CN112508270B - Intelligent material distribution method based on multi-source heterogeneous data in manufacturing process - Google Patents
Intelligent material distribution method based on multi-source heterogeneous data in manufacturing process Download PDFInfo
- Publication number
- CN112508270B CN112508270B CN202011405500.5A CN202011405500A CN112508270B CN 112508270 B CN112508270 B CN 112508270B CN 202011405500 A CN202011405500 A CN 202011405500A CN 112508270 B CN112508270 B CN 112508270B
- Authority
- CN
- China
- Prior art keywords
- route
- ant
- optimal
- algorithm
- road
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
- G06Q10/047—Optimisation of routes or paths, e.g. travelling salesman problem
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
- G06Q10/083—Shipping
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Human Resources & Organizations (AREA)
- Theoretical Computer Science (AREA)
- Economics (AREA)
- General Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Development Economics (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Operations Research (AREA)
- Biomedical Technology (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Game Theory and Decision Science (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a material intelligent distribution method based on multi-source heterogeneous data in a manufacturing process, which comprises the following steps: the method comprises the following steps: analyzing multi-source heterogeneous data in the manufacturing process, and reading production task information, carrying vehicle information and transportation road information from a multi-source heterogeneous data model; step two: establishing a map model by using the information of the transportation roads, and calculating the shortest transportation road between all the working places by using an improved ant colony algorithm; step three: converting the task allocation problem into a graph theory problem by using production task information, carrying vehicle information and a map model, and performing optimal task allocation on each transport vehicle through the graph theory and an annealing algorithm; step four: according to the optimal task allocation information and the map model for each transport vehicle in the third step, the optimal road planning of each vehicle is realized through an ant colony algorithm; step five: and outputting the optimal task allocation and the optimal road plan of each vehicle. The method can improve the transportation efficiency of the materials and reduce the transportation cost of the materials.
Description
Technical Field
The invention relates to a material intelligent distribution method based on multisource heterogeneous data in a manufacturing process, which aims at the lowest material distribution cost in the manufacturing process, correctly gives a delivery task of a distribution vehicle through production task information, road information and carrying vehicle information, and plans a shortest road for the delivery vehicle. The method is suitable for the fields of intelligent manufacturing and processing, logistics distribution, production scheduling and the like.
Background
Material distribution is an important component of the manufacturing process. Modern products are complex, the processing flow is more, the working place is more, and the manufacturing process has huge material transportation amount. In a general production and manufacturing process, the problems that a certain vehicle is idle for a plurality of days and is not used due to unreasonable distribution of conveying tasks, improper vehicle selection and randomness of path selection occur, and the condition that the certain vehicle is continuously used occurs when a plurality of production tasks are performed. The problem not only can cause the production plan to be delayed, but also can cause the service life of the excessively used vehicle to be influenced, the idle stroke of the vehicle can cause the production efficiency of enterprises to be low, and the production cost of the enterprises is increased. The transportation task of the distribution vehicle is given through multi-source heterogeneous data in the production and manufacturing process, and the shortest road planning for the distribution vehicle is significant for reducing the manufacturing cost of enterprises.
In the prior art, the position of a target is mainly aimed at, the route of the target point is assumed to be a straight line, and the path with the minimum planning cost is realized through an algorithm, so that the method only focuses on the shortest path, and ignores important factors in actual production: road constraints, task constraints and vehicle transport capacity constraints make the delivery algorithm less relevant to practice, which is not beneficial to improving the material delivery process of the manufacturing enterprise. In recent years, due to the development of information technology, the difficulty in acquiring multi-source heterogeneous data in the production process is reduced, and more enterprises establish own multi-source heterogeneous databases to form a multi-source heterogeneous model. How to use multisource heterogeneous data to carry out production process material delivery optimization, consider the problem that production actually faces, vehicle scheduling, the road planning problem that faces among the material delivery process in reality is the problem that awaits solution urgently. The invention provides a material intelligent distribution method based on multi-source heterogeneous data in a manufacturing process by utilizing a multi-source heterogeneous data model in the manufacturing process aiming at the material distribution problem in the manufacturing process.
Disclosure of Invention
In the production and manufacturing process, the material delivery directly influences the production efficiency and the product manufacturing cost, and the intelligent scheduling of production tasks and the intelligent planning of material delivery roads by using the multi-source heterogeneous data model are very important. Therefore, the invention aims to provide a material intelligent distribution method based on multi-source heterogeneous data in a manufacturing process.
The invention provides an intelligent material distribution method based on multi-source heterogeneous data in a manufacturing process. The method mainly utilizes a multi-source heterogeneous data model in the production and manufacturing process to optimize task allocation and road planning in the material distribution process. The method comprises the following specific implementation steps:
the method comprises the following steps: the method comprises the steps of analyzing multi-source heterogeneous data in the manufacturing process, reading production task information, carrying vehicle information and transportation road information from a multi-source heterogeneous data model, and designing a reading method aiming at the multi-source heterogeneous data so that the reading method can be used in the analyzing process.
Step two: and (4) establishing a map model by using the transportation road information in the step one, and calculating the shortest transportation road between all the working places by using an improved ant colony algorithm. So that the distance of this shortest road can be used in the following algorithm as the transport distance between the various workplaces.
Step three: and converting the task allocation problem into a graph theory problem by using the production task information and the carrying vehicle information in the first step and the map model established in the second step, and performing optimal task allocation on each transport vehicle by combining the graph theory and an annealing algorithm.
Step four: and according to the optimal task allocation information for each transport vehicle in the third step and the map model established in the second step, the optimal road planning of each vehicle is realized through an ant colony algorithm.
Step five: and combining the optimization information in the third step and the fourth step, and outputting the optimal task allocation and the optimal road plan of each vehicle under the task constraint, the road constraint and the vehicle carrying capacity constraint.
The "production task information" in step one refers to the sequence and required capacity of the respective working processing tasks.
The "carrier vehicle information" in step one refers to the maximum capacity information that can be carried by each carrier vehicle.
The "transportation road information" in the step one refers to the coordinates of each work place, the intersection coordinates of each road, and the communication condition among each coordinate point.
The "improved ant colony algorithm" in the second step refers to an ant colony algorithm that adds mutation probability and a hierarchical evaporation strategy. The ant colony algorithm is a heuristic algorithm, the shortest path between an ant hole and food is searched by simulating the foraging process of ants, pheromones are released when the ants forage, the ants with short paths have more times of back and forth, the pheromone concentration on the shorter paths is higher, and all ants can walk to the nearest road after several iterations. The variation probability means that the path taken by each ant has a certain probability of variation, the current nearest road is not selected, the randomness is increased, and the ant is prevented from falling into the local optimum. The step evaporation strategy is to compare the path taken by the ants with the current shortest path, and the longer the road is, the faster the pheromone is evaporated, and the convergence speed of the algorithm is increased.
The "map model" described in the second step means that the coordinates of the respective work places are marked on the map, and the calculated shortest distance between the respective work places is connected to the respective work places as a straight line.
The step three of converting the task allocation problem into the graph theory problem refers to converting the task allocation problem into the problem of solving the problem of the maximum passing edge weight in the graph theory when a plurality of vehicles are needed for transportation when a plurality of tasks are needed. The conversion of the problem can be completed by forming a directed graph by using the respective work places as points, the task capacity required by the work places as the weight of the points, the process constraint as sides, and the distance between the transportation work places as the weight of the sides.
The annealing algorithm in the third step refers to a combined optimization heuristic algorithm, the probability of accepting the inferior solution is certain at different temperatures, so that the algorithm can jump out of the local optimal solution, and the probability of accepting the inferior solution is reduced along with the reduction of the temperature, so that the algorithm can be finally converged. The method is characterized in that a new solution is randomly generated by using left-right long-shift on the basis of the graph theory problem, the connectivity of tasks cannot be changed, the quality of the new solution is evaluated according to the cost, and the optimization of task allocation is finally realized.
The ant colony algorithm described in the fourth step refers to adding process constraints in the ant colony algorithm so as to satisfy the sequence requirements while optimizing. The specific method is that when the ant selects the next target point, whether the preorder task target point of the target point has already passed is judged, and if not, the target point cannot be selected as the next target point.
The invention discloses a material intelligent distribution method based on multi-source heterogeneous data in a manufacturing process, which has the advantages that:
(1) the invention relates to a method capable of realizing intelligent material distribution in a production and manufacturing process, wherein a multi-source heterogeneous data model based on the production and manufacturing process comprises production task information, carrying vehicle information and transportation road information, so that the transportation efficiency of materials can be improved, and the transportation cost of the materials can be reduced.
(2) Aiming at the problem of material distribution in the production and manufacturing process, the invention considers the road constraint of material distribution, the capacity constraint of a carrying vehicle and the sequence constraint of production tasks in the actual production process, so that the intelligent distribution algorithm is closer to the actual production process, and the material distribution cost can be effectively reduced.
Drawings
The steps of carrying out the present invention can be more readily understood by referring to the following detailed description in conjunction with the accompanying drawings, in which:
fig. 1a shows the result of finding the shortest path between the workplaces according to the present invention, where 1-14 are coordinates of the intersection and the workplace, the thin line in the figure indicates the existence of the road between the intersection and the workplace, and the thick line is the shortest road found finally. FIG. 1b shows the relationship between the optimal distances and the average distances of each generation during the iteration of the ant colony algorithm, the dotted lines show the average distances of each generation, the solid lines show the optimal distances of each generation, and the average distances do not converge to the optimal distances due to the variation in each iteration;
FIG. 2 illustrates the process of the present invention for translating the task assignment problem into the cut-through edge weight maximization problem of directed graph.
FIG. 3 illustrates the use of the present invention to establish process constraints and find optimal task allocation based on manufacturing process multi-source heterogeneous data;
FIG. 4 illustrates the optimal path for each vehicle under the process constraints given by the present invention in connection with optimal task assignment.
Fig. 5 shows a work flow of the intelligent material distribution method combining manufacturing process multi-source heterogeneous data.
The numbers, symbols and codes in the figures are explained as follows:
1 to 14: the point set converted from the finger work (see fig. 1, 2, 3 and 4)
V 1 、V 2 、V 3 Is a continuous division of the graph.
Detailed Description
The invention provides an intelligent material distribution method based on multi-source heterogeneous data of a manufacturing process. In the production and manufacturing process, the method provided by the invention can be used for realizing intelligent processing task allocation and distribution path optimization.
Examples
Experimental hardware and software conditions:
the hardware environment required for implementing the system is a Windows 10 operating system computer supporting Framework4.5 or more versions. It mainly undertakes data storage and algorithm calculation work. The map of the implementation scene is a rectangle of 1000m × 1000m, and the workplace position is shown in fig. 4.
An intelligent material distribution method based on multi-source heterogeneous data in a manufacturing process comprises the following specific implementation steps (as shown in fig. 5):
the method comprises the following steps: the method comprises the steps of analyzing multi-source heterogeneous data in the manufacturing process, reading production task information, carrying vehicle information and transportation road information from a multi-source heterogeneous data model, and designing a multi-source heterogeneous data reading method to enable the multi-source heterogeneous data reading method to be used in the analysis process, wherein in the embodiment, 10 work places, 3 transportation vehicles exist, the transportation capacity of the vehicles is 15,5 product processes, the work places are numbered to form points, and the processes are connected to form directed edges, so that a directed acyclic graph similar to the attached figure 2 can be formed.
The multi-source heterogeneous data refers to data of different types and different sources, and the multi-source heterogeneous data can be generated by different departments, different work types and different acquisition modes in the production and manufacturing process. In the invention, the used multi-source heterogeneous data mainly comprises the following data: production task information, carrier vehicle information, and haul road information. The production task information belongs to real-time information, generally exists in a semi-structured form, can reflect process constraints and task arrangement, and is the basis for realizing task allocation; the carrying vehicle information belongs to the scene information, generally exists in a structured form, can reflect the carrying capacity of the vehicle, and is an important constraint for realizing task allocation; the transportation road information belongs to scene information, generally exists in a structured form, can produce a map, realizes road constraint, and is an important component for realizing the practical use of an algorithm.
The multi-source heterogeneous data model is a data model with the same data type and structure generated after multi-source heterogeneous data are fused, and data in the data model can be directly extracted.
The method mainly refers to the existing form of the multi-source heterogeneous data in the algorithm after being read, and the algorithm stores task information through a process matrix and a task work place matrix; storing task information through a carrying capacity matrix; storing road information through a coordinate matrix formed by the coordinates of a working place and the coordinates of the intersection and a communication matrix among all coordinate points;
step two: and (3) establishing a map model by using the transportation road information in the step one, wherein the coordinates of part of the working places and the intersections are shown in a table 1, and calculating the shortest transportation road set among the working places by using an improved ant colony algorithm. So that the distance of the shortest road set can be used as the transportation distance between the various workplaces in the following algorithm, and the calculation result of the algorithm is shown in fig. 1a.
In the embodiment, the implementation steps of the "improved ant colony algorithm" are as follows:
the first step, setting the maximum iteration times to 300 times, setting the initial pheromone concentration to 1, setting the number of ants to 22, and setting the total pheromone carried by one ant to 20.
And step two, iteration is carried out, ants are placed at the starting point, and the starting point is added into a taboo table.
And step three, enabling the nodes to randomly climb to nodes which are connected with the current nodes in an edge mode and are not in the taboo table according to the transition probability, adding the nodes after transition into the taboo table, wherein the calculation formula of the transition probability is as follows:
wherein tau is ij Representing the concentration of pheromones from node i to node j; eta ij Is a heuristic function, here represented by the reciprocal of the distance between node i and node j, where j represents all feasible nodes. In this example, the induction factors α =1 and β =4.
And thirdly, judging whether the ant climbs to the terminal, if so, recording the route and the distance of the ant, and if not, returning to the second step to continuously search the next node.
Fourthly, in the embodiment, a crawling route of 0.2 probability ants is subjected to variation, a node in the ant route is randomly selected, a previous point and a next point of the node are found, whether other nodes can pass through from the previous point to the next point is judged, if yes, a variation route is selected, and the route after variation are recorded.
Fifthly, calculating the shortest route of the iteration, recording the route, updating the pheromone according to the route of each ant, and adopting a graded evaporation strategy when updating the pheromone, wherein the longer route is evaporated more quickly, the route evaporation rate of the ant route in the embodiment is 0.3 when the ant route is 1.3 times or less of the shortest route, the route evaporation rate of the ant route in the embodiment is 0.6 when the ant route is 1.3 times to 1.6 times of the shortest route, and the route evaporation rate of the ant route in the embodiment is 0.9 when the ant route is 1.6 times or more of the shortest route.
And sixthly, judging whether the maximum iteration times are reached, if not, emptying the tabu table, returning to the second step, and if the maximum iteration times are reached, searching the shortest path in each generation of optimal solution and outputting the shortest path and the path thereof.
TABLE 1 workplace and intersection coordinates
Step three: and (3) converting the task allocation problem into a continuous division problem for solving the directed acyclic graph by using the production task information, the carrying vehicle information and the map model established in the step two, and showing the coordinates and the required capacity of each working place in a table 2. And after the problem is converted, performing optimal task allocation on each transport vehicle by using an annealing algorithm. The production job information and the derived optimal allocation are shown in FIG. 3.
In the present embodiment, the implementation steps of the "annealing algorithm" are as follows:
the method comprises the steps of firstly, setting the maximum iteration times of 100 times, searching for 20 times each time in an iteration mode, setting the initial temperature of 100 degrees, setting the temperature attenuation coefficient alpha =0.95, setting the initial maximum weight value to be 0, and producing the initial working ground arrangement according to the coordinate number sequence.
And secondly, entering iteration, generating new arrangement through left and right long movement, solving the problem that the new arrangement meets the capacity constraint, calculating the weight increased by the left and right long movement through division when the side weight is maximum, and accepting a new solution if the increased weight is more than 0As an optimal solution, if the added weight is less than 0, then the probability is used
Accepting the new solution as the optimal solution, where Δ C is an increasing weight and t refers to the current temperature.
And thirdly, calculating the total weight of the cut edges and the passing edges of the optimal division under the current arrangement, comparing the total weight with the maximum weight, updating the maximum weight if the total weight is greater than the maximum weight, and recording the arrangement and the division.
And step four, judging whether the circulation times under the current temperature reach 20 times, if so, carrying out the next step, and if not, returning to the step two.
And fifthly, judging whether the iteration times are equal to the maximum generation times, if so, outputting the optimal arrangement and division, and if not, enabling t = t × α, cooling and returning to the second step.
| Numbering | X coordinate/m | Y coordinate/m | Volume/ |
| 1 | 856 | 419 | 3 |
| 2 | 484 | 776 | 1 |
| 3 | 734 | 748 | 1 |
| 4 | 216 | 330 | 2 |
| 5 | 342 | 595 | 1 |
| 6 | 685 | 527 | 3 |
| 7 | 209 | 96 | 1 |
| 8 | 458 | 855 | 1 |
| 9 | 69 | 838 | 1 |
| 10 | 407 | 912 | 1 |
Table 2 coordinates and capacities of the respective jobs.
The expression "converting task allocation problem into continuous partitioning problem for obtaining directed acyclic graph" means that a directed acyclic graph (see fig. 2) is formed by using working places as point sets and using task relationships among the working places as edge sets, and passing through V 1 、V 2 、V 3 The graph is divided into three subgraphs, and the point set contained in the three subgraphs is the work place allocated by three vehicles.
Step four: and (2) according to the optimal task allocation information (shown in the attached figure 3) for each transport vehicle in the third step and the map model (shown in the attached figure 1 a) established in the second step, realizing the optimal road planning of each vehicle through an ant colony algorithm. ( At the moment, the road of each workplace is known, but the process constraint must be met, and the final route can be connected into a closed ring, so that the starting point can not be fixed during the operation of the algorithm in the step, and the starting point can be appointed finally. The mutation probability is added in the second step to solve the problem that the fixed starting point is easy to fall into the local optimal solution, the process has no such concern, and most importantly, the process constraint is considered )
In the embodiment, the implementation steps of "realizing the optimal road planning of each vehicle through the ant colony algorithm" include:
in the first step, the maximum iteration number is set to 300, the initial pheromone concentration is set to 1, the number of ants is set to 22, the total pheromone carried by one ant is set to 20, and the evaporation rate is set to 0.2.
And step two, entering iteration, placing the ants at a random starting point, and adding the starting point into a taboo table.
And step three, enabling the nodes to randomly climb to the next node according to the transition probability, and adding a taboo table into the node after transition. The selectable next node meets two constraints that the node is not in the tabu table and the node of the preorder process is in the tabu table, and the calculation formula of the transition probability is as follows:
wherein tau is ij Representing the concentration of pheromone from the optional node i to the optional node j; eta ij Is a heuristic function, here represented by the reciprocal of the distance between node i and node j, where j represents all feasible nodes. In this example, the induction factors α =1 and β =4.
And fourthly, judging whether the ants have feasible nodes or not, if so, repeating the third step, if not, further judging whether the ants finish all the nodes, if so, recording the route and the route, otherwise, indicating that the ants enter a dead zone, and giving no pheromone punishment to the unrecorded route and the route.
And fifthly, calculating the shortest path of the iteration, recording the path of the iteration, and updating the pheromone according to the path of each ant. And sixthly, judging whether the maximum iteration times are reached, emptying the tabu table if the maximum iteration times are not reached, returning to the second step, searching the shortest route in each generation of optimal solution if the maximum iteration times are reached, and outputting the shortest route and the route.
Step five: and (4) combining the optimization information obtained in the third step and the optimization information obtained in the fourth step, outputting the optimal task allocation and the optimal road plan of each vehicle under the task constraint, the road constraint and the vehicle carrying capacity constraint, and finally generating a result as shown in fig. 4, wherein a solid line represents the path of the first vehicle, and a dotted line represents the path of the second vehicle.
Claims (5)
1. A material intelligent distribution method based on multi-source heterogeneous data in a manufacturing process is characterized in that: the method comprises the following steps:
the method comprises the following steps: analyzing multi-source heterogeneous data in the manufacturing process, reading production task information, carrying vehicle information and transportation road information from a multi-source heterogeneous data model, and designing a reading method aiming at the multi-source heterogeneous data so that the reading method can be used in the analysis process;
step two: establishing a map model by using the transportation road information in the step one, and calculating the shortest transportation road between all the working places by using an improved ant colony algorithm; the distance of the shortest road is used as the transportation distance between all the workplaces in the subsequent algorithm;
step three: converting the task allocation problem into a graph theory problem by using the production task information, the carrying vehicle information and the map model established in the second step, and performing optimal task allocation on each transport vehicle through the graph theory and an annealing algorithm;
step four: according to the optimal task allocation information for each transport vehicle in the third step and the map model established in the second step, the optimal road planning of each vehicle is realized through an ant colony algorithm;
step five: combining the optimization information obtained in the third step and the optimization information obtained in the fourth step, and outputting optimal task allocation and optimal road planning of each vehicle under task constraint, road constraint and vehicle carrying capacity constraint;
the improved ant colony algorithm in the step two is an ant colony algorithm added with variation probability and a hierarchical evaporation strategy; the variation probability means that the path taken by each ant has a certain probability of variation, the current nearest road is not selected, the randomness is increased, and the situation that the ant falls into local optimum is avoided; the step evaporation strategy is to compare the path traveled by the ants with the current shortest path, and the longer the road is, the faster the pheromone is evaporated, and the convergence speed of the algorithm is increased;
the specific steps of the ant colony algorithm improvement are as follows:
the method comprises the steps of firstly, setting the maximum iteration times to be 300 times, setting the initial pheromone concentration to be 1, setting the number of ants to be 22, and setting the total amount of pheromones carried by one ant to be 20;
step two, iteration is carried out, ants are placed at a starting point, and the starting point is added into a taboo table;
and step three, enabling the node to randomly climb to a node which is connected with the current node at an edge and is not in the tabu table according to the transition probability, adding the tabu table into the node after the transition, wherein the calculation formula of the transition probability is as follows:
wherein τ is ij Representing the concentration of pheromones from node i to node j; eta ij Is a heuristic function, here expressed as the reciprocal of the distance between node i and node j, j expressing all feasible nodes; induction factors α =1, β =4;
thirdly, judging whether the ant climbs to the terminal, if so, recording the route and the route of the ant, and if not, returning to the second step to continuously search the next node;
fourthly, carrying out variation on a crawling route of 0.2 probability ants, randomly selecting a node in the ant route, finding a front point and a rear point of the node, judging whether the front point and the rear point can pass through other nodes or not, if yes, selecting a variation route, and recording a varied route and a route;
fifthly, calculating the shortest route of the iteration, recording the route of the iteration, updating the pheromone according to the route of each ant, and adopting a graded evaporation strategy when updating the pheromone, wherein the longer route is evaporated, the route evaporation rate of the ant route is 0.3 when the ant route is 1.3 times lower than the shortest route, the route evaporation rate of the ant route is 0.6 when the ant route is 1.3 times to 1.6 times of the shortest route, and the route evaporation rate of the ant route is 0.9 when the ant route is 1.6 times higher than the shortest route;
and sixthly, judging whether the maximum iteration times are reached, if not, emptying the tabu table, returning to the second step, and if the maximum iteration times are reached, searching the shortest path in each generation of optimal solution and outputting the shortest path and the path thereof.
2. The intelligent material distribution method based on multi-source heterogeneous data of the manufacturing process, according to claim 1, is characterized in that: marking the coordinates of each work place on the map, and connecting the calculated shortest distance between the work places as a straight line to each work place; so that the distance of this shortest road set is used in the following algorithm as the transport distance between the various workplaces.
3. The intelligent material distribution method based on multi-source heterogeneous data in the manufacturing process according to claim 1, characterized in that: converting the task allocation problem into a graph theory problem, namely converting the task allocation problem into a problem of solving the maximum passing edge weight in the graph theory when a plurality of vehicles are needed for transportation when a plurality of tasks are available; taking each working place as a point, taking the task capacity required by the working place as the weight of the point, taking the process constraint as an edge, taking the distance between the transportation working places as the weight of the edge, namely forming a directed graph, and completing the conversion of the problems; and after the problem is converted, performing optimal task allocation on each transport vehicle by using an annealing algorithm.
4. The intelligent material distribution method based on multi-source heterogeneous data in the manufacturing process according to claim 1, characterized in that: the annealing algorithm in the third step is a combined optimization heuristic algorithm, the probability of accepting the inferior solution is certain at different temperatures, so that the algorithm can jump out of the local optimal solution, and the probability of accepting the inferior solution is reduced along with the reduction of the temperature, so that the algorithm can finally converge; the method is characterized in that a new solution is randomly generated by using left-right long-shift on the basis of the graph theory problem, the connectivity of tasks cannot be changed, the quality of the new solution is evaluated according to the cost, and the optimization of task allocation is finally realized; the method comprises the following steps:
the method comprises the steps of firstly, setting maximum iteration times of 100 times, searching times of 20 times each time in an iteration mode, setting an initial temperature of 100 ℃, setting a temperature attenuation coefficient alpha =0.95, setting an initial maximum weight value to be 0, and producing the initial working ground arrangement according to a coordinate numbering sequence;
and secondly, entering iteration, generating new arrangement through left and right long movement, calculating the weight increased by the left and right long movement through division when the weight of the edge is maximum when the new arrangement meets the capacity constraint, accepting a new solution as an optimal solution if the increased weight is greater than 0, and taking probability as the optimal solution if the increased weight is less than 0
Accepting a new solution as an optimal solution, wherein Δ C is an increased weight and t refers to a current temperature;
thirdly, calculating the total weight of the cut edges and the passing edges of the optimal division under the current arrangement, comparing the total weight with the maximum weight, updating the maximum weight if the total weight is greater than the maximum weight, and recording the arrangement and the division;
step four, judging whether the cycle times at the current temperature reach 20 times, if so, carrying out the next step, and if not, returning to the step two;
and fifthly, judging whether the iteration times are equal to the maximum iteration times, if so, outputting optimal arrangement and division, and if not, enabling t = t multiplied by alpha, cooling and returning to the second step.
5. The intelligent material distribution method based on multi-source heterogeneous data of the manufacturing process, according to claim 1, is characterized in that: the ant colony algorithm described in the fourth step is to add process constraints in the ant colony algorithm so that optimization is performed and sequence requirements are met at the same time, and the specific method is to judge whether a preorder task target point of the target point passes through when an ant selects a next target point, and if not, the target point cannot be selected as the next target point.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011405500.5A CN112508270B (en) | 2020-12-04 | 2020-12-04 | Intelligent material distribution method based on multi-source heterogeneous data in manufacturing process |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011405500.5A CN112508270B (en) | 2020-12-04 | 2020-12-04 | Intelligent material distribution method based on multi-source heterogeneous data in manufacturing process |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112508270A CN112508270A (en) | 2021-03-16 |
| CN112508270B true CN112508270B (en) | 2023-04-07 |
Family
ID=74968501
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011405500.5A Active CN112508270B (en) | 2020-12-04 | 2020-12-04 | Intelligent material distribution method based on multi-source heterogeneous data in manufacturing process |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112508270B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2159974A1 (en) * | 2008-08-27 | 2010-03-03 | Alcatel, Lucent | Method and equipment for establishing a connection |
| CN104700251A (en) * | 2015-03-16 | 2015-06-10 | 华南师范大学 | Maximum-minimum ant colony optimization method and maximum-minimum ant colony optimization system for solving vehicle scheduling problem |
| CN110705741A (en) * | 2019-08-19 | 2020-01-17 | 浙江工业大学 | Multi-distribution center vehicle path optimization method based on improved ant colony algorithm |
| CN110852469A (en) * | 2019-08-27 | 2020-02-28 | 广西大学 | Method, system and medium for optimizing design of medicine logistics distribution path |
-
2020
- 2020-12-04 CN CN202011405500.5A patent/CN112508270B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2159974A1 (en) * | 2008-08-27 | 2010-03-03 | Alcatel, Lucent | Method and equipment for establishing a connection |
| CN104700251A (en) * | 2015-03-16 | 2015-06-10 | 华南师范大学 | Maximum-minimum ant colony optimization method and maximum-minimum ant colony optimization system for solving vehicle scheduling problem |
| CN110705741A (en) * | 2019-08-19 | 2020-01-17 | 浙江工业大学 | Multi-distribution center vehicle path optimization method based on improved ant colony algorithm |
| CN110852469A (en) * | 2019-08-27 | 2020-02-28 | 广西大学 | Method, system and medium for optimizing design of medicine logistics distribution path |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112508270A (en) | 2021-03-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109034468B (en) | Logistics distribution path planning method with time window based on cuckoo algorithm | |
| CN110909930A (en) | Goods position distribution method of mobile goods shelf storage system for refrigeration house | |
| CN109764882B (en) | Multi-target vehicle path planning method based on self-adaptive local search chain | |
| US20220156693A1 (en) | Computerized system and method for developing optimized cargo transportation solutions | |
| CN110334838B (en) | AGV trolley cooperative scheduling method and system based on ant colony algorithm and genetic algorithm | |
| CN111709560A (en) | Method for solving vehicle path problem based on improved ant colony algorithm | |
| CN109840551B (en) | Method for optimizing random forest parameters for machine learning model training | |
| CN106651086B (en) | Automatic stereoscopic warehouse scheduling method considering assembly process | |
| CN107886196B (en) | Bicycle scheduling method for goods taking and delivering | |
| CN115454070B (en) | K-Means ant colony algorithm multi-robot path planning method | |
| CN113935452A (en) | Distribution center goods picking path planning method based on ant colony algorithm and genetic algorithm fusion | |
| CN112508270B (en) | Intelligent material distribution method based on multi-source heterogeneous data in manufacturing process | |
| CN114253975B (en) | Load-aware road network shortest path distance calculation method and device | |
| Qu et al. | Analysis of distribution path optimization algorithm based on big data technology | |
| CN107464013A (en) | One kind is applied to the non-traditional layout warehouse aisles Optimization Design of V-type | |
| CN114625137A (en) | AGV-based intelligent parking path planning method and system | |
| CN117875189B (en) | Three-dimensional warehouse space layout method based on GA optimization GRO | |
| Wu et al. | Optimization of Order‐Picking Problems by Intelligent Optimization Algorithm | |
| CN117973657A (en) | Combined optimization method for order batch and picking paths of logistics distribution center | |
| CN117455087B (en) | Logistics energy-saving control method and system based on Internet of things | |
| Jiang et al. | Optimal schedule for agricultural machinery in sequential tasks using a multi-population co-evolutionary non-dominant neighbor immune algorithm | |
| CN116907490A (en) | Multi-robot path planning method based on conflict search | |
| CN116859919A (en) | Multi-AGV path planning method based on deep reinforcement learning and cyclic neural network | |
| CN117035897A (en) | Instant delivery platform rider unilateral method | |
| Guo et al. | [Retracted] Transportation Path Optimization of Modern Logistics Distribution considering Hybrid Tabu Search Algorithm |
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 |