基于改进PSO的洞庭湖水源涵养林空间优化模型

以结构化森林经营思想为理论基础,从与水源林涵养水源、保持水土功能密切相关的林分物种组成(树种混交)、种内及种间竞争、空间分布格局、垂直结构4个方面选择混交度、竞争指数、角尺度、林层指数、空间密度指数、开阔比数作为水源林健康经营和林分空间结构优化的目标函数,建立洞庭湖水源林林分多目标空间优化模型,应用改进的群智能粒子群算法求解林分空间结构优化模型,并针对模型输出的目标树空间结构单元制定周密的经营策略.研究结果表明,优化模型能准确定位林分空间关系的薄弱环节,调控措施能显著改善林分空间结构,有利于促进森林生态系统的正向演替,为恢复洞庭湖水源林生态功能和健康经营提供理论依据和技术支撑.应用优化模型进行水源涵养林健康经营突破了传统森林经营模式,为智能信息技术在森林空间经营中的应用提供了新的思路.

A space optimization model of water resource conservation forest in Dongting Lake based on improved PSO

Forest spatial structure plays an important role in the health and stability of forests, so studying the spatial structure and analyzing the adjustment strategy of Dongting Lake water conservation forest has very important theoretical and practical significance for the restoration of the Dongting Lake forest ecosystem. This research was based on the theoretical principles that underlie structure-based forest management. We established a forest spatial structure optimization model to study the regulation structure and control strategy of the Dongting Lake water conservation forest using the search features of a particle swarm algorithm, and we combined the multi-objective optimization of water conservation forest spatial structure with the Particle Swarm Optimization (PSO)algorithm. Four important aspects of forest stand spatial structure are species composition, intraspecific and interspecific competition, spatial distribution pattern and vertical structure. A number of measures related to these aspects including the degree of mingling, competition index, uniform angle index, stand layer index, spatial density index and open comparison index were selected as the objective functions for water conservation forest health management and optimization of forest spatial structure. In this paper, the entire stand was mapped to the goals solution space of the particle groups and each individual tree was used as a solution in the PSO space. This was done to translate the stand space optimization problem into an optimization process. A PSO optimization algorithm was used that iteratively improved a fitness function, based on the multi-objective function, to identify the optimum solution to the stand spatial structure model. Based on the output of this model, we developed a business strategy for the target tree space structure unit. Based on an analysis of 10 block fixed sample case studies from Changde Hefu national forest park, Taoyuan Guniu mountain, Huangshi, Longtang and the Shejiaping area to the west of Dongting Lake, the research results show that the space optimization model of water resource conservation forest in Dongting Lake is used for forest spatial structure adjustment. The model output with regards to the target tree and its spatial structure unit as the control object was adjusted in terms of tree species composition, competition regulation, horizontal pattern and layering, using single selective cutting, replanting and other management measures. Comparison of the forest stands before and after optimization adjustment showed that the forest spatial structure indexes were very different. The non-spatial structure, tree diameter, species order and age class distribution index remained unchanged. The degree of health increased significantly while the forest layer index remained unchanged. The other three aspects that contribute to the space structure index improved to different degrees. The optimized structure tended to be healthier and more stable while reducing the top target tree competition pressure. From the overall optimization of the stand structure, the optimization model and algorithm can accurately locate weaknesses in stand spatial relationships. Management measures can significantly improve forest spatial structure, which can promote the positive succession of forest ecological systems, become the theoretical basis, and provide technical support for the restoration of Dongting Lake water forest ecological function and health. The optimization model and algorithm proposed in this paper are superior to traditional forest management for water conservation forest health management, and provide a new application of intelligent information technology to forest spatial management.