论文部分内容阅读
为了使传感器能够准确测量，标定是必不可少的，而好的似合曲线能够改善传感器的测量性能。通过对某电涡流位移传感器位移、电压用多项式拟合曲线的特点进行分析，经过试探发现该传感器曲线可转化为直线。用逼近法和最小二乘法求得使曲线转化为直线的各参数的初始值，且偏差更小。再对这些参数以非线性编差最小为条件采用模式搜索法进一步优化，得到比多项式拟合更实用的、偏差更小的拟合曲线。这种处理数据的方法对于求有高次项的非线性传感器逆函数有一定的借鉴意义。
In order for the sensor to be accurately measured, calibration is essential, and a good fit curve can improve the sensor’s measurement performance. Through the analysis of the displacement and voltage of an eddy current displacement sensor using the polynomial fitting curve, the curve of the sensor can be transformed into a straight line after the test. Using approximation method and least square method to get the curve into a straight line of the initial value of each parameter, and the deviation smaller. Then these parameters are further optimized by using the pattern search method based on the minimum nonlinear programming error, and a fitting curve with a smaller deviation than the polynomial fitting is obtained. This method of processing data has certain reference value for finding the inverse function of nonlinear sensors with higher order term.