing cyclone parameters. Select the appropriate hydrocyclone grade grazing quality, preparation of the crude product and fine product to meet the wet, dry weight requirements for the increase of China’s coal preparation industry is of great significance.