CN108255173A - Robot follows barrier-avoiding method and device - Google Patents

Robot follows barrier-avoiding method and device Download PDF

Info

Publication number
CN108255173A
CN108255173A CN201711384862.9A CN201711384862A CN108255173A CN 108255173 A CN108255173 A CN 108255173A CN 201711384862 A CN201711384862 A CN 201711384862A CN 108255173 A CN108255173 A CN 108255173A
Authority
CN
China
Prior art keywords
robot
relationship
speed
barrier
follows
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.)
Pending
Application number
CN201711384862.9A
Other languages
Chinese (zh)
Inventor
张伟民
李明珠
姚卓
梁震烁
杨天奇
黄强
张华�
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Institute of Technology BIT
Original Assignee
Beijing Institute of Technology BIT
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Beijing Institute of Technology BIT filed Critical Beijing Institute of Technology BIT
Priority to CN201711384862.9A priority Critical patent/CN108255173A/en
Publication of CN108255173A publication Critical patent/CN108255173A/en
Pending legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05DSYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00Control of position, course or altitude of land, water, air, or space vehicles, e.g. automatic pilot
    • G05D1/02Control of position or course in two dimensions
    • G05D1/021Control of position or course in two dimensions specially adapted to land vehicles

Abstract

This application discloses a kind of robots to follow barrier-avoiding method and device.This method includes:S1. it obtains and follows three-dimensional coordinate of the barrier based on robotic vision sensor coordinate system in object and environment;S2. skeleton data of the object based on the robotic vision sensor coordinate system is followed described in obtaining;S3. robot and the second position relationship followed between the first position relationship of object and the barrier and robot are determined;S4. the benchmark linear velocity and reference angle speed of the robot traveling are determined according to the first position relationship;S5. a skimulated motion track is determined according to the benchmark linear velocity and reference angle speed;S6. the linear velocity and angular speed of the robot are adjusted according to the second position relationship and skimulated motion track;S7. the step S1 to step S6 is periodically performed successively, persistently carry out object being followed to follow to described.Can effective district partial objectives for and barrier, and realize the effect for hiding obstacle during target is followed.

Description

Robot follows barrier-avoiding method and device
Technical field
This application involves robotic technology fields, and barrier-avoiding method and device are followed in particular to a kind of robot.
Background technology
Mobile robot avoiding obstacles autonomous during being moved to target simultaneously find route to object run It is the basic capacity that intelligent mobile robot needs to have, mobile robot is especially being used for the complex environments such as home services When middle, this ability is just particularly important.
Avoidance technology during following at present is mainly the following:
Using Artificial Potential Field Method, the barrier in set environment generates repulsion field, and object generates gravitational field, repulsion field with Composite force field guiding mobile robot avoiding obstacles in advancing to target of gravitational field.This method is widely used in road In diameter project study, but since all information are changed into single resultant force, and control with the resultant direction due to Artificial Potential Field Method The movement of robot next step so having discarded other valuable information such as distribution of obstacles, makes it in complex environment Path planning scarce capacity in addition be unable to reach target point.
Using the sensors such as ultrasonic wave, infrared, avoidance is carried out by measuring the distance between direction of advance and barrier.This Kind method has preferable effect, but due to sensor existing defects itself in terms of avoidance:For ultrasonic wave, the speed of sound is held Easily interfered by temperature and wind direction, sound can by sound-absorbing face absorb etc.;For infrared, minimum detection distance is too big.These all can Influence judgement of the robot to target and barrier.The problem of most fatal is indistinguishable target and obstacle in this way Object can not be realized and hide obstacle during target is followed.
Invention content
The main purpose of the application is that providing a kind of robot follows barrier-avoiding method and device, to solve the problems, such as.
To achieve these goals, it according to the one side of the application, provides a kind of robot and follows barrier-avoiding method.
Barrier-avoiding method is followed to include according to the robot of the application:
S1. it obtains and follows three-dimensional coordinate of the barrier based on robotic vision sensor coordinate system in object and environment;
S2. skeleton data of the object based on the robotic vision sensor coordinate system, the bone are followed described in obtaining Bone data are depth information of each bone node under the visual sensor coordinate system;
S3. robot and the first position relationship for following object are determined by the skeleton data for following object, The second position relationship between the barrier and robot is determined by the three-dimensional coordinate of the barrier;
S4. the benchmark linear velocity and reference angle speed of the robot traveling are determined according to the first position relationship;Its In, the benchmark linear velocity is the robot to the speed that object is followed to move linearly, the reference angle speed The speed of angle rotation is carried out according to the first position relationship for the robot;
S5. a skimulated motion track is determined according to the benchmark linear velocity and reference angle speed;
S6. the linear velocity and angular speed of the robot are carried out according to the second position relationship and skimulated motion track Adjustment;
S7. the step S1 to step S6 is periodically performed successively, persistently carry out object being followed to follow to described.
Further, robot as the aforementioned follows barrier-avoiding method, described true by the skeleton data for following object Determine robot and the first position relationship for following object, including:
The mean depth letter of the bone node of object is followed according to obtaining the depth information of each bone node Breath;
The robot and the first position relationship for following object are determined according to the mean depth information;Described One position relationship includes:The robot follows the distance between object and angle with described.
Further, robot as the aforementioned follows barrier-avoiding method, described according to determining the mean depth information Robot and the first position relationship for following object, including:
According to the three-dimensional coordinate M (x under the corresponding visual sensor coordinate system of the mean depth information acquisition1,y1,z1);
Further, robot as the aforementioned follows barrier-avoiding method, described according to duration, the N according to the period A velocity bias, the plan linear velocity and plan angular speed, obtain N movement locus;Including:
On the basis of the plan linear velocity v and plan angular speed w, 2n different adjustment is carried out, 2n+1 is obtained A velocity group:(v, w), (v+ δ, w+ δ), (v+2 δ, w+2 δ) ... .. (v+n δ, w+n δ), (v- δ, w- δ), (v-2 δ, w-2 δ) ... .. (v-n δ, w-n δ), wherein N=2n+1, δ are a steady state value;
2n+1 movement locus is obtained according to the 2n+1 velocity group and the duration in the period.
To achieve these goals, it according to the another aspect of the application, provides a kind of robot and follows obstacle avoidance apparatus.
Obstacle avoidance apparatus is followed to include according to the robot of the application:
Three-dimensional coordinate determination unit follows barrier in object and environment to be based on robotic vision sensor for obtaining The three-dimensional coordinate of coordinate system;
Skeleton data acquiring unit described follows object to be based on the robotic vision sensor coordinate system for obtaining Skeleton data, the skeleton data be depth information of each bone node under the visual sensor coordinate system;
Position relationship determination unit determines that robot follows pair with described for passing through the skeleton data for following object The first position relationship of elephant determines the second position between the barrier and robot by the three-dimensional coordinate of the barrier Relationship;
Reference speed determination unit, for determining the benchmark linear speed of the robot traveling according to the first position relationship Degree and reference angle speed;Wherein, the benchmark linear velocity is the robot to the speed that object is followed to move linearly Degree, the reference angle speed carry out the speed of angle rotation for the robot according to the first position relationship;
Skimulated motion track determination unit, for determining a simulation fortune according to the benchmark linear velocity and reference angle speed Dynamic rail mark;
Speed adjustment unit, for the linear speed according to the second position relationship and skimulated motion track to the robot Degree and angular speed are adjusted;
Cycling element, for periodically performing the three-dimensional coordinate determination unit, skeleton data acquiring unit, position successively Relation determination unit, reference speed determination unit, skimulated motion track determination unit and speed adjustment unit are persistently carried out to institute It states and follows following for object.
Further, robot as the aforementioned follows obstacle avoidance apparatus, the position relationship determination unit, including:
Mean depth information acquisition module, for following pair described in being obtained according to the depth information of each bone node The mean depth information of the bone node of elephant;
First position relationship determination module, for determining that the robot is followed with described according to the mean depth information The first position relationship of object;The first position relationship includes:The robot with it is described follow the distance between object with And position relation.
Further, robot as the aforementioned follows obstacle avoidance apparatus, the first position relationship determination module, including:
Three-dimensional coordinate obtains module, according to three under the corresponding visual sensor coordinate system of the mean depth information acquisition Dimension coordinate M (x1,y1,z1);
Positional information calculation module, for according to the three-dimensional coordinate M (x1,y1,z1) determine the distanceAnd angle
Further, robot as the aforementioned follows obstacle avoidance apparatus, the speed adjustment unit, including:
Velocity bias setup module, for presetting N number of speed on the basis of the plan linear velocity and plan angular speed Spend offset;The velocity bias is the plan linear velocity and the fluctuation of speed amount of plan angular speed;
Cycle movement track obtain module, for according to the period duration, N number of velocity bias, the plan Linear velocity and plan angular speed, obtain N movement locus;
The third place relationship obtains module, for determining the N movement locus and institute according to the second position relationship State the third place relationship between barrier;
Optimum movement locus selecting module, for according between the velocity bias, movement locus and the barrier Distance and the direction of the movement locus filter out best movement locus from the N movement locus.
Further, robot as the aforementioned follows obstacle avoidance apparatus, states cycle movement track and obtains module, including:
Speed adjusts submodule, on the basis of the plan linear velocity v and plan angular speed w, carrying out 2n times not Same adjustment, is obtained 2n+1 velocity group:(v, w), (v+ δ, w+ δ), (v+2 δ, w+2 δ) ... .. (v+n δ, w+n δ), (v- δ, W- δ), (v-2 δ, w-2 δ) ... .. (v-n δ, w-n δ), wherein N=2n+1, δ are a steady state value;
Movement locus obtains submodule, for obtaining 2n+1 according to the 2n+1 velocity group and the duration in the period Movement locus.
In the embodiment of the present application, barrier in object and environment is followed to be passed based on robotic vision using S1. acquisitions The three-dimensional coordinate of sensor coordinate system;S2. bone of the object based on the robotic vision sensor coordinate system is followed described in obtaining Bone data;S3. it determines robot and described follows between the first position relationship of object and the barrier and robot Second position relationship;S4. the benchmark linear velocity of the robot traveling and reference angle speed are determined according to the first position relationship Degree;S5. a skimulated motion track is determined according to the benchmark linear velocity and reference angle speed;S6. according to the second position Relationship and skimulated motion track are adjusted the linear velocity and angular speed of the robot;S7. described in periodically performing successively Step S1 to step S6 persistently carries out object being followed to follow to described.Can effective district partial objectives for and barrier, and realize Hide during target is followed the effect of obstacle.
Description of the drawings
The attached drawing for forming the part of the application is used for providing further understanding of the present application so that the application's is other Feature, objects and advantages become more apparent upon.The illustrative examples attached drawing and its explanation of the application is for explaining the application, not Form the improper restriction to the application.In the accompanying drawings:
Fig. 1 is to follow barrier-avoiding method flow diagram according to a kind of robot of embodiment of the application;
Fig. 2 is a kind of method flow schematic diagram of embodiment of step S3 in method according to Fig. 1;
Fig. 3 is a kind of method flow schematic diagram of embodiment of step S32 in shown method according to fig. 2;
Fig. 4 is a kind of method flow schematic diagram of embodiment of step S6 in method according to Fig. 1;
Fig. 5 is a kind of method flow schematic diagram of embodiment of step S62 in shown method according to fig. 2;
Fig. 6 is a kind of method flow schematic diagram of embodiment of step S1 in method according to Fig. 1;
Fig. 7 is the module connection diagram that obstacle avoidance apparatus is followed according to a kind of robot of embodiment of the application;
Fig. 8 is to connect flow chart according to a kind of module of embodiment device of module 3 in Fig. 7 shown devices;
Fig. 9 is to connect flow chart according to a kind of module of embodiment device of module 32 in Fig. 8 shown devices;
Figure 10 is to connect flow chart according to a kind of module of embodiment device of module 6 in Fig. 7 shown devices;
Figure 11 is to connect flow chart according to a kind of module of embodiment device of module 62 in Figure 10 shown devices;
Figure 12 is to connect flow chart according to a kind of module of embodiment device of module 64 in Figure 10 shown devices;And
Figure 13 is to connect flow chart according to a kind of module of embodiment device of module 1 in Fig. 7 shown devices.
Specific embodiment
In order to which those skilled in the art is made to more fully understand application scheme, below in conjunction in the embodiment of the present application The technical solution in the embodiment of the present application is clearly and completely described in attached drawing, it is clear that described embodiment is only The embodiment of the application part, instead of all the embodiments.Based on the embodiment in the application, ordinary skill people Member's all other embodiments obtained without making creative work should all belong to the model of the application protection It encloses.
It should be noted that term " first " in the description and claims of this application and above-mentioned attached drawing, " Two " etc. be the object for distinguishing similar, and specific sequence or precedence are described without being used for.It should be appreciated that it uses in this way Data can be interchanged in the appropriate case, so as to embodiments herein described herein.In addition, term " comprising " and " tool Have " and their any deformation, it is intended that cover it is non-exclusive include, for example, containing series of steps or unit Process, method, system, product or equipment are not necessarily limited to those steps or unit clearly listed, but may include without clear It is listing to Chu or for the intrinsic other steps of these processes, method, product or equipment or unit.
It should be noted that in the absence of conflict, the feature in embodiment and embodiment in the application can phase Mutually combination.The application is described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.
It present embodiments provides a kind of robot and follows barrier-avoiding method.As shown in Figure 1, this method includes steps S1 To step S7:
S1. it obtains and follows three-dimensional coordinate of the barrier based on robotic vision sensor coordinate system in object and environment; Specifically, three-dimensional data is the initial data of object under visual sensor coordinate system obtained by visual sensor, it is three-dimensional Data can be mutually converted with the depth information, and robot obstacle-avoiding then needs to consider those objects under robot coordinate system Position;Therefore it is opposite to obtain camera coordinates system first by accurately fixing position of the visual sensor in robot for this method In the transition matrix of robot coordinate system;Then position of the space object under robot coordinate system is obtained using this transition matrix It puts, the object that will be less than robot height is considered as barrier, otherwise is considered as non-barrier.It is to do plane fortune in view of robot It is dynamic, so three-dimensional data information is passed through projection transform into the grating map of two dimension;Robot is in nethermost midpoint, grid The y positive directions of lattice map represent the front of robot;This method represents reality using the grating map of 500 × 500 resolution ratio The space that 10 meters × 10 meters of border;The small lattice of black are considered as obstacle in the grating map, represent where there is less than machine The object of people's height, white small lattice are then opposite;Available barrier data are just obtained using such grating map;
S2. skeleton data of the object based on the robotic vision sensor coordinate system, the bone are followed described in obtaining Bone data are depth information of each bone node under the visual sensor coordinate system;Due to utilizing depth information meeting merely Human body target is considered as obstacle, therefore skeleton data by reference herein, the grid in region shared by bone is considered as white always, So as to correct barrier data, the differentiation of target and barrier is realized;
S3. robot and the first position relationship for following object are determined by the skeleton data for following object, The second position relationship between the barrier and robot is determined by the three-dimensional coordinate of the barrier;
S4. the benchmark linear velocity and reference angle speed of the robot traveling are determined according to the first position relationship;Its In, the benchmark linear velocity is the robot to the speed that object is followed to move linearly, the reference angle speed The speed of angle rotation is carried out according to the first position relationship for the robot;
S5. a skimulated motion track is determined according to the benchmark linear velocity and reference angle speed;
S6. the linear velocity and angular speed of the robot are carried out according to the second position relationship and skimulated motion track Adjustment;
S7. the step S1 to step S6 is periodically performed successively, persistently carry out object being followed to follow to described.Due to The object followed is all constantly moving, therefore robot is also required to follow in object and environment with fixed cycle frequency acquisition Three-dimensional coordinate of the barrier based on robotic vision sensor coordinate system to obtain real-time volume coordinate information, is provided to follow Data.
Embodiment 2 as shown in Fig. 2, robot as shown in Example 1 follows barrier-avoiding method, is led in the step S3 It crosses the skeleton data for following object and determines robot and the first position relationship for following object, including:
S31. the average depth of the bone node of object is followed according to obtaining the depth information of each bone node Spend information;
S32. the robot and the first position relationship for following object are determined according to the mean depth information;Institute First position relationship is stated to include:The robot follows the distance between object and angle with described.
Embodiment 3, as shown in figure 3, robot as shown in Example 2 follows barrier-avoiding method, root in the step S32 The robot and the first position relationship for following object are determined according to the mean depth information, including:
S321. according to the three-dimensional coordinate M (x under the corresponding visual sensor coordinate system of the mean depth information acquisition1, y1,z1);
S322. according to the three-dimensional coordinate M (x1,y1,z1) determine the distanceAnd angle
Embodiment 4, as shown in figure 4, robot as shown in Example 1 follows barrier-avoiding method, the step S6 according to The second position relationship and skimulated motion track are adjusted the linear velocity and angular speed of the robot, including:
S61. N number of velocity bias is preset on the basis of the plan linear velocity and plan angular speed;The speed is inclined Shifting amount is the plan linear velocity and the fluctuation of speed amount of plan angular speed;
S62. according to the duration in the period, N number of velocity bias, the plan linear velocity and plan angular speed, N is obtained Movement locus;Period described in the step is consistent with the period acquiescence in the step S7;
S63. the third place between the N movement locus and the barrier is determined according to the second position relationship Relationship;
S64. according to the distance between the velocity bias, movement locus and described barrier and the movement locus Direction filter out best movement locus from the N movement locus.
Embodiment 5, as shown in figure 5, robot as shown in Example 4 follows barrier-avoiding method, the step S62 according to It is described according to the duration in the period, N number of velocity bias, it is described plan linear velocity and plan angular speed, obtain N items movement rail Mark;Including:
S621. on the basis of the plan linear velocity v and plan angular speed w, 2n different adjustment is carried out, is obtained 2n+1 velocity group:(v, w), (v+ δ, w+ δ), (v+2 δ, w+2 δ) ... .. (v+n δ, w+n δ), (v- δ, w- δ), (v-2 δ, w-2 δ) ... .. (v-n δ, w-n δ), wherein N=2n+1, δ are a steady state value;Wherein δ for w and v is constant and micro- for one Small amount;Here, v, w and δ are the data of not tape unit, and concrete unit can correspond to after specific data are obtained and enclose i.e. It can;
S622. 2n+1 movement locus is obtained according to the 2n+1 velocity group and the duration in the period;Due to by Step S521 has obtained velocity group, therefore it is that can obtain accordingly that need to only be directed to the velocity group to carry out corresponding mathematical computations Track lines are repeated no more secondary.
In some embodiments, robot as the aforementioned follows barrier-avoiding method, described according to the velocity bias, movement The distance between track and the barrier and the direction of the movement locus filter out best from the N movement locus Movement locus, including:
The velocity bias is evaluated:If offset evaluation function f1, the f1With the velocity bias into anti- Than the velocity bias is smaller, then offset evaluation score f1It is higher;
The distance between the movement locus and the barrier are evaluated:If movement locus evaluation function f2, institute State f2Directly proportional to the minimum distance d of the movement locus and the barrier, the minimum distance d is bigger, then Distance evaluation Score f2It is higher;If the movement locus, which is had passed through on barrier/barrier or the movement locus, some point distance barriers Hinder object close, and be more than preset minimum threshold of distance there are the minimum distance d of any and nearest barrier, then this The evaluation score of movement locus is with regard to relatively low;
The direction of the movement locus is evaluated:If offset evaluation function f3, determine rising for the movement locus Point and the angle between the line direction of terminal and the starting point of the movement locus and the line direction for following object, it is described f3It is inversely proportional with the angle theta, the angle theta is smaller, then direction evaluation score f3It is higher;Why by the movement locus Direction is because following object and robot current location (starting point of movement locus) line and machine as an evaluation element The angle of people's directional velocity to judge robot either with or without towards the direction of target advance;Therefore it is bigger than normal to work as the angle theta, then The robot is represented not towards the direction of object is followed to walk, angle theta on the contrary is less than normal, then it represents that the side of the robotic movement To with following object orientation consistent, just giving higher direction evaluation score f3
According to the offset evaluation score f1, Distance evaluation score f2With direction evaluation score f3Obtain overall merit point Number z:
Z=a*f1+b*f2+c*f3
Wherein, a, b and c are respectively the offset evaluation score f1, Distance evaluation score f2With direction evaluation score f3's Weight coefficient, generally, chosen distance evaluation score f2The major consideration of overall merit score;The overall merit score Z is highest for the best movement locus, and a+b+c=1, it is preferred that can rule of thumb take a=0.4;B=0.3;C= 0.3。
Embodiment 6 as shown in fig. 6, robot follows barrier-avoiding method as described in Example 1, obtains in the step S1 The three-dimensional coordinate for following barrier view-based access control model sensor coordinate system in object and environment in path, including:
S11. the depth image for the current environment for including following object is obtained by visual sensor, the depth image carries For the depth information for following barrier in object and current environment;
S12. object and the barrier is followed to be sat based on the visual sensor according to determining the depth information Mark the three-dimensional coordinate of system.
It can be seen from the above description that the present invention realizes following technique effect:
It should be noted that step shown in the flowchart of the accompanying drawings can be in such as a group of computer-executable instructions It is performed in computer system, although also, show logical order in flow charts, it in some cases, can be with not The sequence being same as herein performs shown or described step.
According to embodiments of the present invention, additionally provide it is a kind of for implementing the device that above-mentioned robot follows barrier-avoiding method, such as Shown in Fig. 7, which includes:
Three-dimensional coordinate determination unit 1 follows barrier in object and environment to be sensed based on robotic vision for obtaining The three-dimensional coordinate of device coordinate system;
Skeleton data acquiring unit 2 described follows object to be based on the robotic vision sensor coordinates for obtaining The skeleton data of system, the skeleton data are depth information of each bone node under the visual sensor coordinate system;
Position relationship determination unit 3 described follows the skeleton data of object to determine that robot is followed with described for passing through The first position relationship of object determines the second between the barrier and robot by the three-dimensional coordinate of the barrier Put relationship;
Reference speed determination unit 4, for determining the datum line of the robot traveling according to the first position relationship Speed and reference angle speed;Wherein, the benchmark linear velocity robot follows object to move linearly to described Speed, the reference angle speed carry out the speed of angle rotation for the robot according to the first position relationship;
Skimulated motion track determination unit 5, for determining a simulation according to the benchmark linear velocity and reference angle speed Movement locus;
Speed adjustment unit 6, for the line according to the second position relationship and skimulated motion track to the robot Speed and angular speed are adjusted;
Cycling element 7, for periodically performing the three-dimensional coordinate determination unit, skeleton data acquiring unit, position successively Relation determination unit, reference speed determination unit, skimulated motion track determination unit and speed adjustment unit are put, is persistently carried out pair It is described to follow following for object.
Specifically, the modules in the present embodiment realize that the detailed process of its function can be found in the implementation of method shown in Fig. 1 Associated description in example, details are not described herein again.
In some embodiments, as shown in figure 8, robot follows obstacle avoidance apparatus as the aforementioned, the position relationship determines Unit 3, including:
Mean depth information acquisition module 31, for being followed according to the acquisition of the depth information of each bone node The mean depth information of the bone node of object;
First position relationship determination module 32, for according to the mean depth information determine the robot with it is described with With the first position relationship of object;The first position relationship includes:The robot follows the distance between object with described And position relation.
In some embodiments, as shown in figure 9, the first position relationship determination module 32, including:
Three-dimensional coordinate obtains module 321, according under the corresponding visual sensor coordinate system of the mean depth information acquisition Three-dimensional coordinate M (x1,y1,z1);
Positional information calculation module 322, for according to the three-dimensional coordinate M (x1,y1,z1) determine the distanceAnd angle
In some embodiments, as shown in Figure 10, robot as the aforementioned follows obstacle avoidance apparatus, the speed adjustment unit 6, including:
Velocity bias setup module 61, it is N number of for being preset on the basis of the plan linear velocity and plan angular speed Velocity bias;The velocity bias is the plan linear velocity and the fluctuation of speed amount of plan angular speed;
Cycle movement track obtain module 62, by according to the period duration, N number of velocity bias, it is described based on Speed of crossing and plan angular speed, obtain N movement locus;
The third place relationship obtain module 63, for according to the second position relationship determine the N movement locus with The third place relationship between the barrier;
Optimum movement locus selecting module 64, for according to the velocity bias, movement locus and the barrier it Between distance and the direction of the movement locus filter out best movement locus from the N movement locus.
Specifically, the modules in the present embodiment realize that the detailed process of its function can be found in the implementation of method shown in Fig. 2 Associated description in example, details are not described herein again.
In some embodiments, as shown in figure 11, robot as the aforementioned follows obstacle avoidance apparatus, states cycle movement track Module 62 is obtained, including:
Speed adjusts submodule 621, on the basis of the plan linear velocity v and plan angular speed w, carrying out 2n times 2n+1 velocity group is obtained in different adjustment:(v, w), (v+ δ, w+ δ), (v+2 δ, w+2 δ) ... .. (v+n δ, w+n δ), (v- δ, w- δ), (v-2 δ, w-2 δ) ... .. (v-n δ, w-n δ), wherein N=2n+1, δ are a steady state value;
Movement locus obtains submodule 622, for being obtained according to the 2n+1 velocity group and the duration in the period 2n+1 movement locus.
Specifically, the modules in the present embodiment realize that the detailed process of its function can be found in the implementation of method shown in Fig. 3 Associated description in example, details are not described herein again.
In some embodiments, as shown in figure 12, robot as the aforementioned follows obstacle avoidance apparatus, the optimum movement locus Selecting module 64, including:
Velocity bias evaluates submodule 641, for evaluating the velocity bias:If offset evaluation function f1,The f1It is inversely proportional with the velocity bias, the velocity bias is smaller, then offset evaluation score f1 It is higher;
Distance evaluation submodule 642, for evaluating the distance between the movement locus and the barrier:If Movement locus evaluation function f2,The f2It is directly proportional to the minimum distance d of the movement locus and the barrier, institute It is bigger to state minimum distance d, then Distance evaluation score f2It is higher;
Submodule 643 is evaluated in direction, is evaluated for the direction to the movement locus:If offset evaluation function f3,Determine the line direction of the beginning and end of the movement locus and the starting point of the movement locus Angle between the line direction for following object, the f3It is inversely proportional with the angle theta, the angle theta is smaller, then side To evaluation score f3It is higher;
Overall merit submodule 644, for according to the offset evaluation score f1, Distance evaluation score f2It is commented with direction Valency score f3Obtain overall merit score z:
Z=a*f1+b*f2+c*f3
Wherein, a, b and c are respectively the offset evaluation score f1, Distance evaluation score f2With direction evaluation score f3's Weight coefficient;The overall merit score z is highest for the best movement locus, and a+b+c=1, it is preferred that according to warp A=0.4 can be taken by testing;B=0.3;C=0.3.
Specifically, the modules in the present embodiment realize that the detailed process of its function can be found in the implementation of method shown in Fig. 4 Associated description in example, details are not described herein again.
In some embodiments, as shown in figure 13, robot as the aforementioned follows obstacle avoidance apparatus, and the three-dimensional coordinate determines Unit 1, including:
Depth information acquistion module 11, for obtaining the depth for the current environment for including following object by visual sensor Image, the depth image provide described in follow the depth information of barrier in object and current environment;
Three-dimensional coordinate determining module 12, for following object and the barrier base described in being determined according to the depth information In the three-dimensional coordinate of the visual sensor coordinate system.
Specifically, the modules in the present embodiment realize that the detailed process of its function can be found in the implementation of method shown in Fig. 5 Associated description in example, details are not described herein again.
Obviously, those skilled in the art should be understood that each module of the above-mentioned present invention or each step can be with general Computing device realize that they can concentrate on single computing device or be distributed in multiple computing devices and be formed Network on, optionally, they can be realized with the program code that computing device can perform, it is thus possible to which they are stored In the storage device by computing device come perform either they are fabricated to respectively each integrated circuit modules or by they In multiple modules or step be fabricated to single integrated circuit module to realize.In this way, the present invention is not limited to any specific Hardware and software combines.
The foregoing is merely the preferred embodiments of the application, are not limited to the application, for the skill of this field For art personnel, the application can have various modifications and variations.It is all within spirit herein and principle, made any repair Change, equivalent replacement, improvement etc., should be included within the protection domain of the application.

Claims (16)

1. a kind of robot follows barrier-avoiding method, which is characterized in that including:
S1. it obtains and follows three-dimensional coordinate of the barrier based on robotic vision sensor coordinate system in object and environment;
S2. skeleton data of the object based on the robotic vision sensor coordinate system, the bone number are followed described in obtaining According to the depth information for each bone node under the visual sensor coordinate system;
S3. robot and the first position relationship for following object are determined by the skeleton data for following object, passed through The three-dimensional coordinate of the barrier determines the second position relationship between the barrier and robot;
S4. the benchmark linear velocity and reference angle speed of the robot traveling are determined according to the first position relationship;Wherein, institute State benchmark linear velocity for the robot to the speed that object is followed to move linearly, the reference angle speed is described Robot carries out the speed of angle rotation according to the first position relationship;
S5. a skimulated motion track is determined according to the benchmark linear velocity and reference angle speed;
S6. the linear velocity and angular speed of the robot are adjusted according to the second position relationship and skimulated motion track It is whole;
S7. the step S1 to step S6 is periodically performed successively, persistently carry out object being followed to follow to described.
2. robot according to claim 1 follows barrier-avoiding method, which is characterized in that described to pass through the object that follows Skeleton data determines robot and the first position relationship for following object, including:
The mean depth information of the bone node of object is followed according to obtaining the depth information of each bone node;
The robot and the first position relationship for following object are determined according to the mean depth information;Described first The relationship of putting includes:The robot follows the distance between object and angle with described.
3. robot according to claim 2 follows barrier-avoiding method, which is characterized in that described to be believed according to the mean depth Breath determines the robot and the first position relationship for following object, including:
According to the three-dimensional coordinate M (x under the corresponding visual sensor coordinate system of the mean depth information acquisition1,y1,z1);
According to the three-dimensional coordinate M (x1,y1,z1) determine the distanceAnd angle
4. robot according to claim 1 follows barrier-avoiding method, which is characterized in that described to be closed according to the second position System and skimulated motion track are adjusted the linear velocity and angular speed of the robot, including:
N number of velocity bias is preset on the basis of the plan linear velocity and plan angular speed;The velocity bias is institute It states plan linear velocity and plans the fluctuation of speed amount of angular speed;
According to the duration in the period, N number of velocity bias, the plan linear velocity and plan angular speed, N items movement rail is obtained Mark;
The third place relationship between the N movement locus and the barrier is determined according to the second position relationship;
According to the direction of the distance between the velocity bias, movement locus and described barrier and the movement locus from The N movement locus filters out best movement locus.
5. robot according to claim 2 follows barrier-avoiding method, which is characterized in that it is described according to described according to the week The duration of phase, N number of velocity bias, the plan linear velocity and plan angular speed, obtain N movement locus;Including:
On the basis of the plan linear velocity v and plan angular speed w, 2n different adjustment is carried out, 2n+1 speed is obtained Degree group:(v, w), (v+ δ, w+ δ), (v+2 δ, w+2 δ) ... .. (v+n δ, w+n δ), (v- δ, w- δ), (v-2 δ, w-2 δ) ... .. (v-n δ, w-n δ), wherein N=2n+1, δ are a steady state value;
2n+1 movement locus is obtained according to the 2n+1 velocity group and the duration in the period.
6. a kind of robot follows obstacle avoidance apparatus, which is characterized in that including:
Three-dimensional coordinate determination unit follows barrier in object and environment to be based on robotic vision sensor coordinates for obtaining The three-dimensional coordinate of system;
Skeleton data acquiring unit described follows bone of the object based on the robotic vision sensor coordinate system for obtaining Bone data, the skeleton data are depth information of each bone node under the visual sensor coordinate system;
Position relationship determination unit determines robot and the object that follows for passing through the skeleton data for following object First position relationship determines that the second position between the barrier and robot is closed by the three-dimensional coordinate of the barrier System;
Reference speed determination unit, for according to the first position relationship determine benchmark linear velocity that the robot advances and Reference angle speed;Wherein, the benchmark linear velocity is the robot to the speed that object is followed to move linearly, institute State the speed that reference angle speed carries out angle rotation for the robot according to the first position relationship;
Skimulated motion track determination unit, for determining a skimulated motion rail according to the benchmark linear velocity and reference angle speed Mark;
Speed adjustment unit, for according to the second position relationship and skimulated motion track to the linear velocity of the robot and Angular speed is adjusted;
Cycling element, for periodically performing the three-dimensional coordinate determination unit, skeleton data acquiring unit, position relationship successively Determination unit, reference speed determination unit, skimulated motion track determination unit and speed adjustment unit, persistently carry out to it is described with With following for object.
7. robot according to claim 1 follows barrier-avoiding method, which is characterized in that the position relationship determination unit, Including:
Mean depth information acquisition module, for following object according to the acquisition of the depth information of each bone node The mean depth information of bone node;
First position relationship determination module, for determining that the robot follows object with described according to the mean depth information First position relationship;The first position relationship includes:The robot follows the distance between object and side with described Position relationship.
8. robot according to claim 7 follows barrier-avoiding method, which is characterized in that the first position relationship determines mould Block, including:
Three-dimensional coordinate obtains module, is sat according to the three-dimensional under the corresponding visual sensor coordinate system of the mean depth information acquisition Mark M (x1,y1,z1);
Positional information calculation module, for according to the three-dimensional coordinate M (x1,y1,z1) determine the distanceAnd angle
9. robot according to claim 6 follows obstacle avoidance apparatus, which is characterized in that the speed adjustment unit, including:
Velocity bias setup module, it is inclined for presetting N number of speed on the basis of the plan linear velocity and plan angular speed Shifting amount;The velocity bias is the plan linear velocity and the fluctuation of speed amount of plan angular speed;
Cycle movement track obtains module, for duration, N number of velocity bias, the plan linear speed according to the period Degree and plan angular speed, obtain N movement locus;
The third place relationship obtains module, for determining the N movement locus and the barrier according to the second position relationship Hinder the third place relationship between object;
Optimum movement locus selecting module, for according between the velocity bias, movement locus and the barrier away from From and the direction of the movement locus filter out best movement locus from the N movement locus.
10. robot according to claim 9 follows obstacle avoidance apparatus, which is characterized in that obtains the cycle movement track Module is obtained, including:
Speed adjusts submodule, different on the basis of the plan linear velocity v and plan angular speed w, carrying out 2n times Adjustment, is obtained 2n+1 velocity group:(v, w), (v+ δ, w+ δ), (v+2 δ, w+2 δ) ... .. (v+n δ, w+n δ), (v- δ, w- δ), (v-2 δ, w-2 δ) ... .. (v-n δ, w-n δ), wherein N=2n+1, δ are a steady state value;
Movement locus obtains submodule, for obtaining 2n+1 items fortune according to the 2n+1 velocity group and the duration in the period Dynamic rail mark.
CN201711384862.9A 2017-12-20 2017-12-20 Robot follows barrier-avoiding method and device Pending CN108255173A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201711384862.9A CN108255173A (en) 2017-12-20 2017-12-20 Robot follows barrier-avoiding method and device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201711384862.9A CN108255173A (en) 2017-12-20 2017-12-20 Robot follows barrier-avoiding method and device

Publications (1)

Publication Number Publication Date
CN108255173A true CN108255173A (en) 2018-07-06

Family

ID=62723472

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201711384862.9A Pending CN108255173A (en) 2017-12-20 2017-12-20 Robot follows barrier-avoiding method and device

Country Status (1)

Country Link
CN (1) CN108255173A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110244791A (en) * 2019-07-11 2019-09-17 北京理工大学 A kind of biped robot's foot power and moment follow-up control method

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102831380A (en) * 2011-06-15 2012-12-19 康佳集团股份有限公司 Body action identification method and system based on depth image induction
JP2014130495A (en) * 2012-12-28 2014-07-10 Mitsubishi Heavy Ind Ltd Movement control device, movement control method, and program
CN104766230A (en) * 2015-04-21 2015-07-08 东华大学 Advertising effect evaluation method based on human skeletal tracking
CN106774303A (en) * 2016-10-12 2017-05-31 纳恩博(北京)科技有限公司 A kind of method for tracing and tracing equipment
CN106774301A (en) * 2016-10-25 2017-05-31 纳恩博(北京)科技有限公司 A kind of avoidance follower method and electronic equipment

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102831380A (en) * 2011-06-15 2012-12-19 康佳集团股份有限公司 Body action identification method and system based on depth image induction
JP2014130495A (en) * 2012-12-28 2014-07-10 Mitsubishi Heavy Ind Ltd Movement control device, movement control method, and program
CN104766230A (en) * 2015-04-21 2015-07-08 东华大学 Advertising effect evaluation method based on human skeletal tracking
CN106774303A (en) * 2016-10-12 2017-05-31 纳恩博(北京)科技有限公司 A kind of method for tracing and tracing equipment
CN106774301A (en) * 2016-10-25 2017-05-31 纳恩博(北京)科技有限公司 A kind of avoidance follower method and electronic equipment

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
张启彬 等: "基于速度空间的移动机器人同时避障和轨迹跟踪方法", 《控制与决策》 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110244791A (en) * 2019-07-11 2019-09-17 北京理工大学 A kind of biped robot's foot power and moment follow-up control method
CN110244791B (en) * 2019-07-11 2020-05-15 北京理工大学 Foot force and moment following control method for biped robot

Similar Documents

Publication Publication Date Title
CN106444780B (en) A kind of autonomous navigation method and system of the robot of view-based access control model location algorithm
Kaufmann et al. Deep drone racing: Learning agile flight in dynamic environments
CN105425791B (en) A kind of the group robot control system and method for view-based access control model positioning
CN104748751B (en) The calculation method of attitude matrix and positioning navigation method based on attitude matrix
Franchi et al. Modeling and control of UAV bearing formations with bilateral high-level steering
Corrales et al. Hybrid tracking of human operators using IMU/UWB data fusion by a Kalman filter
Kriegman et al. Stereo vision and navigation in buildings for mobile robots.
Hrabar 3D path planning and stereo-based obstacle avoidance for rotorcraft UAVs
CN105043396B (en) The method and system of self-built map in a kind of mobile robot room
Rekleitis et al. Multi-robot collaboration for robust exploration
Stelzer et al. Stereo-vision-based navigation of a six-legged walking robot in unknown rough terrain
Hager Task-directed sensor fusion and planning: a computational approach
Snape et al. Independent navigation of multiple mobile robots with hybrid reciprocal velocity obstacles
Fujita et al. Passivity-based dynamic visual feedback control for three-dimensional target tracking: Stability and $ L_ {2} $-gain performance analysis
Kuipers et al. A robot exploration and mapping strategy based on a semantic hierarchy of spatial representations
CN104537829B (en) A kind of intelligent transportation Physical Simulation Platform and the localization method for the intelligent transportation Physical Simulation Platform
Taylor et al. Vision-based motion planning and exploration algorithms for mobile robots
Lumelsky et al. Decentralized motion planning for multiple mobile robots: The cocktail party model
Lee et al. Controlling mobile robots in distributed intelligent sensor network
Agrawal et al. Localization and mapping for autonomous navigation in outdoor terrains: A stereo vision approach
Weiss Vision based navigation for micro helicopters
CN106325270B (en) Intelligent vehicle air navigation aid based on perception and from host computer location navigation
Kriegman et al. A mobile robot: Sensing, planning and locomotion
Matthies et al. Stereo vision-based obstacle avoidance for micro air vehicles using disparity space
Ricks et al. Ecological displays for robot interaction: A new perspective

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