Research and management scheme of water resources in Beijing area

Xuanxuan Qi Xiaojiao Li　Zhonglin Ma

Abstract： Has the drought become an inevitable trend？ With this question put forward， we start this study that the center is the BPNN（BP neural network） method for regional water balance and its application. Then based on Genetic Algorithm fast random search capabilities， we establish programs to optimize water utilization.

Given the poor condition of water resources， how do we solve it？ How to evaluate and forecast it？

For our team， we establish the optimization model based on three main objectives （Industrial benefit， social benefit and environmental benefit）. Considering the limiting conditions， the reasonable intervention scheme is obtained by the fast convergence of Genetic Algorithm. For example，the balance of supply and demand gets better through the optimization of Water Transfer Project in Beijing.

The model can analyze the main factors of water security and make a reasonable assessment. Moreover， it puts forward the optimization solutions to ease the tension of water resources. From a broad perspective， our model is widely applied to solve the ecological security evaluation， rational allocation of resources and dynamics system of urban management.

Key words： BP neural network Pressure-State-Response optimization model

1 The rational allocation model of water resources

1.1 The objective Optimization

Model construction Definition is that supply and demand balance is determined by the water supply and water demand. Therefore， we use the ratio between water quantity and the sum of water supply and demand as criterion， B=Y1/（Y1+Y2）.

Economic goals[1]： Under the premise of meeting the regional minimum water demand， the water efficiency of the industrial sector is the largest.

Where i represents the i-th partition. And j is the j-th type users， namely， they are agriculture， industry and tertiary industry under the simplified model. k stands for k kinds of water source， they are respectively surface water， groundwater and sewage recycling water. WSijk is the k-th source of water supply to the j-th type users water supply in the i-th partition. GWij is the i-th partition unit water use can support industry benefits generated by the j-th type users， the added value of the industry is generally on behalf of industry benefits. lij is the weight of the j-th type users in the i-th partition. It is mainly based on the importance of water， the water use efficiency coefficient of each department and each user to the importance of the water. n，m and l represent respectively the number of partitions， number of users and the types of water supply source.

Social goals[2]： Under the premise of satisfying human and animal drinking water safety， we will obtain the maximum social benefits if there are the revitalization of industry， the development of modern agriculture.

In the formula，WRij is the j-th type users water demand in the i-th partition. Including industry， agriculture and the third industry of water demand， etc.

Environmental goals： The total amount of Sewage disposal after recycle and reuse is the smallest. The environment objective function is as follows：

In the formula， di is sewage treatment rate in the i-th partition， Pij is the j-th type wastewater discharge coefficients in the i-th partition， ri is wastewater reuse rate in the i-th partition.

1.2 The constraint conditions

Water supply capacity constraints

Where represents maximum output capacity of the k-th water source in the i-th partition.

1.3 Solution and analysis of the model

In this paper， the multiple objective model is transformed into a single objective model to solve the model by using the weighted method. Respective weights are as follows： V=（0.322，0.394，0.284）.

From the weight， we can see that social goals is the most important. This is also in line to explore the direction of our water supply and demand balance of this article. We will use the genetic algorithm for optimization solution.

By using genetic algorithm and the single objective linear method， we can obtain the feasible water resources allocation scheme. After three times of debugging， it can be concluded that the convergence of genetic algorithm.The best result of the three training is that convergence is achieved at the 50 iterations， and the convergence value is 3800. We use this value to measure the comprehensive index of the rational allocation of water resources.

Here are the water allocation table of the two regions. From the above we obtain the supply and demand ratio to measure the balance between supply and demand.the balance of A zone： B=0.52

the balance of B zone： B=0.49.

Average equilibrium degree is 0.5. Before the allocation of water supply is poor in Beijing area.It can be seen that the water balance in some places can be alleviated to a certain extent by the partition of the allocation.Therefore， only taking the scheme， the carrying capacity of water resources can be improved by combining the South-to-North Water Diversion Project with industrial and agricultural water resource efficient utilization.

References

[1] Food and Agriculture Organization of the United Nations. FAO Water Resources.

[2] Shaofeng Jia， GuoWang and Jun Xia.Research progress of water circulation in social economic system[J].Geography newspaper，2003，58（2）：255-262

[3] Demin Chen，Xingwang Qiao. A preliminary study on the legal protection of water resources security in China[J].Modern science.2003，25（5）：118-121