摘要
To overcome the shortcomings of the conventional variable fuzzy set assessment (VFSA) method in the dynamic water quality assessment, the functional data analysis (FDA) theory is introduced into VFSA method to develop a dynamic variable fuzzy set assessment (DVFSA) model. The procedure of DVFSA is: (1) generating continuous concentration curves through smoothing method, (2) calculating relative membership curves of indicators, (3) generating comprehensive relative membership curves, (4) making dynamic water quality assessment through calculating the ranking feature curve. DVFSA is the generalization of the conventional VFSA from discrete finite time points into continuous time domain through FDA method. DVFSA keeps the property of VFSA in fuzzy domain constructing and comprehensive classification recognizing; and absorbs the advantages of FDA in dealing with different sampling time and missing values, and representing the varying process of aquatic environment comprehensively and intuitively. Furthermore, DVFSA avoids the potential logical error by banning the least square optimization method in comprehensive relative membership generation. The dynamic water quality condition of Jiangdu hydro-junction in 2013 is assessed using DVFSA and the result shows that it belongs to classification “I” from January to April and “II” in other months. To further improve the water quality condition of Jiangdu hydro-junction, it is suggested to strengthen the environment protection in Lixiahe region and New Tongyang Canal.