Optimal Deployment of Water Resources Based on Multi-Objective Genetic Algorithm

Yong Liu, Yongrui Zhuang, Nan LuInternational Journal of u- and e- Service Science and Technology 2016 · DOI

TL;DR

This paper tackles the critical challenge of optimizing water resource allocation in China, a water-scarce nation facing increasing demand. They propose a novel multi-objective optimization model that simultaneously considers economic, social, and environmental benefits, solved using a combination of the Gray Model for demand prediction and a Genetic Algorithm for allocation. This approach offers a more holistic and sustainable strategy for water management.

What Problem Does This Paper Address?

Freshwater is a finite and increasingly strained resource globally, exacerbated by urbanization, pollution, and climate change. This leads to growing water demands and conflicts. China, despite ranking sixth in total water resources, faces severe per capita scarcity due to uneven distribution and rapid economic development. Traditional water allocation strategies often focus on single objectives, failing to capture the complex interplay of economic growth, social well-being, and environmental sustainability. This paper aims to bridge this gap by developing a comprehensive model that balances these competing needs for optimal water resource deployment.

Methods

The authors employ a two-pronged approach:

  • Gray Model (GM (1,1)): This technique is used for forecasting future water demand and supply. GM is particularly useful for making long-term predictions with limited historical data, which is common in water resource management. They use it to predict China's freshwater demand in 2025 and estimate the reduction in surface and groundwater.
  • Multi-Objective Genetic Algorithm (GA): This is the core of their allocation strategy. GA is a metaheuristic inspired by natural selection and evolution, well-suited for complex, non-linear optimization problems. It allows the model to explore a vast solution space and find multiple optimal trade-offs (Pareto solutions) simultaneously. The GA is used to optimize the allocation of various water sources (surface water, groundwater, desalination, wastewater reuse) to different user categories (industry, agriculture, domestic, others) based on the predicted demands and supplies.

The model incorporates three objective functions:

  1. Economic Benefit: Maximizing the net benefit of water supply, considering supply effectiveness and cost.
  2. Social Benefit: Minimizing water deficit across different user groups.
  3. Environmental Benefit: Minimizing pollution discharge (measured by COD).

These objectives are then combined into a single objective function using a penalty function approach, which is then solved by the GA.

Datasets

The paper utilizes historical water demand data from the China Statistical Yearbook for the years 2004-2010. This data is used to train the Gray Model for future demand prediction. For water supply, they refer to the China Statistical Yearbook for data on surface water and groundwater reduction between 2000 and 2008. Estimates for desalination and wastewater reuse are based on the development of Chinese technologies.

Key Insights and Ideas

  • Holistic Optimization is Crucial: The paper strongly advocates for multi-objective optimization in water resource management, arguing that single-objective approaches are insufficient for achieving sustainable outcomes. Balancing economic, social, and environmental factors is paramount.
  • GA's Power for Complex Trade-offs: The Genetic Algorithm proves to be an effective tool for navigating the complex trade-offs inherent in water allocation. Its ability to explore a wide range of solutions and identify optimal compromises is highlighted.
  • Gray Model for Data-Scarce Forecasting: The successful application of the Gray Model demonstrates its utility for predicting water demand in situations with limited historical data, a common challenge in many regions.
  • Desalination and Reuse as Key Solutions: The model explicitly includes desalination and wastewater reuse as water sources, acknowledging their growing importance in addressing water scarcity, particularly in coastal areas.

Key Results

The study successfully predicts China's water demand for 2025 using the Gray Model, with high precision for most categories. The multi-objective optimization model, solved by the GA, provides an optimal allocation strategy for different water sources to various users. Table 3 presents a sample of these allocations, showing, for instance, that surface water is the dominant source for agricultural and industrial users, while desalination plays a role in meeting domestic and other demands. The paper indicates that the GA converges to a stable solution around the 200th iteration, achieving a maximum objective function value of approximately 280.

Limitations and Future Work

The authors acknowledge several limitations:

  • Simplified Assumptions: The model relies on simplified hypotheses, such as the absence of devastating natural disasters and wars, and assumes unchanged rates of water reduction and uniform costs for desalination and transfer across different areas.
  • Predicting Water Reduction: The rate of water reduction is difficult to predict accurately as it's influenced by numerous factors like climate change and water policy. The current model assumes this rate remains constant.
  • Limited Objective Functions: While the model includes economic, social, and environmental objectives, future work could incorporate additional functions, such as maximizing the profit of a shared basin.

Future research directions include integrating more realistic scenarios, exploring a wider range of objective functions, and developing more efficient solution methods for optimal multi-objective problems. The authors also suggest that the model could be further applied to specific cities to enhance its practical significance.

Summary updated

BibTeX
@article{liu2016optimal,
  author = {Yong Liu and Yongrui Zhuang and Nan Lu},
  title = {{Optimal Deployment of Water Resources Based on Multi-Objective Genetic Algorithm}},
  journal = {International Journal of u- and e- Service Science and Technology},
  year = {2016},
  doi = {10.14257/ijunesst.2016.9.12.06},
}
Figure