Research And Its Application On Hydrological System Fuzzy Uncertainty Analysis Method

This paper describes the development of the Fuzzy set theory proposed by Professor Chen Shouyu, as well as its research and application status in the hydrological system. Based on the evaluation of the current hydrology research status and the hydrological system uncertainty analysis theory, and considering the catholicity of uncertainty in hydrology, the paper covers the academic and applicable research on the fuzzy uncertainty of hydrological system. Followings are the primary achievements in this paper:1. It proposes the weighted average optimization weight calculation method and the iterative calculation model through the analysis on the index value and the expectation value. This method was adopted to calculate the daily, three-day and seven-day average rainfall of the LongFengShan Reservoir drainage area, and corrected the flawness of the traditional Thiessen Polygon method. Using the iterative model to evaluate the difference between the parameters'calculation and verification, both the overall average error and each sample's error are obviously smaller than that by the Thiessen polygon method.The application shows that the model is precise and clear in theory and concise in calculation, and has provided a relevantly simple theory, model and method in drainage rainfall calculation. The model in this paper adopts the expectation output value calculated from the GIS method to optimize the weight. Thanks to the spatial distribution factor in the GIS method, generally the iterative model in this paper will result in a more reasonable weight than the area weight by the Thiessen polygon method. Compare the differences of the iterative optimized method and model verification, and it adopts fuzzy optimized nerval network model to test run-offs of floods, which show the feasibility of the method.2. Among the various hydrological phenomenon, the randomicity and fuzziness usually exist together in the researching objects and issues, and during the uncertainty analysis and calculation it is better not adopt a single method, and shall combine the different methods which is said the coupling. Based on the fuzzy pattern recognition theory by Chen Shouyu, and considering the randomicity and fuzziness in the design flood ascertain, this paper suggests a new coupling thought, to reasonably combine the random analysis and fuzzyanalysis for an overall and practical solution on the uncertainty.The current common way of enlarging the typical flood is to sort out the typical data from the many measured ones, which represents a certainty rule in drainage flows' generation and influx. Due to the highly randomicity of the hydrological phenomenon, the sorted typical course might not happen in the future. Furthermore, the method of enlarging the typical data lacks of the theory foundation because of the different frequency of the peak flow. This paper worked out a frequent standard formula of the flood peak and floodwater, and introduced a fuzzy distinguish method on the basis of the standard model of design flood, to select the actual flood closest in value with the design flood index to be the design flood, which breaks through the method of enlarging the typical flood to get the design flood. The paper also studied by dividing the flood seasons of the Yinhe reservoir, to work out the actual design floodwater among major flood season and post flood season, and to get through a reasonable analysis.3. This paper also researched the relative membership degree in flood season using the fuzzy hydrological method proposed by Professor Chen Shouyu, and calculated the design flood control limited level hydrograph for the fuzzy dynamic process control. The study was based on the four reservoirs such as Yinhe reservoir. The ascertain of the fuzzy design flood control stage hydrograph fetched up the shortages of keeping only one design flood control level during the whole flood season and the hierarchy flood control limited level hydrograph method by experiences.4. Based on the Fuzzy Cluster arithmetic proposed by Professor Che...