CN110187705B - Robot remote control delay elimination control system and method - Google Patents
Robot remote control delay elimination control system and method Download PDFInfo
- Publication number
- CN110187705B CN110187705B CN201910443578.7A CN201910443578A CN110187705B CN 110187705 B CN110187705 B CN 110187705B CN 201910443578 A CN201910443578 A CN 201910443578A CN 110187705 B CN110187705 B CN 110187705B
- Authority
- CN
- China
- Prior art keywords
- control
- time
- data
- robot
- distance
- 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.)
- Expired - Fee Related
Links
- 238000000034 method Methods 0.000 title claims abstract description 21
- 230000008030 elimination Effects 0.000 title claims abstract description 19
- 238000003379 elimination reaction Methods 0.000 title claims abstract description 19
- 238000004891 communication Methods 0.000 claims description 35
- 230000000875 corresponding effect Effects 0.000 claims description 24
- 238000013528 artificial neural network Methods 0.000 claims description 18
- 230000006870 function Effects 0.000 claims description 11
- 238000010801 machine learning Methods 0.000 claims description 10
- 238000004458 analytical method Methods 0.000 claims description 9
- 238000012545 processing Methods 0.000 claims description 8
- 230000001276 controlling effect Effects 0.000 claims description 5
- 230000000694 effects Effects 0.000 abstract description 6
- 230000008859 change Effects 0.000 abstract description 4
- 239000011159 matrix material Substances 0.000 description 7
- 238000012360 testing method Methods 0.000 description 7
- 230000008569 process Effects 0.000 description 5
- 238000012549 training Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 230000005540 biological transmission Effects 0.000 description 3
- 230000004913 activation Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 206010063385 Intellectualisation Diseases 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- YVPYQUNUQOZFHG-UHFFFAOYSA-N amidotrizoic acid Chemical compound CC(=O)NC1=C(I)C(NC(C)=O)=C(I)C(C(O)=O)=C1I YVPYQUNUQOZFHG-UHFFFAOYSA-N 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 230000000007 visual effect 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
- G05D1/0212—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
- G05D1/0221—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory involving a learning process
Landscapes
- Engineering & Computer Science (AREA)
- Aviation & Aerospace Engineering (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Feedback Control In General (AREA)
- Manipulator (AREA)
Abstract
The invention provides a robot remote control time delay elimination control system, which is characterized in that a body control system is taken as a basis, time coordinators are added at a control end and an object end, an object characteristic predictor is designed at the control end, and under the condition of time delay, the output of the predictor is utilized to respectively guide the control end and the object end to generate control quality equal to zero time delay; on the other hand, the invention also provides a robot remote control time delay elimination control method aiming at the control system, the invention does not need to change the body control model of the original system, can effectively eliminate the time delay with any length, and improves the effect of the control engineering.
Description
Technical Field
The invention relates to the fields of electronic technology, automatic control and intellectualization, in particular to a system and a method for eliminating remote control delay of a robot.
Background
With the wide application of robots in important fields such as military, industry and the like, the robot gradually replaces many tasks which can be completed only by depending on manpower. In some application scenarios, the application effect of the robot is greatly influenced by remote control delay caused by a complex environment. Remote control delay often results in the system being unable to accurately track the input quantity of the system, and when the system generates external disturbance, the overshoot of the system is gradually increased, the stability of the system is affected, and even the equipment and personal safety are endangered in serious cases.
Disclosure of Invention
In order to solve the problems, the invention provides a robot remote control time delay elimination control system, wherein a time coordinator and an object characteristic predictor are added at a control end, a time coordinator is added at an object end, and the output of the predictor is utilized to respectively guide the control end and the object end, so that the control effect of zero time delay is generated.
The specific technical scheme is as follows:
a robot remote control time delay elimination control system comprises a control end and an object end, wherein the control end is a part capable of sending out a corresponding control instruction according to sensing information fed back by the object end in the control system; the control end comprises a first signal transceiver, a first time coordinator, an object characteristic predictor, a decider, a controller and a first processor; the object end comprises a second signal transceiver, a second time coordinator, a second processor, a controlled object and a sensor.
Preferably, the first signal transceiver is configured to send a control instruction of the control end to the object end; on the other hand, the sensor data is used for receiving feedback of the object side;
the first time coordinator is used for analyzing the data time and the receiving time of the sensor data fed back by the received object terminal by the control terminal so as to judge the communication time delay; on the other hand, the method is used for setting a time delay threshold value of the system;
the object characteristic predictor is used for predicting the sensor value of any future time point in real time through a machine learning model according to the received sensor data transmitted by the object end;
the decision device is used for deciding the generation basis of the control signal, and the generation basis of the control signal can be real-time sensing data of an object end or a prediction value of an object characteristic predictor;
the controller is used for generating a corresponding control signal according to the sensing data or the prediction value transmitted by the decider and transmitting the control signal to the first processor;
and the first processor is used for sending the control signal to the object end through the first signal transceiver after the control signal is coded.
Preferably, the second signal transceiver is used for receiving the control signal transmitted from the control end on one hand, and is used for sending the real-time data sensed by the sensor to the control end on the other hand;
the second time coordinator is used for accurately recording the time when the sensor senses the data and the data sending time on one hand; on the other hand for timing with the first time coordinator;
the second processor is used for decoding the control instruction and controlling the controlled object to execute corresponding action on the one hand; on the other hand, the device is used for transmitting the data sensed by the sensor to the second signal transceiver;
the controlled object is the final control object of the control system;
the sensor is used for sensing the field sensing data of the robot.
A robot remote control time delay elimination method comprises the following processing steps:
s10, the first time coordinator of the control end and the second time coordinator of the object end are calibrated mutually;
s20, the object end feeds back the sensing data and the data time information of the sensor to the control end;
s30, the control end analyzes the received data packet;
s40 the first time coordinator transmits the analysis data to the object property predictor and the decider respectively;
s50, the object characteristic predictor uses the transmitted analysis data to establish a machine learning model for predicting the distance between the sensing data and the obstacle;
the S60 decider receives the analytic data transmitted by the first time coordinator, calculates the communication delay and decides the distance between the obstacle and the robot, and transmits the distance to the controller to generate a corresponding control instruction;
and S70, the object end receives the control instruction sent by the control end and controls the controlled object to execute corresponding operation.
Preferably, the specific processing steps of step S60 are:
s601 calculating communication delay tl＝tr-ts；
S602, according to the system time delay threshold tau0And communication delay tlDeciding the distance between the obstacle and the robot;
① if tl＞τ0Then, the predicted distance of the neural network in the object characteristic predictor is used as the distance between the obstacle and the robot;
② if tl≤τ0Then, consider the current tdThe communication environment of the moment meets the real-time communication requirement, and the decider sends tdDistance d between robot and obstacle at any momentdAs a decision value for the distance between the robot and the obstacle;
③ the deciding device transmits the deciding value of the distance between the robot and the obstacle to the controller, the controller automatically feeds back to the operator through the force feedback function according to the deciding value of the distance between the robot and the obstacle, the operator sends out a corresponding control instruction to the first signal transceiver through the first processor, and the first signal transceiver transmits the control instruction to the object terminal.
Has the advantages that:
the invention provides a robot remote control time delay elimination control system, which is based on a body control system, wherein time coordinators are added at a control end and an object end, an object characteristic predictor is designed at the control end, and under the condition of time delay, the output of the predictor is utilized to respectively guide the control end and the object end to generate control quality with equal zero time delay. On the other hand, the invention also provides a robot remote control time delay elimination control method aiming at the control system, the invention does not need to change the body control model of the original system, can effectively eliminate the time delay with any length, and improves the effect of the control engineering.
Drawings
FIG. 1 is a block diagram of a robot remote control delay elimination control system according to the present invention;
FIG. 2 is a block diagram of a system for controlling a mobile robot to avoid obstacles according to an embodiment of the present invention;
FIG. 3 is a flow chart of a robot remote control delay elimination control method according to the present invention;
FIG. 4 is a flow diagram of an object property predictor prediction and obstacle distance machine learning;
fig. 5 is a diagram of an RBF neural network structure constructed according to an embodiment of the present invention, wherein a symbol indicates the multiplication of two matrix elements by position;
fig. 6 is a flowchart of a decision process performed by the decision device to decide the distance between the robot and the obstacle.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and examples. The specific embodiments described herein are merely illustrative of the invention and are not to be construed as limiting the invention.
Referring to fig. 1, the robot remote control time delay elimination control system according to the present invention is composed of a control end 10 and an object end 20, wherein the control end 10 is a part of the control system that can send out a corresponding control command according to sensing information fed back by the object end. The control terminal 10 includes a first signal transceiver 11, a first time coordinator 12, an object characteristic predictor 13, a determiner 14, a controller 15, and a first processor 16. The object side 20 comprises a second signal transceiver 21, a second time coordinator 22, a second processor 23, a controlled object 24 and a sensor 25.
The first signal transceiver 11 is configured to send a control instruction of the control end 10 to the target end 20; and on the other hand, for receiving sensor data fed back from the subject end 20.
The first time coordinator 12 is configured to, on one hand, analyze the data time and the receiving time of the sensor data fed back by the control end 10 from the object end 20 to determine the communication delay; and on the other hand, the method is used for setting a time delay threshold value and time correction of the system. The data time of the sensor data refers to the real-time of the data sensed by the sensor 25 of the target 20. The first time coordinator 12 is provided with a timing chip, the timing chip can be selected according to the time delay requirement of the application situation as long as the timing error is lower than the time delay threshold of the application scene of the control system by one order of magnitude (base 10) or more, the time delay threshold refers to that the control system is not added with the time delay elimination control system, the control end 10 and the object end 20 are directly controlled, and under the condition that the control effect just meets the control requirement, the remote control time delay between the control end 10 and the object end 20 is the time delay threshold tau0The time delay threshold value can be set by an operator through the first time coordinator according to the requirements of the use scene. As a case of this embodiment, the timing chip selects the timing precision to be 50ms (millisecond), and the system can satisfy the delay threshold τ0The application range is more than or equal to 500 ms.
It should be noted that, when the system is initialized and there is no additional delay caused by the communication link between the control end and the object end except that the system itself cannot eliminate the delay, the timing chips in the first time coordinator 12 and the second time coordinator 22 need to perform a mutual timing first, so that the starting points of the timing of the first time coordinator 12 and the timing of the second time coordinator 22 are synchronized.
The object characteristic predictor 13 is used for predicting the sensor value of any future time point in real time through a machine learning model according to the received sensor data transmitted from the object end 20. As one aspect of this embodiment, when the remote-controlled robot performs obstacle avoidance, an ultrasonic sensor is used to measure the distance between the sensor and the obstacle, and the object characteristic predictor predicts the value of the sensor by a fitting algorithm of machine learning.
The decision device 14 is used for deciding the generation basis of the control signal, which may be the real-time sensing data of the object terminal 20 or the prediction value of the object characteristic predictor 13. Specifically, the determiner 14 first determines the real-time communication delay determined by the first time coordinator 12 and the delay threshold τ of the system0Comparing, if the real-time communication time delay is less than the time delay threshold tau0The determiner 14 will analyze the sensing data transmitted from the object terminal 20 and transmit the sensing data to the controller 15; if the real-time communication delay exceeds the delay threshold tau0The determiner 14 calls the prediction value in the object property predictor 13 and transmits the prediction value to the controller 15 as a basis for generation of the control signal.
The controller 15 is configured to generate a corresponding control signal according to the sensing data or the predicted value transmitted from the determiner 14, and transmit the control signal to the first processor 16.
The first processor 16 is configured to send the control signal to the target 20 through the first signal transceiver 11 after being subjected to encoding processing, and the specific steps of the encoding processing may refer to the prior art, which is not limited in the present invention.
The second signal transceiver 21 is used for receiving the control signal transmitted from the control terminal 10, and for sending the real-time data sensed by the sensor 25 to the control terminal 10.
The second time coordinator 22 is used for accurately recording the time when the sensor 25 senses data and the data transmission time. And on the other hand for timing with the first time coordinator 12.
The second processor 23, on one hand, is used for decoding the control instruction and controlling the controlled object 24 to execute a corresponding action; on the other hand, for transmitting the data sensed by the sensor 25 to the second signal transceiver 21.
The controlled object 24 is the final controlled object of the control system according to the present invention, and as a case of the embodiment of the present invention, the controlled object 24 may be a control motor.
The sensor 25 is used for sensing the field sensing data of the robot. As a case of the embodiment of the present invention, the sensor 25 may be composed of an ultrasonic sensor that measures a distance between the robot and the obstacle and a sensor that measures a moving speed and an angular speed of the robot.
In this embodiment, before a formal control task is performed, a delay threshold of the system is set by the first time coordinator 12, and in a state where the control terminal 10 and the object terminal 20 have good communication, where the good communication state refers to a state where there is no additional delay caused by a communication link except that the system itself cannot eliminate delay between the control terminal and the object terminal, the first time coordinator 12 sends a timing signal to the first signal transceiver 11 and then to the second signal transceiver 21, after receiving the timing signal, the second signal transceiver 21 sends the timing signal to the second time coordinator 22, and the second time coordinator 22 receives the timing signal to implement timing synchronization with the first time coordinator 12, that is, timing.
After the timing work is finished, the system starts to execute formal control tasks, the sensor 25 senses field sensing data of the object end 20 and sends the field sensing data to the second processor 23 in real time, the second processor 23 reads data time of the sensing data from the second time coordinator 22 while receiving the sensing data and sends the sensing data and corresponding data time to the second signal transceiver 21, the second signal transceiver immediately acquires sending time of the data from the second time coordinator 22 after receiving the sensing data and corresponding data time data, and then the second signal transceiver sends the sensing data, corresponding data time and sending time information to the first signal transceiver 11; after receiving the data, the first signal transceiver 11 transmits a data packet to the first time coordinator 12, and the first time coordinator 12 parses the data to obtain the sensing data, the data time of the sensing data, and the data transmission time, and then obtains the data receiving time by using the record of the first time coordinator 12 to complete the data parsing.
The first time coordinator 12 transmits the parsed data to the object property predictor 13 and the determiner 14, respectively.
The object characteristic predictor 13 uses the received analytic data to establish a controlled object characteristic prediction machine learning model of the sensing data, and after the machine learning model is established, the object characteristic predictor 13 can predict the future sensing data.
The determiner 14 receives the analysis data transmitted from the first time coordinator 12, calculates a communication delay, determines sensing data meeting a control requirement under the current communication delay according to a relationship between the communication delay and a delay threshold set by the system, and transmits the sensing data to the controller 15.
The controller 15 receives the sensing data transmitted by the determiner 14, generates a control command, encodes the control command by the first processor 16, and transmits the encoded control command to the first signal transceiver 11, the first signal transceiver 11 transmits the control command to the second signal transceiver 21, and the control command is decoded by the second processor 23 to control the controlled object 24 to execute a corresponding control action.
On the other hand, the present invention also provides a robot remote control delay elimination control method based on the robot remote control delay elimination control system, as a preferred embodiment of the present invention, a detailed description will be given below by taking the method of the present invention as an example for controlling a mobile robot to avoid an obstacle in real time under the condition of having a delay, referring to fig. 2, an example of the controller 15 of the mobile robot control system of the present embodiment is a phantomm haptic device controller of the american sensor Technologies corporation, the controller may feed back distance values fed back by the sensor 25 in the object end 20 to the operator via a force feedback function on the controller, and can feed back forces with different magnitudes which can be obviously distinguished and sensed to an operator according to the distance between the manipulator and the obstacle, therefore, even if the operator does not have a visual condition, the operator can remotely control the robot to avoid the obstacle by means of force feedback.
Referring to fig. 3, the method of the present invention mainly comprises the following steps:
s10 times the first time coordinator 12 of the control end 10 and the second time coordinator 22 of the object end 20 with each other.
In this embodiment, the specific operations of the first time coordinator 12 and the second time coordinator 22 in calibrating time with each other are as follows: before the system works formally, under the condition that a good communication environment exists between the control end 10 and the object end 20, the first time coordinator 12 sends a timing signal to the first signal transceiver 11 and then to the second signal transceiver 21, after receiving the timing signal, the second signal transceiver 21 sends the timing signal to the second time coordinator 22, and the second time coordinator 22 receives the timing signal to realize timing synchronization with the first time coordinator 12, namely timing. A mutual timing of said first time coordinator 12 and said second time coordinator 22 is achieved.
S20 the target 20 feeds back the sensing data and data time information of the sensor 25 to the control terminal 10.
Specifically, when the sensor 25 senses environmental data (hereinafter referred to as "sensing data"), the second time coordinator 22 forms a data packet by one-to-one correspondence between the timing time and the sensing data, and sends the data packet to the control end 10 through the second signal transceiver 21.
S30 the control terminal 10 parses the received packet.
Specifically, after receiving the data packet, the first signal transceiver 11 of the control end 10 transmits the data packet to the first time coordinator 12, and the first time coordinator 12 parses the data packet to obtain the sensing data DiSensing data DiAt data time tdData transmission time t, which is the time when the packet is transmitted from the destination 20sThen, the record of the first time coordinator 12 is used to obtain the time t at which the data packet is received by the control end 10rAnd completing the analysis of the data packet.
In step S10, after the first time coordinator 12 and the second time coordinator 22 complete timing, the timers of both are cleared, and the time is taken as the time origin of the system, and the time thereafter is taken as the relative time with reference to the time origin, so in the description of the present invention, the times are all relative times unless otherwise specified.
S40 the first time coordinator 12 transmits the parsed data to the object property predictor 13 and the determiner 14, respectively.
The S50 object property predictor 13 uses the transmitted analytic data to build a machine learning model for predicting the distance between the sensing data and the obstacle.
In this embodiment, an RBF (radial basis function) neural network is used to predict the distance between the target 20 and the obstacle at a future time according to the speed, the angular velocity, the distance from the obstacle, and the relative time of the target 20, and with reference to fig. 4, the following steps are specifically adopted:
s501, the object characteristic predictor 13 extracts sample data participating in neural network calculation by using the transmitted analysis data and stores the sample data into a data set of the sample data;
in step S50, the object property predictor 13 receives the analysis data as follows: di＝{vi,wi,diI is a natural number greater than 0, and represents a group number of the data; v. ofiRepresenting the speed of movement, w, of a moving robotiRepresenting the rotational speed of the moving robot, diIndicating the distance of the moving robot from the obstacle. Data DiCorresponding to a data acquisition time tdi. The matrix form of the group of data extracted sample features and stored in the sample data matrix P is [ v [ ]iwitdi]In this way, assuming that the object feature predictor 13 sets data received successively at a time as a data set G, n (n is natural) is randomly selected from the data set GNumber) of data sets, n) to ensure that sample data can provide a sufficient amount of sample data for training of the neural network>100. Then the sample data training set P and the result set T are respectively:
s502, constructing an RBF neural network according to the sample data structure, wherein the network structure refers to the attached figure 5.
In FIG. 5, IW1，1Representing the connection weight matrix of the input layer and the hidden layer, in the RBF neural network of the invention, IW1，1For constant value, IW1，1＝(P′)3×n；
I dist I represents the input training set Pn×3And IW1，1The Euclidean distance, | dist | | - | | | IW1，1-Pn×3||；
b1Indicating that the input layer is connected to the hidden layer by a threshold,
where 0.6 represents the control parameter of the center-to-sample relationship of the radial basis function, b1Is n.
Activation function (i.e. radial basis function) selection
LW2，1Representing the connection weight between the hidden layer and the output layer, LW in this embodiment2，1＝Tn×1；
b2Representing a threshold between the hidden layer and the output layer, the specific value of which will be generated in the calculation process;
liner denotes a linear solving function, and in the present embodiment, the function f (x) ═ x is used;
and S503, training a neural network.
① calculating the output of the activation function
② calculating a threshold b between the hidden layer and the output layer2：
Setting an intermediate matrix variable
The denominator has n columns 1 and a threshold value b2Take the (n +1) th column of matrix X.
③ calculating network output
S504, the training effect of the neural network is tested by using the test set and the result set data.
Randomly selecting N (N) from the data set G of the object property predictor 13>10, natural number) data as test set, denoted PtestThe result set is denoted as Ttest。
① test set input PtestInputting the result into the RBF neural network in the step S503, and recording the result as Tsim
② calculating TsimAnd corresponding result set TtestThe formula is as follows:
③ calculating the decision coefficient R of RBF neural network2The formula is as follows:
in the formula, sum (A) is shown in the specification, wherein A is a column vector and represents the sum of all elements of A; ". denotes multiplying two vector co-sited elements by location; ". 2" denotes squaring each element of the vector;
s505 if the relative error is error and the decision coefficient R2If all meet the actual requirement, the RBF neural network is trained completely, otherwise, the circulation is performedAnd executing steps S501 to S504 until the relative error and the decision coefficient meet the actual requirement.
Specifically, in actual use, a relative error level that can satisfy actual requirements is set to δ, which is 10 in this embodiment-2If error r<δ, the relative error is considered to satisfy the actual requirement. Similarly, a higher decision factor level e is set, in this example 0.95, if R2>0.95, the decision coefficient of the RBF neural network is considered to meet the actual requirement.
The S60 decision unit 14 receives the analytic data from the first time coordinator 12, calculates the communication delay and decides the distance between the obstacle and the robot, and transmits the result to the controller 15 to generate a corresponding control command.
Referring to fig. 6, the specific processing steps are as follows:
s601 calculating communication delay tl＝tr-rs；
S602, according to the system time delay threshold tau0And communication delay tlDeciding the distance between the obstacle and the robot;
① if tl＞τ0The predicted distance of the neural network in the object characteristic predictor 13 is used as the distance of the obstacle from the robot.
The process is as follows: the control terminal 10 is set at t by the decision means 14rAt the moment t is receiveddThe time-of-day sensing data is converted into the form of [ v ] of neural network input features in the object characteristic predictor 13dwdtd]. In order to eliminate remote control delay, the neural network predicts the time tk＝(tr+tl) Is the distance of the obstacle from the robot. The input feature estimate is
Wherein
Represents tkAn estimate of the robot's velocity at the moment,
represents tkThe estimated value of the angular velocity of the robot at the moment is compared with the communication time delay, the change of the velocity and the angular velocity of the robot at two adjacent communication moments can be ignored, so that the change of the velocity and the angular velocity of the robot at the two adjacent communication moments is enabled to be ignored
Then there is Pk＝[vdwdtk]A 1 is to PkThe RBF neural network input to the object feature predictor 13 will obtain tkPredicted value d of distance between obstacle and robot at momentkI.e. tl≤τ0A decision value of the distance between the robot and the obstacle under the condition.
② if tl≤τ0Then, consider the current tdThe communication environment at the moment meets the real-time communication requirement, and the decider 14 sends tdDistance d between robot and obstacle at any momentdAs a decision value for the distance between the robot and the obstacle.
③, the controller 15 automatically feeds back the determined value of the distance between the robot and the obstacle to the operator through the force feedback function according to the determined value of the distance between the robot and the obstacle, the operator sends a corresponding control command to the first signal transceiver 11 through the first processor 16, and the first signal transceiver 11 sends the control command to the object terminal 20.
S70 the object side 20 receives the control command from the control side 10 and controls the controlled object 24 to perform corresponding operations.
The second signal transceiver 21 of the object side 20 receives the control instruction sent by the control side 10, decodes the control instruction by the second processor 23, and then sends the decoded control instruction to the controlled object 24 to execute corresponding operations.
In summary, the system and the method for eliminating robot remote control delay according to the present invention can be realized. When there is a communication delay tlThe control end will be based on tlThe situation after the time window gives control instructions to ensure that the quality of the control process is not delayedThe same is true. When the communication delay is at the delay threshold tau0Within the range, the control is carried out according to a real-time control mode, and the control efficiency is improved.
It is to be understood that the invention is not limited to the specific embodiments described above, but is intended to cover various insubstantial modifications of the inventive process concepts and solutions, or its application to other applications without modification.
Claims (5)
1. A robot remote control time delay elimination control system is characterized by comprising a control end (10) and an object end (20), wherein the control end (10) is a part capable of sending a corresponding control instruction according to sensing information fed back by the object end (20) in the control system; the control end (10) comprises a first signal transceiver (11), a first time coordinator (12), an object characteristic predictor (13), a decision maker (14), a controller (15) and a first processor (16); the object end (20) comprises a second signal transceiver (21), a second time coordinator (22), a second processor (23), a controlled object (24) and a sensor (25);
the first signal transceiver (11) is used for sending a control command of the control end (10) to the object end (20); on the other hand, the sensor data is used for receiving feedback of the object end (20);
the first time coordinator (12) is used for analyzing the data time and the receiving time of the sensor data fed back by the received object end (20) by the control end (10) to judge the communication time delay on one hand; on the other hand, the method is used for setting a time delay threshold value of the system;
the object characteristic predictor (13) is used for predicting the sensor value of any future time point in real time through a machine learning model according to the received sensor data transmitted from the object end (20);
the decision device (14) is used for deciding the generation basis of the control signal;
the controller (15) is used for generating a corresponding control signal according to the sensing data or the prediction value transmitted by the determiner (14) and transmitting the control signal to the first processor (16);
the first processor (16) is used for sending the control signal to the object end (20) through the first signal transceiver (11) after the control signal is subjected to coding processing;
the second signal transceiver (21) is used for receiving a control signal transmitted by the control end (10) on one hand and sending real-time data sensed by the sensor (25) to the control end (10) on the other hand;
the second time coordinator (22) is used for accurately recording the time when the sensor (25) senses the data and the data sending time; on the other hand for timing with the first time coordinator (12);
the second processor (23) is used for decoding the control instruction and controlling the controlled object (24) to execute corresponding action; on the other hand, the device is used for transmitting the data sensed by the sensor (25) to the second signal transceiver (21).
2. The system according to claim 1, wherein the control signal is generated based on real-time sensing data of the object side (20) or a prediction value of the object characteristic predictor (13).
3. The robot remote control delay elimination control system according to claim 1, wherein the controlled object (24) is a final control object of the control system of the present invention; and the sensor (25) is used for sensing the field sensing data of the robot.
4. A robot remote control time delay elimination method is characterized in that the processing steps comprise:
s10, the first time coordinator (12) of the control end (10) and the second time coordinator (22) of the object end (20) are calibrated mutually;
s20, the object side (20) feeds back the sensing data and the data time information of the sensor (25) to the control side (10);
s30, the control end (10) analyzes the received data packet;
s40, the first time coordinator (12) transmits the analysis data to the object characteristic predictor (13) and the decider (14) respectively;
s50 the object characteristic predictor (13) uses the transmitted analysis data to build a machine learning model for predicting the distance between the sensing data and the obstacle;
the S60 decision-making device (14) receives the analysis data transmitted by the first time coordinator (12), calculates the communication delay and decides the distance between the obstacle and the robot, and transmits the distance to the controller (15) to generate a corresponding control instruction;
s70 the object end (20) receives the control instruction from the control end (10) and controls the controlled object (24) to execute the corresponding operation.
5. The robot remote control delay elimination method of claim 4, wherein the specific processing steps of step S60 are:
s601 calculating communication delay
=
；
S602 according to the system delay threshold
And communication delay
Deciding the distance between the obstacle and the robot;
① if
Then, the predicted distance of the neural network in the object characteristic predictor (13) is used as the distance between the obstacle and the robot;
② if
Then it is considered as current
The communication environment of the moment meets the real-time communication requirement, and the decider (14) is used for judging whether the real-time communication requirement is met
Distance between robot and obstacle at any moment
As a decision value for the distance between the robot and the obstacle;
③, the deciding unit (14) transmits the deciding value of the distance between the robot and the obstacle to the controller (15), the controller (15) automatically feeds back to the operator through the force feedback function according to the deciding value of the distance between the robot and the obstacle, the operator sends out a corresponding control instruction to the first signal transceiver (11) through the first processor (16), and the first signal transceiver (11) transmits the control instruction to the object terminal (20).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910443578.7A CN110187705B (en) | 2019-05-27 | 2019-05-27 | Robot remote control delay elimination control system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910443578.7A CN110187705B (en) | 2019-05-27 | 2019-05-27 | Robot remote control delay elimination control system and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110187705A CN110187705A (en) | 2019-08-30 |
| CN110187705B true CN110187705B (en) | 2020-03-31 |
Family
ID=67717912
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910443578.7A Expired - Fee Related CN110187705B (en) | 2019-05-27 | 2019-05-27 | Robot remote control delay elimination control system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110187705B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105856235A (en) * | 2016-06-08 | 2016-08-17 | 江苏若博机器人科技有限公司 | Wireless transmission two-core six-axis crawler type natural gas pipeline robot control system |
| CN106534271A (en) * | 2016-10-21 | 2017-03-22 | 中国科学院力学研究所 | Method and device for remotely manipulating multiple operating objects by multiple operation ends |
| CN107945499A (en) * | 2017-11-30 | 2018-04-20 | 中国科学院力学研究所 | The assessment method and equipment of a kind of remote control system |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7668626B2 (en) * | 2006-08-16 | 2010-02-23 | National Institute Of Information And Communications Technology, Incorporated Administrative Agency | Image displaying method with fluctuation correction and remote control system for moving unit using the method |
| CN101483567B (en) * | 2008-12-19 | 2011-04-13 | 西北电网有限公司 | High voltage electricity transmission line monitoring method based on wireless communication and optical communication |
-
2019
- 2019-05-27 CN CN201910443578.7A patent/CN110187705B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105856235A (en) * | 2016-06-08 | 2016-08-17 | 江苏若博机器人科技有限公司 | Wireless transmission two-core six-axis crawler type natural gas pipeline robot control system |
| CN106534271A (en) * | 2016-10-21 | 2017-03-22 | 中国科学院力学研究所 | Method and device for remotely manipulating multiple operating objects by multiple operation ends |
| CN107945499A (en) * | 2017-11-30 | 2018-04-20 | 中国科学院力学研究所 | The assessment method and equipment of a kind of remote control system |
Non-Patent Citations (4)
| Title |
|---|
| A Framework for the Control of Time-Delayed Telerobotic Systems;Tzyh-Jong Tarn，Kevin Brady;《IFAC Robot Control》;19970930;第30卷;第599-604页 * |
| 人机交互力觉临场感遥操作机器人技术研究;宋爱国;《科技导报》;20151231;第33卷(第23期);第100-109页 * |
| 无时延感的遥操作实时共享控制仿真;李华忠，等;《计算机工程与应用》;20091231;第45卷(第3期);第11-15页 * |
| 激光传感器的机器人运动控制研究;吴艳玲，唐穗欣;《激光杂志》;20171231;第38卷(第1期);第127-130页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110187705A (en) | 2019-08-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Saeidi et al. | A mixed-initiative haptic teleoperation strategy for mobile robotic systems based on bidirectional computational trust analysis | |
| CN108890652B (en) | Transformer substation inspection robot and transformer substation equipment inspection method | |
| CN111444954A (en) | Robot autonomous assembly method based on multi-mode perception and learning | |
| Saeidi et al. | Trust-based mixed-initiative teleoperation of mobile robots | |
| CN109429235B (en) | Beam scanning method, network side equipment and mobile communication terminal | |
| CN110519687B (en) | Accurate positioning method and system suitable for robot and robot | |
| JP7494382B2 (en) | Method and device for identifying control instructions, non-volatile storage medium, processor, electronic device, and articulated robot | |
| CN109940619A (en) | Trajectory planning method, electronic device and storage medium | |
| US20190036594A1 (en) | Radio repeater selection apparatus and machine learning device | |
| JP4572497B2 (en) | Robot controller | |
| CN110187705B (en) | Robot remote control delay elimination control system and method | |
| Maddahi et al. | Position referenced force augmentation in teleoperated hydraulic manipulators operating under delayed and lossy networks: A pilot study | |
| CN112091974B (en) | Remote control operation system of distribution network live working robot and control method thereof | |
| CN110788853B (en) | Motion trajectory generation method, electronic device and computer-readable storage medium | |
| CN1330466C (en) | On-line robot hand and eye calibrating method based on motion selection | |
| KR100960296B1 (en) | Zigbee location recognition method | |
| Liaw et al. | Target prediction to improve human errors in robot teleoperation system | |
| Martínez-Rey et al. | Smart laser scanner for event-based state estimation applied to indoor positioning | |
| KR102447386B1 (en) | Method of executing robot and enforcing safety of robot and system therefor | |
| CN112462765B (en) | Robot, control method and device thereof, and computer-readable storage medium | |
| CN116880341B (en) | High-precision motion control system based on industrial Ethernet bus | |
| KR101402090B1 (en) | A tele-operated robotic path compensation apparatus and method based on detecting user intentions | |
| CN117817666A (en) | Industrial robot intelligence centre gripping control system based on artificial intelligence | |
| WO2024018900A1 (en) | Device control equipment, control system, and control method | |
| JP7282016B2 (en) | robot system |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20200331 Termination date: 20210527 |