Please use this identifier to cite or link to this item:
https://ptsldigital.ukm.my/jspui/handle/123456789/476265Full metadata record
| DC Field | Value | Language |
|---|---|---|
| dc.contributor.advisor | Salwani Abdullah, Assoc. Prof. Dr. | |
| dc.contributor.author | Bahareh Nakisa (P62512) | |
| dc.date.accessioned | 2023-10-06T09:15:30Z | - |
| dc.date.available | 2023-10-06T09:15:30Z | - |
| dc.date.issued | 2014-02-24 | |
| dc.identifier.other | ukmvital:80366 | |
| dc.identifier.uri | https://ptsldigital.ukm.my/jspui/handle/123456789/476265 | - |
| dc.description | As the research grows, multi-robot systems present more advantages and efficiency than single robot system. In multi-robot system, each robot interacts with each other and with its environment. This collaboration enables the multi-robot system to accomplish tasks, which are not executable in single robot system. Multi-robot systems are utilized in activities like object finding, search and rescue. There are many methods that are used on the multi-robot systems. One of the methods is Particle Swarm Optimization (PSO) that was inspired by bird flocking and fish schooling. Although PSO is represented in solving many problems as well as multi-robot systems, it has its shortcomings when placed in the complex environments. One of the defects of the PSO is premature convergence of particles. When particles move in search space, they tend to converge to the same region. This means that there is a decrease in the diversity of the search space when the particles converge to local optima. The swarm is located in the state of stagnation when the agent’s momenta decrease. This situation leads to fitness stagnation in the swarm. This also shows that the PSO does not have the ability to escape from the local optima. Another shortcomings of the PSO is about balancing between exploration and exploitation. The PSO is found to perform well in exploration region as opposed to the exploitation region. This situation causes the increase of undesirable search time.In this study, mobile robots are spread out in complex environment, which also include static obstacles while finding the target the mobile robots interact with each other and must move around the search space without colliding with each other or with any static obstacles placed in the environment. Therefore this study, an algorithm called ATREL-PSO is proposed to solve the problem. This simple and effective algorithm maintains high diversity between the population agents and also creates an efficient balance between exploration and exploitation. A 2D simulator that can show the results of the multi-robot system based on ATREL-PSO algorithm was developed. The simulator showed that the diversity of the swarm increases and search time decreases when more particles are found to escape from the local optima. As a conclusion, ATREL-PSO algorithm performs better and can reach the target less than 10 minutes whiles the PSO algorithm in some cases cannot reach the target in 50 minutes.,Sejajar dengan perkembangan kajian, sistem multi-robot menunjukkan banyak kelebihan dan keberkesanan berbanding sistem robot tunggal. Dalam multi-robot, setiap robot berinteraksi dengan robot lain dan dengan persekitarannya. Koloborasi seperti ini membolehkan sistem multi-robot melaksanakan tugasan yang tidak dapat dilaksanakan oleh sistem robot tunggal. Sistem multi-robot digunakan dalam aktivitI seperti mencari objek, pencarian dan penyelamatan. Terdapat banyak kaedah yang digunakan pada sistem multi-robot. Salah satu daripadanya ialah pengoptimisasian kerumunan partikel atau particle swarm optimization (pso) yang diilhamkan oleh kawanan burung dan kawanan ikan. Walaupun pso digunakan untuk menyelesaikan banyak masalah seperti dalam sistem multi-robot, pendekatan tersebut ada beberapa kekekangan apabila ditempatkan dalam persekitaran kompleks. Salah satu kepincangan pso asas ialah penumpuan pramatang partikel. Semasa partikel bergerak keruang carian, partikel tersebut cenderung tertumpu kesatu kawasan yang sama. Ini bermakna terdapat pengurangan diversiti dalam ruang carian apabila partikel bergerak ke optima tempatan. Kerumunan itu berada dalam keadaan tersendat apabila agen momenta menjadi lebih perlahan. Keadaan ini menjurus pada kekakuan daya kerumunan. Hal ini juga menunjukkan bahawa pso asas tidak mampu melepaskan diri daripada optima tempatan. Kekangan lain pso asas ialah tentang penimbang antara eksplorasi dan eksploitasi. Pso asas didapati lebih berjaya dalam ruang ekplorasi berbanding ekploitasi. Situasi ini menyebabkan peningkatan masa carian yang tidak diingini. Dalam kajian ini, robot-robot mudah-alih tersebar luas dalam persekitaran kompleks yang mana turut termasuk ialah halangan statik semasa mencari sasaran. Robot-robot mudah-alih berinteraksi sesama sendiri dan perlu bergerak serata ruang carian tanpa bertembung dengan robot lain atau halangan statik yang ditempatkan dalam persekitaran. Justeru, kajian ini mencadangkan suatu algorithma yang dinamakan atrel-pso untuk menyelesaikan masalah tersebut. Algoritma mudah dan berkesan ini mengekalkan diversiti tinggi antara agen populasi serta mewujudkan pertimbangan cekap antara eksplorasi dan eksploitasi. Simulator 2 dimensi (2d) yang dapat menunjukkan keputusan sistem multi-robot berdasarkan algoritmaatrel-pso telah dibangunkan. Simulator tersebut menunjukkan bahawa diversiti kerumunan meningkat dan masa carian berkurangan apabila lebih banyak partikel didapati melepaskan diri daripada optima tempatan. Sebagai rumusan, oleh sebab algorit maatrel-pso lebih mampu melaksanakan algoritmaatrel-pso lebih berkesan berbanding algoritma lain dalam aspek koloborasi multi-robot.,Master of Science | |
| dc.language.iso | eng | |
| dc.publisher | UKM, BANGI | |
| dc.relation | Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat | |
| dc.rights | UKM | |
| dc.subject | Mathematical optimization | |
| dc.subject | Artificial intelligence | |
| dc.subject | Universiti Kebangsaan Malaysia -- Dissertations | |
| dc.subject | Dissertations, Academic -- Malaysia | |
| dc.title | Simple diversity guided particle swarm optimization with local search on multi-robot system | |
| dc.type | theses | |
| dc.format.pages | 80 | |
| dc.identifier.callno | Q337.3.N337 2014 3 tesis | |
| Appears in Collections: | Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat | |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| ukmvital_80366+SOURCE1+SOURCE1.0.PDF Restricted Access | 2.33 MB | Adobe PDF |  View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.