| In the context of global warming and accelerated urbanization,frequent extreme weather leads to rainstorms and floods.Therefore,it is particularly important to strengthen the intelligent monitoring and accurate scheduling of urban water system for flood disaster management and prevention.As one of the most critical projects in the river network,the comprehensive monitoring and control of the gate can help the city manager to conduct water resource scheduling and risk control.In the urban river scheduling,the intelligent scheduling scheme and early warning information can be provided by forecasting rainfall and monitoring river level,combining with the physical water network model and intelligent algorithm to forecast the changes of river level and give the sluice pump scheduling suggestions.Therefore,this paper designs a river gate monitoring system and optimizes the water resource scheduling in the study area,providing theoretical basis and methods for flood control and early warning.The main contents and conclusions of this study are as follows:(1)Research on water level recognition method based on machine vision.First of all,the image preprocessing of the water scale image is carried out,then the water scale area in the water scale image is extracted based on the two methods of edge detection and HSV(Hue,Saturation,Value)threshold segmentation.Finally,the water level scale is calculated based on the image contour extraction and gray level function.The small monitoring device was developed and the experimental analysis was carried out.The results of 20 images identification by two methods show that the fitting degree between the image recognition value and the real value of the two methods is higher than 0.98,the relative error of water level identification is lower than 10%,and the absolute error is less than 11 mm.It shows that the method of image recognition of water level scale is feasible and provides a new recognition method for water level monitoring in gate monitoring system.(2)Design the gate monitoring system and improve the PID(Proportion Integration Differentiation)control mode of the gate.The hardware of gate control unit was designed,including PLC(Programmable Logic Controller)controller,water level identification device and gate opening meter.Including software and hardware development and design of water level identification equipment,and installation test;The improved PID control method is adopted to control the gate hoist,which improves the efficiency and stability of gate opening and closing.(3)Study on River level prediction based on SWMM(Storm Water Management Model)model.The SWMM model of the study area was constructed by collecting river network and sub-catchment area data,and the model parameters were calibrated according to the literature and characteristics of the study area.Three rainstorms or more were selected to rate and verify the model.The results show that the absolute error of simulated peak water level is less than10%,and the Nash coefficient is more than 0.5,indicating that the model has good accuracy.Using rainfall forecast and predict water as a boundary,the use of SWMM model to forecast the river water level and validation,prediction results show that the model error within 20 min the water level in the peak time,peak absolute error within 0.23 m,prediction results are good,water level prediction provides conditions for scheduling optimization.(4)The water level scheduling optimization of Guangming Port is studied based on genetic algorithm and SWMM model.Based on the analysis of ecological scheduling and flood control scheduling problems in this region,the genetic algorithm combined with SWMM model was used to study the scheduling optimization.The multi-objective function is transformed into a single objective function according to the weight ratio,and the conventional scheduling and the genetic algorithm optimization scheduling are respectively used to study the sluice pump scheduling along the Guangming Port.The control performance of the two scheduling methods in terms of water level regulation and sluice pump opening and closing is compared and analyzed.The results show that the genetic algorithm has better effect in flood control and ecology.(5)Verify and analyze gate control and scheduling.The experimental results show that the improved PID control method can improve the efficiency of gate control by 10.9% ～ 16.0%compared with the original control.In the rain,the plan was implemented and adjusted according to the plan.After the rain,the dispatching effect was verified according to the actual water level change process of Guangming Port.The results showed that the water level change process was basically consistent with the forecast,the peak time error of water level was 20 min,and the peak absolute error was 0.05 m.The proposed scheduling method has a good performance on the original basis. |
