平行智能风沙防治系统构架与功能——以植物措施为例

风沙活动复杂多变,使得风沙危害治理周期长、难预测、难评估。目前的风沙防治模拟系统主要基于物理建模框架,由于在受控环境中构建,预测结果容易偏离实际系统。本文从风沙运动物理原理着手,结合人工智能技术构建知识和数据共同驱动风沙防治模型。针对系统的复杂性,提出了一种借鉴实际工程与虚拟实验相互学习的基于ACP(人工社会+计算实验+平行执行)平行系统理论的风沙平行防治系统,预测不同风沙防治行为对风沙活动以及防治效果的影响。其中防治行为为植被种植情况,防治效果通过地表荒漠化程度表达。同时对风沙防治效果进行三维可视化展示,为风沙防治工程项目提供实施前、实施期间的决策支持。

Framework and Function of Aeolian Sand Parallel Prevention and Control: Application of Artificial Intelligence Technology in Sand Prevention and Control

Aeolian sand is complex, changeable and hard topredict and assess, and its prevention and control is time-consuming. Currently,the simulation system of aeolian sand prevention and control is mainly based onthe physical-process modeling. The model prediction results are generally faraway from the real results due to the low-accurate measured result of aeoliansand activities. In this paper, a new system based on the ACP (ArtificialSocieties + Computational experiments + Parallel systems) approach wasintroduced. Based on the physical principle of sand movement, the system couldbe used to predict the different windproof effects of aeolian sand by applyingthe artificial intelligence for different managements (such as vegetation planting).The system could provide the decision support for the prevention and controlprojects of aeolian sand hazards through developing the model driven by the knowledgeand data to simulate and predict the performance of windproof of aeolian sandand building the three-dimensional sense visualization.