未知信號估測技術應用在車輛狀態估測與控制

Please use this identifier to cite or link to this item: http://mitustr.lib.mitust.edu.tw:8080/ir/handle/987654321/1189

Title:	未知信號估測技術應用在車輛狀態估測與控制 Application of Unknown Input Filtering to Vehicle State Estimation and Control
Authors:	康森 Kang, Sen
Contributors:	電機與電子工程系
Keywords:	未知信號估測強健觀測器設計車輛強健控制 Unknown input filtering Robust observer design Vehicle robust control
Date:	2018-07-25
Issue Date:	2018-07-31 22:27:40 (UTC+8)
Abstract:	車輛控制需要估測系統來提供精確之車輛動態資訊，其估測效果往往會因為實際車輛和車輛模型間的差異，而產生不同之結果。車輛系統模型包含了許多未知的系統參數或信號，如輪胎對地面的摩擦剛度變化、致動器或感測器失效模式及道路坡度等，此皆會造成系統狀態估測或車輛控制之困難度。若採用忽略此些重要系統參數之策略，則其引發之估測效果或控制結果，將會大打折扣，甚而影響車輛行車之安全性。本論文研究目的即在建立一套未知信號估測方法，藉以解決車輛系統所面臨之各種未知信號的影響，以期建立一最佳車輛強健觀測器和控制器設計方法，用以解決日益複雜之車輛控制器設計問題。本論文研究方法首先推導具道路坡度之車輛運動學模型，藉以了解車輛系統所內含之未知信號。其次，探討具量測失效模式之車輛狀態估測問題，並將其轉換為一等效未知信號估測問題。然後，嘗試應用現有未知信號估測技術，來建構一套有效車輛強健觀測器和控制器設計方法。最後，透過廣泛之模擬結果來驗證研究成果之有效性。本論文研究成果包含有： (1) 提出一最佳車輛強健觀測器和控制器設計方法，以解決具未知信號之車輛系統狀態估測問題。 (2) 針對具未知信號之車輛系統模型，提出一整合式系統建模技巧，以同時解決車輛系統之狀態及未知信號估測問題。 (3) 針對車輛在道路具有坡度所產生之未知信號干擾現象，提出一有效之未知信號估測技術，以同時解決車輛系統之狀態及坡度估測問題。 (4) 針對車輛系統可能存在之量測失效模式，提出一有效之未知信號估測技術，以解決車輛系統之狀態及量測失效值估測問題。 (5) 提出利用線性平方調節器設計方法，來穩定車輛在受到外在環境干擾時所產生之運動作用，以減少車輛傾覆之機會。關鍵字：未知信號估測、強健觀測器設計、車輛強健控制。 Abstract Vehicle control requires an estimation system to provide accurate vehicle state information, which may have different results due to modeling errors. The vehicle system contains many unknown system parameters or signals, such as tire-to-ground frictional stiffness changes, actuator or sensor faults and road bank angle, which can cause system state estimation or vehicle control difficulties. If these unknown parameters are ignored, the performance of obtained estimation or control may be degraded, which may even affect the safety of vehicle driving. The purpose of this thesis is to apply unknown input filtering method to solve the state estimation problem of a vehicle system with unknown inputs, and hence to develop an optimal vehicle robust observer and controller design method to achieve increasingly complex vehicle designs. First, the thesis presents a vehicle system model with road bank angle, which clearly shows the unknown inputs existed in a vehicle system. Second, the vehicle state estimation problem with measurement fault is addressed, which then is converted into an equivalent unknown input filtering problem. Next, using the existing unknown input filtering method, an effective vehicle robust observer and controller design method is developed. Finally, the effectiveness of the research results is validated through extensive simulations. The research results of this thesis include: (1) to propose an optimal vehicle robust observer and controller design method to solve the state estimation problem of a vehicle system with unknown inputs. (2) to propose an integrated system modeling technique to simultaneously solve the state and unknown inputs estimation problem of a vehicle system. (3) to propose an effective unknown input filtering method to compensate for the unknown interference phenomenon caused by the bank angle of the road, and hence to solve the simultaneous vehicle state and road bank angle estimation problem. (4) to propose an effective unknown input filtering method to solve the state and measurement fault estimation problem of a vehicle system and. (5) to propose a LQR regulator design to stabilize the motion of a vehicle system when it is disturbed by the external environment. Keywords: Unknown input filtering, Robust observer design, Vehicle robust control.
Appears in Collections:	[電機與電子工程系] 學位論文

Files in This Item:

File	Description	Size	Format
Application of Unknown Input Filtering to Vehicle State Estimation and Control.pdf	Unknown input filtering, Robust observer design, Vehicle robust control	930Kb	Adobe PDF	1310	View/Open

View Licence

Loading...