【摘要】 为了提高板形模式识别精度，提出了一种基于改进海鸥算法结合Elman网络的板形模式识别方法。将改进的海鸥算法对Elman网络权值阈值进行优化，用于板形模式识别，选取20组数据进行测试，并将结果与基于BP神经网络的板形模式识别和基于传统Elman网络的板形模式识别方法进行比较，结果表明本文算法精度更高、效果更好，均方误差MSE相比其他算法低2个数量级。更多还原

【Abstract】 In order to improve the accuracy of flatness pattern recognition, a flatness pattern recognition method based on modified seagull optimization algorithm(MSOA) and Elman network is proposed. The weight threshold of Elman network is optimized with MSOA and then used for flatness pattern recognition. 20 sets of data are chosen for testing, and the obtained results are compared respectively with the flatness pattern recognition results based on BP neural network and traditional Elman network. It is found that algorithm method proposed in this paper has higher accuracy and better effect, with the mean square error lower than other algorithms by two orders of magnitude.更多还原