摘要
为了有效缓解船舶交通拥堵和提高通航效率,对海洋港口和航道管理提供一个更可靠的数据,设计了一种基于禁忌算法优化神经网络的海洋船舶流量预测的方法;首先,建立了基于禁忌算法优化神经网络的海洋船舶流量预测模型;然后,设计了三层的脊波神经网络结构,提出了采用禁忌算法优化脊波神经网络结构参数的具体方法,从而得到一个初始化的脊波神经网络流量预测模型;然后,采用有标签的训练样本数据集对网络进行训练,将满足误差阈值的训练模型作为最终的预测模型;以某港口为例进行仿真实验,结果表明文中得到的预测结果与真实值较为接近,且与其它方法相比,具有更好的预测效果。
In order to relieve the ship traffic, improve the ship efficiency and provide a reliable data for ocean port and ship management, a traffic prediction method based on tabu algorithm optimizing the neural network. Firstly, a traffic prediction model based on tabu algorithm and neural network is proposed. Then, a three--layer neural back bone network structure is designed, the tabu algorithm is introduced to optimize the parameters of the neural network, therefore, a traffic prediction model is proposed to initialize the back bone neural network. Then, the labeled samples are used to train the neural network, the final prediction model satisfying the threshold is used as the final prediction model. The experiment is operated and its result approaches the true prediction result. Compared with the other methods, our method has better prediction effect.
出处
《计算机测量与控制》
2016年第11期124-126,共3页
Computer Measurement &Control
基金
三亚市院地科技合作项目(2014YD11)
关键词
流量预测
禁忌算法
神经网络
海洋船舶
traffic prediction
tabu algorism
neural network
ocean ship
