CN109872001B - Unmanned vehicle task allocation method based on K-means and discrete particle swarm algorithm - Google Patents
Unmanned vehicle task allocation method based on K-means and discrete particle swarm algorithm Download PDFInfo
- Publication number
- CN109872001B CN109872001B CN201910150788.7A CN201910150788A CN109872001B CN 109872001 B CN109872001 B CN 109872001B CN 201910150788 A CN201910150788 A CN 201910150788A CN 109872001 B CN109872001 B CN 109872001B
- Authority
- CN
- China
- Prior art keywords
- task
- packet
- center
- algorithm
- unmanned vehicle
- 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
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
The invention discloses a multi-unmanned vehicle task allocation method based on K-means and discrete particle swarm optimization, which comprises the following steps: s1, initializing logistics scene information; s2, packaging logistics tasks: determining an optimal packing result by using a K-means algorithm, wherein the packing number is K; s3, taking k available unmanned vehicles, and matching the execution unmanned vehicles of each task package; and S4, determining the task sequence of each unmanned vehicle by using a discrete particle swarm algorithm. The method provided by the invention solves the task allocation problem of the multiple unmanned vehicles by combining a clustering idea and a swarm optimization algorithm, and performs task allocation of the multiple unmanned vehicles by using a discretized particle swarm algorithm, so that the convergence speed of the particle swarm algorithm is high, and the discretized iteration mode enables the algorithm to be more suitable for actual logistics scenes. And the tasks are packed by using the K-means clustering algorithm before the particle swarm algorithm is used, so that the size of the learning space is greatly reduced, and the task distribution efficiency is improved.
Description
Technical Field
The invention relates to the field of logistics distribution optimization, in particular to a task allocation method in the field of unmanned logistics.
Background
With the continuous expansion of the electronic commerce industry and the continuous improvement of labor cost, distribution is taken as a basic link of logistics operation, due to the continuous expansion of coverage rate, data updating is more frequent, allocation tasks are heavier, the service requirement of 'delivery to home' is improved, the intelligent logistics is more prevalent, and the unmanned era of China logistics is opened.
The storage efficiency of the unmanned storehouse can reach several times of the storage efficiency of the traditional beam shelf. According to the information of certain E-commerce logistics statistics, the sorting capacity of the unmanned sorting center can reach 9000 pieces/hour, the efficiency of the package ring section is improved by 4 times, and 180 people can be used for saving manpower in each field on the premise of the same field scale and sorting goods quantity. Unmanned aerial vehicle can shorten the time of traditional artifical delivery several times or even tens of times, and logistics cost also reduces thereupon. However, the air conditioner has the defects that the air conditioner cannot deliver goods in bad weather, the air conditioner cannot avoid artificial damage in the flying process and the like, and cannot be used in large quantities at the present stage. And recent experiments show that the unmanned vehicle can once only send 6 express deliveries, and once charge for a journey of 80 kilometers, can compensate unmanned aerial vehicle delivery shortcoming, and mass production uses. With the development of logistics modernization and intellectualization, the demand of the domestic market for unmanned logistics is further expanded. In a practical sense, the unmanned vehicle system has a wide application field, and can be used almost in an environment where goods need to be transported continuously or repeatedly. The dispatching of a single unmanned vehicle cannot meet the existing large material flow, a multi-unmanned vehicle intelligent dispatching system in unmanned logistics application belongs to a complex dynamic and multi-target discrete system, and the probability of delaying a large amount of waiting is higher when the density of a local map is higher and the number of unmanned vehicles is larger. The realization of mutual coordination and cooperation among unmanned vehicles in an uncertain environment (such as on a road) to reduce conflicts generated during the execution of operation tasks such as warehouse entry and exit becomes a relatively important practical target.
Therefore, in order to achieve the desire of "the last 1 km" with multiple unmanned vehicles, the research on the task allocation problem of multiple unmanned vehicles is very important. At present, a particle swarm algorithm, an ant colony algorithm, a simple scanning method, a branch definition method, a simulated annealing algorithm, a dynamic programming algorithm and the like are mainly adopted in a multi-unmanned vehicle distribution plane, but the algorithms have certain defects at present and mainly comprise the following steps: the method has the advantages of low degree of engagement with the actual logistics background, low convergence speed, long operation time and the like, and the solving result is not necessarily an optimal solution.
Therefore, a more efficient method of multi-unmanned vehicle allocation is needed.
Disclosure of Invention
The purpose of the invention is as follows: aiming at the defects of the prior art, the invention provides a multi-unmanned vehicle task allocation method based on K-means and discrete particle swarm optimization, which realizes more practical and efficient unmanned vehicle task allocation.
The technical scheme is as follows: the invention provides a multi-unmanned vehicle task allocation method with a packing strategy (K-means) and a discrete particle swarm algorithm, which comprises the following steps:
s1, initializing logistics scene information;
s2, packaging logistics tasks: determining an optimal packing result by using a K-means algorithm, wherein the packing number is K;
s3, taking k available unmanned vehicles, and matching the execution unmanned vehicles of each task package;
and S4, determining the task sequence of each unmanned vehicle by using a discrete particle swarm algorithm.
Step S1 includes task information initialization, warehouse information initialization, and unmanned vehicle information initialization.
Step S2 includes:
s21, acquiring the number of task points and the number of available unmanned vehicles, and determining the upper limit of the number of task packages to be packaged;
s22, realizing the packing result of K task packages by using a K-means algorithm;
and S23, calculating the class outer distance, the class inner distance and the similarity according to the packing result of the k task packages, and taking the k value and the packing scheme corresponding to the minimum similarity value, wherein the class outer distance is the sum of the distance values from the center of each task package to the center of the whole task package, the class inner distance is the sum of the distance values from the task points contained in each task package to the center of the task package, and the similarity is the sum of the class outer distance and the class inner distance.
In step S4, the velocity and position of the particle are updated through discretization, and the value of the objective function is calculated, thereby determining a task sequence that minimizes the cost of the unmanned vehicle.
Has the advantages that:
1. the method disclosed by the invention is combined with a clustering thought and a crowd-sourcing optimization algorithm to solve the task allocation problem of the multiple unmanned vehicles, the method aims at improving the efficiency of the task allocation algorithm, in order to accord with the actual logistics scene, the discrete particle swarm algorithm is used for allocating the tasks of the multiple unmanned vehicles, the convergence speed of the particle swarm algorithm is high, and the discretization iteration mode enables the algorithm to be more suitable for the actual logistics background.
2. According to the invention, the tasks are packed by using the K-means clustering algorithm before the particle swarm algorithm is used, so that the size of the learning space is greatly reduced, and the task distribution efficiency is improved.
Drawings
FIG. 1 is an overall flow chart of a multi-unmanned vehicle task allocation method based on K-means and discrete particle swarm optimization;
FIG. 2 is a flow chart for determining an optimal packing scheme using the K-means algorithm;
FIG. 3 is a flow chart for performing a one-time packing scheme using the K-means algorithm;
FIG. 4 is a flow chart of the discrete particle swarm algorithm of the present invention.
Detailed Description
The technical scheme of the invention is further explained by combining the attached drawings.
The multi-unmanned vehicle task allocation method based on K-means and discrete particle swarm optimization firstly packs task targets by proper task packet quantity by utilizing K-means clustering before task allocation, and allocates the multi-unmanned vehicles to a plurality of task packets by using the discrete particle swarm optimization, and the overall process of the method is shown in figure 1.
In step S1, the logistics scene information is initialized.
The method comprises the following steps: loading various warehouse information Storehouse (x, y, n, id), wherein Storehouse x represents the geographical dimensionality of the warehouse, Storehouse y represents the geographical longitude of the warehouse, Storehouse n represents the number of the residual articles of the warehouse, and Storehouse id represents the number of the warehouse;
loading information Car (x, y, v, f, id, visible) of each unmanned vehicle, wherein car.x represents the geographical dimension of the unmanned vehicle, car.y represents the geographical longitude of the unmanned vehicle, car.v represents the driving speed of the unmanned vehicle, car.f represents the remaining available fuel quantity of the unmanned vehicle, car.id represents the number of the unmanned vehicle, and car.visible represents the working state of the unmanned vehicle, when the unmanned vehicle is still executing a task, car.visible is 0, and when the unmanned vehicle is in an idle state, car.visible is 1;
loading information Task (x, y, id) of each Task point, wherein the task.x represents the geographical latitude of the Task point, the task.y represents the geographical longitude of the Task point, and the task.id represents the number of the Task point.
In the logistics scene, a plurality of unmanned vehicles are used for conveying articles in each warehouse to a specified target task point, each unmanned vehicle returns to the warehouse after completing a distribution task, the geographic information of the warehouse is used for calculating the distance between the task and the warehouse, the unmanned vehicles only use longitude and latitude, id and speed (the speed is used for calculating the running time from one place to one task point) in the distribution method, the priority of the tasks can be understood as different tasks at the same position, and other information is mainly used for graphical representation of projects and calculation of specific examples.
And step S2, determining the optimal packing result by using a K-means algorithm, wherein the packing number is K.
The tasks are divided into a proper number of task packages according to the relative position of the task target point and the distance from the warehouse. The task package can be understood as a set of task points divided according to certain conditions, the tasks are packaged, the tasks of the set specify a certain vehicle to complete, and other vehicles do not participate.
Referring to fig. 2, the specific steps include:
step S2.1: acquiring the number n _ Task of Task points and the number n _ Car of available unmanned vehicles, and determining the upper limit k of the number of Task packages to be packagedmax；
Upper limit of number of packets kmaxDetermined by both the number of mission points and the number of available unmanned vehicles. According to the empirical method, the packing number k satisfies
Therefore, when the number of available unmanned vehicles is large, the upper limit of the packing number can be obtained by an empirical method, and when the number of available unmanned vehicles is small, the upper limit of the packing number is determined by the number of available unmanned vehicles, then the upper limit of the packing number k ismaxThe following formula is satisfied:
wherein k ismaxThe upper limit value of the Task packing number is represented, n _ Task represents the total number of the Task points, when a plurality of tasks exist in the same target point, the Task points with the same longitude and latitude but different Task numbers are used for representing, and n _ Car represents the total number of the available unmanned vehicles.
Step S2.2: let the cycle variable k equal to 1;
step S2.3: for each K ═ i, a K-means algorithm is used to realize the packing result of K task packets, wherein, as shown in fig. 3, the specific steps of the K-means algorithm are as follows:
step1 packet center a for randomly initializing k task packets1(x1,y1),a2(x2,y2),...,ak(xk,yk) The package center here refers to the longitude and latitude average of all task points in this package.
step2, for each Task point Task (x, y, id), the Task point is classified as the nearest Task package center aj(xj,yj) In the task package j, the division rule satisfies the following formula:
wherein, labelidPacket partition for Task point Task (x, y, id), aj(xj,yj) Indicating the j-th task packet center, i represents a latitude and a longitude from 1 to 2, min represents a distance formula in brackets, and min is an operation for all tasks to find a nearest ajSub means to specify ajThe index j of (a) is such an operation.
step3 updating packet center a of each task packetj(xj,yj) The attribute value (namely longitude and latitude) of the center of the new packet is the average value of the attribute values of all task points belonging to the task packet, and the following formula is satisfied:
wherein, aj(xj,yj) Indicating a new package center, cjIs a new task packet partition, n _ cjRefer to the new task Package cjThe number of task points contained in the file.
step4, when the change rate of each packet center is smaller than a given value or the maximum iteration number is reached, the packing is ended, otherwise, step2 is returned.
The packet center change rate is an order of magnitude describing the change of the geographical position of each task packet, and the longitude and latitude of each packet center are changed correspondingly each time the packet center is updated, for example, from (60,130) to (59.9,129.99), the change Δ x is 0.1, and Δ y is 0.01, here, the change rate is considered to be equal to 0.1, and the maximum value of the two changes is taken, i.e., the formula can be written as max { Δ x, Δ y }. When the change is small enough to be acceptable, the current packaging is considered to be completed.
Step S2.4: determine if (k ≦ k)max) If the logic value is 0, ending all packing and outputting all packing results, otherwise, k + + and returning to the step S2.3.
Step S2.5: compute full task Package center atotal(xtotal,ytotal)
Wherein, atotalRepresenting the center of all Task points, TaskidRepresenting the longitude and latitude of the task of the id, namely averaging through the longitude and latitude to obtain the central position points of all the tasks in the packet; n _ Task represents the number of Task points.
Step S2.6: calculating the class outer distance of the K task packet packing results obtained by using a K-means algorithm for each K ═ i respectively, namely the sum of the distance values from the center of each task packet to the center of the whole task packet meets the following formula:
wherein L represents the extranormal distance, aiIndicates the ith packet center, atotalIndicating a full task package center. Here the subtraction of the absolute values on the bands means that the distance between two points is determined.
Step S2.7: calculating the class inner distance of each K-i task packet packaging result obtained by using a K-means algorithm, namely the sum of the distance values from the task point contained in each task packet to the center of the task packet, and satisfying the following formula:
wherein D represents the class inner distance, aiIndicates the ith packet center, CiIndicating the ith task package.
Step S2.8: respectively calculating a distance similarity function of K task packet packing results obtained by using a K-means algorithm for each K ═ i, namely the sum of the class outer distance and the class inner distance, and satisfying the following formula:
wherein, F (S, k) represents the distance similarity function value of this packing, and the smaller the value, the better the packing result.
Step S2.9: to save the current best packed result, let temp ═ F (S, k)max)，k＝kmax-1；
Step S2.10: if the logical value is 0, the packing result of k task packets is better than the best current result, and then temp is updated to be F (S, k);
step S2.11: k is k-1;
step S2.12: repeating steps S2.10 and S2.11 until k is equal to 0;
step S2.13: and recording the k value corresponding to temp and the packing scheme.
And step S3, k available unmanned vehicles are taken, corresponding execution unmanned vehicles of each task package are matched randomly, and if the number of the available unmanned vehicles is large, k available unmanned vehicles are selected randomly.
Besides random matching, the unmanned vehicles can be matched according to the distance, so that each unmanned vehicle has an unordered task queue claimed by the unmanned vehicle.
And step S4, determining the task sequence of each unmanned vehicle by using a discrete particle swarm algorithm, and then executing the delivery task according to the sequence determined sequence.
The specific steps of confirming the task sequence comprise:
step S4.1: making a circulation variable i equal to 1;
step S4.2: for ith task package CiAnd determining a task execution sequence of the unmanned vehicle corresponding to the task packet by using a discrete particle swarm algorithm, wherein the discrete particle swarm algorithm comprises the following specific steps as shown in fig. 4:
step1 random generation of a certain number of particles XtConstituting an initialization particle swarm, XtShowing an arrangement mode of task points in the task package;
step2, judging whether the iteration times exceed the upper limit, and if so, outputting a task execution sequence for executing the unmanned vehicle;
step3, acquiring the number N of the task points of the task package, and updating Pbest according to the fitness functiontAnd Gbestt，PbesttIs an individual extremum, referring to the best position of each particle, GbesttIt is an overall extremum, which refers to the best position of the overall particle, and the fitness function is the following two functions, which are our goals to minimize, and in this case, the fitness can be understood as the cost of executing unmanned vehicles, and there can be the following two goals:
wherein runi,jIndicates the time length of the i-th stage to the task point j, Xi,jIndicates whether the ith stage executes the task j, Xi,jThat is, 1 is the ith stage to execute task j, Xi,j0 means that the ith stage does not execute task j, delayi,jRepresenting the delay, performance, of the ith stage executing the jth task pointi,jRepresenting the execution time of the jth task in the ith stage, target minJ1It can be understood that the total travel time is the minimum and the target is two minJ2It is understood that the total task time is the shortest.
In order to solve the target by using the discrete particle swarm algorithm, the method for converting the multiple targets into the single target by using the multiplication-division method is as follows:
minJ＝J1J2
step4, updating the particles, turning to step2, designing a new discretization particle updating mode in order to ensure that the updating of the particles conforms to the actual scene and the task is completely executed, wherein the discretization particle updating mode meets the following expression:
wherein, VtIs a circular right shift vector generated randomly, namely, the circular right shift operation, Pbest, is carried out on the task sequence according to the vectortIs an individual extremum, GbesttIs a global extreme value,
Refers to the set of cyclic right-handed vectors required to translate into individual extrema,
in the same way, R1、R2Respectively, represent the randomness selection cycle right shift vectors. Finally each vehicle gets a sequence of its task executions.
Step S4.3: and if the if (i is less than or equal to k) is judged, if the logic value is 0, the task allocation result is output, otherwise, i + + is carried out, and the step S4.2 is returned.
The background of the invention is an unmanned vehicle task execution sequence in logistics, each number in the sequence is an integer, the situation that the number is still in the integer after iteration is guaranteed, and the original particle swarm algorithm is not applicable, so that a new discretization update can be set for the execution sequence. In addition, because the traditional particle swarm algorithm is slow in iteration speed due to overlarge solution space, the invention discretizes the iteration process of the particle swarm and simultaneously considers the use of a k-means algorithm for task packing, so that the task distribution is divided into two steps of task packing and task execution sequence setting, and apparently, one step is changed into two steps, so that the solution space can be greatly reduced, and the calculation speed is accelerated.
The foregoing is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, various modifications and decorations can be made without departing from the principle of the present invention, and these modifications and decorations should also be regarded as the protection scope of the present invention.
Claims (2)
1. Based onK-meansAnd a multi-unmanned vehicle task allocation method based on a discrete particle swarm algorithm, which is characterized by comprising the following steps:
s1, initializing logistics scene information;
s2, packaging logistics tasks: use ofK-meansThe algorithm determines the best packing result, the packing number of which iskThe method comprises the following steps:
s21, acquiring the number of task points and the number of available unmanned vehicles, and determining the upper limit of the number of task packages to be packaged;
s22, useK-meansAlgorithm implementationkThe packing result of each task package specifically includes:
step 1: random initializationkPacket center for individual task packets
；
step 2: for each task point
Classify it as the nearest task package center
Task bag ofjThe dividing rule satisfies the following formula:
wherein the content of the first and second substances,
as a task point
The packet of (2) is divided into,
x in (1) represents the geographical latitude of the task point, y represents the geographical longitude of the task point, id represents the number of the task point,
is shown asjA task packet center, g from 1 to 2 indicates a latitude and a longitude, and sub indicates the position
Subscript j of (a);
step 3: updating packet centers for each task packet
The attribute value of the new packet center is the average value of the attribute values of all task points belonging to the task packet, and the following formula is satisfied:
wherein the content of the first and second substances,
the new center of the packet is indicated,
indicating the new division of the task package,
refer to a new task Package
The number of task points contained in the task point list;
step 4: when the change rate of each packet center is less than a given value or reaches the maximum iteration number, ending the packaging, otherwise, returning tostep 2；
S23, calculating the class outer distance, the class inner distance and the similarity according to the packing result of the k task packages, and taking the k value and the packing scheme corresponding to the minimum similarity value, wherein the class outer distance is the sum of the distance values from the center of each task package to the center of the whole task package, and the calculation formula is as follows:
wherein L represents the distance between the class boundaries, the subtraction of the absolute values on the bands represents the distance between two points,
is shown asiThe center of each task packet is provided with a task packet center,
representing the full task package center:
，
the latitude and longitude of the task representing the id,
representing the number of task points;
the intra-class distance is the sum of the distance values from the task points contained in each task packet to the center of the task packet, and the calculation formula is as follows:
wherein the content of the first and second substances,Dthe distance between the inner sides of the classes is represented,
is shown asiA task package;
the similarity is the sum of the class outer distance and the class inner distance and meets the following formula:
wherein the content of the first and second substances,
the distance similarity function value of the current packaging is shown, and the smaller the value is, the better the packaging result is;
s3, gettingkThe vehicle can use an unmanned vehicle, and the unmanned vehicle is matched with the execution of each task package;
s4, determining a task sequence of each unmanned vehicle by using a discrete particle swarm algorithm, updating the particle speed and position through discretization, and calculating the value of an objective function, so as to determine the task sequence with the minimum unmanned vehicle payment cost, wherein the objective function comprises:
wherein the content of the first and second substances,
is shown asuEach stage of driving to a task pointvThe length of time of the time period,
is shown asuWhether a phase executes a task pointv，
Is shown asuA phase is executedvThe time delay of each task point is determined,
is shown asuA phase is executedvThe execution time of each task point;
the discretization updating particle speed and position is according to the following updating formula:
wherein the content of the first and second substances,
is a particle, represents the current arrangement mode of the task points in the task package,
is a circular right shift vector which is randomly generated, namely, the circular right shift operation is carried out on the task sequence according to the vector,
is the extreme value of the individual,
is a global extremum and is,
refers to the set of cyclic right-handed vectors required to translate into individual extrema,
refers to the set of cyclic right-handed vectors required to translate to a global extremum,
、
respectively, represent the randomness selection cycle right shift vectors.
2. The method according to claim 1K-meansAnd a multi-unmanned vehicle task allocation method based on a discrete particle swarm algorithm, wherein the step S1 comprises task information initialization, warehouse information initialization and unmanned vehicle information initialization.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910150788.7A CN109872001B (en) | 2019-02-28 | 2019-02-28 | Unmanned vehicle task allocation method based on K-means and discrete particle swarm algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910150788.7A CN109872001B (en) | 2019-02-28 | 2019-02-28 | Unmanned vehicle task allocation method based on K-means and discrete particle swarm algorithm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109872001A CN109872001A (en) | 2019-06-11 |
| CN109872001B true CN109872001B (en) | 2021-03-19 |
Family
ID=66919541
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910150788.7A Active CN109872001B (en) | 2019-02-28 | 2019-02-28 | Unmanned vehicle task allocation method based on K-means and discrete particle swarm algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109872001B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113269480B (en) * | 2020-02-17 | 2022-06-14 | 百度在线网络技术（北京）有限公司 | Task allocation path determining method and device and electronic equipment |
| CN111738619B (en) * | 2020-07-06 | 2023-11-07 | 腾讯科技（深圳）有限公司 | Task scheduling method, device, equipment and storage medium |
| CN112035552B (en) * | 2020-09-02 | 2023-05-09 | 国网河南省电力公司电力科学研究院 | Method and device for predicting severity of voltage sag based on association rule |
| CN112965528B (en) * | 2021-02-22 | 2023-06-30 | 广东电网有限责任公司 | Rescue strategy determination method, device and equipment for disaster points and storage medium |
| CN115065683B (en) * | 2022-07-27 | 2023-12-26 | 多伦互联网技术有限公司 | Vehicle edge network task allocation and unloading method based on vehicle clustering |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6952700B2 (en) * | 2001-03-22 | 2005-10-04 | International Business Machines Corporation | Feature weighting in κ-means clustering |
| CN106990792A (en) * | 2017-05-23 | 2017-07-28 | 西北工业大学 | Mix the multiple no-manned plane collaboration sequential coupling task distribution method of gravitation search algorithm |
| CN107093046A (en) * | 2017-04-21 | 2017-08-25 | 北京京东尚科信息技术有限公司 | Unmanned dispensing vehicle method for allocating tasks, system and unmanned dispensing vehicle |
| CN108932598A (en) * | 2018-05-31 | 2018-12-04 | 广东工业大学 | A kind of grand scale logistic allocator based on cloud platform |
| CN109063779A (en) * | 2018-08-09 | 2018-12-21 | 河海大学常州校区 | A kind of cloud manufacturing recourses cluster k-means clustering method |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102831474B (en) * | 2012-08-06 | 2015-04-22 | 江南大学 | Improved fuzzy C-mean clustering method based on quantum particle swarm optimization |
-
2019
- 2019-02-28 CN CN201910150788.7A patent/CN109872001B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6952700B2 (en) * | 2001-03-22 | 2005-10-04 | International Business Machines Corporation | Feature weighting in κ-means clustering |
| CN107093046A (en) * | 2017-04-21 | 2017-08-25 | 北京京东尚科信息技术有限公司 | Unmanned dispensing vehicle method for allocating tasks, system and unmanned dispensing vehicle |
| CN106990792A (en) * | 2017-05-23 | 2017-07-28 | 西北工业大学 | Mix the multiple no-manned plane collaboration sequential coupling task distribution method of gravitation search algorithm |
| CN108932598A (en) * | 2018-05-31 | 2018-12-04 | 广东工业大学 | A kind of grand scale logistic allocator based on cloud platform |
| CN109063779A (en) * | 2018-08-09 | 2018-12-21 | 河海大学常州校区 | A kind of cloud manufacturing recourses cluster k-means clustering method |
Non-Patent Citations (2)
| Title |
|---|
| UAV协同任务分配的改进DPSO算法仿真研究;王强等;《系统仿真学报》;20140531;第26卷(第5期);第207-208页 * |
| 一种初值优化的K-均值文档聚类算法(英文);陈媛媛等;《江西师范大学学报(自然科学版)》;20080430;第32卷(第2期);第1150-1152页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109872001A (en) | 2019-06-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109872001B (en) | Unmanned vehicle task allocation method based on K-means and discrete particle swarm algorithm | |
| WO2022188388A1 (en) | Smart city dynamic cold-chain logistics scheduling method based on ant colony optimization algorithm | |
| CN109214756B (en) | Vehicle logistics scheduling method and device, storage medium and terminal | |
| CN112001064B (en) | Full-autonomous water transport scheduling method and system between container terminals | |
| CN110807236A (en) | Warehouse logistics simulation system based on multiple robots | |
| CN108154254A (en) | Logistic distribution vehicle dispatching method based on modified A* algorithms | |
| CN109345091B (en) | Ant colony algorithm-based whole vehicle logistics scheduling method and device, storage medium and terminal | |
| CN111191847A (en) | Distribution path planning method and system considering order polymerization degree | |
| CN106355291B (en) | Logistics path planning method based on store point group bisector | |
| Tirado et al. | Heuristics for dynamic and stochastic routing in industrial shipping | |
| CN110097231A (en) | Multiple target objects stream scheduling method and device, logistics system and computer-readable medium | |
| CN112488386B (en) | Logistics vehicle distribution planning method and system based on distributed entropy multi-target particle swarm | |
| CN113848970A (en) | Multi-target collaborative path planning method for vehicle and unmanned aerial vehicle | |
| CN116151499A (en) | Intelligent multi-mode intermodal route planning method based on improved simulated annealing algorithm | |
| Yadav et al. | A heuristics based approach for optimizing delivery schedule of an Unmanned Aerial Vehicle (Drone) based delivery system | |
| CN115796530A (en) | Robot logistics scheduling optimization method for simultaneously delivering and taking goods | |
| Alaia et al. | Optimization of the multi-depot & multi-vehicle pickup and delivery problem with time windows using genetic algorithm | |
| CN114444809A (en) | Data-driven multi-target strip mine card path optimization method | |
| CN114819785A (en) | Air-ground joint trajectory optimization and resource allocation method based on reinforcement learning | |
| CN110930092B (en) | Distribution route adjusting method and device, electronic equipment and storage medium | |
| WO2021144577A1 (en) | Vehicle routing | |
| CN112016750A (en) | Improved method for solving problem of vehicle path with constraint | |
| CN111915074A (en) | Cold-chain logistics vehicle path selection method based on improved particle swarm optimization | |
| CHENG et al. | Artificial bee colony approach to solving the electric vehicle routing problem | |
| CN110264140A (en) | A kind of logistics transportation dispatching method, device and equipment with time window |
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 |