TW200806513A

TW200806513A - A system and method for predicting vehicle movement

Info

Publication number: TW200806513A
Application number: TW95127867A
Authority: TW
Inventors: Tsung-Lin Chen; Ling-Yuan Hsu
Original assignee: Univ Nat Chiao Tung
Priority date: 2006-07-28
Filing date: 2006-07-28
Publication date: 2008-02-01
Also published as: TWI294376B

Abstract

A system and method for predicting vehicle movement are disclosed. The system includes a sensor part, an estimator, and a predictor. The sensor part consisting of three sensors is used for measuring longitudinal velocity and lateral acceleration of mass center and length variations of suspension springs of vehicle. The estimator estimates plural physical quantities for movements of vehicles. The predictor receives the physical quantities, obtained from the estimator, and then predicts the future movements of the vehicle. Thus, the system and method predict movement of the vehicles by measuring three physical quantities and have advantages of low cost and accuracy.

Description

200806513 九、發明說明：【發明所屬之技術領域】本發明係涉及一種車輛動態預測系統與方法，尤其是關於利用三個量測之物理量，搭配利用數值演算法則之一估測杰以推算出相關之物理量，並利用利用數值演算法則之一預測為接受這些物理量而可預測車輛之動態行為。【先前技術】對於車輛的行駛狀況，較先進的車輛電子系統皆會配備有車輛一设警告I置，用以預測車輛於何種環境、何種駕駛月/兄之下，可此造成車輛翻覆，而可適時的提出馨告或是直接予以修正。相關技術如美國專利公報公告號第刪2奶號專利係利用侧傾角感測器、俯仰角速度感測器、縱向加速規、側向加速規與垂直加速規，並利用擴增卡曼遽波器 (Extended Kalman Filter)來估測目前的側傾角與俯仰角，且再利用侧傾角速度與麵角套用至經驗加權函數來、、泳來的側傾角’同樣地利用俯仰角速度與俯仰角套用至經驗加權函數來預測未來的俯仰角。如帛61923〇5號專利採用側傾角感測器、俯仰聽度感·、縱向加速規、侧向加域，、垂直加速規’並细擴增卡曼濾波器來估測目前的侧、角俯fp角ϋ利用估測的侧向力口速度與縱向速度套用至經驗公式來獲得續擺麵職生之偏差值，並湘此偏差 5 200806513 值與侧傾角套用至經驗加權函數來預測未來的侧傾角，而此偏差值與俯仰角套用至纖加權函數來翻未來的俯仰角。如第6556908號專利感測器的選用採用緩向加速規、側向加速規、橫擺角制器、側傾角速度感測器與輪胎速度感測器，經由侧向加與麵肢度_錄驗加偏數來算出過去與目W之侧傾肖速度，並藉由另—經驗加權函數來運用過去與目前之側则速度鱗算其蘭校正後之側傾梯度，並藉由此側傾梯度找出較為可靠性_傾角__ 如第6631317號專利感難包健向加速規、侧向加速規、橫擺角感測器、侧傾角速度感·與輪胎速度感測器，接著側向加速規、軸加速規、侧傾驗度制器、橫擺角速度感測器與輪胎速度感測ϋ所測得之物理量套用至簡化公式來歧暫態_賴敎麵肖，细這兩健與兩個數值運算濾波器來估測出側傾角以判斷翻覆事件。但上述專利皆利用四至五種感測器方能估算出車輛的麵角，喊藉由概加#秘或㈣化公絲預測車輛侧傾角’亚以此物理1來宣告是涵覆L賴測之車輛側傾角皆無考慮道路狀況之影響。【發明内容】鑒於MJi關題，本發日撕欲解決之問題在於提供一種車輛動恶預/縣贿方法，藉料_三滅·所得之物理 200806513 經估算且利包含道路狀況之車輛模型來得出車輛於未來守間的動,％，藉由車輛於未來時間之側傾肖來 : 丕合知菊丨平辆疋200806513 IX. DESCRIPTION OF THE INVENTION: TECHNICAL FIELD The present invention relates to a vehicle dynamic prediction system and method, and more particularly to utilizing three physical quantities of measurement, and using a numerical algorithm to estimate the correlation to calculate correlation The physical quantity, and one of the numerical algorithms, is predicted to accept the physical quantities to predict the dynamic behavior of the vehicle. [Prior Art] For the driving condition of the vehicle, the more advanced vehicle electronic system will be equipped with a warning set for the vehicle to predict the environment and the driving month/brother of the vehicle, which may cause the vehicle to overturn. And you can make a good announcement at a timely time or directly correct it. Related Art, such as the U.S. Patent Publication No. 2, the second patent, utilizes a roll angle sensor, a pitch angular velocity sensor, a longitudinal acceleration gauge, a lateral acceleration gauge, and a vertical acceleration gauge, and utilizes an augmented Karman chopper. (Extended Kalman Filter) to estimate the current roll angle and pitch angle, and then use the roll angle and the face angle to apply the empirical weighting function, and the roll angle of the swim' is the same as the pitch angle and pitch angle applied to the experience. A weighting function to predict the future pitch angle. Rugao 61923〇5 patent uses a roll angle sensor, a pitching sense, a longitudinal acceleration gauge, a lateral acceleration domain, a vertical acceleration gauge, and a finely amplified Karman filter to estimate the current side and angle. The ff angle ϋ uses the estimated lateral force velocity and longitudinal velocity to apply the empirical formula to obtain the deviation value of the continuation of the posture, and the deviation 5 200806513 value and the roll angle are applied to the empirical weighting function to predict the future. The roll angle, and the offset and pitch angles are applied to the fiber weighting function to turn the future pitch angle. For example, the sensor of No. 6556908 adopts a slow acceleration gauge, a lateral acceleration gauge, a yaw angle controller, a roll angle velocity sensor and a tire speed sensor, and is laterally added to the limbs. The addition of the partial number is used to calculate the roll velocity of the past and the target W, and the gradient of the past and the current side is used to calculate the roll gradient after the correction of the current and the current side. Gradient to find more reliable _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ The physical quantity measured by the acceleration gauge, the shaft acceleration gauge, the roll gauge controller, the yaw rate sensor and the tire speed sensing 套 is applied to the simplified formula to distinguish the transient state. Two numerical operation filters are used to estimate the roll angle to determine the flip event. However, the above patents use four to five kinds of sensors to estimate the face angle of the vehicle, and the singularity of the vehicle is predicted by the addition of #秘或(四)化公丝. The vehicle roll angle is not affected by the road conditions. SUMMARY OF THE INVENTION In view of the MJi issue, the problem to be solved by the haircut is to provide a vehicle-initiated/premature bribe method, which is derived from the physics of 2008-0613, which is estimated and included in the vehicle model of the road condition. The movement of the vehicle in the future, %, by the side of the vehicle in the future time: 丕合知菊丨平疋

义因此，為解決上述—種車姉態預啦統的技街問題， j明揭露—種車姉態預測系統，係肋預測—車麵於未來日獨之物行為，此車輛祕刪祕包含—❹彳部、— 估測器與—預測器，其中感測部由下列桃組成： —第-感測器係測量車輛之f心的縱向速度，並產生— 縱向速度信號；—第二感測器係測量車輛之質心的側向加速度’並產生，向加速度信號及—第三感測器_量車懸掛彈簧之長度魏，並產生—紐變化信號。而估測器包含—由一車輛動態模型所建構之非線性狀驗察器、;且接受縱向速度信號、側向加速度信號與長度變化L5虎’亚經由非線性狀態觀察器而估測出車孝兩之一側傾角、一側傾角速度、-側傾角加速度、-俯仰角、-俯仰角、、俯仰角加速度、一橫擺角、一橫擺角速度、一橫擺角加速度、—縱向位移、—縱向加速度、-侧向位移、i侧向速度、-垂直位移、—垂直速度、—垂直加速度與一輪胎角速度。而預·包含—車她雜型，且接受縱向速度信號侧向加速度城、長度變化信號、側傾角、侧傾角速度、侧傾角加速度、鮮卩肖，驗度、料角加速度、讎 200806513 角、橫擺角速度、橫擺角加速度、縱向位移、縱向加速度、側向位移、側向速度、垂直位移、垂直速度、垂直加速度與輪胎角速度，而預測車輛於未來時間之運動行為。車輛動態預測系統更包含—判斷器，係接收側傾角或俯仰角，亦或同時接收兩者，而估算側傾角或俯仰角之變化趨勢而判斷車條未來時暇涵覆雜，錄算出未來車輛之翻覆位置。射麵_鱗角之變化趨勢係為側傾角或俯仰角是否為發散函數’車輛於絲時暇否翻覆狀態。並且觸料透上物估算之_量祕算出未來車輛之翻覆位置。，且第一感測器用以測量車輛之質心的側向加速度，可用測量車輛之質蝴魏歧度储，並鼓 :號’村使估測器產生相同的作用。另亦可用測量貝。的&擺角代賴向加速度，並產生—橫同樣可使估測器產生相同的作用。 4就揭露種車她誠測方法峨簡題，本發明、一車動關測方法，係用以預測—車I峨未來時間之運動仃為’此車輛動態翻方法包含下列步驟：測量車1里車^之質心峽向速度，並產生—縱向速度信龍； =車^之質叫侧向加速度，並產生—侧向加速度信號；車柄之懸掛彈簧之長度變化，並產生-長度變化信號； 200806513 接受縱向速度信號、側向加速度信號與長度變化信號，而估測出車輛之一侧傾角、一側傾角速度、一側傾角加速度、一俯仰角、一俯仰角速度、一俯仰角加速度、一横擺角、一橫擺角速度、-橫擺角加速度、一縱向位移、一縱向加連度、 -側向位移、-側向速度、—垂直位移、—垂直速度、一垂直加速度與-輪胎角速度；及接受縱向速度信號、侧向加速度信號、長度變化錢、側傾角、側傾角速度、舰角加速度、俯仰角、俯仰角速度、俯仰角加速度、橫擺角、橫擺角速度、橫擺角加速度、縱向位移、縱向加速度、側向位移、側向速度、垂直位移、垂直速度、垂直加速度與輪月台角速度，而預測車輛於未來時間之運動行為。此車輛__方法更包含接_傾角，算側傾角之k化趨勢而判斷車輛於未來時間是否翻覆。曰其令側傾角之變化趨勢的步驟係為估算側傾角或俯仰角疋否為發散函數的步驟。 t此車輛祕觸綠更包含魏俯仰肖，⑭算之變化趨勢*觸車祕未來時岐否翻覆。 ”中俯仰肖之變化趨勢的步職為轉俯仰肖是否發散函數的步驟。 — 2車輛動態酬方法更包含估算鱗辅之翻覆位置。其中測量車麵之質心的側向加速度，可關量車輪之質 200806513 ^的知'擺角速度代替，.並產生一橫擺角速度訊號，亦可估算出相同的結果。另亦可制量車輛之質心的橫擺肖代替測量質心的側向加速度’而產生一橫擺角速度信號，同樣可估測异出相同的結果。Therefore, in order to solve the above-mentioned problem of the technical street of the car-like pre-existing system, j Ming exposes the car-state prediction system, which predicts the car surface in the future. - an ankle, an estimator and a predictor, wherein the sensing portion is composed of the following peaches: - the first sensor senses the longitudinal velocity of the f-heart of the vehicle and produces - a longitudinal velocity signal; - a second sense The detector measures the lateral acceleration of the center of mass of the vehicle and generates a signal to the acceleration signal and the third sensor _ the suspension spring, and generates a signal. The estimator comprises a non-linear detector constructed by a vehicle dynamic model; and the longitudinal velocity signal, the lateral acceleration signal and the length variation L5 are estimated by the nonlinear state observer. One of the filial angles, one side inclination speed, one roll angle acceleration, one pitch angle, one pitch angle, one pitch angle acceleration, one yaw angle, one yaw rate, one yaw rate acceleration, longitudinal displacement, - longitudinal acceleration, - lateral displacement, i lateral velocity, - vertical displacement, - vertical velocity, - vertical acceleration and a tire angular velocity. And pre-contains-car her miscellaneous type, and accepts the longitudinal velocity signal lateral acceleration city, length change signal, roll angle, roll angle velocity, roll angle acceleration, fresh 卩, check, material angular acceleration, 雠200806513 angle, The yaw rate, yaw rate acceleration, longitudinal displacement, longitudinal acceleration, lateral displacement, lateral velocity, vertical displacement, vertical velocity, vertical acceleration, and tire angular velocity are used to predict the vehicle's motion behavior in the future. The vehicle dynamic prediction system further includes a judging device that receives the roll angle or the pitch angle, or receives both at the same time, and estimates the trend of the roll angle or the pitch angle to determine the future of the vehicle strip, and records the future vehicle. Overturned position. The change trend of the surface _ scale angle is whether the roll angle or the pitch angle is a divergence function. And the amount of material that is estimated by the amount of material is calculated to calculate the position of the future vehicle. And the first sensor is used to measure the lateral acceleration of the center of mass of the vehicle, and can be used to measure the quality of the vehicle, and the drum is used to produce the same effect. It can also be used to measure shellfish. The & angle of the yaw depends on the acceleration, and the resulting - yaw also causes the estimator to produce the same effect. 4 to expose the car, her method of test 峨峨, the invention, a car moving test method, is used to predict - the movement of the car I 峨 future time is 'this vehicle dynamic flip method includes the following steps: measuring car 1 The car's centroid gorge speed, and produces - longitudinal velocity Xinlong; = car ^ is called lateral acceleration, and produces - lateral acceleration signal; the length of the suspension spring of the handle changes, and produces - length change Signal; 200806513 Accepts the longitudinal velocity signal, the lateral acceleration signal and the length variation signal, and estimates one of the vehicle's roll angle, one side inclination speed, one side inclination acceleration, one pitch angle, one pitch rate, one pitch angle acceleration, A yaw angle, a yaw rate, a yaw rate acceleration, a longitudinal displacement, a longitudinal degree of attachment, a lateral displacement, a lateral velocity, a vertical displacement, a vertical velocity, a vertical acceleration, and a tire Angular velocity; and accept longitudinal velocity signal, lateral acceleration signal, length change money, roll angle, roll angle velocity, ship angle acceleration, pitch angle, pitch angle speed, pitch angle acceleration , yaw angle, yaw rate, yaw rate acceleration, longitudinal displacement, longitudinal acceleration, lateral displacement, lateral velocity, vertical displacement, vertical velocity, vertical acceleration and wheel platform angular velocity, and predicting the movement of the vehicle in the future behavior. The vehicle__ method further includes a slant angle, and calculates a k-direction tendency of the roll angle to determine whether the vehicle overturns in the future time. The step of changing the tendency of the roll angle is a step of estimating whether the roll angle or the pitch angle is a function of divergence. t This vehicle secret touches the green and contains Wei Wei Xiao Xiao, 14 counts the trend of change * when the car touches the future, it will not overturn. The step-by-step of the change trend of the mid-tilt is the step of whether or not the divergence function is transferred to the pitch. 2 The vehicle dynamics method also includes the estimated scale-receiving position. The lateral acceleration of the centroid of the vehicle surface is measured. The quality of the wheel 200806513 ^ knows the 'angle angle speed instead, and produces a yaw rate signal, which can also estimate the same result. It can also measure the lateral acceleration of the center of mass of the vehicle instead of measuring the lateral acceleration of the centroid. 'And a yaw rate signal is produced, and the same result can be estimated.

因此，本發明之一種車輛動態預測系統與方法，可利用二個感測出的物理量’並經由侧器與测器可以預測車輛於未來時間德跡’更進而可侧車輛於行進時是否會翻覆’因此’經由估·的設置而可使必須量測之物:量大幅減少，而降低感測器的成本，另—方面，由於感測器的狀減少’使湘❹⑶失效或產生錯誤信號而造成估測錯蚊情況亦可大大地降低’更增加制的穩定性與精雜。而且由㈣知技術中，其估算之法測係為經驗公式，因此， ^估"'之鼓都不是非常準確’然本發鴨運用完整車輛 _為估算之法則’耻，所估算之财非常準確。而且 :==咖所版吻細，可更增加其估有關本發明的特徵與實作，細說明如下。【實施方式】茲配合圖示作最佳實施例詳請參閱「第1圖」，所示為本太恭日日尨m 明之組合方塊示意圖。本U係-種車輛動態預測系係用M預測一車輛之運動 10 200806513 行為，此車姉態預啦統包含—制部、—估測器細與 -預測$ 5⑽。其中感測部係由下列元件組成·· -乐-感測器110係測量車輛之質心的縱向速度，並產生一縱向速度信號；-第二感測器13G係測量車輛之質心的侧向加速度，並產生—側向加速度信號及—第三感測器15〇係測量車輛之賴彈簧之長度變化，並產生—長度變化信號:其中，第-感測器11{)可為縱向速度感測器，而第二感測器130可為側向加速規，而第三感·⑽可為懸掛距離感測器。而估測器300係包含非線性狀態觀察器31〇，係由完整車輛权型320為基礎所建立，在此利用擴增卡曼滤波器 (jMed Kal麵Filter)為—例子說明，此為利用系統動匕方知式、系麟差和觀測誤差的統計麵機特性、及初始Therefore, the vehicle dynamic prediction system and method of the present invention can utilize two sensed physical quantities 'and can predict whether the vehicle will be over the future time traces via the side and the detectors, and then whether the side vehicles can overturn when traveling. 'Thus' can be measured by the setting of the estimate: the amount is greatly reduced, and the cost of the sensor is reduced, and on the other hand, because the shape of the sensor is reduced, the Xiangxi (3) is invalidated or an error signal is generated. The estimation of the wrong mosquitoes can also greatly reduce the stability and complexity of the 'increased system'. Moreover, from (4) knowing the technology, the estimated method is the empirical formula. Therefore, the estimated drum is not very accurate. However, the hair is used in the complete vehicle _ as the rule of estimation 'shame, the estimated profit Precise. Moreover, the === coffee version of the kiss is fine, which can further increase the evaluation. The features and implementations of the present invention are described in detail below. [Embodiment] For the details of the preferred embodiment, please refer to "Figure 1", which shows the combination of the squares of the day. This U-series vehicle dynamic prediction system uses M to predict the motion of a vehicle. 10 200806513 Behavior, this vehicle is pre-existing, including - Ministry, estimator and - prediction $ 5 (10). Wherein the sensing portion is composed of the following elements: - The music sensor 110 measures the longitudinal velocity of the center of mass of the vehicle and generates a longitudinal velocity signal; - the second sensor 13G measures the side of the center of mass of the vehicle To the acceleration, and to generate a lateral acceleration signal and a third sensor 15 to measure the length change of the vehicle, and to generate a length change signal: wherein the first sensor 11{) can be a longitudinal speed The sensor, while the second sensor 130 can be a lateral acceleration gauge, and the third sense (10) can be a suspension distance sensor. The estimator 300 includes a non-linear state observer 31, which is established on the basis of the complete vehicle weight 320. Here, an augmented Kalman filter (jMed Kalface Filter) is used as an example, and this is utilized. The statistical characteristics of the system, the anatomy, the observation error, and the statistical characteristics of the observation error, and the initial

條件等信息對制數翁行處理，從而得到綠狀態變數之最小誤差估測的-種演算法則。當估測器期接受縱向速度信，、側向加速度信號與長度變化信號，而經由非線性狀態觀察器310而估測出車輛之一側傾角、_側傾角速度、一側 =加逮度、—俯仰角、一崎角速度、—俯仰角加速度、 t丄角、一檢擺角速度、一橫擺角加速度、一縱向位移、 ^縱向加逮度、—側向位移、_侧向速度、—垂直位移、一垂直速度、—垂直加速度與-輪16聽度。 200806513 而預測器5GG係包含一完整車輛模型320，且接受縱向速度信號、m加速度信號、長度變化錢、侧_、側傾角速度、側傾角加速度、俯仰角、俯仰角速度、俯仰角加速度、橫擺角、橫擺角速度、橫擺角加速度、縱向位移、縱向加速度、侧向位移、侧向速度、垂直位移、垂直速度、垂直加速度與輪糾速度，而預測車祕未來時間之運動行為。另-迢路狀況（r〇adcc)nditic)n)働被以“外界干擾” 的方式輸入系統，藉由將道路狀況加入至完整車輛模型320 透過選取適當的感測部，道路狀況働對車輛動態的影響將可被估測器300正確估出。 -罵駛者祕彳τ為（steeringmaneuver) 等訊息傳輸至估測器3GG ’以協助估測器能夠估測出其它的物理量。在估測器3⑽中，我們可以採用以完整車輛模型320，如「第1圖」’域礎所建立之擴增卡曼濾波器。完整車輛权型320包含了車輛六個自由度、懸掛系統位移自由度、以及輪轉速自她的車輛模型，其動態方程式可以表示如下式： ^vehicle mvehicle mvehicle 200806513 = Mx + (ly - L )cDy〇yz lymy = My + (lz - Ix )c〇zg>x = + (lx - ly )ωχωγThe conditional information is processed by the number of lines, so that the minimum error estimate of the green state variable is obtained. When the estimator period accepts the longitudinal velocity signal, the lateral acceleration signal and the length change signal, and estimates the vehicle's roll angle, _ roll rate, side = acceleration, via the nonlinear state observer 310, - pitch angle, one-angle angular velocity, - pitch angular acceleration, t-angle, one oscillating angular velocity, one yaw angular acceleration, one longitudinal displacement, ^ longitudinal acceleration, - lateral displacement, _ lateral velocity, - vertical Displacement, a vertical velocity, - vertical acceleration and - wheel 16 listening. 200806513 The predictor 5GG includes a complete vehicle model 320 and accepts longitudinal velocity signals, m-acceleration signals, length change money, side _, roll angular velocity, roll angular acceleration, pitch angle, pitch angular velocity, pitch angular acceleration, and yaw Angle, yaw rate, yaw rate acceleration, longitudinal displacement, longitudinal acceleration, lateral displacement, lateral velocity, vertical displacement, vertical velocity, vertical acceleration, and wheel correction speed, and predict the motion behavior of the car in the future. In addition - the road condition (r〇adcc) nditic) n) is input into the system in a "outside interference" manner, by adding the road condition to the complete vehicle model 320 by selecting the appropriate sensing unit, the road condition is 働 to the vehicle The dynamic impact will be correctly estimated by the estimator 300. - A message such as "steeringmaneuver" is transmitted to the estimator 3GG' to assist the estimator in estimating other physical quantities. In estimator 3 (10), we can use an augmented Kalman filter built on the complete vehicle model 320, such as the "Fig. 1" domain. The complete vehicle weight 320 contains six degrees of freedom of the vehicle, the degree of freedom of displacement of the suspension system, and the wheel speed from her vehicle model. The dynamic equation can be expressed as follows: ^vehicle mvehicle mvehicle 200806513 = Mx + (ly - L )cDy 〇yz lymy = My + (lz - Ix )c〇zg>x = + (lx - ly )ωχωγ

Ht = Z*i9cos^+iSfi*(^cos^cos^-^sin^sin^)-i 治/ - ~7~ (-了/ - ^brake J ^motor,i ) 1 wheel 其中㈣,A-/2-/2]' I[辛 fff 〇)x - φ-έύηθ, ωγ =^cos^ + ^cos^sin^5 ωζ - -θύηφ + έcosθcosφ 上式參數的定義為： X，J，Z :車輛之質心在縱向(longitudinal direction)、侧向（lateral direct ion)與垂直(vertical direct ion)線性方向上的位移量。 :在尤拉轉換中所使用的三個尤拉角（Euler angles) °Ht = Z*i9cos^+iSfi*(^cos^cos^-^sin^sin^)-i 治/ - ~7~ (- / / ^brake J ^motor,i ) 1 wheel where (four), A- /2-/2]' I[辛 fff 〇)x - φ-έύηθ, ωγ =^cos^ + ^cos^sin^5 ωζ - -θύηφ + έcosθcosφ The above equation is defined as: X, J, Z: The amount of displacement of the center of mass of the vehicle in the linear direction of the longitudinal direction, the lateral direct ion, and the vertical direct ion. : Three Euler angles used in the Euler conversion °

〜，，：：車輛沿著Xj，z三轴旋轉的角速度，其中叫定義為側傾（roll)角速度、％定義為俯仰（pitch)角速度、乂定義為橫擺（yaw)角速度。 A# :第’輪胎上之有效三軸作用力，其中仄，&為非線性輪胎模型之輸出。 °’ ·第7輪胎的縱向黏著力。 rproad W:模擬道路狀況所激發的力。 13 200806513 是由輪胎上之有 •以貝心為車專動中心之車專動動量，欵作用力所引起。汉/ :懸掛系統之彈黃長度變化量。曾^第嘴胎的煞車力矩，藉由駕敬者踩煞車力道所計异出。~,,:: The angular velocity of the vehicle rotating along the three axes of Xj,z, which is defined as the angular velocity of the roll, % defined as the pitch angular velocity, and yaw as the angular velocity of the yaw. A#: The effective triaxial force on the 'th tire', where 仄, & is the output of the nonlinear tire model. °' · The longitudinal adhesion of the 7th tire. Rproad W: Simulates the forces that are triggered by road conditions. 13 200806513 It is caused by the special momentum of the car on the tires and the center of the car. Han / : The amount of change in the length of the suspension system. The brake torque of the former tire was calculated by the driver’s stepping on the brake pedal.

計算出第，·輪胎的引擎力矩，藉由駕駛者踩油門力道所以下為車輛常數： :車輔總重量。吨：車輛幾何參數，此為車輛前細輪所蚊長度。咚：車輛幾何參數，此為車輛後軸兩輪所夾之長度。 A:車輛幾何參數，此為車輛質心到前軸之長度。 4 :車輛幾何參數，此為車輛質心到後軸之長度。Calculate the engine torque of the first, tire, by the driver stepping on the throttle force. The following is the vehicle constant: : The total weight of the vehicle. Tons: Vehicle geometry, this is the length of the mosquito in front of the vehicle.咚: Vehicle geometry, this is the length of the two wheels of the rear axle of the vehicle. A: Vehicle geometry, this is the length of the vehicle's center of mass to the front axle. 4: Vehicle geometry, this is the length of the vehicle's center of mass to the rear axle.

:針對於質心之三軸轉動慣量。 :輪胎的轉動慣量。尤：懸掛系統之彈簣彈性係數。乃：懸掛系統之阻尼系統。。··輪胎半徑。因此由上述公式之觀察矩陣可知，把用以測量車輛之質心的側向加速度，可㈣量車輛之質心的橫擺角逮度代 14 200806513 替，並產生-橫擺角速度訊號，亦可使估測器3〇〇產生相同的作用，所以第二感測器130可為横擺角速度陀螺儀。另亦可用測量車輛之質心的橫代t側向加速度，並產生一棒 2角速度信號，同樣可使估測器產生相_作用，所以弟二感測器130可為量測横擺角感測器。乃，車輛動態預測系統更包含—判斷器_，係接收侧: The three-axis moment of inertia for the centroid. : The moment of inertia of the tire. Especially: the elastic modulus of the suspension system. It is: the damping system of the suspension system. . ··The tire radius. Therefore, from the observation matrix of the above formula, it can be known that the lateral acceleration used to measure the center of mass of the vehicle can be replaced by the yaw angle of the center of mass of the vehicle, and the yaw rate signal can be generated. The estimator 3 〇〇 produces the same effect, so the second sensor 130 can be a yaw rate gyroscope. Alternatively, the lateral acceleration of the transverse direction of the vehicle's center of mass can be measured, and a rod 2 angular velocity signal can be generated, which can also cause the estimator to produce a phase effect, so the second sensor 130 can measure the yaw angle. Detector. However, the vehicle dynamic prediction system further includes a -determiner _, the receiving side

傾角’而估算麵角之變化趨勢而靖車祕未來時間是否翻覆。其中側傾角之變化趨勢係為側傾角是否為發散函數，而判斷車麵於未來時間是否翻覆。並將估測器咖產生之訊 =與下—時序之駕駛者參數而再次提供於下—時序之估測器 ^若當判斷器_判斷車輛於未來時間會翻覆，則產生气 =至一防止祕，以自_整車輛狀況，而避免車輛翻 =或是將峨傳至警•，用以警告駕駛者車輛於未來時 ^翻覆’崎駕駛者料輕車域況，用爾止車辅翻 :亚且’此判斷器_可利用此估算時的車輛位點，而估絲料輛之翻概置。數m 2 1 第5圖」，所示為本發明之模擬中，豕^貫線為車輛在時域中之模擬情形，在這些模擬各、亲韩I:走的逮度為每小時90公里並且在第4秒的時候作二Γ ’這些模擬—共有兩種情況’而這些情況分馬馼者處於不同角度的路面上行驶，從「第2圖」至「第 200806513 5圖」便呈現其模擬結果，於圖中用實線來當作車輛在某- 駕駛行為下的動態響應，虛線為估測器的估測與預測器 500的預測。The inclination angle is used to estimate the change trend of the face angle and whether the future time of the car is overturned. The change trend of the roll angle is whether the roll angle is a divergence function, and whether the car surface is overturned in the future time. And the estimator and the driver's parameters of the lower-time sequence are again provided to the next-time estimator. If the determinator _ judges that the vehicle will overturn in the future time, then the gas is generated to prevent Secret, to _ the whole vehicle situation, and to avoid the vehicle to turn = or to pass the rumor to the police • to warn the driver of the vehicle in the future ^ to overturn the 'saki driver's light car domain conditions, use the car to assist the car : Ya and 'this judger _ can use this estimated vehicle position, and estimate the roll of the car. The number m 2 1 Fig. 5 is shown in the simulation of the present invention, and the line is the simulation of the vehicle in the time domain. In these simulations, the pro-Han I: the catch is 90 kilometers per hour. And in the 4th second, the second ''these simulations—there are two kinds of situations' and these situations are carried out on the roads of different angles, from "2nd picture" to "200806513 5 picture" The simulation results are shown in the figure as the dynamic response of the vehicle under a certain driving behavior, and the dotted line is the estimation of the estimator and the prediction of the predictor 500.

為了姐較__與未來麟之間的差異，預測H =將不會去縣辦間點的未來絲翻，岐會故意地十成Ik著％間而輪出目前的車辅動態。除此之外，由三個 11G、13G、150測得的量（f心之側向加速度與縱向 2以及騎科之紐變化）會故意地在5.1秒前進入並助輸_且對細時_進行估算，並在51秒後方=110、130、150關掉，進行車輛動咖 7㈣_定的轉f角度’是模擬車輛在固定的戶^丨固定方向盤轉彎角度）的動態行為。若在5〜8秒中態行為(如圖中虛線，沒有感_ 代表^我們^車輛在¥域中的模擬（如圖中實線）相符合，即 -日她。7用此估測器細與預測器500的架構，在某車輛動姑位或者是汝第2圖」盘「第角度、側傾“峨示車輛之方向盤也、&擺肖對時間關細以及車向位移對時__。之二轴方滑順地轉Mm ，秘代表車私平坦的路面上做 "㈣之車_ _嶋統可成功地預測車輛 16 200806513 2來時間之動態行為，_在時間8秒之後在車柄動態行力4預測器5GG的輸財明_不同，這是因為δ秒之後， j者i：又了方向麵肢，並且如翻地預湘5⑼無法传知。然而吴中不足的是車輛橫擺角並無法相當吻合。如「第4圖」與「第5圖」所示，係顯示車輛之方向盤角度、側傾角與橫擺角對時間關係圖以及車輛質心之三轴方 • =位移對時__。此情況為車輛在斜坡上做滑塌地轉弯’此斜坡為沿著路面縱向軸偏_25度的斜坡，本發明之車輛動恶預測系統依然是可作用的，且此情況是經過特殊設汁，其所有的操作情況與第一種情況不同的是僅僅在於車輛行驗於不同的路面上。此情況的重要性在於車輛行敬於斜坡上，而此斜坡亦是引起車輛翻覆的要件，此外，在這個狀況下，車輛橫擺角不同於其他情況，而可成功地被估測且預測 • 到’這是因為侧向加速度在斜坡上時含有車輛橫擺角此項動恶，因此車輛橫擺角可藉由側向加速度來估測到。而且可預估車柄於接近七秒時翻覆。請參閱「第6圖」、「第7圖」與「第8圖」，所示為本备明之方法流程圖。如「第6圖」所示，本發明之一種車輛動恶預測方法，係用以預測一車輛於未來時間之運動行為，此車輛動態預測方法包含下列步驟：首先測量車輛之質心的縱向速度及側向加速度與車輛 17 200806513 、二:广之長_化’並產生一縱向速度信號、-側向加與—紐變化錢（步驟_);接受縱向速度信〜侧向加速度信號與長度變化信號，而估測出車輛之一侧 =、—側傾角速度、—侧傾角加速度、-俯仰角、-俯仰 /度、—俯仰角加速度、一橫擺角、-橫擺角速度、一橫擺角加速度、—縱向位移、—縱向加速度、-側向位移、一侧向速度、—垂直位移、—垂直速度、—蚊加速度與一輪月。角速度（步驟9Q1);接受縱向速度信號、側向加速度信號、長度變化信號、側傾角、麵角速度、側傾角加速度、俯仰角、俯仰角速度、俯仰角加速度、橫擺角，角速度、橫擺角加速度、縱向位移、縱向加速度、側向位移、側向速度、垂直位移、垂直速度、垂直加速度與輪㈣速度，而預測車車；未來％間之運動行為（步驟9〇3)。接收側傾角，而估算侧傾角之變化趨勢而判斷車輛於未來時間是否翻覆（步: 聊）。如「第7圖」所示’經由上述步驟_至步驟_後，接收俯仰角，而估算俯仰角之變化趨勢而判斷車輛於未來時間是否翻覆（步驟907)。如「第δ圖」所示，經由上述步驟麵至步驟903後，並且可預估車輛於未來時間的翻覆位置 (步驟909)。其中麵角與俯仰角之變化趨勢的步驟係為估算側傾角與俯仰角是否為發散函數的步驟。 18 200806513 其中測里車輛之質心的側向加速度，可用測量車輛之質心的橫擺肖速度储，並赶—麵角速度職，亦可估算出車輛的所有相_理量。另亦可制量車輛之質心的樺: 角代替測量質心的側向加速度，而產生-橫⑽速度信號: 同I可佑心^車輛輯有細物理量。In order to compare the difference between __ and future Lin, it is predicted that H = will not go to the future of the county office, and will deliberately take the current car-related dynamics. In addition, the amount measured by the three 11G, 13G, 150 (the lateral acceleration of the f-heart and the change of the longitudinal 2 and the riding of the branch) will deliberately enter and assist before 5.1 seconds. _ Estimate and turn off after 51 seconds = 110, 130, 150, and the dynamic behavior of the vehicle's mobile coffee 7 (four) _ fixed turn angle 'is the simulated vehicle's fixed steering wheel turn angle in a fixed household. If the behavior is in the middle of 5~8 seconds (as shown in the dotted line in the figure, there is no sense _ on behalf of ^ we ^ vehicle in the ¥ domain simulation (solid line in the figure), ie - day she. 7 use this estimator The structure of the fine and predictor 500, in the case of a certain vehicle or in the second figure, "the angle, the roll", the steering wheel of the vehicle, the time and the displacement of the vehicle __. The second axis smoothly turns to Mm, and the secret represents the car on the privately flat road. The vehicle is _ _ 嶋可 can successfully predict the dynamic behavior of the vehicle 16 200806513 2 _ at 8 seconds After that, in the handle dynamic force 4 predictor 5GG loses money _ different, this is because after δ seconds, j i: again the direction of the limbs, and if the turn to pre-Xiang 5 (9) can not be known. However, Wu Zhong deficiency The yaw angle of the vehicle does not match very well. As shown in Figure 4 and Figure 5, it shows the steering angle of the vehicle, the roll angle and yaw angle versus time, and the three axes of the vehicle's center of mass. Square • = displacement versus time __. This is the case where the vehicle makes a slip on the slope. This slope is along the road surface. The slope of the vehicle _25 degrees, the vehicle motion prediction system of the present invention is still applicable, and the situation is that after special juice setting, all the operation conditions are different from the first case only in that the vehicle inspection is different. On the road surface. The importance of this situation lies in the fact that the vehicle is on the slope. This slope is also a requirement for the vehicle to overturn. Moreover, in this situation, the vehicle yaw angle is different from other conditions and can be successfully estimated. Measure and predict • to 'this is because the lateral acceleration on the slope contains the vehicle yaw angle, so the vehicle yaw angle can be estimated by the lateral acceleration. And the handle can be estimated to be close. Please refer to "Figure 6", "Figure 7" and "Figure 8" for a flow chart of the method of the present invention. As shown in Figure 6, a vehicle of the present invention is moved. The bad prediction method is used to predict the behavior of a vehicle in the future. The vehicle dynamic prediction method includes the following steps: First, measuring the longitudinal velocity and lateral acceleration of the vehicle's center of mass with the vehicle 17 200806513, 2: wide Long _ _ ' and generate a longitudinal velocity signal, - lateral plus - New change money (step _); accept longitudinal velocity signal ~ lateral acceleration signal and length change signal, and estimate one side of the vehicle =, - Rolling angular velocity, - roll angle acceleration, - pitch angle, - pitch / degree, - pitch angle acceleration, one yaw angle, - yaw rate, one yaw rate acceleration, - longitudinal displacement, - longitudinal acceleration, - lateral Displacement, lateral velocity, vertical displacement, vertical velocity, mosquito acceleration and one-month. angular velocity (step 9Q1); longitudinal velocity signal, lateral acceleration signal, length change signal, roll angle, face angular velocity, roll angle Acceleration, pitch angle, pitch rate, pitch angle acceleration, yaw angle, angular velocity, yaw rate acceleration, longitudinal displacement, longitudinal acceleration, lateral displacement, lateral velocity, vertical displacement, vertical velocity, vertical acceleration, and wheel (four) velocity, And predict the car; the future movement behavior between the % (step 9 〇 3). The roll angle is received, and the trend of the roll angle is estimated to determine whether the vehicle is overturned in the future (step: chat). As shown in Fig. 7, after the above steps _ to _, the pitch angle is received, and the trend of the pitch angle is estimated to determine whether the vehicle is overturned in the future (step 907). As shown in the "δth diagram", after the above steps are passed to step 903, the vehicle overturning position at a future time can be estimated (step 909). The step in which the variation of the face angle and the pitch angle is a step of estimating whether the roll angle and the pitch angle are divergence functions. 18 200806513 The lateral acceleration of the centroid of the vehicle can be stored by measuring the yaw rate of the vehicle's center of mass, and can also estimate the phase of the vehicle. Alternatively, the car's center of mass can be measured: the angle instead of measuring the lateral acceleration of the centroid, and the - horizontal (10) speed signal: the same as I can help the vehicle.

、雖然本發明以前述之較佳實施例揭露如上，然其並非用以限林發明，任何熟習姆技藝者，在不_本發明之精㈣fc圍内’當可作些許之更動與潤飾，因此本發明之專利保護範_視本朗#_之^料概_界定者為準。【圖式簡單說明】昂1圖」係顯示本發明之組合方塊示意圖；弟2圖」係顯示本發明車輛之方向盤角度、麵角角對時間關係圖；〜田、月匕關::圖」係顯示本發明之車輛質心之三軸方向位移對時間罘4圖」係顯示本發明車輛之方向盤角度、側傾角與角辦時_鋼；、5 ▲ &第5圖」軸林發明之車姉心之三軸方向位移對時間係顯示本發明之方法流乐6圖」、「第7圖」與「第8圖」程圖。 19 200806513 【主要元件符號說明】 110第一感測器 130第二感測器 150第三感測器 3⑽估測器 310非線性狀態觀察器 320完整車輛模型 400道路狀況 410駕駛者操作行為 500預測器 600判斷器 700防止系統Although the present invention has been disclosed above in the preferred embodiments of the foregoing, it is not intended to limit the invention, and any skilled artisan may make some changes and refinements in the absence of the invention. The patent protection of the present invention is based on the definition of the product. [A Brief Description of the Drawings] Ang 1 diagram "shows a schematic block diagram of the present invention; a brother 2 diagram" shows a steering wheel angle, a face angle angle versus time diagram of the vehicle of the present invention; ~田,月匕关::图" The display of the three-axis direction displacement versus time 罘4 of the vehicle centroid of the present invention is shown in the steering wheel angle, the roll angle and the angle of the vehicle of the present invention. _ steel; 5 ▲ & Fig. 5 The three-axis direction displacement versus time of the vehicle is shown in the method of the present invention, which is shown in Fig. 6, "Fig. 7" and "Fig. 8". 19 200806513 [Description of main component symbols] 110 first sensor 130 second sensor 150 third sensor 3 (10) estimator 310 nonlinear state observer 320 complete vehicle model 400 road condition 410 driver operation behavior 500 prediction 600 determiner 700 prevents the system

2020

Claims

200806513 十、申請專利範圍： 1.-種車—態預測系統，係用以預測—車輛之運動行為，該車辅動態預剛系統包含：感測部’係由下列元件組成：苐一感 >則器，係測量該車輛之質心的縱向速度’並產生一縱向速度信號；鐵刪器，係測量該車輛之質200806513 X. Patent application scope: 1.- Kind of vehicle-state prediction system is used to predict the motion behavior of the vehicle. The vehicle auxiliary dynamic pre-rigid system includes: The sensing part is composed of the following components: 苐一感&gt The device measures the longitudinal velocity of the center of mass of the vehicle and generates a longitudinal velocity signal; the iron-cutting device measures the quality of the vehicle.

速度’並產生一側向加速度信號；及第二感測器，係測量該車輛之懸掛彈筈之長 ^變化，並產生一長度變化信號；.八估測裔，係包含一由一車輛動態模型為基礎所建 /泉丨生狀恶觀察器，且接受該縱向速度信號、該侧城^度彳與錢度變化信號，而經由該非線性狀態 I㈣估測出該車輛之—侧傾角一側傾角速度、一速度、_俯仰角、—俯仰角速度一俯仰角加速度、只肥、一橫擺角逮度、一橫擺角加速度、一縱向位移、一縱向加速度、一側Θ 、向位和、一侧向速度、一垂直位移、一垂直速度、一番古一、w 罝加速度與一輪胎角速度；及係包含該車輛動態模型，且接受該縱向傾角、該側傾度信號、該長度變化信號、該側仰角速度、該崎肖^，_加速度、該俯仰角、該俯該橫擺角加速度、^=、該橫㈣ '該橫擺角速度、該側向速度、該垂^位移、該縱向加速度、該側向位移、立移、該垂直速度、該垂直加速度 21 200806513 與該輪胎角速度，而預測該車輛之動態。 2·如申凊專利範圍第i項所述之車輛動態預測系統，更包含一判斷器，係接收該側傾角，而估算該侧傾角之變化趨勢而判斷該車輛是否翻覆狀態。 3·如申凊專利範圍第2項所述之車輛動態預測系統，其中該侧傾角變化趨勢係為該側傾角是否為發散函數，而判斷該車輛是否翻覆狀態。 _ 4·如申請專利範圍第2項所述之車輛動態預測系統，其中 5亥判斷器估算該車輛之未來翻覆位置。 5·如申請專利範圍第丨項所述之車輛動態預測系統，更包含一判斷器，係接收該俯仰角，而估算該俯仰角之變化趨勢而判斷該車輛是否翻覆狀態。 6·如申請專利範圍第4項所述之車輛動態預測系統，其中忒俯仰角之變化趨勢係為該俯仰角是否為發散函數，而判斷該車輛是否翻覆狀態。 Φ 7·如申請專利範圍第5項所述之車輛動態預測系統，其中該判斷器估算該車輛之未來翻覆位置。 8.如申請專利範圍第1項所述之車輛動態預測系統，該非線性狀態觀察器係為一擴增卡曼濾波器。 9 ·種車輛動恶預測糸統，係用以預測一車輛之運動行為，該車輛動態預測系統包含：一感測部，係由下列元件組成：一第一感測器，係測量該車輛之質心的縱向速 22 200806513 度’並產生一縱向速度信號；一第二感測器，係測量該車輛之橫擺角速度，並產生一橫擺角速度信號；及一第三感測器，係測量該車輛之懸掛彈簧之長度變化，並產生一長度變化信號·，估測為，係包含一由一車輛動態模型為基礎所建立之非線性狀態觀察器，且接受該縱向速度信號、該橫擺角速度信號與該長度變化信號，而該非線性狀態觀察器估測出該車輛之—側傾角、一侧傾角速度、一側傾角加速度、一俯仰角、一俯仰角速度、一俯仰角加速度、一橫才成角、一松擺角加速度、一縱向位移、一縱向加速度、一側向位移、一侧向速度、一側向加速度、一垂直位移、一垂直速度、一垂直加速度與一輪胎角速度；及一預測器，係包含該車輛動態模型，且接受該縱向速度信號、該橫擺角速度、該長度變化信號、該侧傾角、該側傾角速度、該侧傾角加速度、該俯仰角、該俯仰角速度、該俯仰角加速度、該橫擺角、該橫擺角加速度、該縱向位移、该縱向加速度、該侧向位移、該侧向速度、該侧向加速度、該垂直位移、該垂直速度、該垂直加速度與該輪胎角速度，而預測該車輛之動態。 10·如申請專利範圍第9項所述之車輛動態預測系統，更包含一判斷器，係接收該側傾角，而估算該侧傾角之變化趨勢而判斷該車輛是否翻覆狀態。 23 200806513 11·如申請專利範圍第10項所述之車輛動態預測系統，其中該侧傾角之變化趨勢係為該側傾角是否為發散函數’而判斷該車輛是否翻覆狀態。 ’12·如申睛專利範圍第1〇項所述之車輛動態預測系統，其中該判斷器估算該車輛之未來翻覆位置。 13·如申請專利範圍第9項所述之車輛動態預測系統，更匕含判辦态，係接收该側傾角，而估算該侧傾角之變 _ 化趨勢而判斷該車輛是否翻覆狀態。 14·如申請專利範圍第13項所述之車輛動態預測系統，其中該側傾角之變化趨勢係為該侧傾角是否為發散函數，而判斷該車輛是否翻覆狀態。 15.如申請專利範圍第13項所述之車輛動態預測系統，其中5亥判斷裔估异該車輛之未來翻覆位置。 16·如申請專利範圍第9項所述之車輛動態預測系統，該非線性狀恶觀祭益係為一擴增卡曼濾波器。 • I7·一種車輛動態預測系統，係用以預測一車輛之動態，該車輛動態預測系統包含：一感測部，係由下列元件組成：一第一感測器，係測量該車輛之質心的縱向速度，並產生一縱向速度信號；一第二感測器，係測量該車輛之橫擺角，並產生一橫擺角信號；及一第二感測裔’係測量該車輛之懸掛彈簧之長 24 200806513 度變化，並產生一長度變化信號；估測為，係包含一由—車輛動態模型為基礎所建立之非線性狀態觀察器，且接受該縱向速度信5虎、橫擺角信號與該長度變化信號，而該非線性狀態觀察器估測出該車輛之一側傾角、一側傾角速度、-侧傾角加速度、俯仰角、-俯仰角速度、—俯仰角加速度、一橫擺角速度、 1擺角加速度、-縱向位移、—縱向加速度、—側向位移、一側向速度一側向加速度、—垂直位移、—垂直速度、一垂直加速度與一輪胎角速度；及又、—預測益，係包含該車輛動態模型，且接受該縱速度^號 '該橫擺角信號、該長度變化信號、該側傾相傾角速度、該侧傾角加速度、該俯仰角、該俯仰声該俯仰角加速度、該橫擺角速度、該橫擺角加速度、= 位移、該縱向加速度、該側向位移、_向速度、該Λ目二 =迷度、縣直位移、該垂直速度、該垂直加速度盘^ 輪胎角速度，而預測該車輛之動熊。 ^ W 申請專利範圍第Π項所述之料動__ 判斷器，係接收該側傾角，而估算該 j 化趨勢而判斷該車輛是否翻覆狀能。巧<交 1申請專利範圍第18項所之料動態預測中該側傾角之變化趨勢係為該側傾角是否為發：數’而判斷該車輛是否翻覆狀態。 @㈡ 2〇.如申請專利範圍第18項所述之6車幸兩動態預測系統，其 25 200806513 中該判斷器估算該車輛之未來翻覆位置。 21. 如申請專利範圍第17項所述之車輛動態預測系統，更包含一判斷器，係接收該俯仰角，而估算該俯仰角之變化趨勢而判斷該車輛是否翻覆狀態。 22. 如申請專利範圍第21項所述之車輛動態預測系統，其中該俯仰角之變化趨勢係為該俯仰角是否為發散函數，而判斷該車輛是否翻覆狀態。 23. 如申請專利範圍第21項所述之車輛動態預測系統，其中該判斷器估算該車輛之未來翻覆位置。 24. 如申請專利範圍第17項所述之車輛動態預测系統，該非線性狀態觀察器係為一擴增卡曼濾波器。 25. —種車輛動態預測方法，係用以預測一車輛之動態，該車輛動態預測方法包含下列步驟：測量該車輛之質心的縱向速度及側向加速度與該車輛之懸掛彈簧之長度變化，並產生一縱向速度信號、一側向加速度信號與一長度變化信號；接受該縱向速度信號、該侧向加速度信號與該長度變化信號，而估測出該車輛之一側傾角、一側傾角速度、一侧傾角加速度、一俯仰角、一俯仰角速度、一俯仰角加速度、一橫擺角、一橫擺角速度、一橫擺角加速度、一縱向位移、一縱向加速度、一侧向位移、一侧向速度、一垂直位移、一垂直速度、一垂直加速度與一輪胎角速度；及 26 200806513 接受該縱向速度信號、該侧向加速度信號、該長度’支化传號、該侧傾角、該侧傾角速度、該侧傾角加速度、該俯仰角、該俯仰角速度、該俯仰角加速度、該橫擺角、該橫擺角速度、該橫擺角加速度、該縱向位移、該縱向力口速度、該侧向位移、該側向速度、該垂直位移、該垂直速度、該垂直加速度與該輪胎角速度，而預測該之動態。 ^ 2 6 ·如申請專利範圍第2 5項所述之車輛動態預測方法，其中於預測該車輛之動態的步驟後，更包含接故該側傾角，而估算該側傾角之變化趨勢而判斷該車輛是 ' 麵的步驟。翻 27·如申請專利範圍第26項所述之車輛動態預測方法，其中該側傾角之變化趨勢的步驟係為估算該側傾角是否為發散函數的步驟。 2 8 ·如申請專利範圍第2 5項所述之車輛動態預測方法，其中於預測該車輛之動態的步驟後，更包含預估該車輛之未來翻覆位置的驟。 29·如申請專利範圍第25項所述之車輛動態預測方法，其中於預測該車輛之動態的步驟後，更包含接收該俯仰角，而估算該俯仰角之變化趨勢而判斷該車輛是否翻覆狀態。 30.如申請專利範圍第29項所述之車輛動態預測方法，其中該俯仰角之變化趨勢的步驟係為估算該俯仰角是否 27 200806513 為發散函數的步驟。 31. —種車輛動態預測方法，係用以預測一車輛之動態，该車輛動態預測方法包含：测i該車|兩之質心的縱向速度及橫擺角速度與該車輛之懸掛彈簧之長度變化，並產生一縱向速度信號、一橫擺角速度信號與一長度變化信號；接受該縱向速度信號、該橫擺角速度信號與該長 _ 度變化k號，而估測出該車輛之一侧傾角、一侧傾角速度、侧傾角加速度、一俯仰角、一俯仰角速度、一俯仰角加速度、一橫擺角、一橫擺角加速度、一縱向位移、一縱向加速度、一側向位移、一側向速度、一側向加速度、一垂直位移、一垂直速度、一垂直加速度與一輪胎角速度；及接受该縱向速度信號、該橫擺角速度、該長度變化信號、該側傾角、該侧傾角速度、該侧傾角加速度、 • 該俯仰角、該俯仰角速度、該俯仰角加速度、該橫擺角、該檢擺角加速度、該縱向位移、該縱向加速度、該側向位移、該側向速度、該侧向加速度、該垂直位移、該垂直速度、名垂直加速度與該輪胎角速度，而預測該車輛之動態。 32·如申請專利範圍第31項所述之車輛動態預測方法，其中於預測該車輛之動態的步驟後，更包含接收該側傾角，而估算該側傾角之變化趨勢而判斷該車輛是否翻 28 200806513 覆狀態的步驟。 3 3.如申請專利範圍第3 2項所述之車輛動態預測方法，其中該側傾角之變化趨勢的步驟係為估算該側傾角是否為發散函數的步驟。 34. 如申請專利範圍第31項所述之車輛動態預測方法，其中於預測該車輛之動態的步驟後，更包含預估該車輛於未來時間之一翻覆位置的步驟。 35. 如申請專利範圍第31項所述之車輛動態預測方法，其中於預測該車輛之動態的步驟後，更包含接收該俯仰角，而估算該俯仰角之變化趨勢而判斷該車輛是否翻覆狀態的步驟。 36. 如申請專利範圍第35項所述之車輛動態預測方法，其中該俯仰角之變化趨勢的步驟係為估算該俯仰角是否為發散函數的步驟。 37. —種車輛動態預測方法，係用以預測一車輛之動態，該車輛動態預測方法包含下列步驟：測量該車輛之質心的縱向速度及橫擺角與該車輛之懸掛彈簧之長度變化，並產生一縱向速度信號、一橫擺角信號與一長度變化信號；接受該縱向速度信號、橫擺角信號與該長度變化信號，而估測出該車輛之一侧傾角、一侧傾角速度、一侧傾角加速度、一俯仰角、一俯仰角速度、一俯仰角加速度、一橫擺角速度、一橫擺角加速度、一縱向位移、一縱 29 200806513 向加速度、一侧向位移、一侧向速度、一侧向加速度、一垂直位移、一垂直速度、一垂直加速度與一輪胎角速度；及接受該縱向速度信號、該橫擺角、該長度變化信號、該側傾角、該侧傾角速度、該側傾角加速度、該俯仰角、該俯仰角速度、該俯仰角加速度、該橫擺角速度、該橫擺角加速度、該縱向位移、該縱向加速度、該側向位移、該侧向速度、該侧向加速度、該垂直位移、該垂直速度、該垂直加速度與該輪胎角速度，而預測該車輛之動態。 38. 如申請專利範圍第37項所述之車輛動態預測方法，其中於預測該車輛之動態的步驟後，更包含接收該侧傾角，而估算該侧傾角之變化趨勢而判斷該車輛是否翻覆狀態的步驟。 39. 如申請專利範圍第38項所述之車輛動態預測方法，其中該側傾角之變化趨勢的步驟係為估算該侧傾角是否為發散函數的步驟。 40. 如申請專利範圍第37項所述之車輛動態預測方法，其中於預測該車輛之動態的步驟後，更包含預估該車輛於未來時間之一翻覆位置的步驟。 41. 如申請專利範圍第3 7項所述之車輛動態預測方法，其中於預測該車輛之動態的步驟後，更包含接收該俯仰角，而估算該俯仰角之變化趨勢而判斷該車輛是否翻 30 200806513 覆狀態的步驟。 42.如申請專利範圍第41項所述之車輛動態預測方法，其中該俯仰角之變化趨勢的步驟係為估算該俯仰角是否為發散函數的步驟。Speed 'and produces a lateral acceleration signal; and a second sensor that measures the length of the suspension of the vehicle and produces a length change signal; Based on the model, the longitudinal velocity signal, the lateral velocity, and the change signal of the money are accepted, and the side-tilt side of the vehicle is estimated via the nonlinear state I(4). Inclination speed, a speed, _ pitch angle, - pitch rate - pitch rate acceleration, only fat, a yaw angle catch, a yaw rate acceleration, a longitudinal displacement, a longitudinal acceleration, one side 向, a directional sum, a lateral velocity, a vertical displacement, a vertical velocity, a natural one, a w 罝 acceleration, and a tire angular velocity; and a dynamic model of the vehicle, and accepting the longitudinal inclination, the roll signal, the length change signal, The side elevation speed, the slanting acceleration, the pitch angle, the pitch yaw rate acceleration, ^=, the transverse (four) 'the yaw rate, the lateral velocity, the vertical displacement, the longitudinal direction Speed, the lateral displacement, vertical shift, the vertical speed, the vertical acceleration and the angular velocity of the tire 21200806513, predicted dynamics of the vehicle. 2. The vehicle dynamics prediction system of claim i, further comprising a determiner for receiving the roll angle and estimating a trend of the roll angle to determine whether the vehicle is overturned. 3. The vehicle dynamics prediction system according to claim 2, wherein the tendency of the roll angle change is whether the roll angle is a divergence function, and whether the vehicle is overturned is determined. _ 4. The vehicle dynamic prediction system according to claim 2, wherein the 5 hai judging device estimates the future tumbling position of the vehicle. 5. The vehicle dynamic prediction system according to the scope of claim 2, further comprising a determiner for receiving the pitch angle and estimating a trend of the pitch angle to determine whether the vehicle is overturned. 6. The vehicle dynamic prediction system according to claim 4, wherein the change trend of the 忒 pitch angle is whether the pitch angle is a divergence function, and whether the vehicle is overturned. Φ 7. The vehicle dynamics prediction system of claim 5, wherein the determiner estimates a future overturned position of the vehicle. 8. The vehicle dynamics prediction system of claim 1, wherein the non-linear state observer is an augmented Kalman filter. 9) A vehicle propulsion prediction system for predicting a vehicle's motion behavior. The vehicle dynamic prediction system includes: a sensing unit, which is composed of the following components: a first sensor that measures the vehicle The longitudinal velocity of the centroid 22 200806513 degrees 'and produces a longitudinal velocity signal; a second sensor that measures the yaw rate of the vehicle and produces a yaw rate signal; and a third sensor that measures The length of the suspension spring of the vehicle changes, and a length change signal is generated, which is estimated to include a nonlinear state observer established on the basis of a vehicle dynamic model, and receives the longitudinal velocity signal, the yaw The angular velocity signal and the length change signal, and the nonlinear state observer estimates the vehicle's roll angle, one side tilt speed, one side tilt acceleration, one pitch angle, one pitch rate, one pitch angle acceleration, and one horizontal Angle, a loose angle acceleration, a longitudinal displacement, a longitudinal acceleration, a lateral displacement, a lateral velocity, a lateral acceleration, a vertical displacement a vertical velocity, a vertical acceleration, and a tire angular velocity; and a predictor comprising the vehicle dynamic model and receiving the longitudinal velocity signal, the yaw angular velocity, the length variation signal, the roll angle, the roll rate, The roll angle acceleration, the pitch angle, the pitch angle speed, the pitch angle acceleration, the yaw angle, the yaw angle acceleration, the longitudinal displacement, the longitudinal acceleration, the lateral displacement, the lateral velocity, the lateral direction The acceleration, the vertical displacement, the vertical velocity, the vertical acceleration, and the tire angular velocity are used to predict the dynamics of the vehicle. 10. The vehicle dynamics prediction system of claim 9, further comprising a determiner that receives the roll angle and estimates a change trend of the roll angle to determine whether the vehicle is overturned. The vehicle dynamics prediction system according to claim 10, wherein the tendency of the roll angle changes is whether the roll angle is a divergence function and whether the vehicle is overturned. The vehicle dynamics prediction system of claim 1, wherein the determiner estimates a future overturned position of the vehicle. 13. The vehicle dynamics prediction system according to claim 9 is further characterized in that the roll angle is received, and the roll angle is estimated to determine whether the vehicle is overturned. The vehicle dynamics prediction system according to claim 13, wherein the trend of the roll angle is whether the roll angle is a divergence function, and whether the vehicle is overturned. 15. The vehicle dynamics prediction system of claim 13, wherein the 5th judging panel estimates the future reversal position of the vehicle. 16. The vehicle dynamic prediction system according to claim 9, wherein the nonlinear ecstasy is an augmented Kalman filter. • I7· A vehicle dynamic prediction system for predicting the dynamics of a vehicle. The vehicle dynamic prediction system comprises: a sensing unit consisting of the following components: a first sensor that measures the center of mass of the vehicle Longitudinal speed and produces a longitudinal velocity signal; a second sensor that measures the yaw angle of the vehicle and produces a yaw angle signal; and a second sensing descent 'measures the suspension spring of the vehicle Length 24 200806513 degrees change, and produces a length change signal; estimated as comprising a nonlinear state observer based on the vehicle dynamic model, and accepts the longitudinal velocity signal 5 tiger, yaw angle signal And the length change signal, and the nonlinear state observer estimates a roll angle, a side roll speed, a roll roll acceleration, a pitch angle, a pitch angle speed, a pitch angle acceleration, a yaw rate, and 1 Swing angle acceleration, - longitudinal displacement, - longitudinal acceleration, - lateral displacement, lateral acceleration on one side, vertical displacement, vertical velocity, vertical addition Speed and a tire angular velocity; and again, predictive benefit, including the vehicle dynamic model, and accepting the longitudinal velocity ^' the yaw angle signal, the length variation signal, the roll phase tilt velocity, the roll angle acceleration The pitch angle, the pitch sound, the pitch angle acceleration, the yaw rate, the yaw rate acceleration, the = displacement, the longitudinal acceleration, the lateral displacement, the _direction speed, the eye 2 = the degree, the county straight The displacement, the vertical velocity, the vertical acceleration disk, the tire angular velocity, and the predicted bear of the vehicle. ^ W The application of the __ judging device described in the scope of the patent application is to receive the roll angle and estimate the trend to determine whether the vehicle is overturned. In the dynamic prediction of the material of item 18 of the patent application, the trend of the change of the roll angle is whether the roll angle is the number: the number is judged whether the vehicle is overturned. @(二) 2〇. As claimed in the application for the scope of the patent, the six car two dynamic prediction system, the 25 200806513 in the judger to estimate the future position of the vehicle. 21. The vehicle dynamics prediction system according to claim 17, further comprising a determiner that receives the pitch angle and estimates a change trend of the pitch angle to determine whether the vehicle is overturned. 22. The vehicle dynamics prediction system according to claim 21, wherein the change tendency of the pitch angle is whether the pitch angle is a divergence function, and whether the vehicle is overturned. 23. The vehicle dynamics prediction system of claim 21, wherein the determiner estimates a future override position of the vehicle. 24. The vehicle dynamics prediction system of claim 17, wherein the nonlinear state observer is an augmented Kalman filter. 25. A vehicle dynamic prediction method for predicting a vehicle dynamics, the vehicle dynamic prediction method comprising the steps of: measuring a longitudinal velocity and a lateral acceleration of a centroid of the vehicle and a length change of a suspension spring of the vehicle, And generating a longitudinal velocity signal, a lateral acceleration signal and a length variation signal; receiving the longitudinal velocity signal, the lateral acceleration signal and the length variation signal, and estimating a roll angle and a side inclination speed of the vehicle , one-sided tilt acceleration, one pitch angle, one pitch angle speed, one pitch angle acceleration, one yaw angle, one yaw rate, one yaw rate acceleration, one longitudinal displacement, one longitudinal acceleration, one side displacement, one side Speed, a vertical displacement, a vertical speed, a vertical acceleration, and a tire angular velocity; and 26 200806513 accepting the longitudinal velocity signal, the lateral acceleration signal, the length 'branch mark, the roll angle, the roll rate The roll acceleration, the pitch angle, the pitch rate, the pitch angle acceleration, the yaw angle The yaw rate, the yaw rate acceleration, the longitudinal displacement, the longitudinal force port velocity, the lateral displacement, the lateral velocity, the vertical displacement, the vertical velocity, the vertical acceleration, and the tire angular velocity are predicted The dynamics. The vehicle dynamics prediction method according to claim 25, wherein after the step of predicting the dynamics of the vehicle, the method further comprises: picking up the roll angle, and estimating the trend of the roll angle to determine the trend The vehicle is a 'face' step. The vehicle dynamics prediction method according to claim 26, wherein the step of changing the roll angle is a step of estimating whether the roll angle is a divergence function. The vehicle dynamics prediction method described in claim 25, wherein after the step of predicting the dynamics of the vehicle, the step of estimating the future overturning position of the vehicle is further included. The vehicle dynamic prediction method according to claim 25, wherein after the step of predicting the dynamics of the vehicle, the method further comprises receiving the pitch angle, and estimating a trend of the pitch angle to determine whether the vehicle is overturned. . The vehicle dynamics prediction method according to claim 29, wherein the step of changing the pitch angle is a step of estimating whether the pitch angle is 27 200806513 as a divergence function. 31. A vehicle dynamic prediction method for predicting the dynamics of a vehicle, the vehicle dynamic prediction method comprising: measuring a longitudinal speed and a yaw rate of a centroid of the vehicle and a change of a length of a suspension spring of the vehicle And generating a longitudinal velocity signal, a yaw rate signal and a length change signal; receiving the longitudinal velocity signal, the yaw rate signal and the length _ degree change k, and estimating a roll angle of the vehicle, One side inclination speed, roll angle acceleration, one pitch angle, one pitch rate, one pitch angle acceleration, one yaw angle, one yaw rate acceleration, one longitudinal displacement, one longitudinal acceleration, one side displacement, one side speed a lateral acceleration, a vertical displacement, a vertical velocity, a vertical acceleration, and a tire angular velocity; and receiving the longitudinal velocity signal, the yaw angular velocity, the length variation signal, the roll angle, the roll rate, the side Inclination acceleration, • the pitch angle, the pitch rate, the pitch angle acceleration, the yaw angle, the yaw angle acceleration, the longitudinal direction The displacement, the longitudinal acceleration, the lateral displacement, the lateral velocity, the lateral acceleration, the vertical displacement, the vertical velocity, the nominal vertical acceleration, and the tire angular velocity are used to predict the dynamics of the vehicle. 32. The vehicle dynamic prediction method according to claim 31, wherein after the step of predicting the dynamics of the vehicle, the method further comprises receiving the roll angle, and estimating a trend of the roll angle to determine whether the vehicle is turned over. 200806513 Steps to overwrite the status. 3. The vehicle dynamics prediction method according to claim 3, wherein the step of changing the roll angle is a step of estimating whether the roll angle is a divergence function. 34. The vehicle dynamics prediction method of claim 31, wherein the step of predicting the dynamics of the vehicle further comprises the step of estimating a vehicle overturning position at a future time. The vehicle dynamic prediction method according to claim 31, wherein after the step of predicting the dynamics of the vehicle, the method further comprises receiving the pitch angle, and estimating a trend of the pitch angle to determine whether the vehicle is overturned. A step of. The vehicle dynamics prediction method according to claim 35, wherein the step of changing the pitch angle is a step of estimating whether the pitch angle is a divergence function. 37. A vehicle dynamic prediction method for predicting a vehicle dynamics, the vehicle dynamic prediction method comprising the steps of: measuring a longitudinal velocity and a yaw angle of a centroid of the vehicle and a length change of a suspension spring of the vehicle, And generating a longitudinal speed signal, a yaw angle signal and a length change signal; receiving the longitudinal speed signal, the yaw angle signal and the length change signal, and estimating a roll angle, a side tilt speed of the vehicle, One side inclination acceleration, one pitch angle, one pitch angle speed, one pitch angle acceleration, one yaw rate, one yaw rate acceleration, one longitudinal displacement, one longitudinal direction 29 200806513 acceleration, lateral displacement, lateral velocity, a lateral acceleration, a vertical displacement, a vertical velocity, a vertical acceleration, and a tire angular velocity; and receiving the longitudinal velocity signal, the yaw angle, the length change signal, the roll angle, the roll rate, the roll angle Acceleration, the pitch angle, the pitch rate, the pitch angle acceleration, the yaw rate, the yaw rate acceleration The longitudinal displacement, the longitudinal acceleration, the lateral displacement of the lateral speed, the lateral acceleration, the vertical displacement, the vertical speed, the vertical acceleration and the angular velocity of the tire, the vehicle of the predicted dynamics. 38. The vehicle dynamic prediction method according to claim 37, wherein after the step of predicting the dynamics of the vehicle, the method further comprises receiving the roll angle, and estimating a trend of the roll angle to determine whether the vehicle is overturned. A step of. 39. The vehicle dynamics prediction method according to claim 38, wherein the step of changing the roll angle is a step of estimating whether the roll angle is a divergence function. 40. The vehicle dynamics prediction method of claim 37, wherein the step of predicting the dynamics of the vehicle further comprises the step of estimating a vehicle overturning position at a future time. The vehicle dynamic prediction method according to claim 37, wherein after the step of predicting the dynamics of the vehicle, the method further comprises receiving the pitch angle, and estimating a trend of the pitch angle to determine whether the vehicle is turned over. 30 200806513 Steps to overwrite the status. The vehicle dynamics prediction method according to claim 41, wherein the step of changing the pitch angle is a step of estimating whether the pitch angle is a divergence function.

3131