Research On Remote Hydrological Real-time Flow Measurement System And Low-latency Algorithm

Real-time monitoring of hydrology and water resources is the main way to understand and reveal hydrological phenomena,an extremely important means of preventing flood disasters,and it is related to people’s livelihood.Through real-time hydrological monitoring data,accurate prediction and judgment of water conditions can be made,which provides prerequisites for flood control and drought relief and water resources management and protection,and effectively protects people’s livelihood.At present,most of the hydrological flow measurement uses artificial lead fish to measure the flow,but it must be controlled manually on the spot.The data after flow measurement is only stored in the local computer.In case of danger,to achieve data overall planning of the whole watershed,the real-time nature of flow measurement data can not be guaranteed and no longer has reference value.Therefore,the construction of remote real-time flow measurement system has become an inevitable requirement of intelligent hydrological information monitoring system.This topic analyzes and studies the shortcomings of the current flow measurement system and combines the characteristics of the Yellow River Basin,introduces Internet technology and remote control technology into the hydrological flow measurement system,and designs a remote hydrological real-time flow measurement system.The flow platform establishes a connection channel to realize unmanned remote flow measurement.The study of the water surface velocity coefficient is an indispensable content in the hydrological flow measurement work.Through the analysis,it is concluded that the water surface velocity coefficient is related to the characteristics of different rivers.The analysis of the water surface velocity coefficient of different rivers should fully consider the characteristics of the river and adjust measures according to local conditions,and finally determine the surface velocity coefficient through different analysis methods.Combining with the characteristics of current measuring platform,the time stamp technology is introduced to mark the time of the system command,and a low-latency algorithm is proposed to predict and compensate the network delay of the real-time flow measurement system.It can self-correct learning,in-depth fitting,complete delay prediction,and add the predicted network delay combined with improved PID control to the real-time flow measurement system for network delay compensation.Through MATLAB simulation and realtime experimental verification in the tributaries of the Yellow River,it shows that the wavelet neural network-PID algorithm has a high prediction rate for the delay,and the system network delay can be effectively maintained at about 200 ms,which is 20% shorter than that before compensation,and can effectively Improve the network delay of the system,improve the robustness of the control system,and meet the requirements of real-time control of hydrological flow measurement.The system is currently running well in each hydrological station in the Yellow River Basin,which solves the difficulty that the hydrological station cannot automatically upload the hydrological data,improves the real-time performance of unmanned flow measurement,reduces the personnel operations in the flow measurement process,and greatly improves the construction level of hydrological digitization.