Modeling And Optimization Of Pressure Control System For Water Distribution Network

Water distribution network(WDN)is a large scale system with wide geographical distribution range,many intelligent equipment,large amount of information transmission,and the actual water demand of the consumers is real-time and constantly changing.With the expansion of city scale,the growth of resident population and the improvement of people's living standards,it not only meets the water demand and water pressure of consumers,but also realizes energy saving,and then improves the economic benefit,which is a great challenge to the optimal operation of WDN.In recent years,the optimal operation of WDN is mainly through the adoption of reasonable pressure control strategy to achieve the purpose of saving energy.The topological structure of WDN studied in this paper is that the water flow provided by the water source is first stored in the tank through the lifting pressure of the pumping station,and then the water is provided to the water demand area.There is a nonlinear dynamic characteristic between the water head of the node and the water flow of the adjustable hydraulic element in the pressure control system of WDN.This paper mainly focuses on the modeling and optimization of pressure control system,and the main contents are as follows:· The network topology of WDN is analyzed,and it is considered that there must be a dynamic hydraulic element in its network topology,that is,storage tank.The nonlinear dynamic relationship of the pressure control system between the water head of demand node and the flow of the adjustable hydraulic element is effectively expressed by Wiener model.The Wiener model of the pressure control system is divided into linear dynamic module and nonlinear static module according to the core hydraulic componet tank,and the two parts are connected in series.Based on the linear dynamic module of Wiener model,the nonlinear dynamic optimization problem of water head at water demand node is transformed into a linear optimal scheduling problem for tank level.And then the water head of the node is calculated iteratively based on the nonlinear static module of Wiener model.Wiener model of pressure control system is a more accurate modeling method,and the optimization calculation time is reduced at the same time.· At any control time,the switching condition of optimal scheduling is to judge the relationship between the water head value in the predicted time domain and the expected pressure region.If there is no possibility of going beyond the zone,the goal of optimal scheduling is to minimize the action of adjustable hydraulic elements.But if there is the possibility of exceeding the zone,the goal of optimal scheduling is to quickly converge the water head of the consumer node back to the zone.The strategy achieves a series of control effects.First,even if measurable disturbance changes dramatically,nodal water head should be kept within a predefined pressure range.Second,the change of adjustable hydraulic element is minimized.Third,once water head of node exceeds predetermined pressure range,and it is driven to predefined zone as quickly as possible.The control strategy ensures the zone control of the nodal water head and realizes the purpose of saving energy.· Considering the high complexity of centralized optimization control in large scale WDN,a distributed zone predictive control algorithm is proposed.In each subsystem of WDN,the zone predictive control strategy is adopted.The optimization trajectory of the tank level is introduced as the optimization variable,which meets the expected pressure range constraints,and attempts to track the output trajectory of the mechanism model of the pressure control system in order to ensure the existence of feasible solutions in the whole optimization process.The augmented Lagrangian formula is applied to distributed coordination strategy.The parallel coordination scheme is used to realize the synchronous optimization of each subsystem,and the suboptimal state of the whole WDN is realized.The distributed zone predictive control not only ensures that the water head of each subsystem is controlled within the desired pressure range,but also ensures the relative independence of the subsystems,which is convenient for the separate control of the subsystems.· The pipe network system which is consistent with the actual WDN operation condition is built on the hydraulic software EPANET,and the distributed zone prediction strategy of nodal water head is implemented.From the pump flow,water demand and water head of node provided by EPANET operation,the practicability of the distributed zone predictive control strategy for the nodal water head is verified.