Study Of The Networks Development On Hydrologic Observation And Forecasting

Hydrological information collection and hydrological forecasting are the important parts of the non-projective measures in flood control and disaster prevention. This thesis has made a study on the hydrological information collection, processing and real-time forecasting of Fengle Reservoir in Yellow Mountain and achieved some valuable conclusions.The construction of data base system is the important part of system development. The system optimization is the target of the data base system of hydrological forecasting, of which the function of tables is definite, the structure is specific, the pattern of data is reasonable and the table is divided into minimum to promote the data access efficiency. In memory process, SQL sentences are put into the sever port to accelerate the speed of network system access and maintain the data base better. The flip-flop raises the real time of data base and the relevant forecasting scheduling table is renewed with the renewal of real-time table. Here the INSERT pattern of flip-flop is applied.Hydrologic information collection software makes use of MSCoom Control to receive the rain and water information form ports into computers, then recodes it and corrects the wrong information and marks it, finally displays the processed data on the interface. Hydrological information inquisition software has a function of real-time information inquisition and historic information inquisition and outputting the information tables. GSM Module improves the function of this system and reporting rain and water information by short message is a reliable and effective way.Hydrologic forecasting network software applies the traditional single-machine forecasting software into network and explores the hydrological forecasting with B/S structure which includes some main functions such as rain and water information inquisition, parameters optimization, real-time forecasting, historic flood simulation and so on. Any computer on the Internet can execute the above functions if inputting user identity and password, which provides the great convenience for national-wide hydraulic experts to make decisions on flood scheduling on other places.This thesis discuss by two production of run-off Methods, one of which is the traditional Xin' Anjiang Model-"production of run-off after full storage" Method, divides water sources and calculates the amount of affluence by Mcskingum Method, the other one of which is the mixed production of run-off Method. Finally the results by these two methods were compared and reached a conclusion. As for the calculation of the most sensitive early soil moisture in Xin' Anjiang Model, I advanced a creative calculation method by myself and got a satisfactory result.The system made use of two different kinds of methods for correction. AR Model method is of definitive concept and less interference factors, thus its correction result is satisfactory. Nerve-network method is used widely in recent years. Although mainly used for forecasting, here it is used for real-time correction and get a good result.