论文部分内容阅读
三维力传感器作为测力平台的核心元件,其测量精度直接影响测力平台的使用效果,而维间耦合问题是影响精度的主要方面。文章首先讨论了三维力传感器传统的静态标定方法在消除耦合误差方面的应用,并在方法缺陷分析的基础上,提出了新的三维力解耦方法——基于BP神经网络的解耦方法,继而对两种方法进行误差分析,验证了神经网络方法在多维力传感器解耦中的可行性和优越性。
As the core component of the force-measuring platform, the three-dimensional force sensor directly affects the using effect of the dynamometer platform, and the coupling between the two dimensions is the main aspect of the accuracy. The paper firstly discusses the application of the traditional static calibration method of 3D force sensor in eliminating coupling errors. Based on the analysis of the method defects, a new three-dimensional force decoupling method based on BP neural network decoupling method is proposed The error analysis of the two methods verifies the feasibility and superiority of the neural network method in the decoupling of multidimensional force sensors.