Six-axis accelerometer is a kind of inertial sensors measured dynamic parameters in the field of aerospace, robots and the like, and now researches on it are still at exploratory stage at home and abroad. A new scheme of using parallel mechanism as elastic structure and piezoelectric ceramic as sensitive element of the transducer is presented. The latter is also used to serve as pure-slide pair of parallel mechanism and traditional spherical pair is replaced by flexure spherical hinge. Three coordinates are set at mass-block, outer shell and ground, respectively, and their analytical relationships of position and attitude are established according to the theory of multi-body system. Based on Newton-Euler method, the dynamic equation of the system was established. By introducing aided angular velocity and using quaternion to describe rotation matrix and angular velocity vector, the problem was transformed into two initial value problems of first-order ordinary differential equations which was solved by improved Euler method, thus six-axis accelerations could be entirely decoupled. Method of generalized inverse for matrix was used to treat the violated constraint problem of quaternion, and an effective revised algorithm with repeatable calculation was presented. Based on the established decoupled model and algorithm, a numerical example was computed. The results were well consistent with the simulation ones in ADAMS, and the relative error between them was acceptable. Meanwhile , the time for data processing was no more than that of sampling. At last, the experiment was done, which shows the rationality of the proposed scheme and the reliability of the mathematical model.%提出一种新的基于并联机构的压电式六维加速度传感器的设计方案,即用并联机构充当传感器的弹性体结构,压电陶瓷同时充当传感器的敏感元件和并联机构的移动副,弹性球铰链代替传统的球面副.分别在质量块、外壳及地面上固结坐标系,基于多体系统理论推导两两之间位姿关系的解析表达式.运用牛顿-欧拉法建立系统的动力学方程,通过引入辅助角速度并用四元数来描述旋转矩阵和角速度矢量,将其转化成两个一阶常微分方程组的初值问题,并运用改进的欧拉算法进行求解,实现了六维加速度的完全解耦.针对四元数的违约问题,运用矩阵广义逆理论提出了一种有效的可重复运行的修正算法.利用建立的解耦模型及其算法对实例进行了求解,数据处理总时间小于采样点时间跨度,且理论计算值与ADAMS仿真数据吻合得很好；另外,通过试验进一步验证了设计方案的合理性以及数学模型的可靠性.
展开▼
