摘要
针对农作物的生长特点,基于物理动力学方程,建立了农作物产量预测的状态空间模型,并结合反馈粒子滤波算法的优点,对某地区番茄的产量进行预测。仿真结果表明,与标准粒子滤波算法相比,基于反馈粒子滤波的农作物产量短期预测曲线更加逼近于实际产量曲线,预测精度有了一定的提高。
According to the growth characteristics of crops, the state space model of crops production short-term forecasting has been established based on the physical dynamics equation. So the model with feedback particle filter algorithm has been applied to production forecasting for the tomato of an area. The results of the simulation show that the short-term forecasting curve of crops production based on FPF is closer to the actual production curve. So the forecasting accuracy has been improved compared with standard particle filter.
出处
《武汉轻工大学学报》
CAS
2014年第4期46-49,共4页
Journal of Wuhan Polytechnic University
关键词
农产品产量
短期预测
反馈粒子滤波
crops production
short -term forecasting
feedback particle filter.
