Node Flow Calibration For Dynamic Hydraulic Model Of Water Supply Pipe Network

The dynamic hydraulic model of the urban water supply pipe network can help the water technicians to master the real-time operation status of the water supply pipe network,and provide technical support in daily dispatching and long-term planning work.Accurate hydraulic model is the basic guarantee for informationalized water supply,but how to improve the efficiency of calibration of hydraulic model and ensure the accuracy of calibration is an urgent problem to be solved.Based on the dynamic hydraulic model,the improved ant colony algorithm is used to optimize the arrangement of monitoring points to provide a large amount of monitoring data for the hydraulic model calibration,and considering the tank level monitoring data and instrument measurement error,so as to establish the calibration objective function,and then useing the immune genetic algorithm to calculate,getting accurate pipe network dynamic hydraulic model,and establishing accurate platform for subsequent research.The essence of the hydraulic model node flow calibration calculation is to obtain the node flow value by continuously fine-tuning the monitoring point simulation error value.Therefore,it is first necessary to ensure the unformity and sufficiency of the monitoring point to avoid the model calibration ill development because of monitoring value.The monitoring point optimization layout problem is transformed into the objective function maximum value problem.Then,he improved ant colony algorithm is used to calculate the monitoring point.After a scientific arrangement of monitoring points,a large number of monitoring values are obtained to calibrate the hydraulic model.As an important equipment for balancing water pressure and regulating water volume,the tank level simulation error directly affects the hydraulic model calibration accuracy.In order to more accurately calibrate the hydraulic model,this paper considers the tank level simulation error and instrument measurement error,and uses the pressure error monitoring point,the flow monitoring point and the water level monitoring point to simulate the node flow,and establish the node flow calibration objective function.In view of the large scale of the pipe network,the initial convergence speed of the algorithm is slow,the initial antibody is improved,and the improved immune genetic algorithm is used to check the node flow.Finally,an example of the optimization of the monitoring points and the model calibration project for the C city water supply network is carried out.A total of 91 flow monitoring points and 55 pressure monitoring points were arranged,and the monitoring points were evenly arranged.Through the analysis of the node traffic calibration process,the results show that the immune genetic algorithm calculates the global optimal solution,and the result is stable.Compare and analyze the simulated values and measured values of the optional six nodes and two tanks in the water supply network in 96 periods.The pressure relative simulation error is 98.96% within±1m,the maximum simulation error of the water level is 0.329 m,and the maximum relative error of the node flow is 5.78%.The calibration accuracy results meet the accuracy requirements proposed by the predecessors.Therefore,the node flow calibration method have certain reference value for the optimization of the actual pipe network calibration model.