Study On Simulation Analysis Of Water Consumption Based On Multi-agent

The total water consumption which had increased to 609.5 billion m3 in 2014, in addition to inefficient use of water resources, water waste and unreasonable water utilization structure, has aggravated serious water resources shortage, adversely influencing economic development. Dynamic simulation of water consumption and investigation of variation laws and its impact factors are keys for theoretical and decision support for the regulation and control of total water consumption desired to be addressed immediately at current stage of China. Accordingly, a model named MA-UHWDS is built on the basis of dynamic analysis of water consumption as well as multi-agent based modelling (ABM). And then the example study is conducted in this study. The main results obtained are as follows:(1) Based on total water consumption, water utilization structure and economic development of Shandong Province from 1995 to 2012, the vector auto-regression (VAR) model is adopted to reveal the dynamic relationship between industrial total water consumption and water utilization structure through variance decomposition and impulse response function. The results show that the water utilization structure could be further transformed and optimized, and the gap compared with high-income countries could be narrowed down via decreasing the proportion of agricultural water, while increasing the proportion of domestic water and properly controlling that of industrial water.(2) This study points out that the water consumption deviation between the simulated and the actual ones is caused by weak physical interpretation capability of some simulation and prediction approaches, thus ignoring initiative of each element in the simulation process. The prospect of the application of ABM to simulation of water consumption is huge, on the one hand, the formation of water consumption under the influence of human activity is an evolving complex process, which relates to several aspects such as natural condition, ecological environment and social economy. On the other hand, the overall evolution process of system could be reflected by description of the interactions among agents and between agent and environment due to the advantage of modelling from bottom to top and microcosmic to macroscopic.(3) A model named MA-UHWDS, which can analyze variation trend and impact factors of water demand of urban household is proposed. By the application of the proposed model to Qingdao City, the marginal propensity to consume water, basic water demand, and per capita water consumption of different income households are estimated with the statistics from 2003 to 2012, and then the simulation and prediction results of water consumption per capita and total water demand of urban household from 2013 to 2022 under various scenarios are analyzed quantitatively. The results show that the average share of expenditure for water in living expenditure from 2003 to 2012 is just 0.395%, and the share is so small that the public awareness on water conservation is weak; water consumption per capita of the lowest income household from 2003 to 2012 are all below the basic water demand except 2008, while that of the highest is 14.16m3 higher than the basic one, therefore both the payment ability of low income household and the overuse of high should be considered synthetically in the reform of water price; compared with the water demand per capita in scenario P1IhS1E1 in 2022, that of seven kinds of income households in scenario PhI1S1E1 decrease 16.7%,18.4%,18.1%,16.1%, 15.0%,17.8% and 13.8%, respectively. Therefore water price and disposable income could obviously impact the household water demand. MA-UHWDS provides new research idea and quantitative tool for the regulation and control of total water consumption.