一种新的伺服围压算法及其在离散颗粒循环双轴模拟中的应用  

A New Servo-control Method and Its Application to Simulation of Cyclic Biaxial Tests of Particle Assemblies

作　　者：金炜枫[1] 管亚君[1] 杨建辉[1] 曹宇春[1] 

机构地区：[1]浙江科技学院土木与建筑工程学院,杭州310023

出　　处：《科技通报》2016年第3期126-131,共6页Bulletin of Science and Technology

基　　金：国家自然科学基金项目(51408547);浙江省自然科学基金项目(LQ14E080009);浙江省公益技术应用研究项目(2014C33037);浙江省水利厅科技计划项目(RC1517)

摘　　要：联系离散颗粒集合体宏细观动力参数的循环双轴试验需要有效的伺服围压算法,针对已有算法仅将颗粒集合体等效为弹簧,同时没有基于自动控制理论获得伺服控制方程,这里将颗粒集合体试样和伺服墙等效成弹簧和质量块耦合系统,然后建立对应的状态方程和伺服压力目标函数,然后用动态规划得到这个最优控制问题的Hamilton-Jacobi-Bellman方程,并推导了此非线性偏微分方程的解析解,从而得到伺服力的闭环反馈控制函数。最后通过数值试验表明相对已有算法,新的伺服围压算法每一时步所需计算量更少、固结至指定围压的速度更快、伺服围压波动误差更小,验证了此伺服闭环反馈控制函数的有效性。Numerical cyclic biaxial tests utilized to relate dynamic micro and macro parameters of particle assemblies require effective servo-control algorithms. However, the existing servo-control algorithm only simulates the particle assembly as a spring. Besides it does not adopt any optimal control method, thus there is no optimal control function obtained. So that a new algorithm is developed by establishing a spring-mass model which simulate the side wall as a dummy mass and the particle assembly as a spring, and by utilizing dynamic programming the corresponding HJB equation is obtained. Then the analytical solution is gained and the servo force function is obtained subsequently. Compared with the existing servo algorithm, the new algorithm needs less computation cost during every time step, more quickly achieves consolidation of the particle assembly to a specified cell pressure, and causes less deviation from the designated cell pressure under dynamic load.

关 键 词：离散元 伺服围压 循环双轴 

分 类 号：TU435[建筑科学—岩土工程]