Please use this identifier to cite or link to this item:
https://ptsldigital.ukm.my/jspui/handle/123456789/486817Full metadata record
| DC Field | Value | Language |
|---|---|---|
| dc.contributor.advisor | Mamun Ibne Reaz, Prof. Dr. | |
| dc.contributor.author | Mohammad Marufuzzaman (P68081) | |
| dc.date.accessioned | 2023-10-11T02:25:45Z | - |
| dc.date.available | 2023-10-11T02:25:45Z | - |
| dc.date.issued | 2017-04-24 | |
| dc.identifier.other | ukmvital:86175 | |
| dc.identifier.uri | https://ptsldigital.ukm.my/jspui/handle/123456789/486817 | - |
| dc.description | Smart home technologies encourage and allow elderly and physically challenged people to live longer in their own homes. An inhabitant performs a unique pattern or sequence of tasks repeatedly in a home environment. Therefore, in order to build a smart home for people who need special care, it is necessary to recognize the unique pattern of tasks for predicting the next service. Thus, a reliable prediction algorithm is the key to an intelligent network of home appliances, and the algorithm should response faster to execute the decision. In this research, an activity prediction system is proposed, which consists of an algorithm called sequence prediction via all discoverable episodes (SPADE) for inhabitant activity prediction. In a smart environment, multiple agents such as time, location and user need to be considered to predict the next activity. Therefore, this research utilized a two-room house as a testbed, which employed all the smart home agents. The home identifies different users and locations where the time and the state of the home appliances were recorded. All the appliances of the testbed were networked together and all the activity information was stored in a dataset. Noise filter was applied to remove the unwanted data and generate the sequences. Dual state properties of home appliance were utilized to extract episodes from the sequence. Classified episodes were processed and were arranged in a finite order Markov model. A method motivated by Prediction by Partial Matching (PPM) algorithm was applied to select the next event from the frequencies of variable length episodes. Finally, the next state with location was predicted by the decision agent. Multiple user activity datasets were used for validation and accuracy with execution time was calculated. Moreover, the algorithm was tested using a wellknown smart home dataset from MavLab for performance comparison. Experimental results showed that the designed smart home successfully identified all the agents related to user activity. The algorithm extracted 689 predictions and their location with the highest accuracy of 90%. The total execution time was 94 seconds, which was lower compared with the previous research works. Similar results have been found for testbed datasets too. A hardware prototype was designed using Raspberry Pi neral-purpose input output (GPIO) interfaces of Raspberry Pi 2 B were used to communicate with the prototype testbed and were able to execute the next activities. Since human behavior shows natural temporal patterns, multiple agents based solution can be used to predict next event more accurately, which will assist in future home automation.,Certification of Master's/Doctoral Thesis" is not available | |
| dc.language.iso | eng | |
| dc.publisher | UKM, Bangi | |
| dc.relation | Faculty of Engineering and Built Environment / Fakulti Kejuruteraan dan Alam Bina | |
| dc.rights | UKM | |
| dc.subject | Smart home | |
| dc.subject | Algorithm | |
| dc.subject | Intelligent network | |
| dc.subject | Prediction system | |
| dc.subject | Universiti Kebangsaan Malaysia -- Dissertations | |
| dc.title | Design and implementation of sequence prediction algorithm using multiple agents for smart home | |
| dc.type | Theses | |
| dc.format.pages | 123 | |
| dc.identifier.barcode | 002681(2017) | |
| Appears in Collections: | Faculty of Engineering and Built Environment / Fakulti Kejuruteraan dan Alam Bina | |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| ukmvital_86175+SOURCE1+SOURCE1.0.PDF Restricted Access | 379.02 kB | Adobe PDF |  View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.