智能识别方法在板形识别中的应用及发展趋势

对国内外关于板形模式识别技术的研究现状和发展趋势进行了综述。通过比较,提出了传统的基于最小二乘法的多项式拟合法存在的不足,并对模糊分类、神经网络、遗传算法、混沌优化等智能识别方法在板形模式识别中所具有的优势进行了归纳和总结。最后,对智能方法在板形识别问题中的应用以及板形识别技术的发展趋势进行了展望,为板形检测环节得到理想板形信号提供理论研究方法,并将逐步应用于钢铁板形控制的工业过程中。

Application and Development Trend of Intelligent Recognition Methods for Flatness Recognition

The recent research progress and development trend of flatness pattern recognition are presented.The disadvantages of traditional least square method(LSM) for flatness pattern recognition and the advantages of intelligent recognition methods of fuzzy theory,neural network,genetic algorithm and chaotic optimization are found by comparison.Also,both the application of intelligent methods for flatness recognition and the development trend of the technology have been presented.The method of theory research has ...