二维力传感器动态补偿器设计

为改善扁环式二维力传感器动态特性以及消除其维间耦合现象严重的问题,提出了用神经网络逆系统方法去构造该二维传感器的动态补偿器,并由此设计了由微分器和静态神经网络构成的神经网络逆动态补偿器.通过对原传感器的数据采集和数据处理得到了静态神经网络的训练样本,用所得到的训练样本对静态神经网络进行了训练,训练后的静态神经网络即可用来和微分器构成所期望的动态补偿器.将该动态补偿器串接在扁环式二维力传感器之后,进而构成复合的测量系统.进一步地仿真实验表明,经过动态补偿器的信号远好于未补偿的原传感器输出信号.由此可见,利用该方法设计的动态补偿器所组成的复合测量系统在消除维间耦合现象的同时、也大大改善了系统的动态特性.

Design of dynamic compensator for two-dimension force sensor

To improve the dynamic performance and eliminate the coupling between dimensions FOR the two-dimension force sensor,an artificial neural network(ANN) inverse system method is applied to construct a dynamic compensator for the sensor,which is composed of several differentiators and a static ANN.The training samples of the static ANN were obtained through data acquisition and processing, with which the static neural network was trained and the ANN inverse compensator was constructed together with the differentiators.By cascading the compensator with the original sensor,a compounded measurement system was achieved.According to the simulating result,the compensated signals of the compounded system are far better than the signals of the original sensor and it can be concluded that the compounded measurement system achieves the desired result.