以一种比较优秀的林火预测模型作为相关研究的基础,首先分析了林火预测建模的基本方法、数据处理的基本流程和数据处理算法3个方面的可并行性,提出了一种适用于林火预测的3级并行计算模型,并给出了该模型的形式化描述和基本实现算法;然后采用开放遥感平台实现了该模型,并描述了实现算法和数据处理工作流的建模方法,同时讨论了该实现的优化方法;最后搭建了实验平台,验证了该模型的可行性,对比分析了单机和几种优化方案的实现效果.实验结果表明,应用该模型可以大大提高林火预测的实时性.
The paper uses a relatively good forest fire forecast model as the base for relevant study. First, the parallelism of some characteristics of this model is analyzed. These characteristics include the modeling method of forest fire forecast, the flow of data processing and the data processing algorithm. Then, a three-step parallel computing model for forest fire forecast is proposed the formal description and the basic implementation algorithm of the model are described. And then, the model is implemented on the basis of OpenRS. The implementation algorithm and the modeling method of data processing workflow for the implementation are given. In the same time, the optimization method for the implementation is discussed. Finally, the test bed for parallel computing is set up to verify the feasibility of the model. A comparative analysis is done to the effect of a uniprocessing program and several parallel processing programs. The experimental results show that the model can greatly improve the real time performance of forest fire forecast.