Establishing optimal forest harvesting regulation with continuous approximation
Establishing optimal forest harvesting regulation with continuous approximation
复制标题
DOI：
10.1016/j.orp.2020.100158
复制
发表时间：
2020
影响因子：
2.5
通讯作者：
Kai Moriguchi;Tatsuhito Ueki;M. Saito
中科院分区：
文献类型：
--
作者：
Kai Moriguchi;Tatsuhito Ueki;M. Saito
虽然利用线性规划技术建立了最优森林采伐规则（区域森林更新计划），但其建模限制在实践中仍然是一个挑战。在这项研究中，我们开发了一个替代框架，利用连续逼近技术建立最优的森林采伐调节。将调节问题转化为求再生区域最优平滑控制问题，克服了包含非线性或更复杂模型的优化模型难以求全局最优的问题。在此基础上，建立了有效保证林分收敛到稳定状态和禁止滥伐的七维优化模型。利用该模型，我们在日本长野县建立了最优森林采伐法规。利用模拟退火法寻找最优解。基于极值理论的分析以及与网格搜索解的比较表明，最优解可能足够接近全局最优解。最佳解决方案建议控制木材供应，以保持早期的高价格，因为供需原木价格模型和使用净现值作为目标函数。由于该地区的盈利能力较低，解决方案建议尽可能推迟目标状态的实现。
While optimal forest harvesting regulations (regional forest regeneration schedules) have been established using the linear programming technique, its modeling restriction is still a challenge in practice. In this study, we developed an alternative framework for establishing an optimal forest harvesting regulation using the continuous approximation technique. The regulation problem was reformulated as a problem to find optimal smooth control of regeneration area, thereby overcoming the difficulties in finding the global optimum for the optimization models that include nonlinear or more complex models. A seven-dimensional optimization model was developed involving efficient assurance of the convergence to a stable forest state and prohibition of clearcutting in immature stands. Using this model, we established optimal forest harvesting regulations in Nagano Prefecture, Japan. Simulated annealing was utilized to explore optimal solutions. Analyses based on extreme value theory and comparison with solutions produced by grid search indicated that the best solutions may be sufficiently close to global optima. The best solutions suggested controlling timber supply to keep prices high in the early years, due to the supply–demand log price model and the use of net present value as the objective function. Because of the low profitability of the region, the solution suggested delaying the achievement of the target state as far as possible.