害虫治理中的回归优化控制

The Regression Method for Optimal Control of Pests

在病虫害控制问题中,建立了从控制参数到过程分布的线性映射.基于这个映射,将个体模型中带有机会约束的优化控制问题转化为方便求解的条件极值问题.模拟显示,无论从运行时间,还是最终目标函数的取值,该方法都优于流行的近似矩封闭方法.

We consider optimal control policy for individual-based models in which the time evolution of the species abundances is described by a continuous-time Markov chain. For the master equation that is typically difficult to solve and its approximation is basi- cally with high dimension thus hard to embedded in the existing optimal procedure. We propose a linear functional approximation between the control parameters and the moments of the underlying process, based on which the control strategy can be formulated as a constrained optimization problem. It is validated more optimal compared with the existing methods.