摘要
针对未来时刻交通流量的预测问题,在考虑历史时刻车流量数据、日期属性、天气、降雨量等多方面影响因素的同时,提出一种考虑多方面影响因素的长短期记忆神经网络(LSTM*)模型。实验表明在对短期车流量进行预测时,LSTM*模型的准确性优于现有的基线方法;对长期车流量进行预测时,基于粒子群算法改进的长短期记忆神经网络(PSO-LSTM*)模型的准确性优于LSTM*模型。
In order to predict the future traffic flow,a long short-term memory neural network(LSTM*)model is proposed,which takes into account many influencing factors such as traffic flow data,date attribute,weather and rainfall at historical time.The experimental results show that the accuracy of LSTM*model is better than the existing baseline method when predicting short-term traffic flow.When predicting long-term traffic flow,the accuracy of the long short-term memory neural network model based on particle swarm optimization algorithm(PSO-LSTM*)is better than that of LSTM*model.
作者
丁梓琼
汤广李
张波涛
卢自宝
Ding Ziqiong;Tang Guangli;Zhang Botao;Lu Zibao(School of Physics and Electronic Information,Anhui Normal University,Wuhu 241000,China;Anhui Provincial Engineering Laboratory on Information Fusion and Control of Intelligent Robot,Wuhu 241000,China)
出处
《网络安全与数据治理》
2023年第8期52-58,共7页
CYBER SECURITY AND DATA GOVERNANCE
基金
国家自然科学基金面上项目(62071005)
安徽省自然科学基金项目(2008085MF199)。
