A novel method for improving the accuracy of parallel robots based on efficient identification of kinematic parameters

Parallel robots are increasingly used in precision manufacturing and measurement applications. However, factors such as assembly errors and gear clearances can cause the positioning error of the end effector to reach several millimeters, which fails to meet the precision requirements for high-precision machining and measurement. To enhance the positioning accuracy of parallel robots, parameter calibration has become an essential technical approach. As the accuracy of parallel robot systems may gradually decline during long-term operation, regular calibration of the kinematic parameters is particularly important. An efficient and high-precision parameter calibration method for robots is urgently required to avoid disrupting the operation of the measurement system. This paper proposes a two-step method for identifying kinematic parameters based on the kinematic error model of parallel robots, ensuring both calibration accuracy and significant computational efficiency. In the first step, the L2 Regularized Least Squares (L2 RLS) method is used to pre-identify the kinematic parameters, obtaining suboptimal parameter values. Subsequently, the suboptimal values serve as the initial values for the iterative (IT) method, which refines the parameter identification through an iterative process to obtain the optimal parameter values. Numerical simulations have verified the effectiveness of the proposed method, ensuring calibration accuracy while significantly improving computational efficiency. Experimental results further demonstrate that applying the L2 RLS-IT method proposed in this paper reduces the maximum positioning error of the parallel robot from 1.872 mm to 0.294 mm. This significant improvement not only proves the effectiveness of the method, but also plays a crucial role in ensuring the long-term stable operation of parallel robots with high precision.