CN113627642A - Stacker path optimization method based on self-adaptive large-scale neighborhood search algorithm - Google Patents
Stacker path optimization method based on self-adaptive large-scale neighborhood search algorithm Download PDFInfo
- Publication number
- CN113627642A CN113627642A CN202110714678.6A CN202110714678A CN113627642A CN 113627642 A CN113627642 A CN 113627642A CN 202110714678 A CN202110714678 A CN 202110714678A CN 113627642 A CN113627642 A CN 113627642A
- Authority
- CN
- China
- Prior art keywords
- solution
- stacker
- task
- operator
- total time
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 48
- 238000005457 optimization Methods 0.000 title claims abstract description 37
- 238000010845 search algorithm Methods 0.000 title claims abstract description 18
- 230000000694 effects Effects 0.000 claims abstract description 23
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 17
- 230000008569 process Effects 0.000 claims abstract description 17
- 238000003780 insertion Methods 0.000 claims abstract description 14
- 230000037431 insertion Effects 0.000 claims abstract description 14
- 238000002922 simulated annealing Methods 0.000 claims abstract description 11
- 230000003044 adaptive effect Effects 0.000 claims description 11
- 238000009826 distribution Methods 0.000 claims description 8
- 239000011159 matrix material Substances 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 4
- 230000008859 change Effects 0.000 claims description 3
- 230000007423 decrease Effects 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 238000012545 processing Methods 0.000 claims description 3
- 230000010076 replication Effects 0.000 claims description 3
- 230000006378 damage Effects 0.000 abstract description 3
- 238000012360 testing method Methods 0.000 description 4
- 241000208125 Nicotiana Species 0.000 description 2
- 235000002637 Nicotiana tabacum Nutrition 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 230000002776 aggregation Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 238000000137 annealing Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000007599 discharging Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- XULSCZPZVQIMFM-IPZQJPLYSA-N odevixibat Chemical class C12=CC(SC)=C(OCC(=O)N[C@@H](C(=O)N[C@@H](CC)C(O)=O)C=3C=CC(O)=CC=3)C=C2S(=O)(=O)NC(CCCC)(CCCC)CN1C1=CC=CC=C1 XULSCZPZVQIMFM-IPZQJPLYSA-N 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/04—Constraint-based CAD
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/10—Numerical modelling
Abstract
The invention relates to the technical field of path planning, and discloses a stacker path optimization method based on a self-adaptive large-scale neighborhood search algorithm, wherein at the beginning of the algorithm, a score and a weight are given to each operation operator, a plurality of removal operators and insertion operators are randomly selected according to the score and the weight of the operator during each iteration, the current solution is subjected to 'destruction' and 'reconstruction', a new solution is generated, whether the current solution and the optimal solution are updated by the new solution or not is judged according to the effect of the new solution and a simulated annealing receiving criterion, the optimal solution is output until an iteration condition is met, each operation operator has a certain probability to be selected in the iteration process, a larger search space is provided, the possibility of falling into the local optimal solution is smaller, the possibility of obtaining the global optimal solution is increased, the weight of each operation operator is changed in the iteration process, and the operation operators with good effect are gradually compared and screened in the iteration process, the algorithm has high flexibility, the solving time is shortened, and the efficiency is high.
Description
Technical Field
The invention relates to the technical field of path planning, in particular to a stacker path optimization method based on a self-adaptive large-scale neighborhood search algorithm.
Background
With the rapid development of the logistics transportation industry, the tobacco transfer and distribution requirements are continuously increased, in order to better meet the transportation requirements, more and more tobacco distribution centers select advanced logistics equipment to replace manpower, the cost is reduced, and meanwhile, the operation efficiency is improved. The automatic three-dimensional warehouse is a novel storage mode in the technical field of modern logistics, and the main body of the automatic three-dimensional warehouse comprises a goods shelf, a roadway type stacking crane, a warehouse-in and warehouse-out workbench and an automatic conveying, discharging and operation control system. The high-rise rationalization, automatic access and simple and convenient operation of the warehouse are realized by using stereoscopic warehouse equipment enterprises. In the whole system of the automatic stereoscopic warehouse, the key of the scheduling optimization type improvement of the warehouse operation efficiency of the stacker is realized, so that the research on the scheduling problem of the stacker of the automatic stereoscopic warehouse has important theoretical value and practical significance.
The Chinese patent CN106773686A (published as 2018, 12 and 28) discloses a method for establishing a scheduling path model of a stacker under a same-rail double-vehicle operation mode, the same-rail double-vehicle scheduling path model established by the method takes the time required by two stackers to jointly complete a task sequence as an evaluation standard, a batch of tasks are reasonably distributed to the two stackers by a general distribution principle to ensure that the time required by the task sequence is shortest and avoid collision, when the scheduling path of the stacker is optimized and solved, a Chemical Reaction Optimization (CRO) is adopted, a mode of indirect integer coding of a goods space coordinate by a task number is adopted, the general distribution principle is embedded into each iteration of an algorithm, the distribution result is adjusted, the task sequence is changed, and the optimal solution of the scheduling path model of the stacker is found. In the chemical reaction optimization algorithm, the solution of the solved problem is described by adopting a molecular structure, in the four collision modes, invalid collisions between single molecules and multiple molecules occur nearby the original molecules, so that the solution is easy to fall into local optimization, and the method disclosed by the patent has the disadvantages of low global search capability, low convergence speed and low precision.
Disclosure of Invention
The invention aims to provide a good-flexibility and high-efficiency stacker path optimization method based on a self-adaptive large-scale neighborhood search algorithm.
In order to achieve the aim, the invention provides a stacker path optimization method based on a self-adaptive large-scale neighborhood search algorithm, which comprises the following steps:
s1, establishing a corresponding mathematical optimization model by taking the minimum total time of the stacker for completing all tasks as a target;
s2, sequentially encoding the operation tasks, and generating a time matrix required by the stacker to move between any two goods positions according to the distribution of the goods positions in the warehouse;
s3, taking the sequence of the received tasks as an initial operation sequence, generating an algorithm initial solution, storing the initial solution as a current solution and an optimal solution, and giving initial scores and weights to all operation operators;
s4, randomly selecting a removal operator according to the initial score and the weight given to the operator so as to destroy the current solution; then randomly selecting an insert operator to reconstruct the current solution to generate a new solution;
s5, updating the current solution or the optimal solution by the new solution according to the effect of the new solution and the simulated annealing acceptance criterion: obtaining the total time consumption of the new solution, the current solution and the optimal solution according to the time matrix of the step S2, and if the total time consumption of the new solution is less than the total time consumption of the current solution and the optimal solution, updating the current solution and the optimal solution by using the new solution; if the total time consumption of the new solution is larger than the total time consumption of the optimal solution and smaller than the total time consumption of the current solution, the optimal solution is unchanged, and the current solution is updated by the new solution; if the total time consumption of the new solution is greater than the total time consumption of the optimal solution and the current solution, updating the current solution according to the simulated annealing acceptance criterion;
s6, updating the scores and weights of the corresponding operators according to the effect of the new solution;
and S7, repeating the steps S4, S5 and S6 to iterate until the preset maximum iteration number is reached or the preset maximum non-updating iteration number is reached, stopping iteration, and outputting the stored optimal solution, wherein the optimal solution at the moment is the optimal operation sequence obtained through optimization.
Preferably, in step S4, the removed operator is randomly selected by roulette.
Preferably, in step S4, the operator is removed by the maximal cost-saving removal and the random removal.
Preferably, in step S4, the operator is inserted using a minimum incremental cost insertion and a maximum regret value insertion.
Preferably, in step S1, the objective function of the mathematical optimization model is:
min t＝∑i∈A∑j∈Atijxij，
wherein min t is the total time for the stacker to complete all tasks; t is tijIndicating the time, x, required for the stacker to move from the location of task i to the location of task jijA variable of 0-1 indicates that 1 is selected if the stacker finishes the task j after the task i; otherwise, 0 is selected;
the constraint conditions include:
(4)
the constraint condition (1) indicates that the goods space corresponding to each task is accessed only once and the access degree is equal;
wherein x isjiIs changed from 0 to 1Quantity, which means that if the stacker finishes the task i after the task j, 1 is selected, otherwise, 0 is selected; a represents all tasks and initial position sets, A ═ DU P {0 '}, wherein D represents an in-warehouse task set, P represents an out-warehouse task set, and 0' represents an in-warehouse platform and a replication point thereof;
(5)
constraint (2) represents sub-loop cancellation,
wherein s isiIndicating the position of task i in the entire task sequence, sjIndicating the position of task j in the whole task sequence, M indicates more than 104The number of (1);
the decision variables have the value ranges of:
preferably, the constraint condition further includes:
(6)
the constraint condition (3) represents the change of the loading capacity of the stacker and the load limit of the stacker in the process of completing the task,
wherein u isiIndicating the number of pieces of goods, u, carried by the stacker when it leaves the task i cargo spacejRepresenting the number of pieces of cargo carried by the stacker when it leaves the task j cargo space.
Preferably, in step S2, the calculation method of the time required for the stacker to move between the two positions i and j includes:
wherein, aiLayer in coordinates representing the place corresponding to task i, biColumns in coordinates representing the positions corresponding to task i, l represents the length of each position, h represents the height of each position, vxIndicating the horizontal movement rate, v, of the stackeryRepresenting the vertical movement rate of the stacker.
Preferably, in step S5, when the total time consumption of the new solution is greater than the total time consumption of the optimal solution and the current solution, the method further comprises
Updates the current solution with the new solution, where f(s)new) Representing the total time spent by the new solution, f(s)cur) T is temperature, T represents the total time spent by the current solution0＝0.85*f(sinitial)，f(sinitial) For the initial solution, the temperature is measured with a parameter λ (0) for each generation<λ<1) The rate of (c) decreases.
Preferably, in step S6, if the new solution is better than the optimal solution, the corresponding operator adds σ1Dividing; if the new solution effect is better than the current solution, adding sigma to the corresponding operator2Dividing; if the effect of the new solution is inferior to that of the current solution, but the current solution is updated, adding sigma to the corresponding operator3Dividing; the updated operator score is
According to the following steps:
recalculating operator weights, where ωpRepresenting the weight of an operator p, and theta represents the importance degree of a new round of score when the weight is calculated;
and, according to:
and carrying out normalization processing on the weights.
As a preferred embodiment, σ1＞σ2＞σ3。
Compared with the prior art, the invention has the beneficial effects that:
the invention assigns scores and weights to each operator at the beginning of the algorithm, randomly selects a plurality of operators to remove and insert according to the initial scores and weights assigned to the operators during each iteration, carries out 'destruction' and 'reconstruction' to generate a new solution, judging whether to update the current solution and the optimal solution by the new solution according to the effect of the new solution and the simulated annealing acceptance criterion until the optimal solution is output after the iteration condition is met, in the iteration process, each operation operator has certain probability to be selected, has larger search space and smaller possibility of trapping local optimum, increases the possibility of obtaining global optimum solution, moreover, the weight of each operation operator is changed in the iteration process, and the operation operators with good effect are gradually screened out in the iteration process, so that the flexibility of the algorithm is high, the solving time is shortened, and the efficiency is high.
Drawings
Fig. 1 is a flowchart of a stacker path optimization method according to an embodiment of the present invention.
Fig. 2 is a schematic diagram of a stacker path optimization method according to an embodiment of the present invention.
Fig. 3 is a statistical chart of an exemplary test of the stacker path optimization method according to the embodiment of the present invention.
Detailed Description
The following detailed description of embodiments of the present invention is provided in connection with the accompanying drawings and examples. The following examples are intended to illustrate the invention but are not intended to limit the scope of the invention.
As shown in fig. 1 and fig. 2, a stacker path optimization method based on an adaptive large-scale neighborhood search algorithm in a preferred embodiment of the present invention includes the following steps:
s1, establishing a corresponding mathematical optimization model by taking the minimum total time of the stacker for completing all tasks as a target.
The stacker needs to complete a plurality of tasks, so the total operation time of the stacker is the time for completing all tasks in sequence, the sequence of task completion is different, a plurality of different operation schemes can be formed, the total operation time of each operation scheme is also different, and the embodiment is the operation scheme with the shortest total operation time.
The objective function of the mathematical optimization model in this embodiment is:
min t＝∑i∈A∑j∈Atijxij，
wherein min t is the total time for the stacker to complete all tasks; t is tijIndicating the time, x, required for the stacker to move from the location of task i to the location of task jijA variable of 0-1 indicates that 1 is selected if the stacker finishes the task j after the task i; otherwise, 0 is selected;
the constraint conditions include:
(1)
the constraint condition (1) indicates that the goods space corresponding to each task is accessed only once and the access degree is equal;
wherein x isjiA variable of 0-1 indicates that 1 is taken if the stacker finishes the task i after the task j, and 0 is taken if not; a represents all tasks and initial position sets, A ═ DU P {0 '}, wherein D represents an in-warehouse task set, P represents an out-warehouse task set, and 0' represents an in-warehouse platform and a replication point thereof;
(2)
constraint (2) represents sub-loop cancellation,
wherein s isiIndicating the position of task i in the entire task sequence, sjIndicating the position of task j in the whole task sequence, M indicates more than 104M is a very large number, 10 is taken in this example5；
The decision variables have the value ranges of:
the constraints further include:
(3)
the constraint condition (3) represents the change of the loading capacity of the stacker and the load limit of the stacker in the process of completing the task, wherein uiIndicating the number of pieces of goods, u, carried by the stacker when it leaves the task i cargo spacejRepresenting the number of pieces of cargo carried by the stacker when it leaves the task j cargo space.
S2, sequentially coding the job tasks, and generating a time matrix required by the stacker to move between any two goods according to the distribution of the goods in the warehouse.
And (3) encoding the operation tasks by adopting a sequential encoding mode, wherein the operation scheme is represented as a group of numbered sequences so as to reflect the sequence of the completion of the operation, for example, if a certain task comprises 6 warehouse-in and warehouse-out tasks which are numbered 1-6 respectively, 05316240 represents that the stacker starts from the warehouse-in and warehouse-out platform and then sequentially completes the corresponding tasks in the sequence of 5-3-1-6-2-4 and returns to the warehouse-in and warehouse-out platform.
And then eliminating the sub-loop according to the constraint condition (3) and distributing according to the goods positions in the automatic three-dimensional bin. The operation scheme comprises a plurality of tasks, the essence of the tasks is that the tasks are moved from one goods space to another, an aggregation matrix of time required for moving between any two goods spaces is generated, and according to the tasks in the operation scheme, the moving time of each goods space corresponding to the tasks is summed to obtain the total operation time of the operation scheme.
The method for calculating the time required for the stacker to move between the two positions i and j comprises the following steps:
wherein, aiLayer in coordinates representing the place corresponding to task i, biColumns in coordinates representing the positions corresponding to task i, l represents the length of each position, h represents the height of each position, vxIndicating the horizontal movement rate, v, of the stackeryRepresenting the vertical movement rate of the stacker.
And S3, taking the sequence of the received tasks as an initial operation sequence, generating an algorithm initial solution, saving the initial solution into a current solution and an optimal solution, and simultaneously giving initial scores and weights to all the operation operators.
In the embodiment, initially, pi is given to all operation operatorspWhen the weight of each operator is equal.
S4, randomly selecting a removal operator according to the initial score and the weight given to the operator so as to destroy the current solution; the insertion operator is then randomly selected to "reconstruct" the current solution, generating a new solution.
In this embodiment, the removed operator is randomly selected by a roulette method, the higher the score of the operator is, the larger the weight occupied by the operator is, the higher the probability of being selected is, a specified number of points are removed according to the rule of the selected removed operator, the current solution is "destroyed", and then the removed points are sequentially reinserted according to the rule of the selected inserted operator to form a new solution.
The removal of operators employs maximum cost-effective removal and random removal. The maximum cost-saving removal means sequentially removing points which can reduce the objective function (the total time consumed for completing the task) most, and the specific process is as follows: the distance between the front and rear points in the solution sequence is calculated first, the point which is farthest from the front and rear points, i.e. the point which takes the longest time to move, is compared and selected, and is removed, with the aim of adjusting the operation sequence in the direction of reducing the total time consumption. The random removal refers to randomly selecting a specified number of points to remove the points from the sequence, and aims to add certain randomness in the searching process, avoid the searching from being trapped in local optimum and enable the searching process to cover more operation sequencing possibilities.
The insertion of the operator adopts the minimum increase cost insertion and the maximum regret value insertion. The minimum increase cost insertion means that a point is preferentially inserted at a position where the increase of the objective function (the total time consumed for completing the task) is minimum, and the specific process is as follows: the amount of increase in elapsed time due to the insertion of each point into each position is calculated, the position where the increase in elapsed time is the smallest is compared and selected, and the point is inserted into the position, so that the work order is adjusted in the direction of decreasing the total elapsed time. The maximum regret value insertion refers to preferentially inserting points at positions which enable the increase of the objective function to be minimum and the degree to be less than that of other positions, and the specific process is as follows: the method comprises the steps of firstly calculating the increment of consumed time caused by inserting each point into each position, calculating the difference of the increment of the consumed time when each point is inserted into different positions, comparing and selecting the point with the largest difference, and inserting the point into the position with the smallest increment of the consumed time, wherein the aim is to avoid the condition that the solution quality is influenced by the sequence of the inserted points.
S5, updating the current solution or the optimal solution by the new solution according to the effect of the new solution and the simulated annealing acceptance criterion: obtaining the total time consumption of the new solution, the current solution and the optimal solution according to the time matrix of the step S2, and if the total time consumption of the new solution is less than the total time consumption of the current solution and the optimal solution, updating the current solution and the optimal solution by using the new solution; if the total time consumption of the new solution is larger than the total time consumption of the optimal solution and smaller than the total time consumption of the current solution, the optimal solution is unchanged, and the current solution is updated by the new solution; and if the total time consumption of the new solution is greater than the total time consumption of the optimal solution and the current solution, updating the current solution according to the simulated annealing acceptance criterion.
The simulated annealing acceptance criteria for this example are: when the total time consumption of the new solution is greater than the total time consumption of the optimal solution and the current solution, so as to
Updates the current solution with the new solution, where f(s)new) Representing the total time spent by the new solution, f(s)cur) T is temperature, T represents the total time spent by the current solution0＝0.85*f(sinitial)，f(sinitial) For the initial solution, the temperature is measured with a parameter λ (0) for each generation<λ<1) The rate of (c) decreases.
The new solution, the current solution and the optimal solution are all the operation sequences of all tasks. If the total time consumption of the generated new operation sequence is less than the obtained optimal operation sequence, updating the obtained optimal operation sequence and the operation sequence used by the current iteration by using the new operation sequence; if the total time consumption of the generated new operation sequence is larger than the obtained optimal operation sequence and smaller than the operation sequence used by the current iteration, updating the operation sequence used by the current iteration by using the new operation sequence; if the total time consumption of the generated new operation sequence is larger than the operation sequence used by the current iteration, according to the simulated annealing acceptance criterion, the simulation annealing is carried out
Updates the current iteration-used job order with the new job order.
And S6, updating the scores and weights of the corresponding operators according to the effect of the new solution.
If the effect of the new solution is better than that of the optimal solution, adding sigma to the corresponding operator1Dividing; if the new solution effect is better than the current solution, adding sigma to the corresponding operator2Dividing; if the effect of the new solution is inferior to that of the current solution, but the current solution is updated, adding sigma to the corresponding operator3Dividing; the updated operator score is
σ of the embodiment1＞σ2＞σ3。
According to the following steps:
recalculating operator weights, where ωpRepresenting the weight of an operator p, and theta represents the importance degree of a new round of score when the weight is calculated;
and, according to:
and carrying out normalization processing on the weights.
And S7, repeating the steps S4, S5 and S6 to iterate until the preset maximum iteration number is reached or the preset maximum non-updating iteration number is reached, stopping iteration, and outputting the stored optimal solution, wherein the optimal solution at the moment is the optimal operation sequence obtained through optimization.
The embodiment further includes step S8, and optimizing the effect through an example test.
The goods shelf comprises 2 rows, 42 columns and 10 layers of goods positions, each goods position is 1.5m long and 2 m high, the moving speed of the stacker in the horizontal direction is 1.5 m/s, the moving speed of the stacker in the vertical direction is 0.5 m/s, and the stacker can carry one goods at a time.
Randomly generating a batch of warehouse-in and warehouse-out mixed tasks and corresponding goods space coordinates thereof. Classifying the number of the warehouse-in and warehouse-out tasks into I types according to the number of the warehouse-in and warehouse-out tasks in the calculation example, namely, taking the warehouse-in tasks as the main tasks (the number of the warehouse-in tasks is about 2 times of the number of the warehouse-out tasks); o type, namely taking the ex-warehouse tasks as the main (the number of ex-warehouse tasks is about 2 times of the number of in-warehouse tasks); IO class, i.e. equalization tasks (roughly equal number of in and out of library tasks). The examples contain 6 scales of small, medium and large, the total number of tasks is 30, 100, 200, 300, 400 and 500 respectively, and each scale contains 10 of 3 types of examples.
Table 1 is an example test result table, which shows average total consumption time of tasks completed by the I-type, O-type, and IO-type examples before and after the operation sequence optimization of the stacker, and their difference values. Each table contains 6 scales of operators, and the result corresponding to each scale is the average value of the results of 10 operators on the scale. As can be seen from the table, the optimization of the operation sequence can reduce the total time consumption for completing tasks by more than 20% regardless of the type of the calculation example, and the optimization of the operation sequence is meaningful.
Table 1: results of the example test
The line graphs of the optimization effects of the I, O and IO category algorithms are shown in FIG. 3, and it can be seen from the graph that the optimization effects of the I category algorithms and the O category algorithms are equivalent, while the optimization effect of the IO category algorithms is the best, and for the IO category algorithms, the percentage of time consumption reduction caused by path optimization increases with the increase of the scale of the algorithms.
To sum up, the embodiment of the present invention provides a stacker path optimization method based on an adaptive large-scale neighborhood search algorithm, which assigns a score and a weight to each operator at the beginning of the algorithm, randomly selects a plurality of removal operators and insertion operators according to the initial score and the weight of the operator at each iteration, performs "destruction" and "reconstruction" on the current solution to generate a new solution, and judges whether to update the current solution and the optimal solution with the new solution according to the effect of the new solution and a simulated annealing acceptance criterion until the iteration condition is satisfied and then outputs the optimal solution, wherein each operator has a certain probability to be selected in the iteration process, has a larger search space, has a smaller possibility of being trapped in a local optimal solution, increases the possibility of obtaining a global optimal solution, and the weight of each operator is changed in the iteration process, and the operators with good effect are gradually compared and screened in the iteration process, the algorithm has high flexibility, the solving time is shortened, and the efficiency is high.
The above description is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, various modifications and substitutions can be made without departing from the technical principle of the present invention, and these modifications and substitutions should also be regarded as the protection scope of the present invention.
Claims (10)
1. The stacker path optimization method based on the self-adaptive large-scale neighborhood search algorithm is characterized by comprising the following steps of:
s1, establishing a corresponding mathematical optimization model by taking the minimum total time of the stacker for completing all tasks as a target;
s2, sequentially encoding the operation tasks, and generating a time matrix required by the stacker to move between any two goods positions according to the distribution of the goods positions in the warehouse;
s3, taking the sequence of the received tasks as an initial operation sequence, generating an algorithm initial solution, storing the initial solution as a current solution and an optimal solution, and giving initial scores and weights to all operation operators;
s4, randomly selecting a removal operator according to the initial score and the weight given to the operator so as to destroy the current solution; then randomly selecting an insert operator to reconstruct the current solution to generate a new solution;
s5, updating the current solution or the optimal solution by the new solution according to the effect of the new solution and the simulated annealing acceptance criterion: obtaining the total time consumption of the new solution, the current solution and the optimal solution according to the time matrix of the step S2, and if the total time consumption of the new solution is less than the total time consumption of the current solution and the optimal solution, updating the current solution and the optimal solution by using the new solution; if the total time consumption of the new solution is larger than the total time consumption of the optimal solution and smaller than the total time consumption of the current solution, the optimal solution is unchanged, and the current solution is updated by the new solution; if the total time consumption of the new solution is greater than the total time consumption of the optimal solution and the current solution, updating the current solution according to the simulated annealing acceptance criterion;
s6, updating the scores and weights of the corresponding operators according to the effect of the new solution;
and S7, repeating the steps S4, S5 and S6 to iterate until the preset maximum iteration number is reached or the preset maximum non-updating iteration number is reached, stopping iteration, and outputting the stored optimal solution, wherein the optimal solution at the moment is the optimal operation sequence obtained through optimization.
2. The stacker path optimizing method based on adaptive large-scale neighborhood search algorithm of claim 1, wherein in step S4, the removed operator is randomly selected by roulette method.
3. The stacker path optimization method based on adaptive large-scale neighborhood search algorithm according to claim 1 or 2, wherein in step S4, operator removal adopts maximum cost-saving removal and random removal.
4. The stacker path optimizing method based on adaptive large-scale neighborhood search algorithm of claim 3, wherein in step S4, the insertion of the operator uses the minimum incremental cost insertion and the maximum regret value insertion.
5. The stacker path optimizing method based on adaptive large-scale neighborhood search algorithm according to claim 1, wherein in step S1, the objective function of the mathematical optimization model is:
min t＝∑i∈A∑j∈Atijxij，
wherein min t is the total time for the stacker to complete all tasks; t is tijIndicating the time, x, required for the stacker to move from the location of task i to the location of task jijA variable of 0-1 indicates that 1 is selected if the stacker finishes the task j after the task i; otherwise, 0 is selected;
the constraint conditions include:
(1)
the constraint condition (1) indicates that the goods space corresponding to each task is accessed only once and the access degree is equal;
wherein x isjiIs a variable of 0 to 1, tableIf the stacker finishes the task i after the task j, 1 is selected, otherwise, 0 is selected; a represents all tasks and initial position sets, A ═ DU P {0 '}, wherein D represents an in-warehouse task set, P represents an out-warehouse task set, and 0' represents an in-warehouse platform and a replication point thereof;
(2)
constraint (2) represents sub-loop cancellation,
wherein s isiIndicating the position of task i in the entire task sequence, sjIndicating the position of task j in the whole task sequence, M indicates more than 104The number of (1);
the decision variables have the value ranges of:
6. the stacker path optimization method based on adaptive large-scale neighborhood search algorithm according to claim 5, wherein the constraint condition further comprises:
(3)
the constraint condition (3) represents the change of the loading capacity of the stacker and the load limit of the stacker in the process of completing the task,
wherein u isiIndicating the number of pieces of goods, u, carried by the stacker when it leaves the task i cargo spacejRepresenting the number of pieces of cargo carried by the stacker when it leaves the task j cargo space.
7. The method for optimizing the path of a stacker crane based on the adaptive large-scale neighborhood search algorithm according to claim 5, wherein in step S2, the calculation method of the time required for the stacker crane to move between the two positions i and j is as follows:
wherein, aiLayer in coordinates representing the place corresponding to task i, biColumns in coordinates representing the positions corresponding to task i, l represents the length of each position, h represents the height of each position, vxIndicating the horizontal movement rate, v, of the stackeryRepresenting the vertical movement rate of the stacker.
8. The stacker path optimizing method based on adaptive large-scale neighborhood search algorithm of claim 1, wherein in step S5, when the total time consumption of the new solution is greater than the total time consumption of the optimal solution and the current solution, the method further comprises
Updates the current solution with the new solution, where f(s)new) Representing the total time spent by the new solution, f(s)cur) T is temperature, T represents the total time spent by the current solution0＝0.85*f(sinitial)，f(sinitial) For the initial solution, the temperature is measured with a parameter λ (0) for each generation<λ<1) The rate of (c) decreases.
9. The stacker path optimization method based on adaptive large-scale neighborhood search algorithm according to claim 1, wherein the stacker path optimization method is characterized in thatIf the new solution is better than the optimal solution, the correspondence operator adds σ in step S61Dividing; if the new solution effect is better than the current solution, adding sigma to the corresponding operator2Dividing; if the effect of the new solution is inferior to that of the current solution, but the current solution is updated, adding sigma to the corresponding operator3Dividing; the updated operator score is
According to the following steps:
recalculating operator weights, where ωpRepresenting the weight of an operator p, and theta represents the importance degree of a new round of score when the weight is calculated;
and, according to:
and carrying out normalization processing on the weights.
10. The stacker path optimization method based on adaptive large-scale neighborhood search algorithm of claim 9, wherein σ is1＞σ2＞σ3。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110714678.6A CN113627642A (en) | 2021-06-25 | 2021-06-25 | Stacker path optimization method based on self-adaptive large-scale neighborhood search algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110714678.6A CN113627642A (en) | 2021-06-25 | 2021-06-25 | Stacker path optimization method based on self-adaptive large-scale neighborhood search algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113627642A true CN113627642A (en) | 2021-11-09 |
Family
ID=78378434
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110714678.6A Pending CN113627642A (en) | 2021-06-25 | 2021-06-25 | Stacker path optimization method based on self-adaptive large-scale neighborhood search algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113627642A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115609593A (en) * | 2022-12-02 | 2023-01-17 | 天津大学 | Multi-type robot man-machine cooperative assembly scheduling method and device |
| CN116629586A (en) * | 2023-07-24 | 2023-08-22 | 青岛民航凯亚系统集成有限公司 | Airport guarantee vehicle scheduling method and system based on ALNS |
| CN117474486A (en) * | 2023-11-03 | 2024-01-30 | 北京交通大学 | Method for optimizing construction period of repeated construction engineering |
| CN117687554A (en) * | 2023-12-11 | 2024-03-12 | 上海梅斯医药科技有限公司 | Scale element flexible configuration system and method based on visual simulation scoring |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107025363A (en) * | 2017-05-08 | 2017-08-08 | 中国人民解放军国防科学技术大学 | A kind of adaptive big neighborhood search method of Agile satellite scheduling |
| CN107450563A (en) * | 2017-09-21 | 2017-12-08 | 景德镇陶瓷大学 | A kind of adaptive information feedback population robot path system of selection and system based on more subgroups |
| CN111798067A (en) * | 2020-07-17 | 2020-10-20 | 大连理工大学 | Automatic driving automobile distribution path planning method based on self-adaptive large neighborhood search algorithm |
| CN212181514U (en) * | 2020-04-21 | 2020-12-18 | 广东烟草惠州市有限责任公司 | Intelligent spare part system of tobacco commodity circulation |
| CN112257999A (en) * | 2020-10-10 | 2021-01-22 | 东南大学 | Self-adaptive large-scale neighborhood searching method for large-scale pure electric bus scheduling problem |
| CN112800384A (en) * | 2021-01-25 | 2021-05-14 | 西安工业大学 | GTSP solving algorithm based on self-adaptive large-field search |
-
2021
- 2021-06-25 CN CN202110714678.6A patent/CN113627642A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107025363A (en) * | 2017-05-08 | 2017-08-08 | 中国人民解放军国防科学技术大学 | A kind of adaptive big neighborhood search method of Agile satellite scheduling |
| CN107450563A (en) * | 2017-09-21 | 2017-12-08 | 景德镇陶瓷大学 | A kind of adaptive information feedback population robot path system of selection and system based on more subgroups |
| CN212181514U (en) * | 2020-04-21 | 2020-12-18 | 广东烟草惠州市有限责任公司 | Intelligent spare part system of tobacco commodity circulation |
| CN111798067A (en) * | 2020-07-17 | 2020-10-20 | 大连理工大学 | Automatic driving automobile distribution path planning method based on self-adaptive large neighborhood search algorithm |
| CN112257999A (en) * | 2020-10-10 | 2021-01-22 | 东南大学 | Self-adaptive large-scale neighborhood searching method for large-scale pure electric bus scheduling problem |
| CN112800384A (en) * | 2021-01-25 | 2021-05-14 | 西安工业大学 | GTSP solving algorithm based on self-adaptive large-field search |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115609593A (en) * | 2022-12-02 | 2023-01-17 | 天津大学 | Multi-type robot man-machine cooperative assembly scheduling method and device |
| CN116629586A (en) * | 2023-07-24 | 2023-08-22 | 青岛民航凯亚系统集成有限公司 | Airport guarantee vehicle scheduling method and system based on ALNS |
| CN117474486A (en) * | 2023-11-03 | 2024-01-30 | 北京交通大学 | Method for optimizing construction period of repeated construction engineering |
| CN117687554A (en) * | 2023-12-11 | 2024-03-12 | 上海梅斯医药科技有限公司 | Scale element flexible configuration system and method based on visual simulation scoring |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113627642A (en) | Stacker path optimization method based on self-adaptive large-scale neighborhood search algorithm | |
| CN110909930B (en) | Goods position distribution method of mobile goods shelf storage system for refrigeration house | |
| CN106779153B (en) | Intelligent stereoscopic warehouse goods space allocation optimization method | |
| CN107480922B (en) | Method for establishing goods position distribution scheduling model under two-end type same-rail double-vehicle running mode | |
| CN104408589B (en) | AGV Optimization Schedulings based on Hybrid Particle Swarm | |
| CN114417696B (en) | Automatic stereoscopic warehouse cargo space distribution optimization method based on genetic algorithm | |
| CN110543953A (en) | Multi-target disassembly line setting method under space constraint based on wolf colony algorithm | |
| CN111815233B (en) | Goods position optimization method based on total logistics amount and energy consumption | |
| Al-Hawari et al. | Development of a genetic algorithm for multi-objective assembly line balancing using multiple assignment approach | |
| CN112286152B (en) | Distributed flow shop scheduling method and system with batch delivery constraint | |
| CN113033895A (en) | Multi-source multi-point path planning method, equipment and storage medium | |
| CN113762899A (en) | Mixed algorithm-based three-dimensional cargo boxing method | |
| Shang et al. | Spatial scheduling optimization algorithm for block assembly in shipbuilding | |
| CN113408771B (en) | Article delivery scheduling method, device and storage medium | |
| Zhong et al. | Adaptive autotuning mathematical approaches for integrated optimization of automated container terminal | |
| CN114444843A (en) | Agricultural product green logistics distribution vehicle scheduling method and system based on large-scale variable neighborhood search strategy | |
| CN115935616A (en) | Multi-objective optimization method for scheduling of sequence-dependent flow shop groups of consistent batches | |
| CN110942251B (en) | Batch scheduling method based on joint ant colony algorithm | |
| Hani et al. | Simulation based optimization of a train maintenance facility | |
| CN113960971B (en) | Flexible workshop scheduling method based on behavior decision network particle swarm optimization | |
| CN111507644A (en) | Multi-point unloading constrained three-dimensional multi-container loading method | |
| CN113141272B (en) | Network security situation analysis method based on iteration optimization RBF neural network | |
| Yu et al. | A hybrid genetic algorithm for integrated truck scheduling and product routing on the cross-docking system with multiple receiving and shipping docks | |
| CN112418528B (en) | Amphibious vehicle layout area utilization maximization method based on multi-strategy dynamic adjustment | |
| Kucuk et al. | A new hybrid evolutionary algorithm for three-dimensional packing problems |
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 |