用非线性逆滤波改善传感器动态特性的研究

Researching of Improving Sensor's Dynamic Characteristics by Nonlinear Inverse Filtering

在测量系统中许多传感器存在着严重的非线性静态特性和响应滞后的动态特性,当被测量对象的变化率高于传感器的响应速度时,测量结果与真值之间存在较大的误差。为了补偿这个测量误差,采用了一个由无限响应的IIR滤波器和静态非线性环节构成的非线性滤波器去减小误差。IIR滤波器的系数通过实验数据得到,它是传感器的动态逆模型;非线性静态环节采用模糊小脑神经网络(FCAMC)实现。并通过对热敏电阻动态测量误差的补偿,验证了该方法的有效性。

The sensor has severity static nonlinear characteristic and slow dynamic response.When the rate of change of object is high as compared with the speed of response,the sensor has an error in the measured value.A nonlinear inverse filter is employed.It is composed of an infinite impulse response IIR filter and a static nonlinear block.The coefficient of IIR filter is obtained via experimental data,it is an inverse module of sensor's dynamic characteristic.The static nonlinear block is implemented by the algorithm of using fuzzy cerebellum neural networks(FCMAC).The method can be realized without knowing the dynamic characteristics of sensor and the experimental results which temperature is measured by thermistor show that the method is effective.