CN115454061B - Robot path obstacle avoidance method and system based on 3D technology - Google Patents
Robot path obstacle avoidance method and system based on 3D technology Download PDFInfo
- Publication number
- CN115454061B CN115454061B CN202211051740.9A CN202211051740A CN115454061B CN 115454061 B CN115454061 B CN 115454061B CN 202211051740 A CN202211051740 A CN 202211051740A CN 115454061 B CN115454061 B CN 115454061B
- Authority
- CN
- China
- Prior art keywords
- obstacle avoidance
- dimensional
- robot
- path
- obstacle
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 41
- 238000005516 engineering process Methods 0.000 title claims abstract description 16
- 238000013507 mapping Methods 0.000 claims abstract description 7
- 238000012544 monitoring process Methods 0.000 claims abstract description 7
- 230000006870 function Effects 0.000 claims description 25
- 238000011156 evaluation Methods 0.000 claims description 19
- 238000013527 convolutional neural network Methods 0.000 claims description 18
- 238000005457 optimization Methods 0.000 claims description 17
- 238000013528 artificial neural network Methods 0.000 claims description 13
- 238000001514 detection method Methods 0.000 claims description 4
- 238000004891 communication Methods 0.000 claims description 3
- 238000010276 construction Methods 0.000 claims description 3
- 238000011065 in-situ storage Methods 0.000 claims description 3
- 238000012549 training Methods 0.000 claims description 3
- 238000013135 deep learning Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000013433 optimization analysis Methods 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000009776 industrial production Methods 0.000 description 1
- 230000010355 oscillation Effects 0.000 description 1
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
- G05D1/0257—Control of position or course in two dimensions specially adapted to land vehicles using a radar
-
- 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
- G05D1/0214—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory in accordance with safety or protection criteria, e.g. avoiding hazardous areas
-
- 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)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Aviation & Aerospace Engineering (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
- Manipulator (AREA)
Abstract
The invention discloses a robot path obstacle avoidance method and system based on a 3D technology, and the method comprises the following steps: performing model fitting on the three-dimensional characteristics of the obstacle and the three-dimensional obstacle avoidance parameters of the robot to obtain a robot path obstacle avoidance model representing the mapping relation between the three-dimensional characteristics and the three-dimensional obstacle avoidance parameters; monitoring three-dimensional characteristics of real-time obstacles on a robot travelling path, and obtaining real-time three-dimensional obstacle avoidance parameters of the robot based on the three-dimensional characteristics of the real-time obstacles by utilizing a robot path obstacle avoidance model. According to the method, the three-dimensional characteristics of the obstacle and the three-dimensional obstacle avoidance parameters of the robot are subjected to model fitting to obtain the robot path obstacle avoidance model representing the mapping relation between the three-dimensional characteristics and the three-dimensional obstacle avoidance parameters, so that the robot can perform real-time autonomous obstacle avoidance according to the real-time three-dimensional obstacle avoidance parameters, and an autonomous obstacle avoidance algorithm based on the depth model has the capability of autonomous learning in an obstacle avoidance scene, and the problems of complicated three-dimensional reconstruction and path planning are avoided.
Description
Technical Field
The invention relates to the technical field of robot obstacle avoidance, in particular to a robot path obstacle avoidance method and system based on a 3D technology.
Background
With the continuous development of technology, mobile robots have been widely used in various fields including life service, industrial production, military, entertainment, etc. Robotics design control, mechanics, computers, and other disciplines. The navigation and obstacle avoidance capability of the robot is an important index reflecting the intelligence of the robot.
Khatib builds an artificial potential field in Real-timeObstatalcloidefor handmade mobile robot, so that an abstract repulsive force and attractive force are generated on a robot in the artificial potential field by an obstacle and a target point, and the robot is controlled to avoid the obstacle. The path planned by the artificial potential field method has the advantages of safety and smoothness, and is convenient for the robot to actually execute control, but the artificial potential field method is easy to trap into a local extremum, so that the problem of path oscillation is generated when the path approaches an obstacle or a target point. Nicolai et al estimate continuously transformed point cloud data based on a deep convolutional neural network in deep learning laser BaserdOdometyEstimation, thereby realizing path planning and obstacle avoidance of robots. While existing sensors have been able to provide large amounts of accurate environmental data, positioning methods have been difficult to efficiently resolve these large amounts of data and obstacle avoidance algorithms have been overly complex.
Disclosure of Invention
The invention aims to provide a robot path obstacle avoidance method based on a 3D technology, which aims to solve the technical problem that an obstacle avoidance algorithm is too complex in the prior art.
In order to solve the technical problems, the invention specifically provides the following technical scheme:
a robot path obstacle avoidance method based on a 3D technology comprises the following steps:
step S1, detecting three-dimensional characteristics of an obstacle positioned on a robot travelling path by using a three-dimensional radar carried by the robot, and optimally analyzing three-dimensional obstacle avoidance parameters of the robot according to the three-dimensional characteristics of the obstacle;
s2, performing model fitting on the three-dimensional characteristics of the obstacle and the three-dimensional obstacle avoidance parameters of the robot to obtain a robot path obstacle avoidance model representing the mapping relation between the three-dimensional characteristics and the three-dimensional obstacle avoidance parameters;
and S3, monitoring three-dimensional characteristics of the real-time obstacle on a robot travelling path, and obtaining real-time three-dimensional obstacle avoidance parameters of the robot based on the three-dimensional characteristics of the real-time obstacle by utilizing a robot path obstacle avoidance model so as to realize real-time autonomous obstacle avoidance of the robot according to the real-time three-dimensional obstacle avoidance parameters.
As a preferred embodiment of the present invention, the detection of three-dimensional features of an obstacle located on a travel path of a robot by using a three-dimensional radar mounted on the robot includes:
detecting three-dimensional coordinates of an obstacle positioned on a travel path of the robot by using a three-dimensional radar carried by the robot, and respectively extracting two boundary coordinates of the obstacle near the two path boundaries of the travel path;
extracting three-dimensional coordinates from two path boundary coordinates to two edge coordinates of an obstacle on a traveling path respectively to serve as three-dimensional coordinates of a plurality of obstacle avoidance selectable spaces, and marking the plurality of obstacle avoidance selectable spaces as an AA 'space and a B' space respectively, wherein A represents a left boundary of the traveling path, A 'represents a left boundary of the obstacle, B' represents a right boundary of the traveling path, and B represents a right boundary of the obstacle;
and taking the three-dimensional coordinates of the obstacle and the three-dimensional coordinates of a plurality of obstacle avoidance selectable spaces as the three-dimensional characteristics of the obstacle.
As a preferred solution of the present invention, the optimizing and analyzing the three-dimensional obstacle avoidance parameters of the robot according to the three-dimensional characteristics of the obstacle includes:
calculating three-dimensional volumes of the obstacle avoidance selectable spaces by utilizing three-dimensional coordinates of the obstacle avoidance selectable spaces in the three-dimensional features, comparing the three-dimensional volumes of the plurality of obstacle avoidance selectable spaces with the three-dimensional volumes of the robot, wherein,
if the three-dimensional volumes of the AA 'space and the B' space are larger than or equal to the three-dimensional volume of the robot, the AA 'space and the B' space are used as effective obstacle avoidance spaces;
if the three-dimensional volume of the AA ' space is larger than or equal to the three-dimensional volume of the robot and the three-dimensional volume of the B ' space is smaller than the three-dimensional volume of the robot, taking the AA ' space as an effective obstacle avoidance space;
if the three-dimensional volume of the B ' B space is larger than or equal to the three-dimensional volume of the robot and the three-dimensional volume of the AA ' space is smaller than the three-dimensional volume of the robot, taking the B ' B space as an effective obstacle avoidance space;
if the three-dimensional volumes of the AA 'space and the B' space are smaller than the three-dimensional volume of the robot, taking the AA 'space and the B' space as non-effective obstacle avoidance spaces;
when at least one effective obstacle avoidance space exists, setting an optimization function of an obstacle avoidance path in each effective obstacle avoidance space according to the highest path smoothness and the shortest path length, wherein the optimization function of the obstacle avoidance path is as follows:
in the method, in the process of the invention,to characterize the optimal function for the highest path smoothness of the obstacle avoidance path, +.>To characterize the shortest path length of the obstacle avoidance path, an optimization function, p i The three-dimensional coordinate of the ith path point in the obstacle avoidance path is obtained, N is the total number of the path points in the obstacle avoidance path, and i is a metering constant;
setting a solving constraint condition of an optimization function of the obstacle avoidance path, wherein the solving constraint condition is as follows:
in the formula, |p i A| is p i Distance to A, |p i A' |is p i Distance to A' |p i B|is p i Distance to B, |p i B' |is p i The distance to B ', O is characterized as the center point of the robot, C is the left boundary of the robot, A' is the right boundary of the robot, |OC| is the distance from O to C, |OC '| is the distance from O to C', and [ AA '']Is AA 'space, [ BB ]']Is BB' space;
solving based on the optimization functions of the solving constraint conditions and the obstacle avoidance paths to obtain an obstacle avoidance path of an effective obstacle avoidance space;
and taking the effective obstacle avoidance space and the obstacle avoidance path as three-dimensional obstacle avoidance parameters of the robot.
As a preferable scheme of the invention, when no effective obstacle avoidance space exists, the robot is stopped in situ to perform early warning.
As a preferable scheme of the invention, the construction of the robot path obstacle avoidance model comprises the following steps:
taking the three-dimensional characteristics of the obstacle as an input item of a CNN neural network, taking the three-dimensional obstacle avoidance parameters of the robot as an output item of the CNN neural network, and carrying out model training by utilizing the CNN neural network based on the input item and the output item to obtain a robot path obstacle avoidance model, wherein the model expression of the robot path obstacle avoidance model is as follows:
Out[Y]=CNN(In[X]);
in the formula, out [ Y ] is the three-dimensional obstacle avoidance parameter of the robot, in [ X ] is the three-dimensional characteristic of the obstacle, and CNN is a CNN neural network.
As a preferred aspect of the present invention, the monitoring of three-dimensional characteristics of real-time obstacles on a travel path of a robot includes:
taking an obstacle closest to the robot in the travelling direction on a travelling path as a real-time obstacle, detecting three-dimensional coordinates of the obstacle by using a three-dimensional radar carried by the robot, and respectively extracting three-dimensional coordinates of a plurality of obstacle avoidance selectable spaces of the real-time obstacle;
and taking the three-dimensional coordinates of the real-time obstacle and the three-dimensional coordinates of the multiple obstacle avoidance selectable spaces as three-dimensional characteristics of the real-time obstacle.
As a preferred solution of the present invention, the obtaining real-time three-dimensional obstacle avoidance parameters of a robot based on three-dimensional features of real-time obstacles by using a robot path obstacle avoidance model includes:
inputting three-dimensional characteristics of the real-time obstacle into a robot path obstacle avoidance model, and outputting real-time three-dimensional obstacle avoidance parameters of the robot by the robot path obstacle avoidance model, wherein the real-time three-dimensional obstacle avoidance parameters comprise real-time effective obstacle avoidance space and real-time obstacle avoidance path of the robot.
As a preferable mode of the present invention, the three-dimensional feature detection distance of the obstacle of the robot is larger than the safety braking distance of the robot.
As a preferable scheme of the invention, when the number of the real-time obstacle avoidance paths is more than 1, setting the evaluation weight of the path smoothness and the path length, and carrying out weighted summation on the path smoothness and the path length to obtain a selection function for selecting the real-time obstacle avoidance paths, wherein the selection function is as follows:
wherein, jarge is the evaluation value of the real-time obstacle avoidance path,path smoothness evaluation value for real-time obstacle avoidance path, < ->The path length evaluation value W, V for the real-time obstacle avoidance path is the evaluation weight of the path smoothness and the path length respectively;
and the robot performs real-time autonomous obstacle avoidance according to the real-time obstacle avoidance path with the highest evaluation value.
As a preferable scheme of the invention, the invention provides an obstacle avoidance system of the robot path obstacle avoidance method based on the 3D technology, which comprises a three-dimensional radar and a data processor, wherein the three-dimensional radar is in communication connection with the data processor, the three-dimensional radar is used for detecting three-dimensional characteristics of an obstacle, the data processor is used for receiving the three-dimensional characteristics of the obstacle, and three-dimensional obstacle avoidance parameters of the robot are obtained based on the three-dimensional characteristics of the obstacle by utilizing a robot path obstacle avoidance model, and the robot path obstacle avoidance model is arranged in the data processor.
Compared with the prior art, the invention has the following beneficial effects:
according to the method, the three-dimensional characteristics of the obstacle and the three-dimensional obstacle avoidance parameters of the robot are subjected to model fitting to obtain the robot path obstacle avoidance model representing the mapping relation between the three-dimensional characteristics and the three-dimensional obstacle avoidance parameters, so that the robot can perform real-time autonomous obstacle avoidance according to the real-time three-dimensional obstacle avoidance parameters, and an autonomous obstacle avoidance algorithm based on the depth model has the capability of autonomous learning in an obstacle avoidance scene, and the problems of complicated three-dimensional reconstruction and path planning are avoided.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below. It will be apparent to those of ordinary skill in the art that the drawings in the following description are exemplary only and that other implementations can be obtained from the extensions of the drawings provided without inventive effort.
FIG. 1 is a flow chart of a robot path obstacle avoidance method provided by an embodiment of the invention;
fig. 2 is a schematic diagram of a three-dimensional feature structure according to an embodiment of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
As shown in fig. 1, the invention provides a robot path obstacle avoidance method based on a 3D technology, comprising the following steps:
step S1, detecting three-dimensional characteristics of an obstacle positioned on a robot travelling path by using a three-dimensional radar carried by the robot, and optimally analyzing three-dimensional obstacle avoidance parameters of the robot according to the three-dimensional characteristics of the obstacle;
detecting three-dimensional features of an obstacle located on a travel path of a robot by using a three-dimensional radar mounted on the robot, the three-dimensional features including:
as shown in fig. 2, three-dimensional coordinates of an obstacle located on a travel path of a robot are detected by using a three-dimensional radar mounted on the robot, and two boundary coordinates of the obstacle near two path boundaries of the travel path are extracted respectively;
extracting three-dimensional coordinates from two path boundary coordinates to two edge coordinates of an obstacle on a traveling path respectively to serve as three-dimensional coordinates of a plurality of obstacle avoidance selectable spaces, and marking the plurality of obstacle avoidance selectable spaces as an AA 'space and a B' space respectively, wherein A represents a left boundary of the traveling path, A 'represents a left boundary of the obstacle, B' represents a right boundary of the traveling path, and B represents a right boundary of the obstacle;
and taking the three-dimensional coordinates of the obstacle and the three-dimensional coordinates of a plurality of obstacle avoidance selectable spaces as the three-dimensional characteristics of the obstacle.
The method for optimizing and analyzing the three-dimensional obstacle avoidance parameters of the robot according to the three-dimensional characteristics of the obstacle comprises the following steps:
calculating three-dimensional volumes of the obstacle avoidance selectable spaces by utilizing three-dimensional coordinates of the obstacle avoidance selectable spaces in the three-dimensional features, comparing the three-dimensional volumes of the plurality of obstacle avoidance selectable spaces with the three-dimensional volumes of the robot, wherein,
if the three-dimensional volumes of the AA 'space and the B' space are larger than or equal to the three-dimensional volume of the robot, the AA 'space and the B' space are used as effective obstacle avoidance spaces;
if the three-dimensional volume of the AA ' space is larger than or equal to the three-dimensional volume of the robot and the three-dimensional volume of the B ' space is smaller than the three-dimensional volume of the robot, taking the AA ' space as an effective obstacle avoidance space;
if the three-dimensional volume of the B ' B space is larger than or equal to the three-dimensional volume of the robot and the three-dimensional volume of the AA ' space is smaller than the three-dimensional volume of the robot, taking the B ' B space as an effective obstacle avoidance space;
if the three-dimensional volumes of the AA 'space and the B' space are smaller than the three-dimensional volume of the robot, taking the AA 'space and the B' space as non-effective obstacle avoidance spaces;
when at least one effective obstacle avoidance space exists, setting an optimization function of an obstacle avoidance path in each effective obstacle avoidance space according to the highest path smoothness and the shortest path length, wherein the optimization function of the obstacle avoidance path is as follows:
in the method, in the process of the invention,to characterize the optimal function for the highest path smoothness of the obstacle avoidance path, +.>To characterize the shortest path length of the obstacle avoidance path, an optimization function, p i The three-dimensional coordinate of the ith path point in the obstacle avoidance path is obtained, N is the total number of the path points in the obstacle avoidance path, and i is a metering constant;
setting a solving constraint condition of an optimization function of the obstacle avoidance path, wherein the solving constraint condition is as follows:
in the formula, |p i A| is p i Distance to A, |p i A' |is p i Distance to A' |p i B|is p i Distance to B, |p i B' |is p i Distance to B ', O represents the center point of the robot, C is the left boundary of the robot, A' isRight boundary of robot, |oc| is distance from O to C, |oc ' | is distance from O to C, [ AA ' ']Is AA 'space, [ BB ]']For BB' space, i.e. p i The value range of the robot is in the AA 'space or the BB' space, the O point is a planning position point on the obstacle avoidance path of the robot, the center point of the robot is used as the basis of the obstacle avoidance path, the condition that the left and right boundaries exceed or collide with the left and right boundaries of the effective obstacle avoidance space when the robot moves on the obstacle avoidance path is ensured, and the moving safety of the robot is ensured;
solving based on an optimization function of solving the constraint condition multiple obstacle avoidance paths to obtain an obstacle avoidance path of an effective obstacle avoidance space;
and taking the effective obstacle avoidance space and the obstacle avoidance path as three-dimensional obstacle avoidance parameters of the robot.
And when no effective obstacle avoidance space exists, stopping the robot in situ for early warning.
The obstacle avoidance path is determined by using the optimization method of the path smoothness and the path length, so that the path smoothness ensures small adjustment amount of deflection movement while the robot successfully moves on the obstacle avoidance path to realize the obstacle avoidance function, thereby ensuring small movement fluctuation degree of the robot, improving the stability of the obstacle avoidance process of the robot, ensuring small path length, ensuring high obstacle avoidance efficiency of the robot, completing obstacle avoidance with the shortest travel and the shortest time, and realizing the trinity of obstacle avoidance safety, stability and efficiency.
S2, performing model fitting on the three-dimensional characteristics of the obstacle and the three-dimensional obstacle avoidance parameters of the robot to obtain a robot path obstacle avoidance model representing the mapping relation between the three-dimensional characteristics and the three-dimensional obstacle avoidance parameters;
the construction of the robot path obstacle avoidance model comprises the following steps:
taking the three-dimensional characteristics of the obstacle as an input item of a CNN neural network, taking the three-dimensional obstacle avoidance parameters of the robot as an output item of the CNN neural network, and carrying out model training by utilizing the CNN neural network based on the input item and the output item to obtain a robot path obstacle avoidance model, wherein the model expression of the robot path obstacle avoidance model is as follows:
Out[Y]=CNN(In[X]);
in the formula, out [ Y ] is the three-dimensional obstacle avoidance parameter of the robot, in [ X ] is the three-dimensional characteristic of the obstacle, and CNN is a CNN neural network.
The three-dimensional obstacle avoidance parameters obtained through optimization analysis and the three-dimensional characteristics of the obstacle are utilized to perform model fitting, so that the deep learning capacity of the neural network can be utilized, the three-dimensional obstacle avoidance parameters which accord with the safety, stability and efficiency of obstacle avoidance can be predicted through the three-dimensional characteristics of the obstacle, therefore, the three-dimensional obstacle avoidance parameters can be obtained through the three-dimensional characteristics directly only by using the robot path obstacle avoidance model, the planning of an obstacle avoidance path is not required to be performed by repeating the optimization analysis process or a traditional path planning method (an artificial potential field, an A-type path planning method and the like), the path planning effect is ensured, and meanwhile, the obstacle avoidance path planning efficiency is improved.
And S3, monitoring three-dimensional characteristics of the real-time obstacle on a robot travelling path, and obtaining real-time three-dimensional obstacle avoidance parameters of the robot based on the three-dimensional characteristics of the real-time obstacle by utilizing a robot path obstacle avoidance model so as to realize real-time autonomous obstacle avoidance of the robot according to the real-time three-dimensional obstacle avoidance parameters.
Monitoring three-dimensional characteristics of real-time obstructions on a robot travel path, comprising:
taking an obstacle closest to the robot in the travelling direction on a travelling path as a real-time obstacle, detecting three-dimensional coordinates of the obstacle by using a three-dimensional radar carried by the robot, and respectively extracting three-dimensional coordinates of a plurality of obstacle avoidance selectable spaces of the real-time obstacle;
and taking the three-dimensional coordinates of the real-time obstacle and the three-dimensional coordinates of the multiple obstacle avoidance selectable spaces as three-dimensional characteristics of the real-time obstacle.
Obtaining real-time three-dimensional obstacle avoidance parameters of the robot based on three-dimensional features of real-time obstacles by using a robot path obstacle avoidance model, wherein the method comprises the following steps:
inputting three-dimensional characteristics of the real-time obstacle into a robot path obstacle avoidance model, and outputting real-time three-dimensional obstacle avoidance parameters of the robot by the robot path obstacle avoidance model, wherein the real-time three-dimensional obstacle avoidance parameters comprise real-time effective obstacle avoidance space and real-time obstacle avoidance path of the robot.
The three-dimensional feature detection distance of the obstacle of the robot is greater than the safety braking distance of the robot.
When the number of the real-time obstacle avoidance paths is greater than 1, setting evaluation weights of path smoothness and path length, and carrying out weighted summation on the path smoothness and the path length to serve as a selection function for selecting the real-time obstacle avoidance paths, wherein the selection function is as follows:
wherein, jarge is the evaluation value of the real-time obstacle avoidance path,path smoothness evaluation value for real-time obstacle avoidance path, < ->For the real-time path length evaluation value of the obstacle avoidance path, W, V is the evaluation weight of the path smoothness and the path length, w+v=1, W e [0,1 ]],V∈[0,1]The user can set according to the needs, the higher the W setting is, the higher the V setting is, the more the path length of the obstacle avoidance path is, and the real-time obstacle avoidance path with the highest evaluation and more in line with the needs of the user can be selected from the real-time obstacle avoidance paths;
and the robot performs real-time autonomous obstacle avoidance according to the real-time obstacle avoidance path with the highest evaluation value.
Based on the robot path obstacle avoidance method, the invention provides an obstacle avoidance system which comprises a three-dimensional radar and a data processor, wherein the three-dimensional radar is in communication connection with the data processor, the three-dimensional radar is used for detecting three-dimensional characteristics of an obstacle, the data processor is used for receiving the three-dimensional characteristics of the obstacle, three-dimensional obstacle avoidance parameters of a robot are obtained by utilizing a robot path obstacle avoidance model based on the three-dimensional characteristics of the obstacle, and the robot path obstacle avoidance model is arranged in the data processor.
According to the method, the three-dimensional characteristics of the obstacle and the three-dimensional obstacle avoidance parameters of the robot are subjected to model fitting to obtain the robot path obstacle avoidance model representing the mapping relation between the three-dimensional characteristics and the three-dimensional obstacle avoidance parameters, so that the robot can perform real-time autonomous obstacle avoidance according to the real-time three-dimensional obstacle avoidance parameters, and an autonomous obstacle avoidance algorithm based on the depth model has the capability of autonomous learning in an obstacle avoidance scene, and the problems of complicated three-dimensional reconstruction and path planning are avoided.
The above embodiments are only exemplary embodiments of the present application and are not intended to limit the present application, the scope of which is defined by the claims. Various modifications and equivalent arrangements may be made to the present application by those skilled in the art, which modifications and equivalents are also considered to be within the scope of the present application.
Claims (7)
1. The robot path obstacle avoidance method based on the 3D technology is characterized by comprising the following steps of:
step S1, detecting three-dimensional characteristics of an obstacle positioned on a robot travelling path by using a three-dimensional radar carried by the robot, and optimally analyzing three-dimensional obstacle avoidance parameters of the robot according to the three-dimensional characteristics of the obstacle;
s2, performing model fitting on the three-dimensional characteristics of the obstacle and the three-dimensional obstacle avoidance parameters of the robot to obtain a robot path obstacle avoidance model representing the mapping relation between the three-dimensional characteristics and the three-dimensional obstacle avoidance parameters;
step S3, monitoring three-dimensional characteristics of real-time obstacles on a robot travelling path, and obtaining real-time three-dimensional obstacle avoidance parameters of the robot based on the three-dimensional characteristics of the real-time obstacles by utilizing a robot path obstacle avoidance model so as to realize real-time autonomous obstacle avoidance of the robot according to the real-time three-dimensional obstacle avoidance parameters;
the method for detecting three-dimensional characteristics of an obstacle located on a travel path of a robot by using a three-dimensional radar mounted on the robot includes:
detecting three-dimensional coordinates of an obstacle positioned on a travel path of the robot by using a three-dimensional radar carried by the robot, and respectively extracting two boundary coordinates of the obstacle near the two path boundaries of the travel path;
extracting three-dimensional coordinates from two path boundary coordinates to two edge coordinates of an obstacle on a traveling path respectively to serve as three-dimensional coordinates of a plurality of obstacle avoidance selectable spaces, and marking the plurality of obstacle avoidance selectable spaces as an AA 'space and a B' space respectively, wherein A represents a left boundary of the traveling path, A 'represents a left boundary of the obstacle, B' represents a right boundary of the traveling path, and B represents a right boundary of the obstacle;
taking the three-dimensional coordinates of the obstacle and the three-dimensional coordinates of a plurality of obstacle avoidance selectable spaces as three-dimensional features of the obstacle;
the method for optimizing and analyzing the three-dimensional obstacle avoidance parameters of the robot according to the three-dimensional characteristics of the obstacle comprises the following steps:
calculating three-dimensional volumes of the obstacle avoidance selectable spaces by utilizing three-dimensional coordinates of the obstacle avoidance selectable spaces in the three-dimensional features, comparing the three-dimensional volumes of the plurality of obstacle avoidance selectable spaces with the three-dimensional volumes of the robot, wherein,
if the three-dimensional volumes of the AA 'space and the B' space are larger than or equal to the three-dimensional volume of the robot, the AA 'space and the B' space are used as effective obstacle avoidance spaces;
if the three-dimensional volume of the AA ' space is larger than or equal to the three-dimensional volume of the robot and the three-dimensional volume of the B ' space is smaller than the three-dimensional volume of the robot, taking the AA ' space as an effective obstacle avoidance space;
if the three-dimensional volume of the B ' B space is larger than or equal to the three-dimensional volume of the robot and the three-dimensional volume of the AA ' space is smaller than the three-dimensional volume of the robot, taking the B ' B space as an effective obstacle avoidance space;
if the three-dimensional volumes of the AA 'space and the B' space are smaller than the three-dimensional volume of the robot, taking the AA 'space and the B' space as non-effective obstacle avoidance spaces;
when at least one effective obstacle avoidance space exists, setting an optimization function of an obstacle avoidance path in each effective obstacle avoidance space according to the highest path smoothness and the shortest path length, wherein the optimization function of the obstacle avoidance path is as follows:the method comprises the steps of carrying out a first treatment on the surface of the In (1) the->To characterize the optimization function with the highest path smoothness of the obstacle avoidance path,to characterize the shortest path length optimization function of the obstacle avoidance path,p i is the first in the obstacle avoidance pathiThree-dimensional coordinates of the path points, N is the total number of the path points in the obstacle avoidance path,iis a metering constant;
setting a solving constraint condition of an optimization function of the obstacle avoidance path, wherein the solving constraint condition is as follows:or->The method comprises the steps of carrying out a first treatment on the surface of the In the subp i A| isp i Distance to A, |p i A' |isp i Distance to A' |p i B isp i Distance to B, |p i B' |isp i Distance to B ', O represents the center point of the robot, C is the left boundary of the robot, A' is the right boundary of the robot, |OC| is the distance from O to C, |OC ' | is the distance from O to C ' [ AA ] ']Is AA 'space, [ BB ]']Is BB' space; solving based on the optimization functions of the solving constraint conditions and the obstacle avoidance paths to obtain an obstacle avoidance path of an effective obstacle avoidance space;
taking the effective obstacle avoidance space and the obstacle avoidance path as three-dimensional obstacle avoidance parameters of the robot;
the robot path obstacle avoidance model construction comprises the following steps:
taking the three-dimensional characteristics of the obstacle as an input item of a CNN neural network, taking the three-dimensional obstacle avoidance parameters of the robot as an output item of the CNN neural network, and carrying out model training by utilizing the CNN neural network based on the input item and the output item to obtain a robot path obstacle avoidance model, wherein the model expression of the robot path obstacle avoidance model is as follows: out [ Y ] =cnn (In [ X ]); in the formula, out [ Y ] is the three-dimensional obstacle avoidance parameter of the robot, in [ X ] is the three-dimensional characteristic of the obstacle, and CNN is a CNN neural network.
2. The robot path obstacle avoidance method based on 3D technology of claim 1, wherein: and when no effective obstacle avoidance space exists, stopping the robot in situ for early warning.
3. The robot path obstacle avoidance method based on 3D technology of claim 1, wherein: the monitoring of three-dimensional characteristics of real-time obstacles on a path of travel of a robot comprises:
taking an obstacle closest to the robot in the travelling direction on a travelling path as a real-time obstacle, detecting three-dimensional coordinates of the obstacle by using a three-dimensional radar carried by the robot, and respectively extracting three-dimensional coordinates of a plurality of obstacle avoidance selectable spaces of the real-time obstacle;
and taking the three-dimensional coordinates of the real-time obstacle and the three-dimensional coordinates of the multiple obstacle avoidance selectable spaces as three-dimensional characteristics of the real-time obstacle.
4. A robot path obstacle avoidance method based on 3D technology as claimed in claim 3, wherein: the method for obtaining real-time three-dimensional obstacle avoidance parameters of the robot based on the three-dimensional characteristics of the real-time obstacle by using the robot path obstacle avoidance model comprises the following steps:
inputting three-dimensional characteristics of the real-time obstacle into a robot path obstacle avoidance model, and outputting real-time three-dimensional obstacle avoidance parameters of the robot by the robot path obstacle avoidance model, wherein the real-time three-dimensional obstacle avoidance parameters comprise real-time effective obstacle avoidance space and real-time obstacle avoidance path of the robot.
5. The 3D technology-based robot path obstacle avoidance method of claim 4, wherein the three-dimensional feature detection distance of the obstacle of the robot is greater than the safety braking distance of the robot.
6. The 3D technology-based robot path obstacle avoidance method according to claim 5, wherein when the number of real-time obstacle avoidance paths is greater than 1, an evaluation weight of path smoothness and path length is set, and the path smoothness and the path length are weighted and summed as a selection function for selecting the real-time obstacle avoidance paths, wherein the selection function is as follows:the method comprises the steps of carrying out a first treatment on the surface of the Wherein, jarge is the evaluation value of the real-time obstacle avoidance path, < >>Path smoothness evaluation value for real-time obstacle avoidance path, < ->The path length evaluation value W, V for the real-time obstacle avoidance path is the evaluation weight of the path smoothness and the path length respectively;
and the robot performs real-time autonomous obstacle avoidance according to the real-time obstacle avoidance path with the highest evaluation value.
7. The obstacle avoidance system of the 3D technology-based robot path obstacle avoidance method according to any one of claims 1 to 6, comprising a three-dimensional radar and a data processor, wherein the three-dimensional radar is in communication connection with the data processor, the three-dimensional radar is used for detecting three-dimensional characteristics of an obstacle, the data processor is used for receiving the three-dimensional characteristics of the obstacle, and obtaining three-dimensional obstacle avoidance parameters of the robot based on the three-dimensional characteristics of the obstacle by using a robot path obstacle avoidance model, and the robot path obstacle avoidance model is built in the data processor.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211051740.9A CN115454061B (en) | 2022-08-31 | 2022-08-31 | Robot path obstacle avoidance method and system based on 3D technology |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211051740.9A CN115454061B (en) | 2022-08-31 | 2022-08-31 | Robot path obstacle avoidance method and system based on 3D technology |
Publications (2)
Publication Number | Publication Date |
---|---|
CN115454061A CN115454061A (en) | 2022-12-09 |
CN115454061B true CN115454061B (en) | 2024-03-29 |
Family
ID=84301377
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202211051740.9A Active CN115454061B (en) | 2022-08-31 | 2022-08-31 | Robot path obstacle avoidance method and system based on 3D technology |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN115454061B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117762149B (en) * | 2024-02-22 | 2024-05-17 | 本溪钢铁(集团)信息自动化有限责任公司 | Slag dragging robot control method, device, equipment and medium |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106949893A (en) * | 2017-03-24 | 2017-07-14 | 华中科技大学 | The Indoor Robot air navigation aid and system of a kind of three-dimensional avoidance |
CN109116841A (en) * | 2018-07-23 | 2019-01-01 | 昆明理工大学 | A kind of path planning smooth optimization method based on ant group algorithm |
CN109144072A (en) * | 2018-09-30 | 2019-01-04 | 亿嘉和科技股份有限公司 | A kind of intelligent robot barrier-avoiding method based on three-dimensional laser |
CN113031597A (en) * | 2021-03-02 | 2021-06-25 | 南京理工大学 | Autonomous obstacle avoidance method based on deep learning and stereoscopic vision |
CN113759900A (en) * | 2021-08-12 | 2021-12-07 | 中南大学 | Inspection robot track planning and real-time obstacle avoidance method and system based on obstacle area prediction |
-
2022
- 2022-08-31 CN CN202211051740.9A patent/CN115454061B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106949893A (en) * | 2017-03-24 | 2017-07-14 | 华中科技大学 | The Indoor Robot air navigation aid and system of a kind of three-dimensional avoidance |
CN109116841A (en) * | 2018-07-23 | 2019-01-01 | 昆明理工大学 | A kind of path planning smooth optimization method based on ant group algorithm |
CN109144072A (en) * | 2018-09-30 | 2019-01-04 | 亿嘉和科技股份有限公司 | A kind of intelligent robot barrier-avoiding method based on three-dimensional laser |
CN113031597A (en) * | 2021-03-02 | 2021-06-25 | 南京理工大学 | Autonomous obstacle avoidance method based on deep learning and stereoscopic vision |
CN113759900A (en) * | 2021-08-12 | 2021-12-07 | 中南大学 | Inspection robot track planning and real-time obstacle avoidance method and system based on obstacle area prediction |
Non-Patent Citations (6)
Title |
---|
一种基于A*算法的空间多自由度机械臂路径规划方法;宗成星;陆亮;雷新宇;赵萍;;合肥工业大学学报(自然科学版);20170228(第02期);全文 * |
基于人工势场法的AUV避障算法研究综述;郭银景;刘琦;鲍建康;徐锋;吕文红;;计算机工程与应用;20200430(第04期);全文 * |
基于改进遗传算法的移动机器人路径规划;叶晔;岑豫皖;包家汉;;安徽工业大学学报(自然科学版);20071015(第04期);全文 * |
基于模糊改进人工势场法的机器人避障方法研究;游文洋;章政;黄卫华;;传感器与微***;20160131(第01期);全文 * |
多移动机器人编队及避障控制方法研究;陈警;《CNKI优秀硕士学位论文全文库》;20220615;全文 * |
水果采摘机器人自主寻径避障轨迹优化研究―基于启发式智能算法;袁铸;申一歌;;农机化研究;20170701(第07期);全文 * |
Also Published As
Publication number | Publication date |
---|---|
CN115454061A (en) | 2022-12-09 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Ayawli et al. | Mobile robot path planning in dynamic environment using Voronoi diagram and computation geometry technique | |
CN109782779B (en) | AUV path planning method in ocean current environment based on population hyperheuristic algorithm | |
CN111427047B (en) | SLAM method for autonomous mobile robot in large scene | |
CN109947136B (en) | Collaborative active sensing method for unmanned aerial vehicle group rapid target search | |
Mujahad et al. | Closest Gap based (CG) reactive obstacle avoidance navigation for highly cluttered environments | |
CN106873599A (en) | Unmanned bicycle paths planning method based on ant group algorithm and polar coordinate transform | |
Levine et al. | Information-rich path planning with general constraints using rapidly-exploring random trees | |
Saulnier et al. | Information theoretic active exploration in signed distance fields | |
CN109597425A (en) | Navigation of Pilotless Aircraft and barrier-avoiding method based on intensified learning | |
CN114442621A (en) | Autonomous exploration and mapping system based on quadruped robot | |
CN115454061B (en) | Robot path obstacle avoidance method and system based on 3D technology | |
CN116661469B (en) | Robot track error correction method and system | |
CN113391633A (en) | Urban environment-oriented mobile robot fusion path planning method | |
CN109947129A (en) | Rotor wing unmanned aerial vehicle paths planning method based on Dijkstra and improvement particle swarm algorithm | |
Oliveira et al. | Three-dimensional mapping with augmented navigation cost through deep learning | |
CN114879660B (en) | Robot environment sensing method based on target drive | |
CN115639823A (en) | Terrain sensing and movement control method and system for robot under rugged and undulating terrain | |
CN114186112B (en) | Robot navigation method based on Bayesian optimization multiple information gain exploration strategy | |
Amigoni et al. | A mobile robot mapping system with an information-based exploration strategy | |
Yuhang et al. | An Adaptive evolutionary multi-objective estimation of distribution algorithm and its application to multi-UAV path planning | |
CN115826586B (en) | Path planning method and system integrating global algorithm and local algorithm | |
CN111427368A (en) | Improved multi-target collision-prevention driving method for unmanned intelligent vehicle | |
CN115237117B (en) | Mobile robot path planning method based on variation self-adaptive particle swarm algorithm | |
CN108107881A (en) | A kind of new orchard operation machine people's device and method | |
CN116466701A (en) | Unmanned surface vessel energy consumption minimum track planning method |
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 |