Resource Production Under Uncertainty: A Stochastic Control Approach to Timber Harvest Scheduling
用线性二次高斯方法解决不确定性下的木材采伐调度问题，在斯坦尼斯劳斯国家森林案例中模拟显示管理偏好和目标函数设定对采伐水平及不确定性成本影响显著。
Abstract Most empirical natural resource management problems are intertemporal optimization problems under uncertainty. The Linear‐Quadratic‐Gaussian (LQG) method derives an approximate stochastic solution to such problems. The error in modeling resource dynamics and error in observing resource stocks are incorporated explicitly into the LQG problem solution. The LQG is outlined briefly and applied to the harvest‐scheduling problem on the Stanislaus National Forest. The importance of knowing management preferences precisely is shown by stochastic simulations. In particular, harvest levels and the cost of a given source of uncertainty are shown to be quite sensitive to the objective function specification.