Research And Development On The Urban Road Waterlogging Pre-alarming And Supervising Equipment Based On Smart Lamp Poles

With the development of urbanization in China,urban climate and underlying surface conditions have changed significantly,In addition,the construction of urban drainage network system lags behind and the planning of drainage system is not reasonable.Once encountering severe weather such as typhoon and rainstorm,it will cause urban road waterlogging which will seriously affect the travel of urban residents and greatly threaten the safety of human life,property of the country and the people.At present,the supervising methods of urban road waterlogging at home and abroad are mainly divided into two categories: equipment monitoring and data mining analysis.The equipment monitoring mainly uses various water level sensors deployed on urban roads to supervise waterlogging,this kind of supervising methods mostly use a single or single source sensor,which can only measure the depth at the sensor measurement site,therefore,it is difficult to accurately measure the maximum depth or effective depth of the urban road waterlogging area,and it is easy to produce misstatement or false alarm;Data mining analysis can infer and estimate the possibility and depth of urban road waterlogging by analyzing the cause and process of urban road waterlogging’s formation,combining with the surface elevation data of urban road,road drainage pipe network and etc.This kind of supervising method needs all high-precision digital elevation data of urban roads and information of road drainage network,which will lead to the problems of poor data acquisition,incomplete data,complex data,bad data update and insufficient recognition accuracy.With the support of vertical projects and combining with the application scenario of smart lamp pole,the paper puts forward some solutions against the problems mentioned above by developing a kind of urban road waterlogging pre-alarming and supervising equipment based on smart lamp pole,The equipment integrates heterogeneous multi-source and multimodal data such as the area of urban road waterlogging sensed by camera monitoring,the depth of urban road single point waterlogging sensed by pressure level sensor and the rainfall sensed by tipping bucket type rainfall sensor,Combining with the calculation model for the maximum depth of waterlogging proposed in this paper,it can accurately calculate the maximum depth of waterlogging,Then,based on the calculation model for the maximum depth of waterlogging,the perception data of waterlogging and the obtained meteorological forecast information,the depth of waterlogging predicted by the equipment in the next period of time will be accurate.The main work of this paper is as follows:First of all,the architecture of the smart lamp pole Io T system is discussed,based on the four-tier architecture of the conventional smart street lamp Io T system,an edge computing layer is added at the edge of the network of its transmission layer.The urban road waterlogging pre-alarming and supervising equipment developed in this paper is used as the edge computing layer equipment ——the edge smart node,and the traditional "Cloud,Channel,Device" model of the Internet of things is transformed into "Cloud,Channel,Edge,Device" model,on this basis,the scheme design of the equipment is completed from three parts: system structure,hardware composition and software framework.Next,the generation process of urban road waterlogging is studied and analyzed,then,based on the theory and calculation of waterlogging,the formation process of waterlogging is modeled,the calculation model and prediction model of the maximum depth of waterlogging are deduced and constructed,and the least square parameter estimation method of multiple linear regression model is used to locally complete the on-line identification of model parameters based on real-time waterlogging monitoring data.Then,the multi-source heterogeneous sensor sensing technology of waterlogging monitoring is studied to provide technical support for the use of the maximum depth calculation model and prediction model of waterlogging.Waterlogging monitoring mainly perceives the area of waterlogging,the depth of urban road single point waterlogging and rainfall.The combination of HSV color space recognition and support vector machine(SVM)texture feature recognition is used to identify the waterlogging area and realize the monitoring and perception of the waterlogging area.Based on the sensor technology,pressure level sensor and tipping bucket rain sensor are used to measure the depth of urban road single point waterlogging and rainfall.Finally,the realization of urban road waterlogging monitoring and pre-alarming is completed.An experimental verification platform was set up in the laboratory to verify waterlogging area perception,urban road single point waterlogging depth perception in the waterlogging monitoring function of the equipment and the maximum depth of waterlogging calculation model,prediction model and waterlogging pre-alarming,which verified the effectiveness and accuracy of the urban road waterlogging pre-alarming and supervising equipment.