基于混合加速粒子群算法的捕食者-食饵模型参数估计

Parameters Estimation for Predator-Prey Model Based on the Algorithm of Hybrid Acceleration Particle Swarm Optimization

首先应用四阶紧致差分格式,对捕食者-食饵模型的参数估计转化为相应的优化问题,然后将加速粒子群算法与Nelder-Mead算法相结合,提出混合加速粒子群算法,用来优化模型参数.数值实验中,借助四阶龙格-库塔算法对模型进行仿真计算,与观测值比较表明该算法是可行的.

The parameter estimation problem of prey-predator model was transformed into the corresponding optimization problem by the use of fourth-order compact difference scheme in the paper.The hybrid algorithm of acceleration particle swarm optimization and Nelder-Mead algorithm was proposed to optimize the parameters of model.In the numerical experiments,Fourth-order Runge-Kutta algorithm was used to compute the model,and comparing the simulation results with observed values.The results of comparing show that the algorithm is feasible.