Research On Network Layout Optimization Of Water Quality Monitoring And Early-warning System For Water Distribution Systems

With the fast development of society and economy, water pollution problem is increasingly serious. There is an urgent need to establish a perfect water quality monitoring and Early-warning system, which provides real-time water quality monitoring data and protects the drinking water safety of people. Water distribution system has a large number of nodes, while the budget is limited by government finance, it’s necessary to study the monitoring sensor network layout optimization problems, promoting the work efficiency of monitoring network.For the foregoing requirements, this article focuses on the problem of monitoring sensor network layout optimization for water distribution systems, the research content and main innovative points are as follows:(1) The problem of monitoring sensor network layout optimization for water distribution systems is kindly divided into two types, in which models’accuracy plays an important role. In this paper, we put forward population coverage model (PCM) and pollution incidents detecting time shortest model (CDTSD), for solving the shortages of the existing models. In the population coverage model, water age ratio between two junctions and partial coverage concept are used to calculate the similarities between every two junctions. And this model is solved in the situation of Multi-Demand pattern. When it comes to pollution incidents detecting time shortest model, human factors are excluded to ensure the objective accuracy of model; parallel PSO algorithm based on multi thread mechanism is used to increase the speed of solving the model.(2) Considering that water distribution systems are simplified treated in most models, we put forward a method based on uncertainty theory to test the effectiveness of layout scheme provided by layout optimal models. The Monte Carlo method is used to simulate pollution incidents in water distribution systems; damage index of each pollution incident is calculated by simulation program. The probability of damage index below threshold λ is used to test the effectiveness of layout scheme.(3) At last, we put forward a Multi-Objective model, which combines population coverage model, pollution incidents detecting time shortest model and some else models, to meet different needs. Different objective is handled by normalizing method to transformed to uniform scale; Analytic hierarchy process method (AHP) is used to determine objective weight; parallel PSO algorithm based on multi thread mechanism is used to solve the model.