Research On Nonlinear Model For City River Water Quality

Nonlinear is the universal phenomenon about pollution water quality, the migration and transformation of pollutants in water is easy to be affected by a variety of internal or external factors. City river is a special kind of water environment, it is easy to be polluted by city sewage, domestic sewage and other types of pollution, at some polluted section, there is few oxygen in water, the process of pollutant’s migration and transformation in water has non-linear characteristic. In this paper, some studies include water quality model of nonlinear compensation, non-linear differential equation water quality model, water quality analysis of low oxygen river, impulse model of river aeration are researched.1. According to prediction errors of water quality, chaos theory alwalys to be used to analysis water quality, the errors is reconstructed to a new phase space with chaos theory, and neuron number of neural network layer is given by the phase space, base on it, an error compensation model of the neural network is established, that’s breakthrough a single water quality model limitations, the results show that, the relative prediction error is reduced from2.4%to1.65%, it show that the neural network error compensation model is effective.2. According to low dissolved oxygen phenomenon of urban heavy polluted river, the assumptions of SP model and Shastry model are analyzed, on the base of it, assumptions is modified to establish a new nonlinear differential equation model, and the results is Ersp=0.829, Er,shastry,=0.2498Er,hc=0.2426by simulate with Matlab, it show that the new model has more effectivly.3. In order to analyze the state of nonlinear water quality differential equation, methed of differential equations has used to analysis the behavior of City river nonlinear differential equation model, and some results are given, firstly, the City river water quality model has a unique solution in its solution space, it means the model can be used for water quality’s prediction. Secondly, the stability of the City river model is dependent on its parameters.4. On the application of City river water quality model, positive solution should be proved to avoid negative phenomenon of the water quality prediction, so the existence of positive solutions of a model condition is discussed in the paper, the conclusion is that the City river water quality model has positive solutions when the model parameter is greater than zero, it means BOD, DO concentration will not be negative, it also means that the model is effectively in water quality prediction; the other conclusion is that the concentration of BOD will eventually approach to zero, and oxygen (DO) will eventually approach to saturation when there is enough distance, it means that the water will self-purification by itself ability of purification.5. On the management of River aeration process, impulse models base on the SP model and City river model have been established to describe aeration process, its average relative error is5.21%by simulated with practical rive data, it means that the describe is validity; at the same time, SP model’s average relative error is3.114%(less than5.21%), it means that the prediction accuracy of water quality will be influenced by impulse process.In this paper, a new nonlinear model has been established about nonlinear characteristic of pollutant migration, conversion in City river,but there is also some questions to be studied in the further.(1) Components of City river model should to be increased, in the paper, only BOD and DO has been studied, in the future more components water quality model should be studied if it necessed.(2) In the pape, only one-dimensional water quality model has been established on City river water quality, in the future, the two-dimensional or three-dimensional City river water quality model should been studied.(3) Due to the complex of water environment, there is many assumptions should be revised, such as the description of river aeration process, the assumptions of water quality model, the migration and transformation of pollutants in water and so on, all of it will improve model’s prediction accuracy.