Research And Application On The Double Standard Flood Control Forecast Operation Mode

Reservoirs as an important flood control structural measure play an important role in flood resources utilization. It has been found that the scientific design of reservoir flood control operation is the key to construct reservoirs safely and economically and play designing benefit in real-time operation. Reservoir flood control operation mode without forecast information always delays the opportunity to release or impound, increases flood risk and wastes flood resources. To increase flood resource volumes and improve flood control and utilizable benefits, the flood control forecast operation mode is studied in the past two decades except the theory and method of dynamic control of the limited water level, which involves forecast operation mode based on flood forecasting or two-stage rainfall forecasting and classified forecast operation mode etc. While the absence of considering characteristics of design flood hydrographs and regulate performance of reservoirs, these methods only consider the flood forecast accuracy and risk. So the research and supplement of the theory and method of the forecast operation mode is not only the demand for flood control and beniefit, but also the development trend of the reservoir flood control operation theory. In this dissertation, the methods including the principle of maximum entropy, information entropy, stochastic simulation technique and fuzzy optimization model etc. are employed into the theory and method of the reservoir flood control forecast operation mode. The characteristics of flood forecast errors, designing method and risk analysis of double standard flood control forecast operation mode and a real-time operation optimization model based on double standard flood control forecast operation rule are studied. The main researches and results are as follows:(1) At present, researches on the distribution of flood forecast error are very few and inconsistent views. The flood forecast error distribution model based on the principle of maximum entropy is established, and the flood forecast errors of nine reservoirs in humid and sub-humid regions are analyzed by proposed model. The results show that the flood forecast error obeys the normal distribution approximately in humid and semi-humid regions.(2) To employ flood forecast in flood control forecast operation safely and economically, this dissertation puts forward the definitions and numerical solutions of the flood forecast error domain, the reservoir flood control forecast operation acceptable error domain and its reliability. Taking Huanren and Baiguishan reservoirs as examples, the general rules of the reservoir flood control forecast operation acceptable error domain and its reliability are summarized which are the foundations for the design of the double standard flood control forecast operation rule.(3) The basic condition of designing the flood control forecast operation rule is the forecast accuracy achieving the published standard according to the "Standard for hydrological information and hydrological forecasting SL250-2000". That is single standard method, which cannot consider the design flood hydrographs and the regulating characteristics of the reservoir and the security constraints of the upstream and downstream flood control capacities. Moreover, this method lacks the reliability standard of the safe flood control operation. Therefore, the double standard flood control forecast operation rule is employed by combining the forecast accuracy and the reliability of the acceptable error domain. This method is proved to be feasible by two examples of Huanren and Baiguishan reservoirs.(4) The problem of existing risk analysis on flood control forecast operation mode is that they consider a few of risk factors. So, this dissertation proposes integrated risk analysis model of flood control forecast operation mode. Causes and probability distributions of four uncertainties including hydrological uncertainty, hydraulic uncertainty, stage-storage uncertainty and time-delay uncertainty are analyzed. And integrated risk analysis model of flood control forecast operation mode is presented based on Monte Carlo stochastic simulation theory, and this model is appropriate to single and combined factors. Taking Huanren and Baiguishan reservoirs as examples, risk rates of single and double standard flood control forecast operation modes are calculated. The results show that hydrological uncertainty is mainly risk source in most cases, and the double standard forecast operation mode is safe to reservoirs and their lower reaches than conventional operation mode and the single standard flood forecast operation mode.(5) In order to combine the double standard forecast operation rule and real-time flood control operation system, a real-time operation optimization model based on double standard flood control forecast operation rule is established. In this model, the double standard forecast operation rule regards as the first feasible operation scheme, and other schemes are establish by human-computer interaction. The highest water level, peak discharge, generating capacity, waste water volume and terminal water level are selected as indices, and modified entropy method, method of two elements contrasting and combination weighting approach based on game theory are adopted to calculate weight. In the end, fuzzy optimization model is introduced to optimize schemes. At the same time, decision support system of a real-time operation optimization model based on double standard flood control forecast operation rule is proposed. This decision support system could optimize schemes quickly. Lastly, the conclusions and problems need to be further studied are given at the end of the dissertation.