Research And Implementation Of Motion Monitoring System Based On Real-time Task Scheduling And Recommendation Algorithm

With the improvement of people's living standards,exercise and health have become the focus of attention,and exercise monitoring systems have become increasingly popular.At present,the sports monitoring system on the market is designed for the monitoring of physical fitness of school students.On the one hand,the system focuses on data collection through wearable hardware devices.There are few users,slow data updates,and low real-time system problems;On the other hand,the system lacks intelligent analysis of user movement data,usually only for simple data collection.To this end,this paper studies the real-time task scheduling and recommendation algorithm in the sports monitoring system,and then implements a set of sports monitoring system applied to fitness parks.The main work of this article includes the following:(1)In view of the shortcomings of the existing data collection system via Bluetooth base stations,this paper introduces a dynamic priority preemptive scheduling algorithm(Slack-Deadline-First,SDF)based on task margin and deadline to design and improve the existing task scheduling mechanism The core scheduling algorithm in,effectively avoids the phenomenon of multi-task conflict,and eases the pressure of data communication and data forwarding at the base station node.(2)Aiming at the problems of traditional matrix factorization recommendation technology,this paper proposes to apply implicit semantic model to SVD matrix factorization recommendation algorithm.Considering that the traditional Latent Factor Model(LFM)is only suitable for explicit feedback data,it is not suitable for implicit feedback data.In order to make better use of implicit feedback data and make it more accurate to push motion guidance to different users,this paper introduces the idea of weighting on the basis of implicit semantic model and proposes a weighted matrix decomposition based on implicit semantic model Recommended algorithm.Experiments show that this method has a high recommendation accuracy.(3)Based on the three-layer basic architecture of the Internet of Things,a motion monitoring system design scheme based on real-time task scheduling and recommendation algorithms is proposed,and the system is fully elaborated from three aspects: system architecture,logical structure,and functional modules.The paper finally gives the realization process of the system.The system was deployed in Jiangyan Chengdong Sports Park and achieved good results.