CN115469671B - Robot path planning method based on aggregation-replacement strategy - Google Patents
Robot path planning method based on aggregation-replacement strategy Download PDFInfo
- Publication number
- CN115469671B CN115469671B CN202211295725.9A CN202211295725A CN115469671B CN 115469671 B CN115469671 B CN 115469671B CN 202211295725 A CN202211295725 A CN 202211295725A CN 115469671 B CN115469671 B CN 115469671B
- Authority
- CN
- China
- Prior art keywords
- individuals
- fitness
- path
- aggregation
- strategy
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 23
- 238000005265 energy consumption Methods 0.000 claims abstract description 26
- 230000002776 aggregation Effects 0.000 claims abstract description 23
- 238000004220 aggregation Methods 0.000 claims abstract description 23
- 230000002068 genetic effect Effects 0.000 claims abstract description 9
- 238000012163 sequencing technique Methods 0.000 claims abstract description 8
- 230000035772 mutation Effects 0.000 claims abstract description 5
- 239000011159 matrix material Substances 0.000 claims description 8
- 238000013459 approach Methods 0.000 claims description 5
- 238000009826 distribution Methods 0.000 claims description 4
- 230000002708 enhancing effect Effects 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 238000006073 displacement reaction Methods 0.000 claims 4
- 238000004088 simulation Methods 0.000 description 5
- 241000283153 Cetacea Species 0.000 description 4
- 230000000052 comparative effect Effects 0.000 description 4
- 238000002474 experimental method Methods 0.000 description 3
- 241000287127 Passeridae Species 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 230000004927 fusion Effects 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 238000013524 data verification Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000009191 jumping Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000010845 search algorithm Methods 0.000 description 1
- 238000002922 simulated annealing Methods 0.000 description 1
- 238000005728 strengthening Methods 0.000 description 1
- 238000002759 z-score normalization Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
- G05D1/02—Control of position or course in two dimensions
- G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
- G05D1/0212—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
- G05D1/0223—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory involving speed control of the vehicle
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
- G05D1/02—Control of position or course in two dimensions
- G05D1/021—Control of position or course in two dimensions specially adapted to land vehicles
- G05D1/0212—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory
- G05D1/0221—Control of position or course in two dimensions specially adapted to land vehicles with means for defining a desired trajectory involving a learning process
Landscapes
- Engineering & Computer Science (AREA)
- Aviation & Aerospace Engineering (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Feedback Control In General (AREA)
Abstract
The invention discloses a robot path planning method based on an aggregation-replacement strategy, which comprises the following steps: 1) Initializing a population and various variables; 2) Designing a fitness function based on the path length, the angle and the energy consumption, sequencing the fitness of individuals, and selecting an optimal individual; 3) Reserving the first j individuals after the fitness sequencing, and updating the remaining n-j individuals by using an aggregation strategy; 4) Updating the randomly selected m individuals by using a replacement strategy; 5) Judging whether a termination condition is met, if not, increasing population diversity by using a genetic algorithm crossover operator and a mutation operator; 6) And (3) iterating according to the steps 2) to 5) until stopping, deleting redundant path points of the paths generated after the iterating is stopped, and outputting the optimal paths. The invention balances the path length, the angle and the energy consumption cost through a new fitness function, combines an aggregation strategy, and strengthens the searching capability of a solution space; and a replacement strategy is provided, so that the algorithm can rapidly escape from the local extremum and the convergence speed can be increased.
Description
Technical Field
The invention relates to a path planning method, in particular to a robot path planning method based on an aggregation-replacement strategy.
Background
With the continuous development of control technology and artificial intelligence, robots are widely used in various fields such as production, military, life, and the like. The path planning is critical in the selection of the path search algorithm. As the autonomous demand of robots in various fields increases, conventional path planning algorithms are insufficient to meet the requirements of various fields. As a necessary condition for the robot to perform various tasks, the path planning algorithm plays an irreplaceable role for the travel of the robot.
One method for improving the algorithm performance is to design a corresponding strategy to enhance the corresponding algorithm capability aiming at the problems existing in the algorithm; or introducing a related mathematical model to solve the defect problem. And the other method is to realize complementary advantages by fusing with other algorithms.
For both of these methods, the domestic researchers have conducted related studies. The algorithm is improved, more, the algorithm convergence speed, population diversity and other self characteristics are slightly enhanced, and the applicability is low; in addition, for algorithm fusion, the fused algorithm can realize advantage complementation, the capability is greatly enhanced, the domestic algorithm fusion focuses on improvement of performance, and the influence of comprehensive factors such as path length, angle, robot energy consumption and the like on path planning is less considered; focusing on improvements in population diversity, less attention is paid to the ability to enhance the algorithm search.
Disclosure of Invention
The invention aims to: aiming at the problems existing in the prior art, the invention provides a path planning method based on an aggregation-replacement strategy, which is applied to path planning of a mobile robot, so that the path searching capability can be effectively enhanced, the adopted path searching algorithm has the convergence speed and the capability of jumping out of local optimum, the final planning path length and convergence time of the traditional genetic algorithm are greatly shortened, and meanwhile, the path searching algorithm is prevented from sinking into local optimum.
The technical scheme is as follows: the invention provides a robot path planning method based on an aggregation-replacement strategy, which is characterized by comprising the following steps:
1) Initializing a population and various variables, wherein the total number of individuals in the initialized population is n;
2) Designing an fitness function based on the path length, the path angle and the robot energy consumption, and calculating fitness values of n individuals; sorting the fitness values of individuals from small to large, selecting n individuals before sorting, and selecting the optimal individual gb; the optimal individual gb is the individual with the minimum fitness value;
3) Reserving the first j individuals after the fitness sequencing, and updating the remaining n-j individuals by utilizing a fitness aggregation strategy; dividing the n-j individuals into individuals with poor fitness and individuals with medium fitness by threshold comparison or percentage division according to the fitness values of the n-j individuals; guiding individuals with moderate fitness in the remaining n-j individuals to approach to the optimal individual (i.e. the individual with the smallest fitness value) gb through an aggregation strategy, and enabling the individuals with poor fitness to randomly explore a solution space;
Preferably, the individuals with the fitness value less than n/2 in the ordering position of the n-j individuals in the whole population are defined as the individuals with medium fitness, otherwise, the individuals with poor fitness are defined as the individuals with poor fitness;
4) Updating m individuals (10% -20% of the total number of population individuals) selected randomly by using a replacement strategy, exploring the area not searched by the solution space by using the replacement strategy, and enhancing the population diversity;
5) Judging whether a termination condition is reached, if so, ending the iteration; otherwise, generating new n individuals by utilizing a crossover operator and a mutation operator of a genetic algorithm, wherein the total number of individuals in the population is 2n, calculating the fitness of 2n individuals, and keeping the former n individuals with good fitness (namely small fitness value) to enter the next iteration;
6) And (3) iterating according to the steps (2) - (5) until the iteration is finished, deleting redundant path points of the paths generated after the iteration is finished, and outputting the optimal paths.
Further, the step 2) specifically includes the following:
2-1) definition of initial fitness
F(n)＝α1f1(n)+α2f2(n) (1)
In the formula (1), f 1 (n) is the path length, and f 2 (n) is the path angle energy consumption penalty; alpha 1、α2 is the weight of the path length and path angle energy consumption penalty respectively;
2-2) calculating Path Length
Where (2), N represents the number of path nodes, x i represents the abscissa of the i-th path node P i(xi,yi), and y i represents the ordinate of the i-th path node P i(xi,yi);
2-3) calculating path angle energy consumption penalty
Defining path nodes P1(x1,y1)、P2(x2,y2)、P3(x3,y3) as 3 points on a single path angle, wherein a, b and c are the side lengths of a triangle formed by the 3 points respectively, and the corresponding path angle theta is as follows:
Defining energy consumption penalty values at different path angles:
The path angle energy consumption penalty is:
2-4) normalization
Calculating the mean mu and standard deviation sigma of f 1(n)、f2 (n):
Normalizing the initial fitness F (n) to obtain a dimensionless fitness F' (n):
2-5) fitness ranking
And (5) arranging n individuals in the population from small to large according to the fitness, and selecting the optimal individual gb.
Further, in step 3), the specific steps of updating the remaining individuals using the aggregation policy are as follows:
3-1) preserving 10% -20% of individuals with the top fitness ranking, and traversing the rest individuals by using an aggregation strategy;
3-2) judging the ranking position of the fitness of each remaining individual in the whole population;
3-3) if the fitness ordering position of the remaining individuals is smaller than n/2, updating the position of the remaining individuals by using a formula (12); otherwise, randomly updating the location of the remaining individuals using equation (13);
In the formulas (12) and (13), Is the best position occupied by the individual with the highest iterative fitness, and X worst represents the global worst solution position of the previous generation; a represents a matrix of 1xD, where each element is randomly assigned 1 or-1, and A +＝AT(AAT)-1; q is a random number subject to normal distribution, L is a matrix of 1xD dimension with each element being 1;
3-4) repeating step 3-2) -step 3-3) until all remaining individuals of the population have been traversed.
Further, in step 4), the specific steps of updating the randomly selected m individuals using the replacement policy are as follows:
4-1) randomly selecting 10% -15% of individuals m t from the population after the aggregation strategy is acted;
4-2) randomly generating one feasible path m tr from the entire solution space;
4-3) respectively calculating the fitness of m t,mtr according to the fitness function;
4-4) comparing the fitness size of the two m t,mtr;
4-5) if f (m t)<f(mtr) indicates that the generated path is better than the original path, then replacing path m t with path m tr; otherwise, a poor path m tr is probabilistically accepted according to the metapolis criterion and the original path m t is replaced with path m tr;
Wherein the metapolis criterion is specifically expressed as follows:
In the formula (14), k is the number of iterations, E (k) is the state in the kth iteration, T is the initial temperature set in the algorithm, and p is the probability value that the algorithm accepts the new state.
4-6) Repeating steps 4-2) -4-5) until traversal m t ends.
Advantageous effects
Compared with the prior art, the invention has the following remarkable progress:
(1) And comprehensively considering factors such as path length, path angle and robot energy consumption, and providing a new fitness function to balance the relationship among the path length, the angle and the robot energy consumption.
(2) The aggregation strategy is introduced to strengthen the searching capability of the solution space, and meanwhile, the algorithm is realized to rapidly escape from the local extremum by utilizing the replacement strategy, so that the algorithm is prevented from sinking into the local optimum, and the convergence speed is increased. And the position updating means of cooperation of the aggregation strategy and the replacement strategy is adopted, so that the performance of the algorithm is enhanced to the greatest extent.
In summary, the invention can greatly shorten the path planning length and the convergence time, has stronger path searching capability, simultaneously gives consideration to the convergence speed, can jump out of local optimum, and has good running efficiency obviously superior to the existing algorithm.
Drawings
FIG. 1 is a flow chart of path planning in accordance with the present invention;
FIG. 2 is a convergence curve of the comparative GA algorithm;
FIG. 3 is a plot of the convergence of the algorithm of comparative example GAWOA;
FIG. 4 is a converging curve according to an embodiment of the present invention;
FIG. 5 is a path trace of the comparative GA algorithm;
FIG. 6 is a path trace of the comparative example GAWOA algorithm;
FIG. 7 is a path trace of an embodiment of the present invention;
In fig. 5 to 7, white lines represent path trajectories.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. The following description of at least one exemplary embodiment is merely exemplary in nature and is in no way intended to limit the invention, its application, or uses. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
A robot path planning method based on an aggregation-replacement strategy comprises the following steps:
1) Initializing a population and each variable, wherein the total number of individuals in the initialized population is n.
2) Designing a fitness function based on the correlation among the path length, angles formed by every three continuous path nodes and the robot energy consumption, and calculating the fitness values of n individuals according to the fitness function; and sorting the fitness, selecting the first n individuals with small fitness values as a new population, and selecting the optimal individual (namely, the individual with the smallest fitness value) gb.
3) The first j individuals after the fitness sequencing are reserved, the rest n-j individuals are updated by utilizing an aggregation strategy, and the aggregation strategy mainly has the functions of guiding individuals with moderate fitness among the n-j individuals to approach to individuals with optimal fitness, and enabling individuals with poor fitness to randomly explore a solution space.
The invention divides the fitness of the individual into three grades of poor, medium and excellent, and the smaller the fitness value is, the better the fitness of the individual is. Specifically, a threshold value can be preset, and the fitness value of each individual is compared with the threshold value to divide the fitness quality of different individuals; the fitness may also be ordered first, with the fitness level being divided by percentage. In this embodiment, an individual with a ranking position of less than n/2 in the overall population of the fitness value among n-j individuals is defined as an individual with moderate fitness, and the rest are individuals with poor fitness.
4) Updating m randomly selected individuals by using a replacement strategy, wherein the m individuals account for 10% -20% of the total number of the population individuals; the replacement strategy is used for exploring the area not searched by the solution space on one hand, and is used for enhancing population diversity and avoiding the algorithm from falling into local optimum on the other hand.
5) Judging whether a termination condition is reached, wherein in the embodiment, the termination condition is the maximum iteration number; judging whether the maximum iteration times are reached, if so, ending the iteration, otherwise, generating new n individuals by utilizing a genetic algorithm crossover operator and a mutation operator successively, wherein the total number of individuals in the population is 2n, recalculating the fitness of the 2n individuals, sequencing the fitness, and reserving the first n individuals with good fitness (namely small fitness value) to enter the next iteration.
6) And (3) iterating according to the steps (2) - (5) until the iteration is finished, deleting redundant path points of the paths generated after the iteration is finished, and outputting the optimal paths.
The following is a detailed description of the key steps of the present invention.
Fitness function design
(1) Definition of initial fitness
F(n)＝α1f1(n)+α2f2(n) (1)
In the formula (1), f 1 (n) is the path length, and f 2 (n) is the path angle energy consumption penalty; alpha 1、α2 is the weight of the path length and path angle energy consumption penalty, respectively.
(2) Calculating path length
Where (2), N represents the number of path nodes, x i represents the abscissa of the i-th path node P i(xi,yi), and y i represents the ordinate of the i-th path node P i(xi,yi).
(3) Calculating path angle energy consumption penalty
In order to reduce the energy consumption of the robot, a path angle energy consumption penalty function is applied to the fitness function, so that the algorithm preferentially screens paths with small path angle penalty.
Defining path nodes P1(x1,y1)、P2(x2,y2)、P3(x3,y3) as 3 points on a single path angle, wherein a, b and c are the side lengths of a triangle formed by the 3 points respectively, and the corresponding path angle theta is as follows:
Defining energy consumption penalty values at different path angles:
The path angle energy consumption penalty is:
(4) Normalization processing
According to the importance of the index, corresponding weight is given, and it can be seen that when the method is used as a fitness function, the path is shortest, and the smooth path with small path angle penalty is the optimal path; according to equation (1), it is known that the optimal path has the smallest fitness value, and preferably, the weights α 1、α2 are respectively 0.6 and 0.4.
Since the two evaluation indexes f 1(n)、f2 (n) have different dimensions and dimension units, and the path angle energy consumption penalty decreases in measurement capability as the map scale increases. Therefore, the invention uses a Z-score normalization method to eliminate the dimension influence between two indexes.
Calculating the mean mu and standard deviation sigma of f 1(n)、f2 (n):
Normalizing the initial fitness F (n) to obtain a dimensionless fitness F' (n):
Adaptive ranking and location update
The invention initializes the population number as n, so the total number of individuals in the population is n when the first fitness is ordered, and each individual in the population is arranged from small to large according to the fitness; after updating the individual position by using the aggregation-replacement strategy, judging whether the maximum iteration times are reached; if the iteration times are reached, ending the iteration; otherwise, generating new n individuals by using a crossover operator and a mutation operator, wherein the total number of the individuals in the population is 2n; and calculating the fitness of the 2n individuals, and after sequencing, keeping the first n individuals with good fitness to enter the next iteration.
Aggregation-replacement strategy
The invention provides an aggregation-replacement strategy, which is used for cooperatively strengthening the performance of an algorithm. The aggregation strategy mainly guides individuals with moderate fitness to approach to the position of the individual with optimal fitness, enables the individual with poor fitness to randomly explore a solution space, and strengthens the searching capability of an algorithm; the replacement strategy explores the area not searched by the solution space on one hand, enhances the population diversity on the other hand, and avoids the algorithm from sinking into local optimum.
The complete steps of the aggregation-permutation strategy are as follows:
s1: after population fitness ordering, the embodiment reserves 10% of individuals with the fitness ordering at the front, and uses an aggregation strategy to act on the rest of individuals, namely, the rest 90% of individuals are used as aggregation individuals;
S2: judging the positions of the aggregated individuals, wherein the positions refer to the ranking of the fitness of the individuals in the whole population;
s3: if the fitness ordering position of the rest individuals is smaller than n/2, updating the positions of the rest individuals by using a formula (12); otherwise, randomly updating the location of the remaining individuals using equation (13);
s4: repeating the step S2-the step S3 until all the rest individuals of the population are traversed;
S5: randomly selecting 10% of individuals m t from the population after the aggregation strategy is acted;
s6: starting traversing m t;
s7: randomly generating a feasible path m tr from the entire solution space;
S8: respectively calculating the fitness of m t,mtr according to the designed fitness function;
S9: comparing the fitness of the two m t,mtr;
S10: if f (m t)<f(mtr), it indicates that the generated path is superior to the original path, then path m t is replaced with path m tr; otherwise, generating paths which are not as same as the original paths, probability-accepting a worse path m tr according to the Metropolis criterion in order to avoid the excessive singleness of the later algorithm population, and replacing the original path m t with the path m tr;
S11: and repeating the steps S7-S10 until the traversal is finished.
(1) Aggregation policy
To enhance the search capabilities of the algorithm, a higher fitness is obtained, and the follower individuals will approach the most fitness individuals, strive to reach this location, or gather around this location. The aggregation strategy is to update the individual position to the vicinity of the global optimal individual position by using the distance between the individual to be updated and the optimal individual gb as a step length, and the update formula is shown as formula (12); the later individuals with moderate fitness use the distance between the own position and the worst individual as step length, and update the position to the random update of the solution space as (13).
In the formulas (12) and (13),Is the best position occupied by the individual with the highest fitness for this iteration, and X worst represents the global worst solution position of the previous generation. A represents a matrix of 1xD, where each element is randomly assigned 1 or-1, and a +＝AT(AAT)-1. Q is a random number subject to normal distribution, and L is a matrix of 1xD dimension with 1 for each element.
The specific steps of the aggregation strategy are as follows in algorithm 1:
In algorithm 1, a represents a matrix of 1xD, where each element is randomly assigned 1 or-1, and a +＝AT(AAT)-1, Q is a random number following normal distribution, and L is a matrix of 1xD dimension with 1 for each element.
(2) Replacement strategy
In order to avoid the algorithm to be in local optimum, the invention provides a replacement strategy based on the simulated annealing idea Metropolis criterion; the metapolis criteria are specifically expressed as follows:
in the formula (14), k is the number of iterations, E (k) is the state in the kth iteration, T is the initial temperature set in the algorithm, and p is the probability value of the algorithm accepting the new state;
The specific implementation process of the replacement strategy is as follows in algorithm 2:
In order to illustrate the beneficial effects of the invention, the effectiveness and applicability of the invention to the path planning problem are verified by comparing the performances of the invention and other algorithms under different simulation environments. In the invention, simulation experiments are carried out on a traditional Genetic Algorithm (GA), a sparrow algorithm (SSA), a whale algorithm (WOA), a genetic whale algorithm (GAWOA) and a genetic whale algorithm (GASSA) on a Matlab R2020b simulation platform, and the parameter settings of the algorithms are shown in table 1.
Table 1 experimental parameter settings
In order to obtain the validity of the accurate data verification algorithm, the invention independently carries out 30 experiments on 5 algorithms. Simulation experiments are carried out on a map with the size of 100x100, and the iteration number of the experiment is 100. The primary path trace contrast and convergence curves obtained by the algorithm are shown in fig. 2-7.
The experimental results obtained are shown in table 2:
Table 2 comparison of 100×100 simulation experiment results
As can be seen from fig. 2-4 and table 2, GASSA is superior to the other 4 algorithms for the seek path length, compared to the genetic algorithm, GAWOA algorithm, sparrow algorithm, whale algorithm, 33%, 11%, 40%, 26%, respectively; for the convergence speed, the method is superior to other algorithms, and is respectively reduced by 27%, 45%, 24% and 21%; for the optimal path node number, compared with other algorithms, the method reduces the found paths by 26%, 22%, 2% and 32% respectively; in the experiment, the length of the found path of the method, the iteration times of the algorithm and the node number of the path are superior to those of other algorithms.
Therefore, the final path length sought by the invention is greatly reduced, the time efficiency is overall superior, and the sinking of local optimum can be effectively avoided.
The foregoing examples illustrate only a few embodiments of the invention and are described in detail herein without thereby limiting the scope of the invention. It should be noted that it will be apparent to those skilled in the art that several variations and modifications can be made without departing from the spirit of the invention, which are all within the scope of the invention.
Claims (5)
1. The robot path planning method based on the aggregation-replacement strategy is characterized by comprising the following steps of:
1) Initializing a population and various variables, wherein the total number of individuals in the initialized population is n;
2) Designing an fitness function based on the path length, the path angle and the robot energy consumption, and calculating fitness values of n individuals; sorting the fitness values of individuals from small to large, selecting n individuals before sorting, and selecting the optimal individual gb; the optimal individual gb is the individual with the minimum fitness value; the method specifically comprises the following steps:
2-1) definition of initial fitness
F(n)＝α1f1(n)+α2f2(n) (1)
In the formula (1), f 1 (n) is the path length, and f 2 (n) is the path angle energy consumption penalty; alpha 1、α2 is the weight of the path length and path angle energy consumption penalty respectively;
2-2) calculating Path Length
Where (2), N represents the number of path nodes, x i represents the abscissa of the i-th path node P i(xi,yi), and y i represents the ordinate of the i-th path node P i(xi,yi);
2-3) calculating path angle energy consumption penalty
Defining path nodes P1(x1,P1)、P2(x2,y2)、P3(x3,y3) as 3 points on a single path angle, wherein a, b and c are the side lengths of a triangle formed by the 3 points respectively, and the corresponding path angle theta is as follows:
Defining energy consumption penalty values at different path angles:
The path angle energy consumption penalty is:
2-4) normalization
Calculating the mean mu and standard deviation sigma of f 1(n)、f2 (n):
Normalizing the initial fitness F (n) to obtain a dimensionless fitness F' (n):
2-5) fitness ranking
The n individuals in the population are arranged according to the fitness from small to large, and the optimal individual gb is selected;
3) Reserving the first j individuals after the fitness sequencing, and updating the remaining n-j individuals by utilizing a fitness aggregation strategy; dividing the n-j individuals into individuals with poor fitness and individuals with medium fitness by threshold comparison or percentage division according to the fitness values of the n-j individuals; guiding individuals with moderate fitness to approach to the optimal individuals gb through an aggregation strategy, and enabling individuals with poor fitness to randomly explore a solution space;
4) Updating m randomly selected individuals by using a replacement strategy, exploring areas not searched in a solution space by using the replacement strategy, and enhancing population diversity;
5) Judging whether a termination condition is reached, if so, ending the iteration; otherwise, generating new n individuals by utilizing a crossover operator and a mutation operator of a genetic algorithm, wherein the total number of the individuals in the population is 2n, calculating the fitness of the 2n individuals, sequencing the fitness values of the 2n individuals from small to large, and keeping the former n individuals with small fitness values to enter the next iteration;
6) And (3) iterating according to the steps 2) to 5) until the iteration is finished, deleting redundant path points of the paths generated after the iteration is finished, and outputting the optimal paths.
2. The method for planning a path of a robot based on an aggregation-displacement strategy according to claim 1, wherein in the step 3), the individuals with the fitness value of n-j less than n/2 in the ranking position of the whole population are defined as the individuals with moderate fitness, otherwise, the individuals with poor fitness are defined as the individuals with poor fitness.
3. The method for planning a path of a robot based on an aggregation-replacement strategy according to claim 2, wherein in step 3), the specific step of updating the remaining individuals with the aggregation strategy is as follows:
3-1) preserving 10% -20% of individuals with the top fitness ranking, and traversing the rest individuals by using an aggregation strategy;
3-2) judging the ranking position of the fitness of each remaining individual in the whole population;
3-3) if the fitness ordering position of the remaining individuals is smaller than n/2, updating the position of the remaining individuals by using a formula (12); otherwise, randomly updating the location of the remaining individuals using equation (13);
In the formulas (12) and (13), Is the best position occupied by the individual with the highest iterative fitness, and X worst represents the global worst solution position of the previous generation; a represents a matrix of 1xD, where each element is randomly assigned 1 or-1, and A +＝AT(A AT)-1; q is a random number subject to normal distribution, L is a matrix of 1xD dimension with each element being 1;
3-4) repeating step 3-2) -step 3-3) until all remaining individuals of the population have been traversed.
4. A method for planning a path for a robot based on an aggregation-displacement strategy as claimed in claim 3, wherein in step 4), the specific step of updating the randomly selected m individuals with the displacement strategy is as follows:
4-1) randomly selecting 10% -15% of individuals m t from the population after the aggregation strategy is acted;
4-2) randomly generating one feasible path m tr from the entire solution space;
4-3) respectively calculating the fitness of m t,mtr according to the fitness function;
4-4) comparing the fitness size of the two m t,mtr;
4-5) if f (m t)<f(mtr) indicates that the generated path is better than the original path, then replacing path m t with path m tr; otherwise, a poor path m tr is probabilistically accepted according to the metapolis criterion and the original path m t is replaced with path m tr;
4-6) repeating steps 4-2) -4-5) until traversal m t ends.
5. The method for planning a path of a robot based on an aggregation-displacement strategy according to claim 4, wherein in the step 4-5), the metapolis criterion is specifically expressed as follows:
In the formula (14), k is the number of iterations, E (k) is the state in the kth iteration, T is the initial temperature set in the algorithm, and p is the probability value that the algorithm accepts the new state.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211295725.9A CN115469671B (en) | 2022-10-21 | 2022-10-21 | Robot path planning method based on aggregation-replacement strategy |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211295725.9A CN115469671B (en) | 2022-10-21 | 2022-10-21 | Robot path planning method based on aggregation-replacement strategy |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115469671A CN115469671A (en) | 2022-12-13 |
| CN115469671B true CN115469671B (en) | 2024-05-28 |
Family
ID=84337068
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211295725.9A Active CN115469671B (en) | 2022-10-21 | 2022-10-21 | Robot path planning method based on aggregation-replacement strategy |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115469671B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108364069A (en) * | 2018-01-23 | 2018-08-03 | 广东工业大学 | A kind of robot welding paths planning method based on genetic algorithm |
| CN111780759A (en) * | 2020-05-28 | 2020-10-16 | 南京邮电大学 | Mobile robot path planning method based on improved genetic algorithm |
| WO2021022637A1 (en) * | 2019-08-06 | 2021-02-11 | 南京赛沃夫海洋科技有限公司 | Unmanned surface vehicle path planning method and system based on improved genetic algorithm |
| CN116295398A (en) * | 2022-12-30 | 2023-06-23 | 江苏海洋大学 | Path optimization method based on cooperative aggregation and branch deviation |
-
2022
- 2022-10-21 CN CN202211295725.9A patent/CN115469671B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108364069A (en) * | 2018-01-23 | 2018-08-03 | 广东工业大学 | A kind of robot welding paths planning method based on genetic algorithm |
| WO2021022637A1 (en) * | 2019-08-06 | 2021-02-11 | 南京赛沃夫海洋科技有限公司 | Unmanned surface vehicle path planning method and system based on improved genetic algorithm |
| CN111780759A (en) * | 2020-05-28 | 2020-10-16 | 南京邮电大学 | Mobile robot path planning method based on improved genetic algorithm |
| CN116295398A (en) * | 2022-12-30 | 2023-06-23 | 江苏海洋大学 | Path optimization method based on cooperative aggregation and branch deviation |
Non-Patent Citations (2)
| Title |
|---|
| 基于聚集—置换策略的机器人路径规划方法;徐晨晨 等;江苏海洋大学学报（自然科学版）;20230331;全文 * |
| 基于高适应度值遗传算法的AGV最优路径规划;赵大兴;余明进;许万;;计算机工程与设计;20170616(第06期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115469671A (en) | 2022-12-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112380008B (en) | Multi-user fine-grained task unloading scheduling method for mobile edge computing application | |
| CN107992051B (en) | Unmanned vehicle route planning method based on improved multi-target particle swarm algorithm | |
| CN110657816B (en) | Vehicle path problem planning method with hard time window based on firework algorithm | |
| CN109542103A (en) | A kind of robot welding paths planning method based on fireworks particle swarm algorithm | |
| CN109764882B (en) | Multi-target vehicle path planning method based on self-adaptive local search chain | |
| CN110738365B (en) | Flexible job shop production scheduling method based on particle swarm algorithm | |
| CN113341889B (en) | Distributed blocking flow workshop scheduling method and system with assembly stage and energy consumption | |
| CN110986960B (en) | Unmanned aerial vehicle flight path planning method based on improved clustering algorithm | |
| CN107122844A (en) | A kind of Multipurpose Optimal Method and system being combined based on index and direction vector | |
| CN108803332A (en) | Based on the paths planning method for improving biogeography | |
| CN113850385A (en) | Coarse and fine granularity combined neural network pruning method | |
| CN112272120B (en) | Virtual network mapping algorithm optimization method, equipment and storage medium | |
| CN113570112B (en) | Optimization algorithm for solving vehicle path problem with time window considering cooperation | |
| CN113052537A (en) | Logistics vehicle low-carbon route planning method based on heuristic particle swarm optimization | |
| CN115481727A (en) | Intention recognition neural network generation and optimization method based on evolutionary computation | |
| CN112070418A (en) | Weapon target allocation method of multi-target whale optimization algorithm | |
| CN115469671B (en) | Robot path planning method based on aggregation-replacement strategy | |
| CN111340303A (en) | Route planning method for travelers based on novel mixed frog-leaping algorithm | |
| CN114065896A (en) | Multi-target decomposition evolution algorithm based on neighborhood adjustment and angle selection strategy | |
| JP2020106957A (en) | Optimization calculation method, optimization calculation apparatus and optimization calculation program | |
| CN116341860A (en) | Vehicle dispatching optimization method based on improved NSGA-II algorithm | |
| CN116295398A (en) | Path optimization method based on cooperative aggregation and branch deviation | |
| CN110222055A (en) | The single-wheel core value maintaining method of multiple edge update under a kind of Dynamic Graph | |
| CN115936317A (en) | Suspension cable-stayed cooperative system scheme evaluation method and system | |
| CN110263906B (en) | Asymmetric negative correlation search method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |