介绍了处理方形分离器原始数据的一种新方法 ,提出了方形分离器基于改进后的 BP算法的神经网络运行优化模型 ,并借助该模型用相应的优化准则对方形分离器的运行参数进行了优化 ,得出了不同入口颗粒浓度下的优化风速 ,为该方形分离器的运行优化提供了参照依据 A new method of processing the square data of the square separator is introduced. A neural network operation optimization model of the square separator based on the improved BP algorithm is proposed. With the aid of the model, the operation parameters of the square separator are optimized Optimization, obtained at different inlet particle concentration of the optimized wind speed, which provides a reference for the operation of the square separator optimization