QQ客服 官方微博 反馈留言

	FLOPS: An Efficient and High-precisionPrediction on Available Parking Spaces in a LongTime-span
	Min sun; Zhongmin Li; Lei Peng; Huiyun Li; Xiangyan Fang
	2018
会议日期	2018
会议地点	Hawaii，USA
英文摘要	Available parking spaces prediction is an important AI technology in the parking guidance system (PGS). The current prediction technologies based on time series and neural networks can achieve quite high accuracy in the short-term prediction. However, with the increase of the prediction steps or time-span, the accuracy of the prediction is dramatically reduced, and lose most of the information in process. In this paper, we propose a prediction method, named FLOPS, that can keep the characteristics of data change in the long time-span. This method uses the idea of fuzzy information granulation (FIG) to obtain the feature data sets. Then a long short-term memory (LSTM) network is trained to predict the future feature data set. Finally working with the spline interpolation, the predicted feature data set is to reconstruct the full continuous curve of the available parking spaces in the time-span. The simulation shows that FLOPS has higher accuracy than the traditional method when predicting over the long time-span. Also, the computing efficiency of FLOPS is higher than the traditional prediction method under the condition of the approximate prediction accuracy.
内容类型	会议论文
源URL	[http://ir.siat.ac.cn:8080/handle/172644/13731]
专题	深圳先进技术研究院_集成所
推荐引用方式 GB/T 7714	Min sun,Zhongmin Li,Lei Peng,et al. FLOPS: An Efficient and High-precisionPrediction on Available Parking Spaces in a LongTime-span[C]. 见:. Hawaii，USA. 2018.

个性服务

查看访问统计

相关权益政策

暂无数据

收藏/分享

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。