In this paper,a finite element method is proposed to investigate multiple solutions of the Navier-Stokes equations for an unsteady,laminar,incompressible flow in a porous expanding channel.Dual or triple solutions for the fixed values of the wall suction Reynolds number R and the expansion ratioαare obtained numerically.The computed multiple solutions for the symmetric flow are validated by comparing them with approximate analytic solutions obtained by the similarity transformation and homotopy analysis method.Unlike previous works,our method deals with the Navier-Stokes equations directly and thus has no similarity and other restrictions as in previous works.Finally we use the method to study multiple solutions for three cases of the asymmetric flow(which has not been studied before using the similarity-type techniques).
为了提高产品质量的稳定性和可靠性,利用机器学习方法实现产品质量在线监控、在线优化和在线预设定,是钢铁企业目前亟待解决的关键技术.针对企业需求,提出基于软超球体算法的产品质量异常在线识别和异常原因诊断方法、基于流形学习的工艺参数在线优化方法和基于多变量统计过程控制的工艺规范制定方法.通过将上述方法进行系统集成,并利用工业互联网技术和大数据分析方法,研发了产品质量在线智能监控系统.目前该系统已在钢铁企业十余条生产线上推广应用,质量在线判定的准确率达到99.2%,在线检测时间不到0.1 s.
