A new constitutive model of a magneto-rheological fluid actuator using an extreme learning machine method

Recent progress on the simulation technology of magnetic fluid

“…Bahiuddin等人 [19] 络为代表的"黑盒子"型机器学习无法给出本构模型 的 显 式 表 达 式 . 基 于 进 化 算 法 (evolutionary algorithm)的机器学习可以得到显示表达式 [71] , 但其形式 会因具体问题而异, 本构模型不具有普适性.…”

“…图 6 (网络版彩色)磁性液体数值模拟的典型结果. (a) Monte Carlo模拟的内部结构图 [18] ; (b) 机器学习模拟的流动曲线 [19] Figure 6 (Color online) Typical results of simulations of magnetic fluid. (a) Microstructures obtained from Monte Carlo simulation [18] ; (b) flow curves obtained from machine learning simulation [19] 评 述 表 1 磁性液体常用数值模拟方法的适用范围及优缺点 [20] ; (b) 流动曲线 [20] ; (c) 粒径、 壁厚、 质量分数对空心Fe 3 O 4 磁流变液表观黏度的影响 [20] ; (d) 体积分数、反磁颗粒比例对顺-反磁颗粒磁性液体正应力的影响 [80] Figure 7 (Color online) Simulation results of mechanical properties of novel magnetic fluid.…”

“…(a) Monte Carlo模拟的内部结构图 [18] ; (b) 机器学习模拟的流动曲线 [19] Figure 6 (Color online) Typical results of simulations of magnetic fluid. (a) Microstructures obtained from Monte Carlo simulation [18] ; (b) flow curves obtained from machine learning simulation [19] 评 述 表 1 磁性液体常用数值模拟方法的适用范围及优缺点 [20] ; (b) 流动曲线 [20] ; (c) 粒径、 壁厚、 质量分数对空心Fe 3 O 4 磁流变液表观黏度的影响 [20] ; (d) 体积分数、反磁颗粒比例对顺-反磁颗粒磁性液体正应力的影响 [80] Figure 7 (Color online) Simulation results of mechanical properties of novel magnetic fluid. (a) Shear stress versus magnetic flux density curves of hollow Fe 3 O 4 MR fluid [20] ; (b) flow curves [20] ; (c) influences of particle diameter, wall thickness, and weight fraction on the apparent viscosity of hollow Fe 3 O 4 MR fluid [20] ; (d) the magneto-induced normal stress of magnetic fluid with different volume fractions and ratios of diamagnetic particles [80] ( ) 图 8 (网络版彩色)传统磁性液体复杂力学行为的数值模拟结果.…”

Comparing the linear and logarithm normalized extreme learning machine in flow curve modeling of magnetorheological fluid

“…Another model is Herschel Bulkley that tried to overcome the disadvantage of Bingham plastic model, which is the high accuracy for a narrow shear rate range [18], [19]. In general, all models can only be applicable to one value of the magnetic fields [20], [21]. If one wants to cover other magnetic fields, a polynomial model of yield stress needs to be applied [22], [23].…”

“…Extreme Learning Machine (ELM) is known for its solution in terms of the quicker training, more accuracy value and highest possibility to gain accepted generalization than the classic methods. The existing method to predict MR fluid behavior using ELM has been proposed in [20] to predict various rheological parameters. Despite the high accuracy at some values and ranges, the prediction needs to be improved.…”