混合式智慧型磁浮避震系統之設計與實現

CCUR > College of Natural Science and Engineering > College of Engineering > Department of Mechanical Engineering > thesis > Item 987654321/33595

Please use this identifier to cite or link to this item: https://irlib.pccu.edu.tw/handle/987654321/33595

Title:	混合式智慧型磁浮避震系統之設計與實現 Design and Implementation of Hybrid Intelligent Maglev Suspension System
Authors:	潘韋宏
Contributors:	機械工程學系數位機電碩士班
Keywords:	磁浮避震器類神經網路最小均方演算法基因演算法模糊控制滑動模式 Magnetic suspension system Artificial neural network Least-mean-square algorithm Genetic algorithm Fuzzy control Sliding mode
Date:	2016-06
Issue Date:	2016-08-15 14:09:21 (UTC+8)
Abstract:	磁浮系統是透過非接觸的方式，運用電磁力使載物平台具有漂浮、轉動或移動等行為的系統，其優點可有效降低機械式接觸造成的震動、摩擦和損耗，延長機械設備的壽命、減少維修頻率並可降低噪音等。如將磁浮系統應用於避震器，並安裝於足型機器人及交通運輸載具上即可克服惡劣的道路狀況，不但於崎嶇路面行走移動時，不會產生劇烈震動，能使足型機器人準確的採集數據、影像捕獲及夾爪定位，亦能提升交通載具的乘坐舒適性。然而磁浮避震系統具有高度非線性動態的特性，使得傳統控制器較難以使用。為克服此問題並加快系統反應時間，本論文提出兩種混合式智慧型磁浮避震控制系統，首先，以最小均方演算法或基因演算法來離線訓練類神經網路，做為系統之主控制器，緊接著，加入一種適應摸糊滑動控制器，做為系統的輔助控制器，以克服路面之崎嶇不平。最後計算出適當的激磁電流，藉以產生反抗磁力，以抵抗重力變化。為驗證所提方法之有效性，研究過程將此控制器以MATLAB模擬，實驗室的磁浮平台實驗及實作之磁浮避震器雛形在六足機器人之應用，從實驗數據來看，所提之磁浮避震器架構確有穩定的追蹤能力及較短的反應時間等效果。 Magnetic levitation system makes loading platform a floating, rotating or moving system by non-contact and electromagnetic force. Its advantages include effectively reducing the mechanical vibration, friction and wearing loss, prolonging the life of mechanical equipment and decreasing the fixing frequency and cutting down the noise, etc. If the magnetic levitation system applies into suspension vibrator and installs on foot robot, then it will not vibrate severely while moving on a bumpy road. Therefore, the magnetic suspension system (MSS) can make the collecting data, capturing images and gripping actions of foot robot accurately but also promote the riding comfort of transportation. However, the MSS possesses highly nonlinear dynamical characteristics such that the conventional controllers are difficult to use. In order to deal with this problem and to accelerate the speed of the system performance, two hybrid intelligent MSS controllers are proposed in this thesis. First, a genetic algorithm is utilized to train the neural network offline as the main controller of system. Then, an adaptive fuzzy sliding mode controller is added as an auxiliary controller of system. Finally, the appropriate electric current is calculated and sent to produce the resistance to magnetic force and resist the variety of gravity. Some MATLAB simulated results, experimental results and one implemented MSS prototype are provided to verify the effectiveness of the proposed architecture. From these results, the proposed MSS scheme possesses some features of stable tracking performance and shorter response time.
Appears in Collections:	[Department of Mechanical Engineering ] thesis

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	262	View/Open

View Licence

Loading...