基于RBF神經(jīng)網(wǎng)絡(luò)的糧庫溫度預(yù)測

由于糧庫溫度是非線性的時間序列，文章提出了基于RBF神經(jīng)網(wǎng)絡(luò)的糧庫溫度預(yù)測模
型。該模型優(yōu)于傳統(tǒng)的糧庫溫度分析方法，又避免了BP算法容易陷入局部極小點和收斂速度慢的缺點。根據(jù)實驗的仿真結(jié)果顯示，該模型對于糧庫溫度的預(yù)測效果較好。
關(guān)鍵詞：糧庫溫度，徑向基函數(shù)神經(jīng)網(wǎng)絡(luò)，非線性時間序列
Abstract: Because of the grain depot temperature is a nonlinear time series, this paper presents the grain depot temperature forecast model based on tradition basic function neural networks. The model is not only better than the tradition stock technical analysis method but also is avoiding the defects which relapse into the partial smallest point and convergence rate is little of BP algorithm. The simulation results of the experiment show that model is efficient to forecast the trend of grain depot temperature.
Keywords: Grain depot temperature, RBF neural network，nonlinear time series