论文部分内容阅读
针对捕食模型、利用已知观测数据、在非线性最小二乘准则下,建立了基于增广的广义卡尔曼滤波模型来解决噪声背景下的高精度参数估计问题,并予以了验证.提出以相平面方程为约束的初值搜索算法,利用Matlab优化工具箱、深度搜索和剪枝加速等技术来提高搜索速度.还建立了二重规划模型以解决观测时间有误差时的高精度估计问题,并对时间误差作了正态分布检验.
Aiming at the predator-prey model, using the known observation data, a generalized Kalman filter model based on augmented nonlinear is established to solve the problem of high-precision parameter estimation under noisy background under nonlinear least squares criterion. Plane equation is a constrained initial value search algorithm, and the search speed is improved by using Matlab optimization toolbox, depth search and pruning acceleration etc. A dual programming model is also established to solve the problem of high-precision estimation with error in observation time. The time error was normal distribution test.