论文部分内容阅读
光电位置敏感器件 (PSD)是一种可直接对其光敏面上光斑位置进行检测的光电器件 ,基于光电位置敏感器件可以构成多种非接触的高精度动态位移监测仪器。根据光电位置敏感器件的原理及特征方程 ,分析了存在背景光时它的输出信号是非线性的。提出一种基于神经网络的光电位置敏感器件背景光非线性补偿方法。利用神经网络具有逼近任意非线性函数的特点 ,通过训练使神经网络建立在不同背景光下光电位置敏感器件输出与其标准值之间的非线性映射关系 ,对其实现全程跟踪补偿。计算机仿真表明 ,该方法能有效地消除背景光的影响 ,在神经网络的输出端得到期望的线性输出
Photoelectric position sensitive device (PSD) is a kind of optoelectronic device that can directly detect the spot position on the photosensitive surface. Based on the photoelectric position sensitive device, it can form a variety of non-contact high precision dynamic displacement monitoring instruments. According to the principle and characteristic equation of photoelectric position sensitive device, it is analyzed that the output signal is nonlinear in the presence of background light. A non-linear compensation method for background light of photoelectric position sensitive device based on neural network is proposed. The neural network has the characteristics of approximating any non-linear function. Through training, the neural network is established based on the non-linear mapping relationship between the outputs of photoelectric position sensitive devices and their standard values under different background light, and the tracking compensation is realized. Computer simulation shows that this method can effectively eliminate the influence of background light and get the desired linear output at the output of neural network