A Study On Hidden Markov Model Based Empty-Nest Home Security Monitoring

China has been entering aging society with the improvement of medical standards and implementation of population policy.Meanwhile, our country economy has caused the separation of the aging parents and their children who have jobs in the city, which makes empty-nest home becoming primary type for Chinese old people. Therefore, it is significant that how to take care of these elders is becoming one way to reduce the burden of the society and family.By doing some researches on household care products, technology and living condition of the elderly in empty-nest family. This paper focuses on the monitoring algorithm of home security monitoring. The following work has been accomplished:First, In view of the existing security monitoring algorithm’s high complexity and difficulty to realize, this paper suggests RRT algorithm for abnormal state detection. This algorithm can be applied on various equipments without considering the difference between them. Using Exponential Smoothing to optimize the collected data and establish efficient individual models adapting the old people in different regions with kinds of habits.Second, the physiology decline and chronic disease of the elderly is difficult to realize through daily observation. This paper suggests a HMM based method to judge physiology change. Through building models separately of usual and unsual condition, we can get the matching deree of the observed behavior sequence and trained models in order to judge the old person’s stateto find the risk of disease threat, reduce the social and families’ burden and improve the quality of live.Finally, based on the theories above, we use the simulation experiment to prove that the RRT and HHM based security monitoring algorithm feasible and effective.