CAJ | 학술논문

基于BP人工神经网络，研究高密度聚乙烯(PE–HD)、低密度聚乙烯（PE–LD）材料的振动挤出加工过程，建立螺杆转速、振动及口模诸因子与挤出胀大、压力和功率等输出特性参数间精确、高效、简洁的非线性映射关系，为挤出加工参数的优化和挤出加工质量的控制提供依据。结果证明，基于BP神经网络模型能够很好地预测聚合物挤出加工特性参数，所建网络具有精确、高速、自适应等特点。
기우BP인공신경망락，연구고밀도취을희(PE–HD)、저밀도취을희（PE–LD）재료적진동제출가공과정，건립라간전속、진동급구모제인자여제출창대、압력화공솔등수출특성삼수간정학、고효、간길적비선성영사관계，위제출가공삼수적우화화제출가공질량적공제제공의거。결과증명，기우BP신경망락모형능구흔호지예측취합물제출가공특성삼수，소건망락구유정학、고속、자괄응등특점。
Based on BP neural network,the vibration extrusion process of high density polyethylene(PE–HD) and low density polyethylene(PE–LD) were studied. A high efficiency,simple and precise nonlinear mapping relationship were established,which were about the screw speed,vibration and die as the input factor and the extrusion swell,pressure and power as the output factor relations. Provide the basis for the optimization of extrusion process parameters and controlling the quality of extrusion processing. The experimental results show that the BP neural network model can well predict the processing properties of polymer extrusion parameters. The neural network model has the advantages of high precision,high speed,adaptive characteristics.