基于模糊信息融合的气固流化床流型及其转换的识别研究

Study on identifying flow regime and transition based on fuzzy information fusion in gas-solid fluidized bed

本文在模糊集理论的基础上介绍了多传感器、多参数识别气固流化床流型的信息融合模型。将压力脉动信号的算法复杂性Cn、涨落复杂性Cf和香农熵En作为融合的特征参数,进行特征层的多参数融合;根据特征参数建立了过渡流型的隶属度函数;对多个传感器的特征层识别结果进行决策层融合,得到了多传感器对不同流化状态的最终识别结果。实验结果表明,采用香农熵特征参数能较好地解决鼓泡与湍动2种流化状态转换的识别;应用多传感器、多参数数据融合对流态化不同流型及其转换的识别能得到较好的效果。

Based on fuzzy logical theory, an information fusion model with multi-sensor and multi-parameter is proposed and applied to identify the flow regimes in gas-solid fluidized beds. Algorithmic complexity, fluctuation complexity and Shannon entropy of the pressure fluctuation time series are employed as the fused characteristic parame- ters to carry out multi-parameter fusion. Membership functions are established to identify the state transition between different flow regimes according to the characteristic parameters. After the multi decisions of each sensor are fused at the decision level, the final identification results for different flow regimes can to be obtained. Experimental results indicate that Shannon entropy characteristic parameter can be used to better solve the problem of identifying flow regime transition of bubbling and turbulent motion. Multi-sensor, multi-parameters data fusion method can be applied to identify different fluidization regimes and transition between them and achieve better results.