CN115256395B - Model uncertain robot safety control method based on control obstacle function - Google Patents
Model uncertain robot safety control method based on control obstacle function Download PDFInfo
- Publication number
- CN115256395B CN115256395B CN202210989731.8A CN202210989731A CN115256395B CN 115256395 B CN115256395 B CN 115256395B CN 202210989731 A CN202210989731 A CN 202210989731A CN 115256395 B CN115256395 B CN 115256395B
- Authority
- CN
- China
- Prior art keywords
- robot
- model
- control
- representing
- joint
- 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 65
- 230000008569 process Effects 0.000 claims abstract description 26
- 230000001133 acceleration Effects 0.000 claims description 15
- 239000011159 matrix material Substances 0.000 claims description 10
- 230000010354 integration Effects 0.000 claims description 7
- 230000004888 barrier function Effects 0.000 claims description 5
- 230000005484 gravity Effects 0.000 claims description 3
- 238000013499 data model Methods 0.000 claims 1
- 239000012636 effector Substances 0.000 description 8
- 238000004088 simulation Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 230000003993 interaction Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J9/00—Programme-controlled manipulators
- B25J9/16—Programme controls
- B25J9/1674—Programme controls characterised by safety, monitoring, diagnostic
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
- B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
- B25J9/00—Programme-controlled manipulators
- B25J9/16—Programme controls
- B25J9/1602—Programme controls characterised by the control system, structure, architecture
- B25J9/1605—Simulation of manipulator lay-out, design, modelling of manipulator
Landscapes
- Engineering & Computer Science (AREA)
- Robotics (AREA)
- Mechanical Engineering (AREA)
- Automation & Control Theory (AREA)
- Feedback Control In General (AREA)
- Numerical Control (AREA)
Abstract
The invention relates to a model uncertain robot safety control method based on a control obstacle function, belongs to the technical field of robot control, and particularly relates to a method for realizing safety control based on the control obstacle function under the condition that a robot dynamics model is uncertain. The invention adopts a Gaussian process regression method, uses less data to quickly obtain a more accurate model estimation value, and utilizes the learning capacity of the Gaussian process to obtain an uncertain part of the model with a certain confidence, thereby completing the safety control of the robot by adopting a method based on a control obstacle function.
Description
Technical Field
The invention relates to a model uncertain robot safety control method based on a control obstacle function, belongs to the technical field of robot control, and particularly relates to a method for realizing safety control based on the control obstacle function under the condition that a robot dynamics model is uncertain.
Background
In the current robot technical field, security is receiving more and more attention as one of the most important performances in the process of robot interaction with the environment. In order to meet the safety requirement, safety constraint conditions, such as avoidance of obstacles in the track tracking process, contact force limitation during man-machine interaction to ensure human body safety, and the like, need to be considered when designing the controller. The control design method commonly used at present comprises model predictive control, an artificial potential field method and the like. In addition, safety control methods based on control obstacle functions have recently been developed with higher computational efficiency compared to model predictive control and less conservation compared to artificial potential field methods.
The control methods all depend on an accurate dynamic model of the robot, but the robot is a strong coupling, nonlinear system and even has interference problems, so that the accurate model is difficult to establish. Therefore, the safety effect of the controller designed by the method cannot be guaranteed.
Disclosure of Invention
The technical solution of the invention is as follows: the method combines track planning and track tracking tasks, and under the condition that a robot dynamics model is inaccurate, the model uncertainty part is estimated through Gaussian process regression, so that the design of a safety controller is completed based on the control obstacle function, and the safety of the tasks is guaranteed in the whole process.
The technical scheme of the invention is as follows:
The model uncertain robot safety control method based on the control obstacle function comprises the following steps:
Firstly, establishing a robot dynamics estimation model;
Secondly, obtaining the robot joint speed meeting the safety constraint, and carrying out numerical integration on the robot joint speed to obtain the safe robot joint position;
Thirdly, tracking the safe robot joint position obtained in the second step to obtain the robot joint speed, the robot joint acceleration and the robot joint moment in the tracking process;
Fourthly, tracking the robot dynamics estimation model established in the first step to obtain model joint acceleration in the tracking process;
fifthly, calculating a difference value between the robot joint acceleration obtained in the third step and the model joint acceleration obtained in the fourth step to obtain model error data;
sixth, carrying out Gaussian process regression according to the robot joint speed, the robot joint acceleration and the robot joint moment obtained in the third step and the model error data obtained in the fifth step to obtain a robot dynamics accurate model;
Seventh, according to the robot dynamics accurate model obtained in the sixth step, obtaining a control instruction meeting safety constraint, and carrying out safety control on the robot by the obtained control instruction meeting the safety constraint, thereby completing uncertain robot safety control based on the model of the control obstacle function.
In the first step, the established robot dynamics estimation model is as follows:
Wherein, Representing the positions of n joints,/>Is an inertial matrix of the robot and,Representing the centrifugal and Coriolis force matrices,/>Representing gravity term,/> M (q)/>, respectivelyEstimated value of G (q)/>Is a model error; u is the control input.
In the second step, a quadratic programming model is adopted to obtain the robot joint speed meeting the safety constraint.
The quadratic programming model is as follows:
qmin≤q≤qmax
Wherein, Representing the robot joint speed satisfying the safety constraint, q representing the current robot joint position; /(I)Representing the current robot joint speed,/>Indicating the expected joint velocity, e=x (q) -x d (t) indicating the robot tip position tracking error, t indicating the current time, γ >0, γ being a constant, being a set value,/>Representing a pseudo-inverse matrix; x (q) is the current robot end position, x d (t) is the desired robot end position,/>A derivative of x d (t); /(I)Is the derivative of x (q) and/>J y (q) represents a robot jacobian matrix; q min denotes a robot minimum joint position, q max denotes a robot maximum joint position, h (q) denotes a control obstacle function,/>Represents the derivative of h (q), α (h (q)) represents a monotonically increasing function, and α (0) =0.
For a pair ofAnd carrying out numerical integration to obtain the safe robot joint position q safe.
In the third step, a PID control model is adopted to track the safe robot joint position, and in the fourth step, a PID control model is adopted to track a robot dynamics estimation model.
The PID control model is input as follows
Wherein k p,kd,ki is a control parameter to be designed, in the tracking process, the control time [0, t * ] is discretized into N times t j, j=1, 2, & gt, N, the current robot joint position q (t j) is fed back at each time, the control input is updated to complete the track tracking, in the whole control process, the joint position q (t j) at each time, the joint speedPID control input u pid(tj) as input value of Gaussian regression training data/>Model error dataOutput value as training data/>And save the dataset [ X train,Ytrain ].
In the sixth step, the method for performing Gaussian process regression comprises the following steps:
Each time 100 groups of continuous data are selected Training is carried out, and model errors under the current robot joint position are predicted according to training results, wherein the prediction results are expressed as follows: /(I) Wherein represents the prediction mean, delta represents the prediction variance, k is related to the selection of confidence intervals, the prediction result based on model error data,Is the i-th subset of the saved dataset [ X train,Ytrain ].
In the seventh step, a trajectory tracking quadratic programming model is adopted to obtain a control instruction meeting the safety constraint.
The track tracking quadratic programming model is as follows:
umin≤u≤umax
wherein k 1>0,k2>0,Lf and L g represent robot edges AndIn (2), where h (q) is a second order control barrier function,/>, is providedRepresenting a model error derivative; u safe is a control instruction that satisfies a safety constraint, and u min,umax represents a maximum value and a minimum value of the control input, respectively.
Compared with the prior art, the invention provides a method for realizing safety control based on a control obstacle function under the condition that a robot dynamics model is inaccurate, and the method has the following remarkable advantages:
(1) According to the model uncertain robot safety control method based on the control obstacle function, the track planning task and the track tracking task in the robot safety control are combined, the track tracking is realized by solving the safety control instruction after the position of the robot safety joint is obtained, the safety constraint is met in the whole process, and the safety is stronger;
(2) In the model uncertain robot safety control method based on the control obstacle function, the secondary planning model is used, so that the joint position of the robot safety can be obtained under the condition of not depending on the robot model;
(3) In the model uncertain robot safety control method based on the control obstacle function, a Gaussian process training data storage method is used, and compared with randomly selected training data, a model error estimation result has higher confidence coefficient;
(4) In the model uncertain robot safety control method based on the control obstacle function, the method for predicting the model error by using the Gaussian process can quickly obtain the model error estimated value by using less data, and obtain the robot dynamics accurate model with higher confidence;
(5) In the model uncertain robot safety control method based on the control obstacle function, the trajectory tracking quadratic programming model is used, and the constraint conditions related to the control obstacle function are combined with the robot model error estimated value to realize safe trajectory tracking, so that the calculation efficiency is high and the implementation is easy.
Drawings
FIG. 1 is a diagram of simulation results of a trajectory planning of an end effector of a robotic arm in an embodiment;
FIG. 2 is a graph comparing simulation results of motion trajectories of an end effector of a mechanical arm under the condition of safety constraint in an embodiment;
FIG. 3 is a diagram of control input simulation results for ensuring security in an embodiment;
Fig. 4 is a schematic diagram of a three degree of freedom planar mechanical arm structure.
Detailed Description
The invention is further described below with reference to the drawings and examples.
The model uncertain robot safety control method based on the control obstacle function comprises the following steps:
Firstly, establishing a robot dynamics estimation model;
Secondly, acquiring the robot joint speed meeting the safety constraint by adopting a quadratic programming model, and carrying out numerical integration on the robot joint speed so as to acquire the safe robot joint position;
Thirdly, tracking the safe robot joint position obtained in the second step by adopting a PID control model to obtain the robot joint speed, the robot joint acceleration and the robot joint moment in the tracking process;
tracking the robot dynamics estimation model established in the first step by adopting a PID control model to obtain model joint acceleration in the tracking process;
fifthly, calculating a difference value between the robot joint acceleration obtained in the third step and the model joint acceleration obtained in the fourth step to obtain model error data;
sixth, carrying out Gaussian process regression according to the robot joint speed, the robot joint acceleration and the robot joint moment obtained in the third step and the model error data obtained in the fifth step to obtain a robot dynamics accurate model;
Seventh, the precise robot dynamics model obtained in the sixth step is subjected to a trajectory tracking quadratic programming model to obtain a control instruction meeting safety constraint, the obtained control instruction meeting the safety constraint is used for carrying out safety control on the robot, and the safety control of the robot is completed based on the model uncertainty of the control obstacle function.
In the first step, the robot dynamics estimation model is built as follows:
Wherein, Representing the positions of n joints,/>Is an inertial matrix of the robot and,Representing the centrifugal and Coriolis force matrices,/>Representing gravity term,/> M (q)/>, respectivelyEstimated value of G (q)/>Is a model error; u is a control input;
in the second step, the quadratic programming model is as follows:
qmin≤q≤qmax
Wherein, Representing the robot joint speed satisfying the safety constraint, q representing the current robot joint position; /(I)Representing the current robot joint speed,/>Indicating the expected joint velocity, e=x (q) -x d (t) indicating the robot tip position tracking error, t indicating the current time, γ >0, γ being a constant, being a set value,/>Representing a pseudo-inverse matrix; x (q) is the current robot end position, x d (t) is the desired robot end position,/>A derivative of x d (t); /(I)Is the derivative of x (q) and/>J y (q) represents a robot jacobian matrix; q min denotes a robot minimum joint position, q max denotes a robot maximum joint position, h (q) denotes a control obstacle function,/>Represents a derivative of h (q), α (h (q)) represents a monotonically increasing function, and α (0) =0;
For a pair of Performing numerical integration to obtain a safe robot joint position q safe;
In the third step and the fourth step, the PID control model is input
Wherein k p,kd,ki is a control parameter to be designed, in the tracking process, the control time [0, t * ] is discretized into N times t j, j=1, 2, & gt, N, the current robot joint position q (t j) is fed back at each time, the control input is updated to complete the track tracking, in the whole control process, the joint position q (t j) at each time, the joint speedPID control input u pid(tj) as input value of Gaussian regression training data/>Model error dataOutput value as training data/>And save the dataset [ X train,Ytrain ];
in the sixth step, the method for performing Gaussian process regression comprises the following steps:
Each time 100 groups of continuous data are selected Training is carried out, and model errors under the current robot joint position are predicted according to training results, wherein the prediction results are expressed as follows: /(I) Wherein represents the prediction mean, delta represents the prediction variance, k is related to the selection of confidence intervals, the prediction result based on model error data,Is the ith subset in the saved dataset [ X train,Ytrain ];
In the seventh step, the trajectory tracking quadratic programming model is:
umin≤u≤umax
wherein k 1>0,k2>0,Lf and L g represent robot edges AndIn (2), where h (q) is a second order control barrier function,/>, is providedRepresenting a model error derivative; u safe is a control instruction that satisfies a safety constraint, and u min,umax represents a maximum value and a minimum value of the control input, respectively.
Examples
The invention selects the obstacle avoidance problem in the motion process of the three-degree-of-freedom planar mechanical arm end effector as an embodiment, and the detailed description is as follows in combination with the accompanying drawings:
The three-degree-of-freedom planar mechanical arm shown in fig. 4 is selected, and the structural parameters of each joint are as follows: m 1=1kg,m2=0.8kg,m3=0.6kg,L1=1m,L2=0.8m,L3 = 0.6m. The mechanical arm end effector reaches the end point (-1.44,1.84) along a straight line from the start point (-0.12,2.4), a circular obstacle with the radius of 0.1m exists at the point (-1, 2), and in order to ensure the end effector safety, the end effector needs to avoid the obstacle and deviate from the original track as little as possible.
In order to achieve this task, the specific steps are as follows:
a. Constructing a three-degree-of-freedom planar mechanical arm dynamics estimation model:
Wherein, Set as a third order identity matrix,/>And/>
B. Based on the control obstacle function, the following quadratic programming method is adopted to obtain the movement speed
qmin≤q≤qmax
Wherein,
γ=diag[27,8],/>h(q)=(x1(q)+1)2+(x2(q)-2)2-0.01,α(h(q))=h(q),qmin=-50,qmax=50.
For a pair ofPerforming numerical integration to obtain a safe robot joint position q safe;
As shown in fig. 1, when the robot arm moves at an angle q safe, the path completely avoids the obstacle and is minimally deviated from the original trajectory.
C. The articulation angle q safe is tracked using the input of the PID control model,
Wherein k p=diag[800,800,200],kd=diag[90,10,11],ki = diag [10,10,0.35]. The control time [0,20] is discretized into N times t j, j=1, 2..n, n=2000, and at each time the current robot joint position q is fed back (t j), the control input is updated to complete the trajectory tracking. Joint position q (t j), joint velocity at each moment throughout the control processInput u pid(tj of PID control model) as input value of gaussian regression training data
Model error dataSaved as output value of training data/>When there is a small error in the PID control results, it is possible to place the end effector in an unsafe position, as shown in fig. 2 by the motion profile of the PID control model alone.
D. For the saved dataset [ X train,Ytrain ], 100 consecutive sets of data are selected at a timeModel training is carried out by adopting gpml tool boxes, and a predicted average value/>, of model errors, is obtained according to training resultsAnd variance delta.
E. On the basis of a PID control model, the following trajectory tracking quadratic programming model is adopted, and a control barrier function h (q) = (x 1(q)+1)2+(x2(q)-2)2 -0.01 is added as a constraint solution safety control instruction u safe:
umin≤u≤umax
Where k 1=20,k2 =30, k=2, at which point the confidence is 95%. L f and L g represent robot edges And/>Li Daoshu,/>umin=[-50 -50 -50]T,umax=[50 50 50]T。
Through the steps, the control command u safe meeting the safety constraint is solved, so that the motion trail simulation result is shown in fig. 2, the end effector completely avoids the barrier and meets the safety requirement, and fig. 3 shows that u safe is in the section [ u min,umax ] and the problem of supersaturation of joint angles is not caused.
Claims (6)
1. The model uncertain robot safety control method based on the control obstacle function is characterized by comprising the following steps:
Firstly, establishing a robot dynamics estimation model;
Secondly, obtaining the robot joint speed meeting the safety constraint, and carrying out numerical integration on the robot joint speed to obtain the safe robot joint position;
Thirdly, tracking the safe robot joint position obtained in the second step to obtain the robot joint speed, the robot joint acceleration and the robot joint moment in the tracking process;
Fourthly, tracking the robot dynamics estimation model established in the first step to obtain model joint acceleration in the tracking process;
fifthly, calculating a difference value between the robot joint acceleration obtained in the third step and the model joint acceleration obtained in the fourth step to obtain model error data;
sixth, carrying out Gaussian process regression according to the robot joint speed, the robot joint acceleration and the robot joint moment obtained in the third step and the model error data obtained in the fifth step to obtain a robot dynamics accurate model;
seventh, according to the robot dynamics accurate model obtained in the sixth step, obtaining a control instruction meeting safety constraint, and carrying out safety control on the robot by the obtained control instruction meeting the safety constraint, so as to complete uncertain robot safety control based on the model of the control obstacle function;
In the second step, a quadratic programming model is adopted to obtain the robot joint speed meeting the safety constraint;
the quadratic programming model is as follows:
Wherein, Representing robot joint speed meeting safety constraints,/>Representing a current robot joint position; /(I)Representing the current robot joint speed,/>Representing the desired joint velocity,/>Representing robot tip position tracking error,/>Representing the current time,/>,/>Is a constant, is a set value,/>Representing a pseudo-inverse matrix; /(I)For the current robot end position,/>To expect robot tip position,/>For/>Is a derivative of (2); /(I)For/>Derivative of/>;/>Representing a robot jacobian matrix; /(I)Representing the minimum joint position of the robot,/>Representing the maximum joint position of the robot,/>Representing a control obstacle function,/>Representation/>Derivative of/(I)Represents a monotonically increasing function, and/>;
In the seventh step, a trajectory tracking quadratic programming model is adopted to obtain a control instruction meeting the safety constraint;
the track tracking quadratic programming model is as follows:
Wherein, ,/>, />And/>Representing robot edge/>AndLi Daoshu at this time/>Is a second order control barrier function,/>Representing a model error derivative; /(I)Is a control instruction meeting security constraints,/>,/>Representing the maximum and minimum values of the control input, respectively.
2. The model uncertainty robot safety control method based on the control obstacle function according to claim 1, characterized in that:
In the first step, the established robot dynamics estimation model is as follows:
Wherein, Representation/>The position of the joints,/>Is the inertial matrix of the robot,/>Representing the centrifugal and Coriolis force matrices,/>Representing gravity term,/>,/>,/>Respectively/>,,/>Estimate of/>Is a model error; u is the control input.
3. The model uncertainty robot safety control method based on the control obstacle function according to claim 2, characterized in that:
For a pair of Performing numerical integration to obtain safe robot joint position/>。
4. The model uncertainty robot safety control method based on the control obstacle function according to claim 3, wherein:
In the third step, a PID control model is adopted to track the safe robot joint position, and in the fourth step, a PID control model is adopted to track a robot dynamics estimation model.
5. The model uncertainty robot safety control method based on the control obstacle function according to claim 4, wherein:
The PID control model is input as follows
Wherein,,/>,/>To design control parameters, control time/>, is controlled in the tracking processDiscrete into/>Individual moments/>,/>At each moment, feeding back the current robot joint position/>Updating the control input to complete trajectory tracking, joint position/>, at each moment throughout the control processJoint velocity/>PID control inputInput value/>, as gaussian regression training dataModel error dataOutput value as training data/>And save the data set。
6. The model uncertainty robot safety control method based on the control obstacle function according to claim 5, wherein:
in the sixth step, the method for performing Gaussian process regression comprises the following steps:
Each time 100 groups of continuous data are selected Training is carried out, and model errors under the current robot joint position are predicted according to training results, wherein the prediction results are expressed as follows: /(I),/>Wherein the predicted mean value,/>, is representedRepresenting prediction variance,/>In connection with the selection of the confidence interval, based on the prediction result of the model error data,For saved dataset/>/>And a subset.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210989731.8A CN115256395B (en) | 2022-08-17 | 2022-08-17 | Model uncertain robot safety control method based on control obstacle function |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210989731.8A CN115256395B (en) | 2022-08-17 | 2022-08-17 | Model uncertain robot safety control method based on control obstacle function |
Publications (2)
Publication Number | Publication Date |
---|---|
CN115256395A CN115256395A (en) | 2022-11-01 |
CN115256395B true CN115256395B (en) | 2024-06-18 |
Family
ID=83752985
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210989731.8A Active CN115256395B (en) | 2022-08-17 | 2022-08-17 | Model uncertain robot safety control method based on control obstacle function |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN115256395B (en) |
Families Citing this family (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116880213B (en) * | 2023-08-16 | 2024-02-13 | 北京航空航天大学 | Unmanned aerial vehicle anti-interference safety control method and related products |
CN117260736B (en) * | 2023-11-02 | 2024-04-02 | 北京长木谷医疗科技股份有限公司 | Mechanical arm dynamics self-adaptive control method and device and electronic equipment |
CN117572810B (en) * | 2024-01-16 | 2024-05-10 | 大连海事大学 | Mechanical arm safety cooperative control system based on control obstacle function |
CN117687342B (en) * | 2024-01-31 | 2024-05-17 | 中国科学技术大学 | Robot safety control method based on non-conservative probability error bound of Gaussian process |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114571469A (en) * | 2022-05-05 | 2022-06-03 | 北京科技大学 | Zero-space real-time obstacle avoidance control method and system for mechanical arm |
Family Cites Families (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106647282B (en) * | 2017-01-19 | 2020-01-03 | 北京工业大学 | Six-degree-of-freedom robot trajectory planning method considering tail end motion error |
CN107121977B (en) * | 2017-06-02 | 2019-07-16 | 南京邮电大学 | Mechanical arm actuator failures fault-tolerant control system and its method based on double-layer structure |
CN113843793B (en) * | 2021-09-21 | 2023-07-21 | 兰州大学 | Mobile redundant mechanical arm model predictive control method with obstacle avoidance function |
CN113954077B (en) * | 2021-11-15 | 2023-03-24 | 天津大学 | Underwater swimming mechanical arm trajectory tracking control method and device with energy optimization function |
CN114721412B (en) * | 2022-03-16 | 2023-06-16 | 北京理工大学 | Unmanned aerial vehicle track tracking obstacle avoidance method based on model predictive control |
-
2022
- 2022-08-17 CN CN202210989731.8A patent/CN115256395B/en active Active
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114571469A (en) * | 2022-05-05 | 2022-06-03 | 北京科技大学 | Zero-space real-time obstacle avoidance control method and system for mechanical arm |
Non-Patent Citations (1)
Title |
---|
漂浮基空间机器人自适应RBF网络终端滑模控制;郭胜鹏;李东旭;孟云鹤;范才智;;动力学与控制学报;20141220(第04期);全文 * |
Also Published As
Publication number | Publication date |
---|---|
CN115256395A (en) | 2022-11-01 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN115256395B (en) | Model uncertain robot safety control method based on control obstacle function | |
CN111413966B (en) | Progressive model prediction unmanned planning tracking cooperative control method | |
Wang et al. | A hybrid visual servo controller for robust grasping by wheeled mobile robots | |
CN102825603B (en) | Network teleoperation robot system and time delay overcoming method | |
Mueller et al. | Iterative learning of feed-forward corrections for high-performance tracking | |
CN108614560B (en) | Tracking control method for visual servo performance guarantee of mobile robot | |
US11281208B2 (en) | Efficient teleoperation of mobile robots via online adaptation | |
CN114274147B (en) | Target tracking control method and device, mechanical arm control equipment and storage medium | |
Matsubara et al. | Sequential intention estimation of a mobility aid user for intelligent navigational assistance | |
CN104898659B (en) | A kind of man-machine coordination control method based on model prediction | |
Zein et al. | Enhanced teleoperation using autocomplete | |
Fröhlich et al. | Contextual tuning of model predictive control for autonomous racing | |
Nfaileh et al. | Formation control of multiple wheeled mobile robots based on model predictive control | |
Huang | Control approach for tracking a moving target by a wheeled mobile robot with limited velocities | |
Gaskett et al. | Reinforcement learning for visual servoing of a mobile robot | |
CN116604565A (en) | Force guiding control method and system for variable admittance of robot | |
Prasad et al. | Multi-agent Polygon Formation using Reinforcement Learning. | |
CN117021066A (en) | Robot vision servo motion control method based on deep reinforcement learning | |
Zhang et al. | Vision-based moving target interception with a mobile robot based on motion prediction and online planning | |
Liu et al. | The formation control of mobile autonomous multi-agent systems using deep reinforcement learning | |
CN114667852A (en) | Hedge trimming robot intelligent cooperative control method based on deep reinforcement learning | |
Xu et al. | Leader-follower formation control for multiple mobile robots with velocity constraints | |
Li et al. | Model predictive control for constrained robot manipulator visual servoing tuned by reinforcement learning | |
Zhang et al. | A review of intelligent control algorithms applied to robot motion control | |
CN114375431A (en) | Trajectory simplification method and device for industrial robot, computer storage medium and industrial robot operating platform |
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 |