基于分数阶优化的极限学习机交通流短时预测模型

Short-term traffic flow prediction model based on extreme learning machine with fractional-order optimization

中文关键词  交通流预测  神经网络  分数阶累加  改进分数阶极限学习机

英文关键词  traffic flow prediction  neural networks  fractional-order accumulation  improved fractional-order extreme learning machine

中文摘要

      在智能交通系统中, 交通流预测在交通管理和减少拥堵方面发挥着关键作用. 本文旨在开发一种高效、非 迭代的神经网络方法用于交通流短时预测, 该方法被称为改进分数阶极限学习机(IFra-ELM). 通过分数岭回归算 法, 使得ELM算法在求解输出权重时具有更好的性能, 该方法被称为分数极限学习机(Fra-ELM). 在输入层之后添 加改进分数阶累加层, 将与原始输入进行拼接, 增强了ELM算法的特征提取能力. 通过改进分数阶累加层与分数极 限学习机相结合, 增强了模型对于不同场景下交通流的预测能力和鲁棒性. 实验结果表明, 与传统方法相比, IFraELM模型在交通流预测方面具有优越的性能. 该框架有潜力提高交通预测系统的准确性和效率, 有助于智能交通 基础设施的发展.

英文摘要

      In intelligent transportation systems, traffic flow prediction plays a key role in traffic management and congestion reduction. The current manuscript aims to develop an efficient, non-iterative neural network method for short-term prediction of traffic flow, which is called improved fractional order extreme learning machine (IFra-ELM). Through the fractional ridge regression algorithm, the ELM algorithm has better performance in solving the output weight. This method is called a fractional extreme learning machine (Fra-ELM). An improved fractional-order accumulation layer is added after the input layer to splice it with the original input, which enhances the feature extraction capability of the ELM algorithm. By combining the improved fractional-order cumulative layer with the fractional extreme learning machine, the model’s prediction ability and robustness for traffic flow in different scenarios are enhanced. Experimental results show that compared with traditional methods, the IFra-ELM model has superior performance in traffic flow prediction. This framework has the potential to improve the accuracy and efficiency of traffic prediction systems and contribute to the development of smart transportation infrastructure.