Comprehensive Assessment of Sustainable Utilization of Water Resources Based on RBF Neural Network Evaluation Method

Abstract

Comprehensive assessment of sustainable utilization of regional water resources, including three subsystems: social economy, water resources and eco-environment, is a large complicated and systematic evaluating problem. The selection of the index system for the regional sustainable development of water resources was discussed herein. Furthermore, a neural network methodology for regional water resources evaluation was introduced. The method adopts the Radial Basis Function (RBF) architecture and dynamically penalized rival competitive learning algorithm, which is fast and repetitive, compared with most traditional techniques. Additionally, a new approach to produce training samples, testing samples and examining samples randomly distributed between the critical values was established. The sample used consists of 16 indexes, and by using the proposed methodology the sustainable evaluation of regional water resources could be successfully done. Results showed that from 1994 to 2005, the comprehensive index of water resources sustainable utilization in Zhengzhou city ascended with a fast speed of 0.023 per year, from sustainable development level IV to level III. The calculation with the model showed that, the comprehensive index of water resources sustainable utilization in the future planning years (2010, 2015, and 2020) could be enhanced with a speed of 0.011 per year. A sensitive analysis indicates that it is necessary to increase eco-environmental water consumption rate, industrial output from each cubic meter of water consumption, public greenbelt area per capita, foodstuff output from each cubic meter of water consumption, and to ensure that natural growth rate of population decreases steadily, which are playing an important role in improving sustainable utilization of water resources in Zhengzhou city.