Research On Automatic Detection And Real-time Forecast Method Of Qiantang River Tidal Bore Based On Black Box Modeling Theory

The funnel-shape of Hangzhou Bay forms unique scenery of Qiantang River, which is wellknown as tidal bore. However, as we enjoy the magical glamour of nature, the tidal bore alsobrings severe disaster. Almost every year, some people are drowned by the tidal bore of QiantangRiver. If the forecast of tidal bore could be highly accurate, the efficiency of disaster preventionwould be greatly improved, thus making it convenient for relevant departments to take measuresto prevent accident caused by tidal bore. This paper proposes a black box modeling method, andon this basis we develop an automatic detection and forecast system for Qiantang River tidalbore. The main results are as follows:(1) Forecast method based on statistical law, harmonic analysis and hydrodynamics areintroduced, which proved to be not effective. So it is necessary to explore a forecast methodwhich is self-learning and self-adaptive. A black box modeling based on historical water leveldata is proposed to determine the detailed conditions of judge the arriving of tidal bore, as wellas a real-time forecast of tidal bore by dividing the river into segments which meets therequirements of tidal bore forecast.(2) Aware of the problems exist in historical water level data, such as disturbance andirregularity, we carry on a field research of Qiantang River, coming to a conclusion of thenecessary of adding the velocity detection device, which we design based on Android to gatherthe data from Surface Velocity Radar (SVR) speed sensor. What’s more, we upgrade themonitoring station, including adding the SVR speed sensor as well as adaptor to eliminate theinfluence of turbulent flow.(3) The field data is transmitted to the server on hydrologic station through GPRS/GSMnetwork. Tidal bore forecast software is developed to implement the proposed forecastingmethod, and the forecast result is stored in database for information distribution.(4) To facilitate a quick view of real-time forecasting information of tidal bore, a website isdeveloped, which show the information both in Google map and text form. Meanwhile, theadministrator could correct forecasting information by the website, which facilitates themanagement of forecasting information.(5) The system is run in practice to test the performance of the proposed forecast method.The result shows that it has high accuracy, which satisfies people’s demand of tidal bore forecast.