CN107037808B - Track-finding method for transport vehicle based on sliding mode controller - Google Patents
Track-finding method for transport vehicle based on sliding mode controller Download PDFInfo
- Publication number
- CN107037808B CN107037808B CN201610814708.XA CN201610814708A CN107037808B CN 107037808 B CN107037808 B CN 107037808B CN 201610814708 A CN201610814708 A CN 201610814708A CN 107037808 B CN107037808 B CN 107037808B
- Authority
- CN
- China
- Prior art keywords
- strategy
- sliding mode
- action
- parameter
- transport vehicle
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 54
- 238000001514 detection method Methods 0.000 claims abstract description 18
- 230000009467 reduction Effects 0.000 claims abstract description 10
- 230000009471 action Effects 0.000 claims description 46
- 230000006870 function Effects 0.000 claims description 20
- 230000008569 process Effects 0.000 claims description 15
- 230000007704 transition Effects 0.000 claims description 12
- 238000013507 mapping Methods 0.000 claims description 9
- 230000008859 change Effects 0.000 claims description 8
- 230000005540 biological transmission Effects 0.000 claims description 4
- 238000005070 sampling Methods 0.000 claims description 4
- 230000003247 decreasing effect Effects 0.000 claims description 3
- 239000011159 matrix material Substances 0.000 claims description 3
- 230000000694 effects Effects 0.000 description 4
- 230000033001 locomotion Effects 0.000 description 4
- 230000004044 response Effects 0.000 description 3
- 230000009466 transformation Effects 0.000 description 2
- 241000232299 Ralstonia Species 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000009776 industrial production Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
- G05D1/02—Control of position or course in two dimensions
- G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
- G05D1/0259—Control of position or course in two dimensions specially adapted to land vehicles using magnetic or electromagnetic means
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/04—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- Electromagnetism (AREA)
- Aviation & Aerospace Engineering (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
- Feedback Control In General (AREA)
Abstract
The invention discloses a track-seeking method for a transport vehicle based on a sliding mode controller, which uses the sliding mode controller to control the running track of the transport vehicle so that the running track fits a magnetic track path, and comprises the following steps: (1) acquiring and obtaining the distance error e from the central position of the transport vehicle to the magnetic track by using a detection unitdAnd the deviation angle e of the transport carriage relative to the track pathθ(ii) a (2) According to the distance error edThe size of the epsilon parameter of the corresponding discrete sliding mode approach law in the sliding mode controller is adjusted, so that the epsilon parameter is reduced along with the reduction of the distance error, and the epsilon parameter is a parameter for controlling the approach rate in the discrete sliding mode approach law; (3) according to the determined epsilon parameter and distance error edAngle of deviation eθAnd calculating and adjusting the speed of the left wheel and the right wheel of the conveying vehicle by using a discrete sliding mode approach law so as to adjust the running path of the conveying vehicle. The invention dynamically adjusts the epsilon parameter, so that the AGV can quickly enter a stable tracking state, and after the system is stable, the buffeting is small, even the buffeting is eliminated.
Description
Technical Field
The invention relates to a track-seeking method for a transport vehicle, in particular to a track-seeking method for the transport vehicle by using a sliding mode controller.
Background
The transport vehicle (AGV, automated guided vehicle) is widely used in industrial production such as warehouse logistics, automated manufacturing and the like due to its unmanned control, flexibility and intelligent transportation characteristics, and the trajectory control of AGV is the core of the whole AGV behavior, and the following 3 methods are mainly adopted in the current AGV trajectory control: the AGV motion model is based on a feedback linearization method, but due to nonlinearity of the AGV motion model and complexity and changeability of the surrounding environment, the AGV motion model has larger error and weak robustness; although a method based on the adaptive PID controller can ensure certain robustness, the method has limited adaptability to uncertain systems; based on the sliding mode variable structure control method, the traditional sliding mode variable structure control can generate large buffeting in a system, so that the tracking effect is not ideal.
Therefore, a method for controlling a transport vehicle to control the movement of the transport vehicle is needed.
Disclosure of Invention
The invention aims to provide a track-finding method of a transport vehicle based on a sliding mode controller, which is characterized in that on the basis of a traditional sliding mode variable structure controller, a parameter epsilon of a discrete sliding mode approach law is dynamically adjusted to achieve the purpose of quickly approaching to a track-finding surface at the beginning of track finding as much as possible, and when the track-finding surface is close to stability, buffeting is reduced to approach equivalent control, so that buffeting of sliding mode control is compressed, even buffeting is eliminated, and the tracking effect is relatively ideal.
In order to achieve the purpose, the invention discloses a track searching method of a transport vehicle based on a sliding mode controller, which uses the sliding mode controller to control the running track of the transport vehicle so that the running track of the transport vehicle is attached to a magnetic track path, and comprises the following steps: (1) using detection unit to collectObtaining the distance error e from the central position of the transport vehicle to the magnetic trackdAnd the deviation angle e of the transport carriage relative to the track pathθ(ii) a (2) According to the distance error edThe size of the epsilon parameter of the corresponding discrete sliding mode approach law in the sliding mode controller is adjusted, so that the epsilon parameter is reduced along with the reduction of the distance error, and the epsilon parameter is a parameter for controlling the approach rate in the discrete sliding mode approach law; (3) according to the determined epsilon parameter and distance error edAngle of deviation eθAnd calculating and adjusting the speed of the left wheel and the right wheel of the conveying vehicle by using a discrete sliding mode approach law so as to adjust the running path of the conveying vehicle.
The discussion of the epsilon parameter in the discrete sliding mode approach law can find that when the epsilon parameter is large, the speed of the system approaching the sliding mode surface in the sliding mode control is increased, namely the system can be converged to the sliding mode surface more quickly, the response speed is high, if the epsilon parameter is too large, although the system can be enabled to approach the sliding mode surface quickly, the system is easy to cause great buffeting, when the epsilon parameter is small, the speed of the system approaching the sliding mode surface is reduced, and when the epsilon parameter is extremely small, the sliding mode control is also gradually approached to equivalent control. Therefore, the epsilon parameter is dynamically changed, so that when the distance error of the transport vehicle is large, the transport vehicle is enabled to quickly approach the sliding mode surface to reduce the distance error at a higher speed, and the transport vehicle can quickly enter a stable tracking state. And after the system is stabilized, the system is close to equivalent control to track, and buffeting controlled by a sliding mode is compressed and even eliminated.
Preferably, the step (2) is specifically: calculating the distance error e according to a preset or existing strategy model of epsilon parameterdThe corresponding size of the epsilon parameter, the strategy of the epsilon parameter is the mapping of the epsilon parameter or the parameter change corresponding to each action of the sliding mode controller, and the strategy model of the epsilon parameter is a strategy set of the epsilon parameter. The scheme ensures that the system of the invention has quick response and simple physical realization.
Preferably, the sliding mode controller-based transport vehicle tracking method further comprises a method for determining the strategy model, and comprises the following steps: and (4) running the transport vehicle for a plurality of times in an off-line mode, calculating a strategy pi by utilizing a Markov decision process, and solving an optimal strategy to obtain a strategy model. The scheme enables the method to quickly enter a stable tracking state under different complex earth surface environments, the tracking effect is good under the stable state, the tracking buffeting is small, and the method has strong adaptability and robustness.
Specifically, the method for determining the policy model specifically comprises the following steps: the distance error is Markov-processed, and an epsilon parameter corresponding to each action is established according to a preset rule, so that the epsilon parameter is decreased according to the preset rule during each action; detecting a distance error e using the detection unitdObtaining the parameter of epsilon and the distance error e of each actiondAnd (3) mapping between the two, namely running the transport vehicle for a plurality of times in an off-line mode, and iteratively calculating an optimal strategy according to the strategy to obtain a strategy model.
More specifically, the specific steps of obtaining the strategy pi by using the markov decision process include: (11) detecting the distance error from the central position of the transport vehicle to the magnetic track by using a detection unit, and detecting the maximum deviation e of the magnetic detection unitdmaxThe values are divided into M levels, so that a set of states S-e corresponding to the distance error is establisheddiI is 1, 2 … M, wherein ediTo quantify the error, thereby measuring the distance error edMarkov process; (12) establishing an action set A, representing a primary action a by using the change degree of the epsilon parameter, and setting the primary action a or the epsilon parameter as descending at different speeds according to a preset rule so that the epsilon parameter is reduced along with the reduction of the distance error; (13) a mapping of the state set S to the action set a is obtained.
More specifically, the method for iteratively calculating the optimal strategy according to the strategy specifically comprises the following steps: (A) calculating a return value R corresponding to each action a; (B) and solving a strategy corresponding to the optimal return value according to the optimal principle of the maximum return expectation so as to obtain the optimal strategy.
More specifically, the step (a) specifically includes: designing a return function R ═ C/(Δ e)di+ l) where C, l is a constant, Δ ediFor the variation degree of the distance error of each action, calculating each action pair according to the return functionThe corresponding reward value R.
More specifically, the step (B) specifically includes: determining optimal reward criteria
And (4) calculating a strategy corresponding to the optimal return value according to the optimal return criterion for the discount factor, wherein K is M, and E is a constant.
More specifically, the step (B) is followed by the step of determining the probability of the state transition: when state SiIf action a is takenkSlave state SiTo state SjThen the probability of this transition is the state transition probability (P)sa)ijkSo that each action has a state transition probability matrix
And (7) corresponding.
More specifically, the specific steps of iteratively calculating the optimal strategy according to the strategy to obtain the strategy model include:
defining a value function VπS → R employs the strategy π, with an expected reward in state S:
skthe state at the sampling time, where k is 0, is an initial state s, and γ is the transmission ratio of the reduction gear of the transport vehicle, and is represented in a recursive manner as:
performing a slave value function Vπ(s) Process to strategy π, defining action value function QπS.A → R, to take action a in state S, the expected return of strategy π in other states:
the policy function is obtained by maximizing the action value function:
repeating the steps to obtain an optimal strategy through continuous iteration, thereby obtaining the strategy model.
Preferably, the detecting unit includes a magnetic sensor and a potentiometer, the distance error is acquired by using the magnetic sensor in the step (1), and the deviation angle e is acquired by using the potentiometerθ. Of course, the detection unit may also adopt other detection devices for data acquisition.
Preferably, the step (3) is specifically: according to the determined epsilon parameter and distance error edAngle of deviation eθAnd calculating the control speed of the left wheel and the right wheel of the conveying vehicle according to a discrete sliding mode approach law, calculating the control pressure difference of a left wheel driving motor and a right wheel driving motor of the conveying vehicle according to the control speed, and correspondingly adjusting the voltage of the left wheel and the voltage of the right wheel according to the control pressure difference so as to control the speed of the left wheel and the right wheel of the conveying vehicle and adjust the running path of the conveying vehicle.
Drawings
Fig. 1 is a flowchart of a tracking method of a feed carriage according to a first embodiment of the present invention.
Fig. 2 is a flowchart of a tracking method of a transport vehicle according to a second embodiment of the present invention.
Detailed Description
In order to explain technical contents, structural features, and objects and effects of the present invention in detail, the following detailed description is given with reference to the accompanying drawings in conjunction with the embodiments.
The invention discloses a track-finding method for a conveying vehicle based on a sliding mode controller.
Referring to fig. 1, the tracking method of the transport vehicle includes the steps of: (1) acquiring and obtaining the distance error e from the central position of the transport vehicle to the magnetic track by using a detection unitdAnd the deviation angle e of the transport carriage relative to the track pathθ(ii) a (2) According to the distance error edSize adjustment of said slip-form controlThe size of an epsilon parameter of a corresponding discrete sliding mode approximation rule in the device enables the epsilon parameter to be reduced along with the reduction of the distance error, and the epsilon parameter is a parameter for controlling the approximation rate in the discrete sliding mode approximation rule; (3) according to the determined epsilon parameter and distance error edAngle of deviation eθAnd calculating and adjusting the speed of the left wheel and the right wheel of the conveying vehicle by a discrete sliding mode approach law so as to adjust the running path of the conveying vehicle, namely the determined epsilon parameter and distance error edAngle of deviation eθAnd substituting the speed into a discrete sliding mode approach law to calculate the speed of the left wheel and the right wheel of the conveying vehicle. The invention dynamically changes the epsilon parameter, so that when the distance error of the transport vehicle is large, the transport vehicle is enabled to quickly approach the sliding mode surface to reduce the distance error at a faster speed, thereby enabling the transport vehicle to quickly enter a stable tracking state. And after the system is stabilized, the system is close to equivalent control to track, and buffeting controlled by a sliding mode is compressed and even eliminated.
Wherein the step (2) specifically comprises: calculating the distance error e according to a preset or existing strategy model of epsilon parameterdAnd (4) optimizing the epsilon parameter of the sliding mode controller according to the strategy model according to the corresponding epsilon parameter. The strategy of the epsilon parameter is mapping of epsilon parameter or parameter change corresponding to each action of the sliding mode controller, and the strategy model of the epsilon parameter is a strategy set of the epsilon parameter. The scheme ensures that the system of the invention has quick response and simple physical realization.
Wherein, the detection unit comprises a magnetic sensor and a potentiometer, the distance error is acquired by adopting the magnetic sensor in the step (1), and the deviation angle e is acquired by adopting the potentiometerθ. Of course, the detection unit may also adopt other detection devices for data acquisition.
Referring to fig. 1, the step (3) is specifically: according to the determined epsilon parameter and distance error edAngle of deviation eθCalculating the control speed of the left wheel and the right wheel of the conveying vehicle according to a discrete sliding mode approach law, calculating the control pressure difference of a left wheel driving motor and a right wheel driving motor of the conveying vehicle according to the control speed, and correspondingly adjusting the left wheel and the right wheel according to the control pressure differenceThe wheel voltages thus control the left and right wheel speeds of the transport vehicle to adjust the travel path of the transport vehicle.
In a second embodiment of the present invention, the method for tracking a transport vehicle further comprises a method for determining the strategy model, comprising the steps of: and (4) running the transport vehicle for a plurality of times in an off-line mode, calculating a strategy pi by utilizing a Markov decision process, and solving an optimal strategy to obtain a strategy model. Referring to fig. 2, the method for determining the policy model includes the following specific steps: (21) the distance error is Markov processed, (22) an epsilon parameter corresponding to each action is established according to a preset rule, so that the epsilon parameter is decreased according to the preset rule in each action, namely, a Markov decision process epsilon parameter model is set; (23) detecting a distance error e using the detection unitd(24) acquiring the parameter of epsilon and the distance error e of each actiondAnd (3) mapping between the two strategies, namely running the transport vehicle for a plurality of times in an off-line mode, and iteratively calculating the optimal strategy according to the strategy to obtain a strategy model, namely learning the optimal strategy in an off-line mode to obtain the strategy model.
Wherein the step of Markov procedural distance error comprises: detecting the distance error from the central position of the transport vehicle to the magnetic track by using a detection unit, and detecting the maximum deviation e of the magnetic detection unitdmaxThe values are divided into M levels, so that a set of states S-e corresponding to the distance error is establisheddiI is 1, 2 … M, wherein ediTo quantify the error, thereby measuring the distance error edMarkov procedural.
In the step (22), the step of establishing an epsilon parameter corresponding to each action according to a preset rule includes: and establishing an action set A, representing an action a by using the change degree of the epsilon parameter, and setting the action a or the epsilon parameter of each action as the decrement of different speeds according to a preset rule so that the epsilon parameter is reduced along with the reduction of the distance error. In the step (24), a mapping from the state set S to the action set A is obtained along with the operation of the transport vehicle.
In the step (24), the method for iteratively calculating the optimal strategy according to the strategy specifically includes: (A) calculate each actiona corresponding return value R; (B) and solving a strategy corresponding to the optimal return value according to the optimal principle of the maximum return expectation so as to obtain the optimal strategy. Specifically, the step (a) specifically includes: designing a return function R ═ C/(Δ e)di+ l) where C, l is a constant, Δ ediFor the variation degree of the distance error of each action, a corresponding report value R of each action is calculated according to a report function. Specifically, the step (B) specifically includes: determining optimal reward criteria
And (4) calculating a strategy corresponding to the optimal return value according to the optimal return criterion for the discount factor, wherein K is M, and E is a constant.
Preferably, the step (B) further comprises the step of obtaining the probability of the state transition: when state SiIf action a is takenkSlave state SiTo state SjThen the probability of this transition is the state transition probability (P)sa)ijkSo that each action has a state transition probability matrix
And (7) corresponding.
The process of the invention for optimising a sliding mode controller using a markov decision process, namely the specific step of said step (24), is described: defining a value function VπS → R employs the strategy π, with an expected reward in state S:
skthe state at the sampling time, where k is 0, is an initial state s, and γ is the transmission ratio of the reduction gear of the transport vehicle, and is represented in a recursive manner as:
performing a slave value function Vπ(s) Process to strategy π, defining action value function QπS.A → R, to take action a in state S, the expected return of strategy π in other states:
the policy function is obtained by maximizing the action value function:
repeating the steps to obtain an optimal strategy through continuous iteration, thereby obtaining the strategy model.
In the track tracking of the AGV, a sliding mode controller is widely applied, but for a complex ground environment, the buffeting amplitude of the conventional sliding mode controller is often large at present, and the tracking precision often cannot meet the requirement.
The working principle of the sliding mode controller is described as follows:
for the transport vehicle, the armature voltage of the motor with left and right wheels is Uc,UrThe load constant is TmConstant of reverse electromotive force of kdIf the radius of the wheel is r and the transmission ratio of the speed reducing mechanism is gamma, the following steps are performed:
after Laplace transformation, the following can be obtained:
when the motor speed is changed, the change amount of the voltage is expressed by a delta U, and U is usedcRepresents the reference driving voltage, then UlAnd UrCan be expressed as:
Ul＝Uc+ΔU；
Ur＝Uc-ΔU；
then let the linear speed of the left and right wheels of the AGV be Vl，VrThen it can be obtained:
Vl＝Vc+ΔV；
Vr＝Vc-ΔV；
let the AGV geometric center-to-track error be edThe error between the geometric positive direction of the transport vehicle and the tangential direction of the magnetic track is eθAssuming that at time elapses, the following is satisfied:
the same can be obtained:
the formula is combined to obtain:
where f (t) is random disturbances of system parameter disturbances and external variations, but these disturbances are rooted in changes to Δ V, so defining a Δ VfRepresenting the disturbance speed difference, without loss of generality, the above equation can be expressed as:
it is modeled as
Discrete expression x (k +1) ═ xAx(k)+Bv(k)；
And (3) changing the speed difference into a Ralstonia transform:
after the inverse transformation:
discretizing it into:
to this end, the differential speed of one sampling interval phase can be used to represent the change in two-wheel differential pressure, for Δ V as described abovefThe term can be ignored in actual processing since the later-mentioned sliding mode variable structure control can eliminate such disturbances.
In the sliding mode control process, the track control aims to make the track x (k) of the AGV in the operation process fit the track path r (k) as much as possible, so that a sliding mode surface equation (switching function) of the system is defined as follows:
s(k)＝C(R(k)-x(k))；
so that:
s(k+1)＝C(R(k+1)-x(k+1))＝CR(K+1)-CAx(k)-CBv(k)；
when CB ≠ 0, the signal quality of the signal is improved,
v(k)＝(CB)-1[CR(k+1)-CAx(k)-s(k+1)]；
taking the approximation law of the discrete sliding mode as follows:
s (k +1) -s (k) ═ - δ ts (k) - ε Tsgn (s (k))) wherein δ >0, ε >0,1- ε T > 0;
then:
v(k)＝(CB)-1[CR(k+1)-CAx(k)-(1-δT)s(k)+εTsgn(s(k)]；
the sliding mode controller controls the conveying vehicle by using a discrete sliding mode approach law, which is well known in the art and will not be further described herein.
The above disclosure is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the scope of the present invention, therefore, the present invention is not limited by the appended claims.
Claims (8)
1. A track seeking method of a transport vehicle based on a sliding mode controller controls the running track of the transport vehicle by using the sliding mode controller, so that the running track of the transport vehicle is attached to a track path, and is characterized by comprising the following steps of:
(1) acquiring and obtaining the distance error e from the central position of the transport vehicle to the magnetic track by using a detection unitdAnd the deviation angle e of the transport carriage relative to the track pathθ；
(2) Calculating the distance error e according to a preset or existing strategy model of epsilon parameterdThe corresponding epsilon parameter is reduced along with the reduction of the distance error, the epsilon parameter is a parameter for controlling and controlling the approaching rate in a discrete sliding mode approach law, the strategy of the epsilon parameter is mapping of the epsilon parameter or parameter change corresponding to each action of a sliding mode controller, and the strategy model of the epsilon parameter is a strategy set of the epsilon parameter;
(3) according to the determined epsilon parameter and distance error edAngle of deviation eθCalculating and adjusting the speed of the left wheel and the right wheel of the conveying vehicle by using a discrete sliding mode approach law so as to adjust the running path of the conveying vehicle;
the method for determining the strategy model comprises the following steps: and (4) running the transport vehicle for a plurality of times in an off-line mode, calculating a strategy pi by utilizing a Markov decision process, and solving an optimal strategy to obtain a strategy model.
2. The sliding mode controller-based conveyor car tracking method according to claim 1, wherein the method for determining the strategy model comprises the specific steps of: the distance error is Markov-processed, an epsilon parameter corresponding to each action is established according to a preset rule, so that the epsilon parameter is decreased according to the preset rule during each action, and the detection unit is used for detecting the distanceError edObtaining the parameter of epsilon and the distance error e of each actiondAnd (3) mapping between the two, namely running the transport vehicle for a plurality of times in an off-line mode, and iteratively calculating an optimal strategy according to the strategy to obtain a strategy model.
3. The sliding-mode-controller-based conveying vehicle tracking method according to claim 2, wherein the specific step of calculating the strategy pi by using the Markov decision process comprises the following steps:
(11) detecting the distance error from the central position of the transport vehicle to the magnetic track by using a detection unit, and detecting the maximum deviation e of the magnetic detection unitdmaxThe values are divided into M levels, so that a set of states S-e corresponding to the distance error is establisheddiI is 1, 2 … M, wherein ediTo quantify the error, thereby measuring the distance error edMarkov process;
(12) establishing an action set A, representing a primary action a by using the change degree of the epsilon parameter, and setting the primary action a or the epsilon parameter as descending at different speeds according to a preset rule so that the epsilon parameter is reduced along with the reduction of the distance error;
(13) a mapping of the state set S to the action set a is obtained.
4. The sliding-mode-controller-based transport vehicle tracking method according to claim 3, wherein the method for iteratively calculating the optimal strategy according to the strategy comprises the following specific steps: (A) calculating a return value R corresponding to each action a; (B) and solving a strategy corresponding to the optimal return value according to the optimal principle of the maximum return expectation so as to obtain the optimal strategy.
5. The sliding mode controller-based conveying vehicle tracking method according to claim 4, wherein the step (A) specifically comprises: designing a return function R ═ C/(Δ e)di+ l) where C, l is a constant, Δ ediFor the variation degree of the distance error of each action, a corresponding report value R of each action is calculated according to a report function.
6. The sliding mode controller-based transport vehicle tracking method according to claim 4, wherein the step (B) specifically comprises: determining optimal reward criteria
And (4) calculating a strategy corresponding to the optimal return value according to the optimal return criterion for the discount factor, wherein K is M, and E is a constant.
7. The sliding-mode-controller-based transport vehicle tracking method according to claim 6, wherein the method for iteratively calculating the optimal strategy according to the strategy further comprises the step of obtaining the probability of state transition: when state SiIf action a is takenkSlave state SiTo state SjThen the probability of this transition is the state transition probability (P)sa)ijkSo that each action has a state transition probability matrix
And (7) corresponding.
8. The sliding-mode-controller-based conveying vehicle tracking method according to claim 7, wherein the specific step of iteratively calculating the optimal strategy according to the strategy to obtain the strategy model comprises the following steps:
defining a value function VπS → R employs the strategy π, with an expected reward in state S:
skthe state at the sampling time, where k is 0, is an initial state s, and γ is the transmission ratio of the reduction gear of the transport vehicle, and is represented in a recursive manner as:
performing a slave value functionVπ(s) Process to strategy π, defining action value function QπS.A → R, to take action a in state S, the expected return of strategy π in other states:
the policy function is obtained by maximizing the action value function:
repeating the steps to obtain an optimal strategy through continuous iteration, thereby obtaining the strategy model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610814708.XA CN107037808B (en) | 2016-09-09 | 2016-09-09 | Track-finding method for transport vehicle based on sliding mode controller |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610814708.XA CN107037808B (en) | 2016-09-09 | 2016-09-09 | Track-finding method for transport vehicle based on sliding mode controller |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107037808A CN107037808A (en) | 2017-08-11 |
| CN107037808B true CN107037808B (en) | 2020-04-24 |
Family
ID=59532565
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610814708.XA Active CN107037808B (en) | 2016-09-09 | 2016-09-09 | Track-finding method for transport vehicle based on sliding mode controller |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107037808B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110231041B (en) * | 2018-03-06 | 2021-05-25 | 北京京东尚科信息技术有限公司 | Navigation method and device for lane switching |
| WO2020000127A1 (en) * | 2018-06-25 | 2020-01-02 | 深圳市大疆创新科技有限公司 | Navigation path tracking control method, device, mobile robot and system |
| CN112578671B (en) * | 2020-12-11 | 2022-05-27 | 上海应用技术大学 | AGV track tracking control method based on U model optimization SMC |
| CN112947072B (en) * | 2021-01-28 | 2024-04-05 | 西安建筑科技大学 | Mobile robot driving control method, system and equipment based on sliding mode variable structure |
| CN113051767A (en) * | 2021-04-07 | 2021-06-29 | 绍兴敏动科技有限公司 | AGV sliding mode control method based on visual servo |
| KR102605907B1 (en) * | 2023-04-06 | 2023-11-27 | 세종대학교 산학협력단 | Method and apparatus for adaptive sliding mode attitude control for spacecraft |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103538068A (en) * | 2013-10-12 | 2014-01-29 | 江南大学 | Fuzzy sliding mode trajectory tracking control method for SCARA robot |
| CN104932506A (en) * | 2015-06-09 | 2015-09-23 | 东南大学 | Wheel type moving robot track tracking method based on fast terminal sliding mode |
-
2016
- 2016-09-09 CN CN201610814708.XA patent/CN107037808B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103538068A (en) * | 2013-10-12 | 2014-01-29 | 江南大学 | Fuzzy sliding mode trajectory tracking control method for SCARA robot |
| CN104932506A (en) * | 2015-06-09 | 2015-09-23 | 东南大学 | Wheel type moving robot track tracking method based on fast terminal sliding mode |
Non-Patent Citations (8)
| Title |
|---|
| Variable structure control for affine systems based on sigmoid variable rate reaching law;Yun-long Liu 等;《Proceedings of the 8th World Congress on Intelligent Control and Automation》;20100731;全文 * |
| 一种改进的离散时间系统的变结构控制;翟长连 等;《计算技术与自动化》;19990630(第2期);第27-30页 * |
| 几类不确定时滞系统的滑模变结构控制;孟波;《万方》;20121227;全文 * |
| 基于策略迭代和值迭代的POMDP算法;孙湧 等;《计算机研究与发展》;20081231;全文 * |
| 基于马尔科夫理论的不确定性规划和感知问题研究;柏爱俊;《中国博士学位论文全文数据库信息科技辑》;20150615;全文 * |
| 履带式机器人系统的滑模控制研究;李小龙;《中国优秀硕士学位论文全文数据库信息科技辑》;20160315;全文 * |
| 滑模变结构控制理论及其算法研究与进展;刘金琨 等;《控制理论与应用》;20070630;全文 * |
| 视觉AGV的运动控制策略研究;刘忠强 等;《机械科学与技术》;20150731(第7期);第1072-1074页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107037808A (en) | 2017-08-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107037808B (en) | Track-finding method for transport vehicle based on sliding mode controller | |
| Rokonuzzaman et al. | Review and performance evaluation of path tracking controllers of autonomous vehicles | |
| CN110018632B (en) | Vehicle lane change control method and device | |
| Cui et al. | Extended state observer-based adaptive sliding mode control of differential-driving mobile robot with uncertainties | |
| CN107831761B (en) | Path tracking control method of intelligent vehicle | |
| CN112622903B (en) | Longitudinal and transverse control method for autonomous vehicle in vehicle following driving environment | |
| CN106123922B (en) | The magnetic navigation AGV straight trip route method for correcting error adjusted based on proportion differential | |
| CN108153309B (en) | Control method for tracked robot and tracked robot | |
| CN108646756B (en) | Intelligent automobile transverse control method and system based on segmented affine fuzzy sliding mode | |
| CN108594639B (en) | Reinforced learning-based track tracking control method for hovercraft | |
| CN201993646U (en) | Double closed loop control system of industrial laser guidance AGV (Automatic Guided Vehicle) | |
| CN110989597B (en) | Adaptive path tracking method of integrated fuzzy neural network | |
| Pandey et al. | Modeling and control of an autonomous three wheeled mobile robot with front steer | |
| CN112578671B (en) | AGV track tracking control method based on U model optimization SMC | |
| Chen et al. | Real-time predictive sliding mode control method for AGV with actuator delay | |
| Chen et al. | An adaptive path tracking controller based on reinforcement learning with urban driving application | |
| Yoo | Adaptive tracking and obstacle avoidance for a class of mobile robots in the presence of unknown skidding and slipping | |
| CN110109363B (en) | Neural network self-adaptive control method for wheeled mobile robot formation | |
| Zhou et al. | Modeling and simulation research of heavy-duty AGV tracking control system based on magnetic navigation | |
| Cheng et al. | A game theoretical chassis domain approach to trajectory tracking for automated vehicles | |
| Xie et al. | Parameter self‐learning feedforward compensation‐based active disturbance rejection for path‐following control of self‐driving forklift trucks | |
| CN115079569A (en) | AGV nonlinear active disturbance rejection control method based on longicorn stigma search algorithm | |
| Miah et al. | Linear time-varying feedback law for vehicles with ackermann steering | |
| Li et al. | Path-following control for self-driving forklifts based on cascade disturbance rejection with coordinates reconstruction | |
| Meng et al. | Parking path tracking method based on Kalman filter and fuzzy control |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |