CN117705124A - Route planning method of logistics robot - Google Patents
Route planning method of logistics robot Download PDFInfo
- Publication number
- CN117705124A CN117705124A CN202410161212.1A CN202410161212A CN117705124A CN 117705124 A CN117705124 A CN 117705124A CN 202410161212 A CN202410161212 A CN 202410161212A CN 117705124 A CN117705124 A CN 117705124A
- Authority
- CN
- China
- Prior art keywords
- representing
- layer
- convolution
- output
- map
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 51
- 238000005457 optimization Methods 0.000 claims description 27
- 230000004888 barrier function Effects 0.000 claims description 26
- 230000004913 activation Effects 0.000 claims description 22
- 238000009499 grossing Methods 0.000 claims description 22
- 238000004364 calculation method Methods 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 5
- 239000002699 waste material Substances 0.000 abstract description 2
- 238000000605 extraction Methods 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 3
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
- G01C21/20—Instruments for performing navigational calculations
- G01C21/206—Instruments for performing navigational calculations specially adapted for indoor navigation
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Automation & Control Theory (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Manipulator (AREA)
Abstract
The invention discloses a path planning method of a logistics robot, which belongs to the technical field of path planning and comprises the following steps: s1, acquiring a map of a logistics warehouse, and determining initial position coordinates and storage position coordinates of logistics packages in the map; s2, determining an ideal logistics robot according to initial position coordinates of the logistics packages in the map, and generating an initial distribution path for the ideal logistics robot; and S3, optimizing the initial distribution path according to the initial position coordinates and the storage position coordinates of the logistics packages in the map, and generating an optimized distribution path for the ideal logistics robot. The route planning method of the logistics robot reduces turning inflection points of the ideal logistics robot as much as possible, reduces resource waste and improves distribution efficiency.
Description
Technical Field
The invention belongs to the technical field of path planning, and particularly relates to a path planning method of a logistics robot.
Background
At present, the logistics warehouse is a place where electronic commerce and logistics companies store a large amount of packages to be sent and stored, and with the rapid growth of express business in China, more and more packages are sent to a logistics transfer station, and more human capital is needed. However, with the development of artificial intelligence, the intelligent logistics robot gradually enters the field of vision of the masses, logistics work is gradually combined with the intelligent robot, the logistics robot for transportation at present adopts a manually preset route for delivery, real-time optimization cannot be performed according to the actual warehouse condition, and therefore excessive resource consumption can be caused in the intelligent carrying process of the robot, and the efficiency is reduced.
Disclosure of Invention
The invention provides a path planning method of a logistics robot in order to solve the problems.
The technical scheme of the invention is as follows: the path planning method of the logistics robot comprises the following steps:
s1, acquiring a map of a logistics warehouse, and determining initial position coordinates and storage position coordinates of logistics packages in the map;
s2, determining an ideal logistics robot according to initial position coordinates of the logistics packages in the map, and generating an initial distribution path for the ideal logistics robot;
and S3, optimizing the initial distribution path according to the initial position coordinates and the storage position coordinates of the logistics packages in the map, and generating an optimized distribution path for the ideal logistics robot.
Initial position coordinates, namely the starting point of the logistics robot path planning; and storing position coordinates, namely, the end point of the logistics robot path planning.
Further, in S2, an initial delivery path is generated for the ideal logistics robot using a free space method.
In S2, the free logistics robot closest to the initial position coordinate travel distance is used as the ideal logistics robot. The free space method adopts a predefined basic shape to construct a free space, the free space is expressed as a connected graph, and then path planning is carried out by searching the graph, but the disadvantage is that the complexity of an algorithm is increased when a plurality of obstacles are arranged, and the planned path is not necessarily optimal. One of the characteristics of the logistics warehouse is that the barriers are more, so that the method disclosed by the invention corrects and optimizes the initial distribution path generated by adopting the existing algorithm, and generates the optimal distribution path for logistics packages.
Further, S3 comprises the following sub-steps:
s31, acquiring all barriers in an initial distribution path, and determining transport penalty coefficients of all the barriers;
s32, determining a first obstacle set to be optimized of the initial distribution path according to the transportation penalty coefficients of the obstacles;
s33, determining a second obstacle set to be optimized in the initial distribution path;
s34, constructing a path optimization model, inputting the first obstacle set to be optimized and the second obstacle set to be optimized into the path optimization model, and generating an optimized distribution path for the ideal logistics robot.
The beneficial effects of the above-mentioned further scheme are: in the invention, a plurality of obstacles exist in an initial distribution path, an ideal logistics robot turns (namely turns) when passing through the obstacles, and the logistics robot turns slower, so that the transportation time is increased, and the transportation cost is increased. The invention builds a path optimization model, and performs smooth operation on turning points of the ideal logistics robot in the initial distribution path to complete path optimization.
Further, in S31, the transport penalty coefficient p for the ith obstacle i The calculation formula of (2) is as follows:the method comprises the steps of carrying out a first treatment on the surface of the Wherein alpha is i Represents the turning angle, x of an ideal logistics robot passing through the ith obstacle i Representing the abscissa of the ith obstacle on the map, y i Representing the ordinate, x of the ith obstacle on the map i-1 Representing the i-1 th obstacle on the abscissa of the map, y i-1 Representing the i-1 th obstacle on the ordinate, x of the map i+1 Representing the (i+1) th obstacle on the abscissa, y of the map i+1 Representing the (i+1) th obstacle on the ordinate, X of the map 0 Represents the abscissa of the initial position of the logistic package in the map, Y 0 Representing the ordinate, X of the initial position of the logistic package in the map 1 Representing the abscissa of the storage position of the logistic package in the map, Y 1 Representing the storage position of a physical distribution package in a mapCoordinates, R i Representing the transport costs of an ideal logistics robot past the ith obstacle, < >>Representing an upward rounding.
Further, in S32, the specific method for determining the first obstacle set to be optimized is as follows: taking all barriers with transport penalty coefficients smaller than the transport penalty threshold as a first barrier set to be optimized;
the transport penalty threshold may be set manually or may be a mean of transport penalty coefficients for all obstacles.
In S33, the specific method for determining the second set of obstacles to be optimized is as follows: and taking all the obstacles with the turn times larger than 1 except the first obstacle set to be optimized in the initial distribution path as a second obstacle set to be optimized.
Further, the path optimization model comprises a first convolution layer, a second convolution layer, a first activation layer, a second activation layer and a smoothing layer;
the input end of the first convolution layer is used as a first input end of the path optimization model; the input end of the second convolution layer is used as a second input end of the path optimization model; the output end of the first convolution layer, the first activation layer and the first input end of the smoothing layer are sequentially connected; the output end of the second convolution layer, the second activation layer and the second input end of the smoothing layer are sequentially connected; the first output end of the smoothing layer is used as the first output end of the path optimization model; the second output of the smoothing layer serves as the second output of the path optimization model.
The beneficial effects of the above-mentioned further scheme are: in the invention, a first convolution layer performs feature extraction on a first obstacle set to be optimized, a plurality of feature values are output, and a second convolution layer performs feature extraction on a second obstacle set to be optimized, and a plurality of feature values are output. The first activation layer operates the maximum/small characteristic value output by the first convolution layer, and the second activation layer operates the maximum/small characteristic value output by the second convolution layer. The smoothing layer determines a final smoothing angle based on the outputs of the first and second activation layers. When the ideal logistics robot passes through the obstacle, the smooth angle can be directly adopted, or the smooth angle can be modified based on the turning angle of the ideal logistics robot.
Further, the expression of the output U of the first convolution layer is: u= { U 1 ,…,u m ,…,u M }，The method comprises the steps of carrying out a first treatment on the surface of the Wherein u is 1 Representing the output of a first convolution kernel in a first convolution layer, u m Representing the output of the mth convolution kernel in the first convolution layer, u M Represents the output of the M th convolution kernel in the first convolution layer, M represents the number of convolution kernels of the first convolution layer, and p m A transport penalty coefficient, p, representing the barrier to be treated by the mth convolution kernel in the first convolution layer m-1 A transport penalty coefficient, p, representing the corresponding processing barrier of the m-1 th convolution kernel in the first convolution layer m+1 A transport penalty coefficient, gamma, representing the corresponding processing barrier of the (m+1) th convolution kernel in the first convolution layer m The weight, b, representing the mth convolution kernel in the first convolution layer m Representing the size of the mth convolution kernel in the first convolution layer;
the output V of the second convolution layer is expressed as: v= { v 1 ,…,v n ,…,v N }，The method comprises the steps of carrying out a first treatment on the surface of the In the formula, v 1 Representing the output of the first convolution kernel in the second convolution layer, v n Representing the output of the nth convolution kernel in the second convolution layer, u N Representing the output of the nth convolution kernel in the second convolution layer, N representing the number of convolution kernels of the second convolution layer, c n Indicating the turn number of the barrier to be processed by the nth convolution kernel in the second convolution layer, gamma n The weight, b, representing the nth convolution kernel in the second convolution layer n Representing the size of the nth convolution kernel in the second convolution layer.
Further, the output J of the first active layer 1 The expression of (2) is:the method comprises the steps of carrying out a first treatment on the surface of the Wherein e represents an exponent, max (. Cndot.) represents a maximum value calculation, β 1 Representing the bias of the first active layer, U representing the output of the first convolutional layer, σ (·) representing the activation function, c representing a constant;
output J of second active layer 2 The expression of (2) is:the method comprises the steps of carrying out a first treatment on the surface of the Wherein beta is 2 Representing the bias of the second active layer, V represents the output of the second convolutional layer.
Further, the expression of the output W of the smoothing layer is:the method comprises the steps of carrying out a first treatment on the surface of the Wherein J is 1 Representing the output of the first active layer, J 2 Representing the output of the second active layer, min (·) represents the minimum operation.
The beneficial effects of the invention are as follows: according to the route planning method of the logistics robot, an ideal logistics robot is distributed to complete a distribution task according to the initial position of a map, and an initial distribution route is configured for the ideal logistics robot; in consideration of the characteristic of more barriers of the logistics warehouse, the initial distribution path is corrected and optimized, turning inflection points of an ideal logistics robot are reduced as much as possible, resource waste is reduced, and distribution efficiency is improved.
Drawings
FIG. 1 is a flow chart of a path planning method of a logistics robot;
fig. 2 is a schematic diagram of the structure of the path optimization model.
Detailed Description
Embodiments of the present invention are further described below with reference to the accompanying drawings.
As shown in fig. 1, the present invention provides a path planning method for a logistics robot, which includes the following steps:
s1, acquiring a map of a logistics warehouse, and determining initial position coordinates and storage position coordinates of logistics packages in the map;
s2, determining an ideal logistics robot according to initial position coordinates of the logistics packages in the map, and generating an initial distribution path for the ideal logistics robot;
and S3, optimizing the initial distribution path according to the initial position coordinates and the storage position coordinates of the logistics packages in the map, and generating an optimized distribution path for the ideal logistics robot.
Initial position coordinates, namely the starting point of the logistics robot path planning; and storing position coordinates, namely, the end point of the logistics robot path planning.
In the embodiment of the present invention, in S2, an initial distribution path is generated for an ideal logistics robot by using a free space method.
In S2, the free logistics robot closest to the initial position coordinate travel distance is used as the ideal logistics robot. The free space method adopts a predefined basic shape to construct a free space, the free space is expressed as a connected graph, and then path planning is carried out by searching the graph, but the disadvantage is that the complexity of an algorithm is increased when a plurality of obstacles are arranged, and the planned path is not necessarily optimal. One of the characteristics of the logistics warehouse is that the barriers are more, so that the method disclosed by the invention corrects and optimizes the initial distribution path generated by adopting the existing algorithm, and generates the optimal distribution path for logistics packages.
In an embodiment of the present invention, S3 comprises the following sub-steps:
s31, acquiring all barriers in an initial distribution path, and determining transport penalty coefficients of all the barriers;
s32, determining a first obstacle set to be optimized of the initial distribution path according to the transportation penalty coefficients of the obstacles;
s33, determining a second obstacle set to be optimized in the initial distribution path;
s34, constructing a path optimization model, inputting the first obstacle set to be optimized and the second obstacle set to be optimized into the path optimization model, and generating an optimized distribution path for the ideal logistics robot.
In the invention, a plurality of obstacles exist in an initial distribution path, an ideal logistics robot turns (namely turns) when passing through the obstacles, and the logistics robot turns slower, so that the transportation time is increased, and the transportation cost is increased. The invention builds a path optimization model, and performs smooth operation on turning points of the ideal logistics robot in the initial distribution path to complete path optimization.
In the embodiment of the present invention, in S31, the transport penalty coefficient p of the i-th obstacle i The calculation formula of (2) is as follows:the method comprises the steps of carrying out a first treatment on the surface of the Wherein alpha is i Represents the turning angle, x of an ideal logistics robot passing through the ith obstacle i Representing the abscissa of the ith obstacle on the map, y i Representing the ordinate, x of the ith obstacle on the map i-1 Representing the i-1 th obstacle on the abscissa of the map, y i-1 Representing the i-1 th obstacle on the ordinate, x of the map i+1 Representing the (i+1) th obstacle on the abscissa, y of the map i+1 Representing the (i+1) th obstacle on the ordinate, X of the map 0 Represents the abscissa of the initial position of the logistic package in the map, Y 0 Representing the ordinate, X of the initial position of the logistic package in the map 1 Representing the abscissa of the storage position of the logistic package in the map, Y 1 Representing the ordinate of the storage position of the logistics package in the map, R i Representing the transport costs of an ideal logistics robot past the ith obstacle, < >>Representing an upward rounding.
In the embodiment of the present invention, in S32, a specific method for determining the first to-be-optimized obstacle set is as follows: taking all barriers with transport penalty coefficients smaller than the transport penalty threshold as a first barrier set to be optimized;
the transport penalty threshold may be set manually or may be a mean of transport penalty coefficients for all obstacles.
In S33, the specific method for determining the second set of obstacles to be optimized is as follows: and taking all the obstacles with the turn times larger than 1 except the first obstacle set to be optimized in the initial distribution path as a second obstacle set to be optimized.
In the embodiment of the present invention, as shown in fig. 2, the path optimization model includes a first convolution layer, a second convolution layer, a first activation layer, a second activation layer, and a smoothing layer;
the input end of the first convolution layer is used as a first input end of the path optimization model; the input end of the second convolution layer is used as a second input end of the path optimization model; the output end of the first convolution layer, the first activation layer and the first input end of the smoothing layer are sequentially connected; the output end of the second convolution layer, the second activation layer and the second input end of the smoothing layer are sequentially connected; the first output end of the smoothing layer is used as the first output end of the path optimization model; the second output of the smoothing layer serves as the second output of the path optimization model.
In the invention, a first convolution layer performs feature extraction on a first obstacle set to be optimized, a plurality of feature values are output, and a second convolution layer performs feature extraction on a second obstacle set to be optimized, and a plurality of feature values are output. The first activation layer operates the maximum/small characteristic value output by the first convolution layer, and the second activation layer operates the maximum/small characteristic value output by the second convolution layer. The smoothing layer determines a final smoothing angle based on the outputs of the first and second activation layers. When the ideal logistics robot passes through the obstacle, the smooth angle can be directly adopted, or the smooth angle can be modified based on the turning angle of the ideal logistics robot.
In the embodiment of the present invention, the expression of the output U of the first convolution layer is: u= { U 1 ,…,u m ,…,u M }，The method comprises the steps of carrying out a first treatment on the surface of the Wherein u is 1 Representing the output of a first convolution kernel in a first convolution layer, u m Representing the output of the mth convolution kernel in the first convolution layer, u M Represents the output of the M th convolution kernel in the first convolution layer, M represents the number of convolution kernels of the first convolution layer, and p m A transport penalty coefficient, p, representing the barrier to be treated by the mth convolution kernel in the first convolution layer m-1 Representing the movement of the m-1 th convolution check in the first convolution layer to deal with the obstacleInput penalty coefficient, p m+1 A transport penalty coefficient, gamma, representing the corresponding processing barrier of the (m+1) th convolution kernel in the first convolution layer m The weight, b, representing the mth convolution kernel in the first convolution layer m Representing the size of the mth convolution kernel in the first convolution layer;
the output V of the second convolution layer is expressed as: v= { v 1 ,…,v n ,…,v N }，The method comprises the steps of carrying out a first treatment on the surface of the In the formula, v 1 Representing the output of the first convolution kernel in the second convolution layer, v n Representing the output of the nth convolution kernel in the second convolution layer, u N Representing the output of the nth convolution kernel in the second convolution layer, N representing the number of convolution kernels of the second convolution layer, c n Indicating the turn number of the barrier to be processed by the nth convolution kernel in the second convolution layer, gamma n The weight, b, representing the nth convolution kernel in the second convolution layer n Representing the size of the nth convolution kernel in the second convolution layer.
In an embodiment of the invention, the output J of the first active layer 1 The expression of (2) is:the method comprises the steps of carrying out a first treatment on the surface of the Wherein e represents an exponent, max (. Cndot.) represents a maximum value calculation, β 1 Representing the bias of the first active layer, U representing the output of the first convolutional layer, σ (·) representing the activation function, c representing a constant;
output J of second active layer 2 The expression of (2) is:the method comprises the steps of carrying out a first treatment on the surface of the Wherein beta is 2 Representing the bias of the second active layer, V represents the output of the second convolutional layer.
In the embodiment of the present invention, the expression of the output W of the smoothing layer is:the method comprises the steps of carrying out a first treatment on the surface of the Wherein J is 1 Representing the output of the first active layer, J 2 Representing the output of the second active layer, min (·) represents the minimum operation.
Those of ordinary skill in the art will recognize that the embodiments described herein are for the purpose of aiding the reader in understanding the principles of the present invention and should be understood that the scope of the invention is not limited to such specific statements and embodiments. Those of ordinary skill in the art can make various other specific modifications and combinations from the teachings of the present disclosure without departing from the spirit thereof, and such modifications and combinations remain within the scope of the present disclosure.
Claims (9)
1. The path planning method of the logistics robot is characterized by comprising the following steps of:
s1, acquiring a map of a logistics warehouse, and determining initial position coordinates and storage position coordinates of logistics packages in the map;
s2, determining an ideal logistics robot according to initial position coordinates of the logistics packages in the map, and generating an initial distribution path for the ideal logistics robot;
and S3, optimizing the initial distribution path according to the initial position coordinates and the storage position coordinates of the logistics packages in the map, and generating an optimized distribution path for the ideal logistics robot.
2. The route planning method of a logistics robot of claim 1, wherein in S2, an initial distribution route is generated for the ideal logistics robot using a free space method.
3. The path planning method of a logistics robot of claim 1, wherein the S3 comprises the sub-steps of:
s31, acquiring all barriers in an initial distribution path, and determining transport penalty coefficients of all the barriers;
s32, determining a first obstacle set to be optimized of the initial distribution path according to the transportation penalty coefficients of the obstacles;
s33, determining a second obstacle set to be optimized in the initial distribution path;
s34, constructing a path optimization model, inputting the first obstacle set to be optimized and the second obstacle set to be optimized into the path optimization model, and generating an optimized distribution path for the ideal logistics robot.
4. A route planning method of a logistic robot according to claim 3, wherein in S31, the transport penalty coefficient p of the ith obstacle i The calculation formula of (2) is as follows:the method comprises the steps of carrying out a first treatment on the surface of the Wherein alpha is i Represents the turning angle, x of an ideal logistics robot passing through the ith obstacle i Representing the abscissa of the ith obstacle on the map, y i Representing the ordinate, x of the ith obstacle on the map i-1 Representing the i-1 th obstacle on the abscissa of the map, y i-1 Representing the i-1 th obstacle on the ordinate, x of the map i+1 Representing the (i+1) th obstacle on the abscissa, y of the map i+1 Representing the (i+1) th obstacle on the ordinate, X of the map 0 Represents the abscissa of the initial position of the logistic package in the map, Y 0 Representing the ordinate, X of the initial position of the logistic package in the map 1 Representing the abscissa of the storage position of the logistic package in the map, Y 1 Representing the ordinate of the storage position of the logistics package in the map, R i Representing the transport costs of an ideal logistics robot past the ith obstacle, < >>Representing an upward rounding.
5. The path planning method of the logistic robot according to claim 3, wherein in S32, the specific method for determining the first set of obstacles to be optimized is as follows: taking all barriers with transport penalty coefficients smaller than the transport penalty threshold as a first barrier set to be optimized;
in S33, the specific method for determining the second set of obstacles to be optimized includes: and taking all the obstacles with the turn times larger than 1 except the first obstacle set to be optimized in the initial distribution path as a second obstacle set to be optimized.
6. A method of path planning for a logistics robot as claimed in claim 3, wherein the path optimization model comprises a first convolution layer, a second convolution layer, a first activation layer, a second activation layer and a smoothing layer;
the input end of the first convolution layer is used as a first input end of a path optimization model; the input end of the second convolution layer is used as a second input end of the path optimization model; the output end of the first convolution layer, the first activation layer and the first input end of the smoothing layer are sequentially connected; the output end of the second convolution layer, the second activation layer and the second input end of the smoothing layer are sequentially connected; the first output end of the smoothing layer is used as the first output end of the path optimization model; and the second output end of the smoothing layer is used as the second output end of the path optimization model.
7. The path planning method of a logistics robot of claim 6, wherein the expression of the output U of the first convolution layer is: u= { U 1 ,…,u m ,…,u M }，The method comprises the steps of carrying out a first treatment on the surface of the Wherein u is 1 Representing the output of a first convolution kernel in a first convolution layer, u m Representing the output of the mth convolution kernel in the first convolution layer, u M Represents the output of the M th convolution kernel in the first convolution layer, M represents the number of convolution kernels of the first convolution layer, and p m A transport penalty coefficient, p, representing the barrier to be treated by the mth convolution kernel in the first convolution layer m-1 A transport penalty coefficient, p, representing the corresponding processing barrier of the m-1 th convolution kernel in the first convolution layer m+1 A transport penalty coefficient, gamma, representing the corresponding processing barrier of the (m+1) th convolution kernel in the first convolution layer m The weight, b, representing the mth convolution kernel in the first convolution layer m Representing the size of the mth convolution kernel in the first convolution layer;
the expression of the output V of the second convolution layer is: v= { v 1 ,…,v n ,…,v N }，The method comprises the steps of carrying out a first treatment on the surface of the In the formula, v 1 Representing the output of the first convolution kernel in the second convolution layer, v n Representing the output of the nth convolution kernel in the second convolution layer, u N Representing the output of the nth convolution kernel in the second convolution layer, N representing the number of convolution kernels of the second convolution layer, c n Indicating the turn number of the barrier to be processed by the nth convolution kernel in the second convolution layer, gamma n The weight, b, representing the nth convolution kernel in the second convolution layer n Representing the size of the nth convolution kernel in the second convolution layer.
8. The path planning method of a logistics robot of claim 6, wherein the output J of the first activation layer 1 The expression of (2) is:the method comprises the steps of carrying out a first treatment on the surface of the Wherein e represents an exponent, max (. Cndot.) represents a maximum value calculation, β 1 Representing the bias of the first active layer, U representing the output of the first convolutional layer, σ (·) representing the activation function, c representing a constant;
the output J of the second active layer 2 The expression of (2) is:the method comprises the steps of carrying out a first treatment on the surface of the Wherein beta is 2 Representing the bias of the second active layer, V represents the output of the second convolutional layer.
9. The path planning method of a logistics robot of claim 6, wherein the expression of the output W of the smoothing layer is:the method comprises the steps of carrying out a first treatment on the surface of the Wherein J is 1 Representing the output of the first active layer, J 2 Represents the output of the second active layer, min (·) represents the minimumAnd (5) value operation.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410161212.1A CN117705124B (en) | 2024-02-05 | 2024-02-05 | Route planning method of logistics robot |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410161212.1A CN117705124B (en) | 2024-02-05 | 2024-02-05 | Route planning method of logistics robot |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117705124A true CN117705124A (en) | 2024-03-15 |
| CN117705124B CN117705124B (en) | 2024-05-03 |
Family
ID=90144632
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410161212.1A Active CN117705124B (en) | 2024-02-05 | 2024-02-05 | Route planning method of logistics robot |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117705124B (en) |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150227140A1 (en) * | 2014-02-10 | 2015-08-13 | John Bean Technologies Corporation | Routing of automated guided vehicles |
| US20180364719A1 (en) * | 2017-06-15 | 2018-12-20 | Sap Se | Dynamic Layout Management for Robotics Warehouse System |
| 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 |
| US20210027088A1 (en) * | 2018-03-21 | 2021-01-28 | Guangzhou Xaircraft Technology Co.,Ltd. | Method and device for acquiring boundary of area to be operated, and method for planning operation route |
| CN112835333A (en) * | 2020-12-31 | 2021-05-25 | 北京工商大学 | Multi-AGV obstacle avoidance and path planning method and system based on deep reinforcement learning |
| CN113870246A (en) * | 2021-10-13 | 2021-12-31 | 广东新时空科技股份有限公司 | Obstacle detection and identification method based on deep learning |
| CN114166235A (en) * | 2021-12-06 | 2022-03-11 | 福建工程学院 | Global dynamic smooth path planning method based on optimized A-STAR algorithm |
| CN115877853A (en) * | 2023-03-03 | 2023-03-31 | 天津牛磨王科技有限公司 | Intelligent storage flow path planning system and method |
| CN117058235A (en) * | 2023-08-15 | 2023-11-14 | 哈尔滨工业大学 | Visual positioning method crossing various indoor scenes |
| CN117093009A (en) * | 2023-10-19 | 2023-11-21 | 湖南睿图智能科技有限公司 | Logistics AGV trolley navigation control method and system based on machine vision |
-
2024
- 2024-02-05 CN CN202410161212.1A patent/CN117705124B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150227140A1 (en) * | 2014-02-10 | 2015-08-13 | John Bean Technologies Corporation | Routing of automated guided vehicles |
| US20180364719A1 (en) * | 2017-06-15 | 2018-12-20 | Sap Se | Dynamic Layout Management for Robotics Warehouse System |
| US20210027088A1 (en) * | 2018-03-21 | 2021-01-28 | Guangzhou Xaircraft Technology Co.,Ltd. | Method and device for acquiring boundary of area to be operated, and method for planning operation route |
| 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 |
| CN112835333A (en) * | 2020-12-31 | 2021-05-25 | 北京工商大学 | Multi-AGV obstacle avoidance and path planning method and system based on deep reinforcement learning |
| CN113870246A (en) * | 2021-10-13 | 2021-12-31 | 广东新时空科技股份有限公司 | Obstacle detection and identification method based on deep learning |
| CN114166235A (en) * | 2021-12-06 | 2022-03-11 | 福建工程学院 | Global dynamic smooth path planning method based on optimized A-STAR algorithm |
| CN115877853A (en) * | 2023-03-03 | 2023-03-31 | 天津牛磨王科技有限公司 | Intelligent storage flow path planning system and method |
| CN117058235A (en) * | 2023-08-15 | 2023-11-14 | 哈尔滨工业大学 | Visual positioning method crossing various indoor scenes |
| CN117093009A (en) * | 2023-10-19 | 2023-11-21 | 湖南睿图智能科技有限公司 | Logistics AGV trolley navigation control method and system based on machine vision |
Non-Patent Citations (4)
| Title |
|---|
| SANKUI SUN; CHONGLIN GU; QIAN WAN; HEJIAO HUANG; XIAOHUA JIA: "CROTPN Based Collision-Free and Deadlock-Free Path Planning of AGVs in Logistic Center", 2018 15TH INTERNATIONAL CONFERENCE ON CONTROL, AUTOMATION, ROBOTICS AND VISION, 20 December 2018 (2018-12-20) * |
| 刘毅;: "移动机器人路径规划中的仿真研究", 计算机仿真, vol. 28, no. 06, 30 June 2011 (2011-06-30) * |
| 李立, 魏晓晨: "云环境下的仓储物流机器人路径规划方法研究", 机械设计与制造工程, vol. 52, no. 7, 31 July 2023 (2023-07-31) * |
| 陈至坤;郭宝军;王淑香;: "移动机器人目标路径规划的仿真研究", 计算机仿真, vol. 33, no. 05, 31 May 2016 (2016-05-31) * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117705124B (en) | 2024-05-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110135580B (en) | Convolution network full integer quantization method and application method thereof | |
| CN110930454B (en) | Six-degree-of-freedom pose estimation algorithm based on boundary box outer key point positioning | |
| WO2019196223A1 (en) | Acceleration method and accelerator used for convolutional neural network | |
| CN109214353B (en) | Training method and device for rapid detection of face image based on pruning model | |
| CN111583094B (en) | Image pulse coding method and system based on FPGA | |
| CN115329683B (en) | Aviation luggage online loading planning method, device, equipment and medium | |
| CN112330719A (en) | Deep learning target tracking method based on feature map segmentation and adaptive fusion | |
| CN111144556A (en) | Hardware circuit of range batch processing normalization algorithm for deep neural network training and reasoning | |
| CN112580662A (en) | Method and system for recognizing fish body direction based on image features | |
| US20230351180A1 (en) | Quantization recognition training method of neural network that supplements limitations of gradient-based learning by adding gradient-indipendent update | |
| CN117705124B (en) | Route planning method of logistics robot | |
| CN104504177A (en) | Method for quickly configuring and designing large container crane | |
| CN114841611A (en) | Method for solving job shop scheduling based on improved ocean predator algorithm | |
| CN110275868A (en) | A kind of multi-modal pretreated method of manufaturing data in intelligent plant | |
| CN114201860A (en) | Cigarette raw material warehouse optimal layout method based on group intelligence | |
| CN117391096A (en) | Transformer encoder hardware accelerator based on FPGA and calculation method thereof | |
| CN110531340B (en) | Identification processing method of laser radar point cloud data based on deep learning | |
| CN117361013A (en) | Multi-machine shelf storage scheduling method based on deep reinforcement learning | |
| CN116543291A (en) | Method for realizing CNN (customer premise network) by using FPGA (field programmable gate array) with flexible resource configuration | |
| CN109141438B (en) | Forklift global path planning method | |
| CN113570036A (en) | Hardware accelerator architecture supporting dynamic neural network sparse model | |
| CN113159301B (en) | Image processing method based on binarization quantization model | |
| TWI793951B (en) | Model training method and model training system | |
| US20230114610A1 (en) | Neural network construction method and apparatus having average quantization mechanism | |
| CN117236821B (en) | Online three-dimensional boxing method based on hierarchical reinforcement learning |
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 |