CN117516548B - Path planning method for autonomous mobile robot - Google Patents

Path planning method for autonomous mobile robot Download PDF

Info

Publication number
CN117516548B
CN117516548B CN202311839597.4A CN202311839597A CN117516548B CN 117516548 B CN117516548 B CN 117516548B CN 202311839597 A CN202311839597 A CN 202311839597A CN 117516548 B CN117516548 B CN 117516548B
Authority
CN
China
Prior art keywords
grid
path
potential
population
free
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
Application number
CN202311839597.4A
Other languages
Chinese (zh)
Other versions
CN117516548A (en
Inventor
卢旭
周程
刘军
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Guangdong Polytechnic Normal University
Original Assignee
Guangdong Polytechnic Normal University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Guangdong Polytechnic Normal University filed Critical Guangdong Polytechnic Normal University
Priority to CN202311839597.4A priority Critical patent/CN117516548B/en
Publication of CN117516548A publication Critical patent/CN117516548A/en
Application granted granted Critical
Publication of CN117516548B publication Critical patent/CN117516548B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

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

Landscapes

  • Engineering & Computer Science (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Automation & Control Theory (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)

Abstract

The invention discloses a path planning method of an autonomous mobile robot, which comprises the following steps: scanning a working scene to construct a grid map; setting a potential descriptor on a grid map, calculating free grid potential classification and marking a potential grid; performing k-means clustering on the potential grids according to the distribution positions of the potential grids to obtain alternative clusters; calibrating task points on the grid map according to the alternative clusters and preset task requirements; setting an initial position of the autonomous mobile robot, and carrying out optimizing search on the task point access sequence based on an improved genetic algorithm to generate an optimal access sequence; using ant colony algorithm to segment and plan paths and combining the segmented paths to obtain an optimal combined path; performing curve fitting by using a quadratic uniform B spline curve method to output a final path; and the autonomous mobile robot moves to the task point according to the final path and executes the task. The invention effectively improves the autonomous capacity of the autonomous mobile robot for executing tasks, shortens the path planning time and improves the working efficiency of the autonomous mobile robot.

Description

Path planning method for autonomous mobile robot
Technical Field
The invention belongs to the technical field of robot movement and path planning, and particularly relates to a path planning method of an autonomous mobile robot.
Background
With the rapid development of science and technology, autonomous mobile robots are becoming an indispensable technical assistant in various fields. Such robots have the ability to independently generate a movement path and perform tasks, and path planning technology is one of the keys. The path planning is the basis for successfully executing tasks by the autonomous mobile robot, and relates to the process of reasonably selecting a moving path by the robot according to environment information, so that the target position is reached in an optimal mode.
The current autonomous mobile robot path planning step often comprises a plurality of stages of sensing environment, generating candidate paths, evaluating path feasibility and the like, which puts higher requirements on calculation resources of the robot, consumes a large amount of calculation resources of the robot, and limits the flexibility and response speed of the robot in actual work. At the same time, there is room for improvement in the autonomous capabilities of the robot in performing tasks. Despite significant progress in sensing, planning and executing tasks, there is still little room for improvement in terms of autonomy of task point selection.
Disclosure of Invention
The invention aims to overcome the defects and shortcomings of the prior art, and provides a path planning method of an autonomous mobile robot, which can effectively improve the autonomous capacity of the autonomous mobile robot for executing tasks, and shorten the path planning time, so that the working efficiency of the autonomous mobile robot is improved.
In order to achieve the above purpose, the present invention adopts the following technical scheme:
a path planning method of an autonomous mobile robot, comprising the steps of:
s1, scanning a working scene to construct a grid map, wherein the grid map comprises an obstacle grid and a free grid;
s2, setting a potential descriptor on a grid map, calculating the potential of the free grid, classifying and marking potential grids; the potential descriptor is a grid matrix centered on a free grid;
s3, performing k-means clustering on the potential grids according to the distribution positions of the potential grids to obtain alternative clusters;
s4, calibrating task points on the grid map according to the alternative clusters and preset task requirements;
s5, setting an initial position of the autonomous mobile robot, and carrying out optimizing search on the task point access sequence based on an improved genetic algorithm to generate an optimal access sequence;
s6, segmenting and planning paths by using an ant colony algorithm according to the optimal access sequence and combining the segmented paths to obtain an optimal combined path;
s7, performing curve fitting on the optimal combined path by using a quadratic uniform B spline curve method, and outputting a final path;
and S8, the autonomous mobile robot moves to a task point according to the final path and executes the task.
As a preferred technical solution, the classifying and marking the potential grid specifically comprises:
s201, setting potential descriptors on a grid map, wherein the potential descriptors are used for describing distribution conditions of barrier grids around the grids; the potential descriptor is a grid square matrix with the size of 5 multiplied by 5 by taking a free grid i as a center; setting potential descriptors on free grids at the boundaries of the grid map, and setting the excess parts of the potential descriptors to be free grids if the potential descriptors exceed the grid map; all grids adjacent to the free grid i in the potential descriptor form a first layer; all grids adjacent to the first layer except the free grid i constitute a second layer;
s202, calculating potential D of the free grid marked as potential grid according to potential descriptor of the free grid cap The calculation formula is as follows:
，
wherein, gamma 1 And gamma 2 Is a weight coefficient; c 1 The number of free grids of the first layer in the potential descriptor of the free grid i; c 2 The number of free grids of the second layer in the potential descriptor of the free grid i; s is(s) 1 The number of adjacent barrier grids of the first layer in the potential descriptor of the free grid i; s is(s) 2 The number of adjacent barrier grids of the second layer in the potential descriptor of the free grid i; s is(s) 3 The number of barrier grids of the first layer and barrier grids of the second layer in the potential descriptor of the free grid i are adjacent;
s203 potential D according to the free grid cap Dividing all the free grids in the grid map according to a threshold epsilon, marking potential grids by using tags, and generating a potential grid set; the marking method comprises the following steps:
，
wherein D is cap (i) Potential D for free grid i cap 1 indicates that the free grid i is marked as a potential grid, and 0 indicates that the free grid i is not marked as a potential grid.
As a preferable technical solution, k-means clustering is performed on the potential grids according to the distribution positions of the potential grids to obtain alternative clusters, which specifically are:
s301, determining the number k of clustering clusters by using an elbow method;
s302, setting a grid distance function, wherein the grid distance function is expressed as:
，
wherein alpha and beta are weight coefficients; (x) i ,y i ) Is the coordinates of the free grid i; (x) j ,y j ) Is the coordinates of the free grid j; c (C) i And S is i The number of free grids and the number of barrier grids in the free grid i potential descriptor are respectively; c (C) j And S is j The number of free grids and the number of barrier grids in the free grid j potential descriptor are respectively;
s303, calculating the distance between potential grids according to the coordinates of each potential grid, and clustering by using a k-means clustering algorithm to obtain k candidate clusters.
As a preferable technical solution, the calibrating task points on the grid map according to the candidate cluster and the preset task requirement specifically includes:
s401, judging whether a preset task requirement exists or not; if so, jumping to S402; otherwise, jumping to S403; the preset task demands comprise key work areas and work complexity;
s402, determining whether an alternative cluster exists in a key work area in a preset task demand; if the candidate clusters exist, marking the candidate clusters as key clusters; if the free grids do not exist, dividing the free grids in the range of the key work area into key clusters;
s403, randomly selecting a free grid from each alternative cluster as a task point; if the key clusters exist, selecting a free grid which is positioned at or approximately at the geometric center of the key clusters in each key cluster as a task point, and marking the task point in a grid map.
As an optimal technical scheme, the optimizing search is performed on the task point access sequence based on the improved genetic algorithm to generate an optimal access sequence, which specifically comprises the following steps:
s501, setting an initial position of the autonomous mobile robot in a grid map, and numbering the initial position and task points;
s502, setting an initial population scale m, setting the length of an individual chromosome as the number p of task points, randomly generating a numbered feasible solution which traverses each task point from an initial position and finally returns to the initial position, and forming an alternative population by m numbered feasible solutions;
S503, improving an alternative seed group: randomly exchanging the positions of two task points of each individual in the candidate population, calculating the sum of Euclidean distances of the task point sequences in the individuals, and if the sum of Euclidean distances of the improved individual task point sequences is shorter, updating the individuals and the candidate population to be the initial population;
s504, planning a segmentation combination path of a task point access sequence corresponding to an individual in a grid map;
s505, calculating a fitness value fitness; the fitness value is determined by the length and smoothness of the segment combination path corresponding to the individual, and the calculation formula is as follows:
，
wherein alpha and beta are weight coefficients; (x) i ,y i ) The abscissa of the ith task point in the individual; k is the number of task points; u is the number of corners in the individual; a, a j 、b j And c j A diagonal edge and two adjacent edges of the jth corner respectively; the diagonal edge of the jth corner refers to the diagonal edge of the corner j, which is the line segment obtained by connecting the front corner and the rear corner of the corner j; the two adjacent edges of the jth corner are respectively a line segment obtained by searching forward from the corner j to the previous corner along the sectional combination path and a line segment obtained by searching backward from the corner j to the next corner along the sectional combination path;
S506, iterative selection operation, cross operation and mutation operation are carried out until the end condition is reached to output the optimal task point sequence.
As an optimal technical scheme, when optimizing and searching the task point access sequence based on the improved genetic algorithm, an elite algorithm is adopted to accelerate the convergence rate of the improved genetic algorithm, namely: taking the individual with the optimal fitness value in the population in each iteration as elite individual, so that the individual directly enters the population of the next iteration by skipping the selection operation, the cross operation and the mutation operation;
in the selecting operation in step S506, the roulette algorithm is adopted to calculate the selected probability according to the fitness value of each individual, the larger the fitness value is, the more likely the individual is selected, and the selected probability calculation formula is:
，
wherein p (i) is the probability of being selected for individual i in the current iterative population; i is the individual number; m is the total number of individuals; the fitness (i) is the fitness value of the individual i in the current iterative population;
step S506, when the crossover operation is carried out, an individual crossover probability formula is introduced, and the individual crossover probability is controlled according to the evolution condition; probability of individual crossing p c (i) The formula is:
，
，
wherein p is c0 Is an initial probability; the fitness (i) is the fitness value of the individual i in the current iterative population; fitness * (i) The fitness value of the individual i in the previous iteration population is obtained; n is the iteration number; individuals who have acquired a crossover opportunity perform crossover operations using the POX crossover method;
step S506, when the mutation operation is carried out, an individual mutation probability formula is introduced, and the individual mutation probability is controlled according to the evolution times; the individual variation probability formula is:
，
wherein p is m0 Is an initial probability; n is the iteration number; individuals who have a chance of mutation undergo mutation using a process mutation method;
the determining step of the ending condition is as follows:
introducing population fitness change rateReflecting the potential of population evolution in the current iteration, and the calculation formula is as follows:
，
wherein t is the population in the current iteration; t-1 is the population in the previous iteration; m is the number of individuals in the population; i is the individual number; fitness t (i) The fitness value of the individual i in the population in the current iteration is obtained; fitness t-1 (i) The fitness value of the individual i in the population in the previous iteration is obtained;
setting an adaptive potential threshold omega based on the change of the iteration times, which is expressed as:
，
wherein omega 0 Is the initial value of the self-adaptive potential threshold omega; n is the iteration number;
when the population fitness change rateAnd stopping iterating the improved genetic algorithm when the adaptive potential threshold omega is smaller than the adaptive potential threshold omega.
As an optimal technical scheme, when optimizing and searching the task point access sequence based on the improved genetic algorithm, a strategy prevention algorithm is set to avoid falling into a local optimal state;
The specific process of the strategy prevention algorithm is as follows:
(1) Setting a backtracking population, storing an excellent population in an iterative process of an improved genetic algorithm, and assigning the excellent population as an initial population after the initial population is generated by the improved genetic algorithm;
(2) Setting an adjustment period as T, and carrying out adjustment judgment once every T times of iterative improvement genetic algorithm, wherein the judgment content is the total number of fitness of the population in the current iteration and the backtracking population;
(3) If the total number of the fitness of the population in the current iteration is better than that of the backtracking population, the population in the current iteration is saved to the backtracking population, and the step (2) is returned; if the total number of the fitness of the population in the current iteration is inferior to the total number of the fitness of the backtracking population, entering a step (4);
(4) Defining a counter F, and initially setting zero; if the total number of the fitness of the population in the current iteration is inferior to the total number of the fitness of the backtracking population, the counter F is self-increased; when the counter F reaches the threshold F, the step (5) is entered; if not, returning to the step (2) to continue adjusting and judging;
(5) Copying the backtracking population to the population in the current iteration, setting the counter F to zero, and returning to the step (2) to continue adjusting and judging.
As an optimal technical solution, the method uses an ant colony algorithm to segment and plan paths according to an optimal access sequence and combines the segmented paths to obtain an optimal combined path, which specifically includes:
S601, generating a complete work sequence from an initial position- > task point access sequence- > initial position according to the optimal access sequence; sequentially determining a starting point and a target point of each mobile segment in a complete working sequence to obtain a segment path;
s602, initializing ant colony algorithm parameters, wherein the parameters comprise ant number, maximum iteration times, heuristic factors, expected heuristic factors, total pheromone amount of one-time circulation and pheromone volatilization coefficients;
s603, placing ants at the starting points of the segmented paths, and starting to search the optimal path of the target point;
s604, setting a motion constraint at the obstacle grid in the grid map, namely: if an obstacle grid exists around the free grid where the ant is located, the ant is forbidden to adopt a diagonal direction moving mode when passing between the two grids which are respectively in the horizontal direction and the vertical direction with the feasible grids of the obstacle grid;
s605, acquiring free grid information which is not accessed by ants, and calculating transition probability according to a state transition formula; the transition probability calculation formula is as follows:
，
wherein α is a heuristic factor; beta is a desired heuristic;pheromone intensities on the path from free grid i to free grid j; allowed represents a set of free grids that the ant has not accessed; / >Heuristic functions on the path from free grid i to free grid j;
s606, selecting a next free grid to move by using a roulette algorithm based on the transition probability of the free grid which is not accessed by ants; updating the moving path and path length of ants after moving;
s607, judging whether the ants move to the target point, if so, storing an ant moving path, and waiting for the rest ants to move to the target point; if not, jumping to S605;
s608, when all ants reach the target point, comparing paths traversed by all ants, selecting an optimal path in the iteration, and updating the global pheromone; the global pheromone updating formula is as follows:
，
，
wherein,the pheromone increment on the path from the free grid i to the free grid j in the iteration t is represented; ρ is the volatility coefficient of the self-adaptive pheromone; />Representing the increment of pheromone on the path from the free grid i to the free grid j of the kth ant in the iteration t;
s609, judging whether the ant colony algorithm falls into local optimum, if so, jumping to S610; if not, jumping to S611;
s610, processing local optimal conditions: correcting the size of the path information element, wherein the correction formula is as follows:
，
wherein avg is the global path pheromone average value; And->For correction factor->+/>=1，/>∈(0,0.3)；
S611, judging whether the maximum iteration times are reached, if so, ending the ant colony algorithm, and jumping to S612; if not, jumping to S603 for iteration to perform an ant colony algorithm;
s612, saving the optimal path of the segmented path;
s613, repeating S602 to S612, and when the optimal path of each segmented path is planned, connecting the optimal paths of the segmented paths end to end according to the complete working sequence to form an optimal combined path.
As a preferred embodiment, the heuristic function η ij Expressed as:
，
，
wherein delta 1 And delta 2 Is a weight coefficient; d, d ij The Euclidean distance from free grid i to free grid j; (x) i ,y i ) Position coordinates of the free grid i; (x) j ,y j ) Position coordinates of the free grid j; r is (r) ij The number of turns between the previous grid of the free grid i and the free grid j;a vector pointing to the target point for the start point; />A vector pointing to free grid j for free grid i;
the self-adaptive pheromone volatilization coefficient is self-adaptively adjusted along with the change of the iteration times, and the adjustment formula is as follows:
，
wherein ρ is 0 Is the initial value of ρ; n is the current iteration number; n (N) max The maximum iteration number;
the saidThe improved Ant-Cycle model is used as an Ant pheromone updating model, and the updating formula is as follows:
，
Wherein Q represents the total amount of pheromones in one cycle; l (L) k Represents the path length traversed by ant k; n (N) max The maximum iteration number; sigma is the pheromone delta enhancement factor, expressed as:
，
wherein,is antk, the path length of the last path in the iteration;
the judging step of judging whether the ant colony algorithm falls into local optimum or not comprises the following steps:
if the iteration number does not reach the maximum iteration number and the path length of the optimal path is still changing, judging that the ant colony algorithm does not fall into local optimal; if the iteration number does not reach the maximum iteration number and the path length of the optimal path is not changed, the ant colony algorithm is judged to be in local optimum.
As an optimal technical scheme, the curve fitting is performed on the optimal combination path by using a quadratic uniform B-spline curve method, and a final path is output, specifically:
s701, forming a control point set P (P) by the starting point, the middle point and the end point of each folding line segment on the optimal combination path 0 ,P 1 ,P 2 ) The method comprises the steps of carrying out a first treatment on the surface of the Wherein P is 0 Represents the starting point of a certain folded line segment, P 1 Represents the midpoint of a segment, P 2 Indicating the end point of a certain folded line segment;
s702, bringing a control point set of each folding line segment into a quadratic uniform B spline curve equation to construct a curve of each folding line segment; the quadratic uniform B-spline curve equation represents a linear combination of the set of control points and the basis function, expressed as:
，
Wherein P is i Is the ith control point; t is a parameter defining a domain between control points; b (B) i,2 (t) is the i < 2 > order B-spline basis function, expressed as:
B 0,2 (t) = 1/2(1-t+t 2 )，
B 1,2 (t) = 1/2(-t 2 +2t-1)，
B 2,2 (t) = 1/2t 2 ；
s703, performing curve fitting on the curves of the folded line segments, and outputting a final path.
Compared with the prior art, the invention has the following advantages and beneficial effects:
1. according to the invention, the surrounding environment of the free grid is comprehensively considered, the free grid is classified and clustered, and the autonomous calibration of the task point position is realized by combining with the preset work requirement, so that the autonomous capability of the autonomous mobile robot for executing the work task in the known scene is obviously improved.
2. According to the invention, the improved genetic algorithm is used for optimizing the task point access sequence, and the self-adaptive genetic operator is designed to ensure that the algorithm can still ensure a higher convergence rate when facing a large number of task points, so that the task point optimizing efficiency is improved; meanwhile, the provided local optimal processing scheme and algorithm ending conditions can enable the algorithm to obtain the near-optimal task point access sequence with fewer iteration times, reduce the consumption of calculation resources of the autonomous mobile robot and improve the working efficiency of the robot.
3. The method uses the ant colony algorithm to conduct subsection path planning on the task points, gives consideration to the path smoothness and the target direction information through the heuristic function, and improves the searching efficiency of the algorithm; the provided pheromone updating strategy can ensure that the algorithm has the capability of jumping out of local optimum while searching efficiently.
4. The method uses a quadratic uniform B spline curve method to perform curve fitting on the path, and generates a smooth path convenient for the autonomous mobile robot to move.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flowchart of a path planning method of an autonomous mobile robot according to an embodiment of the present invention.
FIG. 2 is a schematic diagram of setting potential descriptors in an embodiment of the present invention.
Fig. 3 is a grid map after task points are identified in an embodiment of the present invention.
FIG. 4 is a graph of the calculated optimal path for the method of the present invention.
Detailed Description
In order to enable those skilled in the art to better understand the present application, the following description will make clear and complete descriptions of the technical solutions in the embodiments of the present application with reference to the accompanying drawings in the embodiments of the present application. It will be apparent that the described embodiments are only some, but not all, of the embodiments of the present application. All other embodiments, which can be made by those skilled in the art based on the embodiments herein without making any inventive effort, are intended to be within the scope of the present application.
Reference in the specification to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the application. The appearances of such phrases in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. Those of skill in the art will explicitly and implicitly understand that the embodiments described herein may be combined with other embodiments.
As shown in fig. 1, the path planning method of the autonomous mobile robot according to the present embodiment includes the following steps:
s1, scanning a working scene to construct a grid map, wherein the grid map comprises an obstacle grid and a free grid;
s2, setting a potential descriptor on a grid map, calculating the potential of the free grid, classifying and marking potential grids; wherein the potential descriptor is a grid matrix centered on a free grid;
s3, performing k-means clustering on the potential grids according to the distribution positions of the potential grids to obtain alternative clusters;
s4, calibrating task points on the grid map according to the alternative clusters and preset task requirements;
s5, setting an initial position of the autonomous mobile robot, and carrying out optimizing search on the task point access sequence based on an improved genetic algorithm to generate an optimal access sequence;
S6, segmenting and planning paths by using an ant colony algorithm according to the optimal access sequence and combining the segmented paths to obtain an optimal combined path;
s7, performing curve fitting on the optimal combined path by using a quadratic uniform B spline curve method, and outputting a final path;
and S8, the autonomous mobile robot moves to a task point according to the final path and executes the task.
Further, in step S2, the free grids are classified and potential grids are labeled, specifically:
s201, setting potential descriptors on a grid map, wherein the potential descriptors are used for describing distribution conditions of barrier grids around the grids; as shown in fig. 2, the potential descriptor is a grid square matrix with a size of 5×5 centered on the free grid i, the data structure of the potential descriptor can be mapped into a matrix of 5×5, and the free grid and the obstacle grid in the potential descriptor are respectively mapped into the matrix with 0 and 1. Setting potential descriptors on free grids at the boundaries of the grid map, and setting the excess parts as free grids if the potential descriptors exceed the grid map;
the potential descriptor comprises a first layer and a second layer, wherein all grids adjacent to the free grid i constitute the first layer; all grids adjacent to the first layer except the free grid i constitute a second layer; determining a first layer and a second layer by calculating Euclidean distance from each grid to a free grid i in the potential descriptor, wherein the calculation formula is as follows:
，
Wherein layer (i, j) is the layer number of the jth grid in the potential descriptor of the free grid i, 1 is the first layer, 2 is the second layer, none is the layer number, and d (i, j) is the Euclidean distance from the jth grid in the potential descriptor to the free grid i;
s202, calculating potential D of the free grid marked as potential grid according to potential descriptor of the free grid cap The calculation formula is as follows:
，
wherein, gamma 1 And gamma 2 Is a weight coefficient; c 1 The number of free grids of the first layer in the potential descriptor of the free grid i; c 2 The number of free grids of the second layer in the potential descriptor of the free grid i; s is(s) 1 The number of adjacent barrier grids of the first layer in the potential descriptor of the free grid i; s is(s) 2 The number of adjacent barrier grids of the second layer in the potential descriptor of the free grid i; s is(s) 3 The number of barrier grids of the first layer and barrier grids of the second layer in the potential descriptor of the free grid i are adjacent;
s203 potential D according to the free grid cap Dividing all the free grids in the grid map according to a threshold epsilon, marking potential grids by using tags, and generating a potential grid set; the marking method comprises the following steps:
，
wherein D is cap (i) Potential D for free grid i cap 1 indicates that the free grid i is marked as a potential grid, and 0 indicates that the free grid i is not marked as a potential grid.
Further, in step S3, the potential grids are clustered to obtain alternative clusters, which specifically are:
s301, determining the number k of clustering clusters by using an elbow method;
s302, setting a grid distance function, wherein the grid distance function is expressed as:
，
wherein alpha and beta are weight coefficients; (x) i ,y i ) Is the coordinates of the free grid i; (x) j ,y j ) Is the coordinates of the free grid j; c (C) i And S is i The number of free grids and the number of barrier grids in the free grid i potential descriptor are respectively; c (C) j And S is j The number of free grids and the number of barrier grids in the free grid j potential descriptor are respectively;
s303, calculating the distance between potential grids according to the coordinates of each potential grid, and clustering by using a k-means clustering algorithm to obtain k candidate clusters.
Further, the step S4 of calibrating the task point specifically includes:
s401, judging whether a preset task requirement exists or not; if so, jumping to S402; otherwise, jumping to S403; the preset task demands comprise important working areas, working complexity and the like, wherein the important working areas comprise information such as positions of the important working areas, coverage areas of the important working areas and the like;
s402, determining whether an alternative cluster exists in a key work area in a preset task demand; if the candidate clusters exist, marking the candidate clusters as key clusters; if the free grids do not exist, dividing the free grids in the range of the key work area into key clusters;
S403, randomly selecting a free grid from each alternative cluster as a task point; if the key clusters exist, selecting a free grid which is positioned at or approximately at the geometric center of the key clusters in each key cluster as a task point, and marking the task point in a grid map. As shown in fig. 3, 45 task points are marked in a certain grid map.
Further, after the task point is obtained, step S5 performs optimizing search on the access sequence of the task point to generate an optimal access sequence, specifically:
s501, setting an initial position of the autonomous mobile robot in a grid map, and numbering the initial position and task points;
s502, setting an initial population scale m, setting the length of an individual chromosome as the number p of task points, randomly generating a numbered feasible solution which traverses each task point from an initial position and finally returns to the initial position, and forming an alternative population by m numbered feasible solutions;
s503, improving an alternative seed group: randomly exchanging the positions of two task points of each individual in the candidate population, calculating the sum of Euclidean distances of the task point sequences in the individuals, and if the sum of Euclidean distances of the improved individual task point sequences is shorter, updating the individuals and the candidate population to be the initial population;
S504, planning a segmentation combination path of a task point access sequence corresponding to an individual in a grid map;
s505, calculating a fitness value fitness; the fitness value is determined by the length and smoothness of the sectional combined path corresponding to the individual, and the fitness value is lower when the path placement length is longer; the higher the smoothness of the path is, the higher the fitness value is, the smoothness of the path is considered by the sum of corner angles in the path, and the greater the corner angle is, the higher the smoothness of the path is; the calculation formula of the fitness value is as follows:
，
wherein alpha and beta are weight coefficients; (x) i ,y i ) The abscissa of the ith task point in the individual; k is the number of task points; u is the number of corners in the individual; a, a j 、b j And c j A diagonal edge and two adjacent edges of the jth corner respectively; the diagonal edge of the j-th corner refers to the diagonal edge of the corner j, which is the line segment obtained by connecting the front corner and the rear corner of the corner j; the two adjacent edges of the jth corner are respectively a line segment obtained by searching forward from the corner j to the previous corner along the sectional combination path and a line segment obtained by searching backward from the corner j to the next corner along the sectional combination path;
s506, iterative selection operation, cross operation and mutation operation are carried out until the end condition is reached to output the optimal task point sequence.
Further, when the improved genetic algorithm is used for optimizing and searching the task point access sequence, the elite algorithm is adopted to accelerate the convergence rate of the improved genetic algorithm, namely: taking the individual with the optimal fitness value in the population in each iteration as elite individual, so that the individual directly enters the population of the next iteration by skipping the selection operation, the cross operation and the mutation operation;
when the genetic algorithm is improved to perform selection operation, a roulette algorithm is adopted to calculate the selected probability according to the fitness value of each individual, the larger the fitness value is, the more likely the individual is selected, and the selected probability calculation formula is as follows:
，
wherein p (i) is the probability of being selected for individual i in the current iterative population; i is the individual number; m is the total number of individuals; the fitness (i) is the fitness value of the individual i in the current iterative population;
introducing an individual crossover probability formula when the genetic algorithm is improved to carry out crossover operation, and controlling the individual crossover probability according to the evolution condition; probability of individual crossing p c (i) The formula is:
，
，
wherein p is c0 Is an initial probability; the fitness (i) is the fitness value of the individual i in the current iterative population; fitness * (i) The fitness value of the individual i in the previous iteration population is obtained; n is the iteration number; individuals who have acquired a crossover opportunity perform crossover operations using the POX crossover method; initial probability p in this embodiment c0 Set to 1.
Introducing an individual variation probability formula when the genetic algorithm is improved to perform variation operation, and controlling individual variation probability according to the evolution times; the individual variation probability formula is:
，
wherein p is m0 Is an initial probability; n is the iteration number; individuals who have a chance of mutation undergo mutation using a process mutation method; initial probability p in this embodiment m0 Set to 1.
The invention sets the ending condition by introducing the population fitness change rate, and the determining steps of the ending condition are as follows:
introducing population fitness change rateReflecting the potential of population evolution in the current iteration, and the calculation formula is as follows:
，
wherein t is the population in the current iteration; t-1 is the population in the previous iteration; m is the number of individuals in the population; i is the individual number; fitness t (i) The fitness value of the individual i in the population in the current iteration is obtained; fitness t-1 (i) The fitness value of the individual i in the population in the previous iteration is obtained;
to avoid extreme populations falling into infinite iterations, an adaptive potential threshold ω based on the number of iterations is set to decrease with increasing number of evolutions, expressed as:
，
wherein omega 0 Is the initial value of the self-adaptive potential threshold omega; n is the iteration number;
the adaptive potential threshold omega and population fitness change rate The result of the comparison provides a reference for whether the improved genetic algorithm exits the evolution; when the rate of change of population fitness +.>And stopping iterating the improved genetic algorithm when the adaptive potential threshold omega is smaller than the adaptive potential threshold omega.
Furthermore, when the genetic algorithm is improved to perform optimizing search on the task point access sequence, the situation of local optimum may be trapped, so that the strategy is set to prevent the algorithm from trapping in the local optimum state, and the specific process is as follows:
(1) Setting a backtracking population, storing an excellent population in an iterative process of an improved genetic algorithm, and assigning the excellent population as an initial population after the initial population is generated by the improved genetic algorithm;
(2) Setting an adjustment period as T, and carrying out adjustment judgment once every T times of iterative improvement genetic algorithm, wherein the judgment content is the total number of fitness of the population in the current iteration and the backtracking population;
(3) If the total number of the fitness of the population in the current iteration is better than that of the backtracking population, the population in the current iteration is saved to the backtracking population, and the step (2) is returned; if the total number of the fitness of the population in the current iteration is inferior to the total number of the fitness of the backtracking population, entering a step (4);
(4) Defining a counter F, and initially setting zero; if the total number of the fitness of the population in the current iteration is inferior to the total number of the fitness of the backtracking population, the counter F is self-increased; when the counter F reaches the threshold F, the step (5) is entered; if not, returning to the step (2) to continue adjusting and judging;
(5) Copying the backtracking population to the population in the current iteration, setting the counter F to zero, and returning to the step (2) to continue adjusting and judging.
Further, step S6 uses the ant colony algorithm to segment and plan the paths according to the optimal access sequence and combines the segmented paths to obtain an optimal combined path, which specifically includes:
s601, generating a complete work sequence from an initial position- > task point access sequence- > initial position according to the optimal access sequence; sequentially determining a starting point and a target point of each mobile segment in a complete working sequence to obtain a segment path;
s602, initializing ant colony algorithm parameters including ant number, maximum iteration times, heuristic factors, expected heuristic factors, total pheromone amount of one-time circulation, pheromone volatilization coefficients and the like;
s603, placing ants at the starting points of the segmented paths, and starting to search the optimal path of the target point;
s604, setting a motion constraint at the obstacle grid in the grid map, namely: if an obstacle grid exists around the free grid where the ant is located, the ant is forbidden to adopt a diagonal direction moving mode when passing between the two grids which are respectively in the horizontal direction and the vertical direction with the feasible grids of the obstacle grid;
S605, acquiring free grid information which is not accessed by ants, and calculating transition probability according to a state transition formula, wherein the calculation formula is as follows:
，
wherein α is a heuristic factor; beta is a desired heuristic;pheromone intensities on the path from free grid i to free grid j; allowed represents a set of free grids that the ant has not accessed; />Heuristic functions on the path from free grid i to free grid j; />Pheromone intensities on the path from free grid i to unviewed free grid s;heuristic functions on the path from free grid i to free grid s;
s606, selecting a next free grid to move by using a roulette algorithm based on the transition probability of the free grid which is not accessed by ants; updating the moving path and path length of ants after moving;
s607, judging whether the ants move to the target point, if so, storing an ant moving path, and waiting for the rest ants to move to the target point; if not, jumping to S605;
s608, when all ants reach the target point, comparing paths traversed by all ants, selecting an optimal path in the iteration, updating the global pheromone, and updating a formula as follows:
，
，
wherein,the pheromone increment on the path from the free grid i to the free grid j in the iteration t is represented; ρ is the volatility coefficient of the self-adaptive pheromone; / >Representing the increment of pheromone on the path from the free grid i to the free grid j of the kth ant in the iteration t;
s609, judging whether the ant colony algorithm falls into local optimum, if so, jumping to S610; if not, jumping to S611;
s610, processing local optimal conditions: correcting the size of the path pheromone, weakening the oversized path of the pheromone, enhancing the smaller path of the pheromone, and the pheromone correction formula is as follows:
，
wherein avg is the global path pheromone average value;and->For correction factor->+/>=1，/>∈(0,0.3)；
S611, judging whether the maximum iteration times are reached, if so, ending the ant colony algorithm, and jumping to S612; if not, jumping to S603 for iteration to perform an ant colony algorithm;
s612, saving the optimal path of the segmented path;
s613, repeating S602 to S612, and when the optimal path of each segmented path is planned, connecting the optimal paths of the segmented paths end to end according to the complete working sequence to form an optimal combined path.
Further, heuristic function η ij Expressed as:
，
，
wherein delta 1 And delta 2 Is a weight coefficient; d, d ij The Euclidean distance from free grid i to free grid j; (x) i ,y i ) Position coordinates of the free grid i; (x) j ,y j ) Position coordinates of the free grid j; r is (r) ij The number of turns between the previous grid of the free grid i and the free grid j; A vector pointing to the target point for the start point; />A vector pointing to free grid j for free grid i;
in order to improve algorithm searching efficiency, the invention sets the self-adaptive pheromone volatilization coefficient to be self-adaptively adjusted along with the change of iteration times, and the adjustment formula is as follows:
，
wherein ρ is 0 Is the initial value of ρ; n is the current iteration number; n (N) max The maximum iteration number; the value of rho becomes larger along with the increase of the iteration times, the directivity of ants is enhanced, and the convergence speed is further increased;
for the followingThe present application uses an improvement of Ant-CThe mould model is used as an ant pheromone updating model, and the updating formula is as follows:
，
wherein Q represents the total amount of pheromones in one cycle; l (L) k Represents the path length traversed by ant k; n (N) max The maximum iteration number; sigma is the pheromone delta enhancement factor, expressed as:
，
wherein,the path length of the last path of ant k in the iteration is; if the current path is better than the previous path, adding the current pheromone increment; if the current path is inferior to the previous path, the current pheromone increment is reduced.
Also, in order to avoid the ant colony algorithm from sinking into the local optimum, it is also necessary to determine whether the ant colony algorithm is sinking into the local optimum, the steps are:
if the iteration number does not reach the maximum iteration number and the path length of the optimal path is still changing, judging that the ant colony algorithm does not fall into local optimal; if the iteration number does not reach the maximum iteration number and the path length of the optimal path is not changed, the ant colony algorithm is judged to be in local optimum.
Further, after obtaining the optimal combination path, step S7 uses a quadratic uniform B spline curve method to perform curve fitting, and outputs a final path, specifically:
s701, forming a control point set P (P) by the starting point, the middle point and the end point of each folding line segment on the optimal combination path 0 ,P 1 ,P 2 ) The method comprises the steps of carrying out a first treatment on the surface of the Wherein P is 0 Represents the starting point of a certain folded line segment, P 1 Represents the midpoint of a segment, P 2 Indicating the end point of a certain folded line segment;
s702, bringing a control point set of each folding line segment into a quadratic uniform B spline curve equation to construct a curve of each folding line segment; the quadratic uniform B-spline curve equation represents a linear combination of the set of control points and the basis function, expressed as:
，
wherein P is i Is the ith control point; t is a parameter defining a domain between control points; b (B) i,2 (t) is the i < 2 > order B-spline basis function, expressed as:
B 0,2 (t) = 1/2(1-t+t 2 )，
B 1,2 (t) = 1/2(-t 2 +2t-1)，
B 2,2 (t) = 1/2t 2 ；
s703, performing curve fitting on the curves of the folded line segments, and outputting a final path. As shown in fig. 4, the final path diagram of the autonomous mobile robot performing the task in the grid map after curve fitting.
It should be noted that, for the sake of simplicity of description, the foregoing method embodiments are all expressed as a series of combinations of actions, but it should be understood by those skilled in the art that the present invention is not limited by the order of actions described, as some steps may be performed in other order or simultaneously in accordance with the present invention.
The technical features of the above embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.
The above examples are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above examples, and any other changes, modifications, substitutions, combinations, and simplifications that do not depart from the spirit and principle of the present invention should be made in the equivalent manner, and the embodiments are included in the protection scope of the present invention.

Claims (8)

1. A path planning method for an autonomous mobile robot, comprising the steps of:
s1, scanning a working scene to construct a grid map, wherein the grid map comprises an obstacle grid and a free grid;
s2, setting a potential descriptor on a grid map, calculating the potential of the free grid, classifying and marking potential grids; the potential descriptor is a grid matrix centered on a free grid;
the potential grids are classified and marked, specifically:
s201, setting potential descriptors on a grid map, wherein the potential descriptors are used for describing distribution conditions of barrier grids around the grids; the potential descriptor is a grid square matrix with the size of 5 multiplied by 5 by taking a free grid i as a center; setting potential descriptors on free grids at the boundaries of the grid map, and setting the excess parts of the potential descriptors to be free grids if the potential descriptors exceed the grid map; all grids adjacent to the free grid i in the potential descriptor form a first layer; all grids adjacent to the first layer except the free grid i constitute a second layer;
S202, calculating potential D of the free grid marked as potential grid according to potential descriptor of the free grid cap The calculation formula is as follows:
，
wherein, gamma 1 And gamma 2 Is a weight coefficient; c 1 The number of free grids of the first layer in the potential descriptor of the free grid i; c 2 The number of free grids of the second layer in the potential descriptor of the free grid i; s is(s) 1 The number of adjacent barrier grids of the first layer in the potential descriptor of the free grid i; s is(s) 2 The number of adjacent barrier grids of the second layer in the potential descriptor of the free grid i; s is(s) 3 The number of barrier grids of the first layer and barrier grids of the second layer in the potential descriptor of the free grid i are adjacent;
s203 potential D according to the free grid cap Will be covered in the grid mapDividing the free grids according to a threshold epsilon, marking potential grids by using tags, and generating a potential grid set; the marking method comprises the following steps:
，
wherein D is cap (i) Potential D for free grid i cap 1 indicates that free grid i is marked as a potential grid, 0 indicates that free grid i is not marked as a potential grid;
s3, performing k-means clustering on the potential grids according to the distribution positions of the potential grids to obtain alternative clusters;
s4, calibrating task points on the grid map according to the alternative clusters and preset task requirements;
S5, setting an initial position of the autonomous mobile robot, and carrying out optimizing search on the task point access sequence based on an improved genetic algorithm to generate an optimal access sequence;
the optimizing search is carried out on the task point access sequence based on the improved genetic algorithm to generate an optimal access sequence, and the method specifically comprises the following steps:
s501, setting an initial position of the autonomous mobile robot in a grid map, and numbering the initial position and task points;
s502, setting an initial population scale m, setting the length of an individual chromosome as the number p of task points, randomly generating a numbered feasible solution which traverses each task point from an initial position and finally returns to the initial position, and forming an alternative population by m numbered feasible solutions;
s503, improving an alternative seed group: randomly exchanging the positions of two task points of each individual in the candidate population, calculating the sum of Euclidean distances of the task point sequences in the individuals, and if the sum of Euclidean distances of the improved individual task point sequences is shorter, updating the individuals and the candidate population to be the initial population;
s504, planning a segmentation combination path of a task point access sequence corresponding to an individual in a grid map;
s505, calculating a fitness value fitness; the fitness value is determined by the length and smoothness of the segment combination path corresponding to the individual, and the calculation formula is as follows:
，
Wherein alpha and beta are weight coefficients; (x) i ,y i ) The abscissa of the ith task point in the individual; k is the number of task points; u is the number of corners in the individual; a, a j 、b j And c j A diagonal edge and two adjacent edges of the jth corner respectively; the diagonal edge of the jth corner refers to the diagonal edge of the corner j, which is the line segment obtained by connecting the front corner and the rear corner of the corner j; the two adjacent edges of the jth corner are respectively a line segment obtained by searching forward from the corner j to the previous corner along the sectional combination path and a line segment obtained by searching backward from the corner j to the next corner along the sectional combination path;
s506, iteratively performing selection operation, cross operation and mutation operation until reaching an end condition to output an optimal task point sequence;
s6, segmenting and planning paths by using an ant colony algorithm according to the optimal access sequence and combining the segmented paths to obtain an optimal combined path;
s7, performing curve fitting on the optimal combined path by using a quadratic uniform B spline curve method, and outputting a final path;
and S8, the autonomous mobile robot moves to a task point according to the final path and executes the task.
2. The path planning method of an autonomous mobile robot according to claim 1, wherein the k-means clustering is performed on the potential grids according to the distribution positions of the potential grids to obtain alternative clusters, specifically:
S301, determining the number k of clustering clusters by using an elbow method;
s302, setting a grid distance function, wherein the grid distance function is expressed as:
，
wherein alpha and betaIs a weight coefficient; (x) i ,y i ) Is the coordinates of the free grid i; (x) j ,y j ) Is the coordinates of the free grid j; c (C) i And S is i The number of free grids and the number of barrier grids in the free grid i potential descriptor are respectively; c (C) j And S is j The number of free grids and the number of barrier grids in the free grid j potential descriptor are respectively;
s303, calculating the distance between potential grids according to the coordinates of each potential grid, and clustering by using a k-means clustering algorithm to obtain k candidate clusters.
3. The path planning method of an autonomous mobile robot according to claim 1, wherein the calibrating task points on the grid map according to the candidate cluster and the preset task demand specifically comprises:
s401, judging whether a preset task requirement exists or not; if so, jumping to S402; otherwise, jumping to S403; the preset task demands comprise key work areas and work complexity;
s402, determining whether an alternative cluster exists in a key work area in a preset task demand; if the candidate clusters exist, marking the candidate clusters as key clusters; if the free grids do not exist, dividing the free grids in the range of the key work area into key clusters;
S403, randomly selecting a free grid from each alternative cluster as a task point; if the key clusters exist, selecting a free grid which is positioned at or approximately at the geometric center of the key clusters in each key cluster as a task point, and marking the task point in a grid map.
4. The path planning method of an autonomous mobile robot according to claim 1, wherein when the optimizing search is performed on the task point access sequence based on the improved genetic algorithm, an elite algorithm is adopted to accelerate the convergence rate of the improved genetic algorithm, namely: taking the individual with the optimal fitness value in the population in each iteration as elite individual, so that the individual directly enters the population of the next iteration by skipping the selection operation, the cross operation and the mutation operation;
in the selecting operation in step S506, the roulette algorithm is adopted to calculate the selected probability according to the fitness value of each individual, the larger the fitness value is, the more likely the individual is selected, and the selected probability calculation formula is:
，
wherein p (i) is the probability of being selected for individual i in the current iterative population; i is the individual number; m is the total number of individuals; the fitness (i) is the fitness value of the individual i in the current iterative population;
step S506, when the crossover operation is carried out, an individual crossover probability formula is introduced, and the individual crossover probability is controlled according to the evolution condition; probability of individual crossing p c (i) The formula is:
，
，
wherein p is c0 Is an initial probability; the fitness (i) is the fitness value of the individual i in the current iterative population; fitness * (i) The fitness value of the individual i in the previous iteration population is obtained; n is the iteration number; individuals who have acquired a crossover opportunity perform crossover operations using the POX crossover method;
step S506, when the mutation operation is carried out, an individual mutation probability formula is introduced, and the individual mutation probability is controlled according to the evolution times; the individual variation probability formula is:
，
wherein p is m0 Is an initial probability; n is the iteration number; individuals who have a chance of mutation undergo mutation using a process mutation method;
the determining step of the ending condition is as follows:
introduction of populationsRate of change of fitnessReflecting the potential of population evolution in the current iteration, and the calculation formula is as follows:
，
wherein t is the population in the current iteration; t-1 is the population in the previous iteration; m is the number of individuals in the population; i is the individual number; fitness t (i) The fitness value of the individual i in the population in the current iteration is obtained; fitness t-1 (i) The fitness value of the individual i in the population in the previous iteration is obtained;
setting an adaptive potential threshold omega based on the change of the iteration times, which is expressed as:
，
wherein omega 0 Is the initial value of the self-adaptive potential threshold omega; n is the iteration number;
When the population fitness change rateAnd stopping iterating the improved genetic algorithm when the adaptive potential threshold omega is smaller than the adaptive potential threshold omega.
5. The path planning method of an autonomous mobile robot according to claim 1, wherein when the task point access sequence is searched for optimization based on an improved genetic algorithm, a strategy prevention algorithm is set to avoid falling into a locally optimal state;
the specific process of the strategy prevention algorithm is as follows:
(1) Setting a backtracking population, storing an excellent population in an iterative process of an improved genetic algorithm, and assigning the excellent population as an initial population after the initial population is generated by the improved genetic algorithm;
(2) Setting an adjustment period as T, and carrying out adjustment judgment once every T times of iterative improvement genetic algorithm, wherein the judgment content is the total number of fitness of the population in the current iteration and the backtracking population;
(3) If the total number of the fitness of the population in the current iteration is better than that of the backtracking population, the population in the current iteration is saved to the backtracking population, and the step (2) is returned; if the total number of the fitness of the population in the current iteration is inferior to the total number of the fitness of the backtracking population, entering a step (4);
(4) Defining a counter F, and initially setting zero; if the total number of the fitness of the population in the current iteration is inferior to the total number of the fitness of the backtracking population, the counter F is self-increased; when the counter F reaches the threshold F, the step (5) is entered; if not, returning to the step (2) to continue adjusting and judging;
(5) Copying the backtracking population to the population in the current iteration, setting the counter F to zero, and returning to the step (2) to continue adjusting and judging.
6. The path planning method of an autonomous mobile robot according to claim 1, wherein the steps of using an ant colony algorithm to segment and plan paths according to an optimal access sequence and combining the segmented paths to obtain an optimal combined path include:
s601, generating a complete work sequence from an initial position- > task point access sequence- > initial position according to the optimal access sequence; sequentially determining a starting point and a target point of each mobile segment in a complete working sequence to obtain a segment path;
s602, initializing ant colony algorithm parameters, wherein the parameters comprise ant number, maximum iteration times, heuristic factors, expected heuristic factors, total pheromone amount of one-time circulation and pheromone volatilization coefficients;
s603, placing ants at the starting points of the segmented paths, and starting to search the optimal path of the target point;
s604, setting a motion constraint at the obstacle grid in the grid map, namely: if an obstacle grid exists around the free grid where the ant is located, the ant is forbidden to adopt a diagonal direction moving mode when passing between the two grids which are respectively in the horizontal direction and the vertical direction with the feasible grids of the obstacle grid;
S605, acquiring free grid information which is not accessed by ants, and calculating transition probability according to a state transition formula; the transition probability calculation formula is as follows:
，
wherein α is a heuristic factor; beta is a desired heuristic;pheromone intensities on the path from free grid i to free grid j; allowed represents a set of free grids that the ant has not accessed; />Heuristic functions on the path from free grid i to free grid j;
s606, selecting a next free grid to move by using a roulette algorithm based on the transition probability of the free grid which is not accessed by ants; updating the moving path and path length of ants after moving;
s607, judging whether the ants move to the target point, if so, storing an ant moving path, and waiting for the rest ants to move to the target point; if not, jumping to S605;
s608, when all ants reach the target point, comparing paths traversed by all ants, selecting an optimal path in the iteration, and updating the global pheromone; the global pheromone updating formula is as follows:
，
，
wherein,the pheromone increment on the path from the free grid i to the free grid j in the iteration t is represented; ρ is the volatility coefficient of the self-adaptive pheromone; / >Representing the increment of pheromone on the path from the free grid i to the free grid j of the kth ant in the iteration t;
s609, judging whether the ant colony algorithm falls into local optimum, if so, jumping to S610; if not, jumping to S611;
s610, processing local optimal conditions: correcting the size of the path information element, wherein the correction formula is as follows:
，
wherein avg is the global path pheromone average value;and->For correction factor->+/> =1，/> ∈(0,0.3)；
S611, judging whether the maximum iteration times are reached, if so, ending the ant colony algorithm, and jumping to S612; if not, jumping to S603 for iteration to perform an ant colony algorithm;
s612, saving the optimal path of the segmented path;
s613, repeating S602 to S612, and when the optimal path of each segmented path is planned, connecting the optimal paths of the segmented paths end to end according to the complete working sequence to form an optimal combined path.
7. The path planning method of an autonomous mobile robot of claim 6, wherein the heuristic function η ij Expressed as:
，
，
wherein delta 1 And delta 2 Is a weight coefficient; d, d ij The Euclidean distance from free grid i to free grid j; (x) i ,y i ) Position coordinates of the free grid i; (x) j ,y j ) Position coordinates of the free grid j; r is (r) ij The number of turns between the previous grid of the free grid i and the free grid j;a vector pointing to the target point for the start point; />A vector pointing to free grid j for free grid i;
the self-adaptive pheromone volatilization coefficient is self-adaptively adjusted along with the change of the iteration times, and the adjustment formula is as follows:
，
wherein ρ is 0 Is the initial value of ρ; n is the current iteration number; n (N) max The maximum iteration number;
the saidThe improved Ant-Cycle model is used as an Ant pheromone updating model, and the updating formula is as follows:
，
wherein Q represents the total amount of pheromones in one cycle; l (L) k Represents the path length traversed by ant k; n (N) max The maximum iteration number; sigma is the pheromone delta enhancement factor, expressed as:
，
wherein,the path length of the last path of ant k in the iteration is;
the judging step of judging whether the ant colony algorithm falls into local optimum or not comprises the following steps:
if the iteration number does not reach the maximum iteration number and the path length of the optimal path is still changing, judging that the ant colony algorithm does not fall into local optimal; if the iteration number does not reach the maximum iteration number and the path length of the optimal path is not changed, the ant colony algorithm is judged to be in local optimum.
8. The path planning method of an autonomous mobile robot according to claim 1, wherein the curve fitting is performed on the optimal combined path by using a quadratic uniform B-spline curve method, and a final path is output, specifically:
S701, forming a control point set P (P) by the starting point, the middle point and the end point of each folding line segment on the optimal combination path 0 ,P 1 ,P 2 ) The method comprises the steps of carrying out a first treatment on the surface of the Wherein P is 0 Represents the starting point of a certain folded line segment, P 1 Represents the midpoint of a segment, P 2 Indicating the end point of a certain folded line segment;
s702, bringing a control point set of each folding line segment into a quadratic uniform B spline curve equation to construct a curve of each folding line segment; the quadratic uniform B-spline curve equation represents a linear combination of the set of control points and the basis function, expressed as:
，
wherein P is i Is the ith control point; t is a parameter defining a domain between control points; b (B) i,2 (t) is the i < 2 > order B-spline basis function, expressed as:
B 0,2 (t) = 1/2(1-t+t 2 )，
B 1,2 (t) = 1/2(-t 2 +2t-1)，
B 2,2 (t) = 1/2t 2 ；
s703, performing curve fitting on the curves of the folded line segments, and outputting a final path.
CN202311839597.4A 2023-12-29 2023-12-29 Path planning method for autonomous mobile robot Active CN117516548B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202311839597.4A CN117516548B (en) 2023-12-29 2023-12-29 Path planning method for autonomous mobile robot

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202311839597.4A CN117516548B (en) 2023-12-29 2023-12-29 Path planning method for autonomous mobile robot

Publications (2)

Publication Number Publication Date
CN117516548A CN117516548A (en) 2024-02-06
CN117516548B true CN117516548B (en) 2024-04-12

Family

ID=89745954

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202311839597.4A Active CN117516548B (en) 2023-12-29 2023-12-29 Path planning method for autonomous mobile robot

Country Status (1)

Country Link
CN (1) CN117516548B (en)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107368077A (en) * 2017-08-15 2017-11-21 西京学院 A kind of robot path planning method based on GACA algorithm
WO2018176595A1 (en) * 2017-03-31 2018-10-04 深圳市靖洲科技有限公司 Unmanned bicycle path planning method based on ant colony algorithm and polar coordinate transformation
CN109164810A (en) * 2018-09-28 2019-01-08 昆明理工大学 It is a kind of based on the adaptive dynamic path planning method of ant colony-clustering algorithm robot
CN112461247A (en) * 2020-12-16 2021-03-09 广州大学 Robot path planning method based on self-adaptive sparrow search algorithm

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2018176595A1 (en) * 2017-03-31 2018-10-04 深圳市靖洲科技有限公司 Unmanned bicycle path planning method based on ant colony algorithm and polar coordinate transformation
CN107368077A (en) * 2017-08-15 2017-11-21 西京学院 A kind of robot path planning method based on GACA algorithm
CN109164810A (en) * 2018-09-28 2019-01-08 昆明理工大学 It is a kind of based on the adaptive dynamic path planning method of ant colony-clustering algorithm robot
CN112461247A (en) * 2020-12-16 2021-03-09 广州大学 Robot path planning method based on self-adaptive sparrow search algorithm

Also Published As

Publication number Publication date
CN117516548A (en) 2024-02-06

Similar Documents

Publication Publication Date Title
CN109164810B (en) Robot self-adaptive dynamic path planning method based on ant colony-clustering algorithm
CN111310999B (en) Warehouse mobile robot path planning method based on improved ant colony algorithm
CN108399451A (en) A kind of Hybrid Particle Swarm Optimization of combination genetic algorithm
CN110220525A (en) A kind of paths planning method based on potential field ant group algorithm
CN107330902B (en) Chaotic genetic BP neural network image segmentation method based on Arnold transformation
CN108985515B (en) New energy output prediction method and system based on independent cyclic neural network
CN112462803B (en) Unmanned aerial vehicle path planning method based on improved NSGA-II
CN112484732B (en) IB-ABC algorithm-based unmanned aerial vehicle flight path planning method
CN114815802A (en) Unmanned overhead traveling crane path planning method and system based on improved ant colony algorithm
CN108106624A (en) A kind of more people&#39;s Dispatch by appointment paths planning methods and relevant apparatus
CN115099459B (en) Workshop multi-row layout method considering gaps and loading and unloading points
CN111189455B (en) Unmanned aerial vehicle route planning method, system and storage medium
Bai et al. Design and Simulation of a Collision-free Path Planning Algorithm for Mobile Robots Based on Improved Ant Colony Optimization.
CN114253215B (en) Civil cabin door automatic drilling and riveting path planning method based on improved ant colony algorithm
CN117516548B (en) Path planning method for autonomous mobile robot
CN110852435A (en) Neural evolution calculation model
CN117471919A (en) Robot path planning method based on improved pelican optimization algorithm
CN117434950A (en) Mobile robot dynamic path planning method based on Harris eagle heuristic hybrid algorithm
CN115454070B (en) K-Means ant colony algorithm multi-robot path planning method
Laskar et al. A survey on VLSI floorplanning: its representation and modern approaches of optimization
CN115344046A (en) Mobile robot path planning based on improved deep Q network algorithm
CN108921354A (en) A method of the ant colony algorithm for solving TSP problems based on particle group optimizing
CN111460550B (en) Free-form surface structure multi-working-condition Pareto solution set optimization method based on changing niches
Zheng et al. Application of a Parallel Adaptive Cuckoo Search Algorithm in the Rectangle Layout Problem.
GULIASHKI et al. An evolutionary algorithm for integer multicriteria optimization (EVALIMCO)

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