Multi-objective Optimization Design Of Water Distribution Network Based On Monte Carlo Simulation Method

Water distribution network is an important part of urban water supply system.Whether the planning and design of water supply network is scientific is very important to ensure the normal operation of the pipe network system.Due to the large scale and complex design of water distribution network,the traditional design method is based on relevant specifications and experience,which often leads to problems such as low velocity of pipe section,high water pressure of nodes and uneconomic design scheme.In this case,it is of practical significance to seek the optimal design scheme which can not only meet the performance goal of water distribution network,but also reduce the project cost.Based on the existing research results of water distribution network optimization design,a multi-objective optimization model of water distribution network was established with the objective of minimizing the annual cost of water distribution network,the mean value of node surplus head,the mean square error of node surplus head and the variance of node water age.Aiming at the nonlinear,discrete and multi-attribute problems of the model,the shortest path method was combined with Monte Carlo simulation method,firstly,the shortest route method and the emergency water supply requirements were used to carry out pipe network layout and flow predistribution.Under the time-of-use electricity price(TOU)mechanism,the economic velocity calculation formula was used to determine the economic velocity of each pipe segment in combination with the location and type of the pipe segment.the economic velocity of each pipeline was determined initially by using the economic velocity formula and combining the position and type of pipelines,and the range of pipe diameter of each pipeline was determined according to the economic velocity.Then the pipe diameter of each pipeline was simulated by Monte Carlo method,which generated several alternatives.By calling EPANET toolkit in MATLAB,the compensating computation and working condition checking of the alternative schemes were carried out.The schemes that do not meet the constraint conditions or check requirements were directly eliminated,and the water quality of the remaining schemes was simulated.Thus,the flow,head loss,node water pressure and water age of each pipe section can be obtained,and the objective function values can be obtained.On this basis,the near-optimal solution was selected by the topsis method.The water distribution network of a county in Anhui Province is taken as an example,it is proved that this method is superior to the single objective optimization method,and the problem of nonlinear,discrete and multi-attribute in the optimal design of water distribution networks is well solved.At the same time,the multi-objective approximate optimization solutions obtained under the implementation of TOU price mechanism and the non-implementation of TOU price mechanism are compared,it is comprehensively judged that the optimal solution based on time of use price can obtain better optimization results.