基于BP神经网络的三维力传感器静态标定方法研究

Research on Static Calibration for Three-dimensional Force Sensor Based on BP Neural Network

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, the three-dimensional force sensor is the key for the best use of force platform. In order to improve the measurement effectiveness of platform, coupling errors of sensors must be decreased greatly. Firstly, the traditional method is discussed. Then, a new method based BP neural network is put forward. Experiments and simulations show that, BP neural network method has more strength at improving the measure precision than tradi- tional method.