根据二维光电位置敏感器件 (PSD)的工作原理 ,分析了影响PSD线性度的主要因素 ,提出了一种用神经网络对PSD进行非线性修正的方法。以PSD的输入输出数据对作为样本训练的神经网络 ,利用神经网络所具有地能够以任意精度逼近非线性函数的能力 ,实现PSD的输出与实际光点位置之间的映射 ,在神经网络的输出端得到线性响应。该方法的优点是不需要很大的数据存储量即可得到很好的修正效果。结果表明 ,修正后的PSD能在较宽的位置范围内输出高线性度的信号 According to the working principle of two-dimensional photoelectric position sensitive device (PSD), the main factors affecting the linearity of PSD are analyzed and a method of nonlinearly correcting the PSD by neural network is proposed. The input and output data of PSD are used to train the neural network as a sample, and the mapping between the PSD output and the actual position of the light spot is realized by using the ability of the neural network to approximate the nonlinear function with arbitrary precision. At the output of the neural network Get a linear response. The advantage of this method is that it does not require a large amount of data storage to get a good correction. The results show that the modified PSD can output a high linearity signal over a wide range of positions