Abstract To improve tracking precision against complex external forces in the working environment, dynamic model-based control algorithms are widely used in tracking mechanisms. As a foundation of the real-time dynamic control system, this paper focuses on the complete predicted dynamic modelling and dynamic parameter identification of a 2-DoF over-constrained tracking parallel mechanism. Firstly, the kinematic model is obtained by differential mapping of the topology models based on finite and instantaneous screws (FIS). Referred to virtual work principle, dynamic equation is constructed. To reduce the computation cost in practice, a linear model involving inertial and friction parameters is formulated for dynamic identification. A direct identification method is applied and all the parameters are obtained at the same time. Experiments are performed to show the accuracy and efficiency of the identification results. The contributions of the paper lie in 1) a concise method for constructing velocity and acceleration models of components is provided without tedious analysis 2) complete dynamic model of an over-constrained tracking parallel mechanism is developed 3) practical dynamic identification process is carried out compromising between precision and efficiency.

中文翻译：

---

一种跟踪并联机构的动态识别

摘要 为了提高对工作环境中复杂外力的跟踪精度，基于动态模型的控制算法被广泛应用于跟踪机构。作为实时动态控制系统的基础，本文重点研究了2-DoF过约束跟踪并联机构的完整预测动态建模和动态参数识别。首先，通过基于有限和瞬时螺钉（FIS）的拓扑模型的差分映射获得运动学模型。参照虚功原理，构造动力学方程。为了在实践中降低计算成本，建立了一个包含惯性和摩擦参数的线性模型用于动态识别。应用直接识别方法，同时获得所有参数。进行实验以显示识别结果的准确性和效率。本文的贡献在于 1) 提供了一种构建组件速度和加速度模型的简洁方法，无需繁琐的分析 2) 开发了一个过约束跟踪并联机构的完整动力学模型 3) 进行了实际的动态识别过程在精度和效率之间。