CN115730789A - ASRS task scheduling and goods allocation method and system under classified storage - Google Patents
ASRS task scheduling and goods allocation method and system under classified storage Download PDFInfo
- Publication number
- CN115730789A CN115730789A CN202211422100.4A CN202211422100A CN115730789A CN 115730789 A CN115730789 A CN 115730789A CN 202211422100 A CN202211422100 A CN 202211422100A CN 115730789 A CN115730789 A CN 115730789A
- Authority
- CN
- China
- Prior art keywords
- warehouse
- goods
- task
- tasks
- sequence
- 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
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Abstract
The invention discloses an ASRS task scheduling and goods space allocation method and a system under classified storage.A goods shelf is partitioned according to goods characteristics, a model of in-out warehouse task scheduling and goods space allocation considers the dynamic allocation of goods spaces, the model is a dynamic planning model comprising in-out warehouse task stacker allocation, warehouse-in task ordering, warehouse-out task ordering and state transfer of goods space sets, an integer planning model is established by taking a hamming distance matched with a minimized warehouse-in and warehouse-out task as a target to perform warehouse-in task ordering, a warehouse-in task order optimization model is embedded into a cultural genetic algorithm to realize the integral optimization of the problems of in-out warehouse task ordering, allocation and goods space selection, and the local search of the cultural genetic algorithm comprises two goods space exchange operators; the invention optimizes the sequence of the ex-warehouse and the warehouse-in tasks, empty goods positions generated by the ex-warehouse tasks can be used by subsequent warehouse-in operations, and simultaneously, the optimal solution is solved by a method combining global search and local search, thereby reducing the task completion time and the delay time.
Description
Technical Field
The invention relates to a warehouse task scheduling and distributing method and system, in particular to an ASRS task scheduling and goods allocation method under classified storage.
Background
An automatic Storage/Retrieval System (ASRS) has the advantages of high space utilization rate, low labor cost, high cargo warehousing speed and the like, and is generally applied to distribution centers and other fields. The classified storage strategy is one of the storage strategies commonly used in the ASRS, the goods are placed in a partitioned mode by considering goods entering and exiting frequency or attribute characteristics, and the like, and a random storage strategy is adopted in each goods area. The ASRS task sequencing and goods space allocation integrated optimization means that a warehouse-out task list and an empty goods space set are given, the warehouse-out tasks are sequenced, and a proper goods space is selected for the warehouse-in tasks.
At present, a First-Come-First-Served (FCFS) strategy is adopted for a warehousing task sequence, and the influence of warehousing task sequencing under the constraint of a cargo area on the operation efficiency of a stacker, particularly the operation distance of the stacker and the delay of ex-warehouse tasks, is not considered; in the prior art, research is carried out based on a single roadway, an ASRS (automatic sequence reporting system) comprises a plurality of roadways in reality, and few documents consider the problem of stacker allocation based on a multi-roadway global angle at present; in addition, in the prior art, the improvement of ASRS operation efficiency is taken as an optimization target, the condition that the warehouse-out task has cut-off time is less considered, the customer order delivery is delayed due to the fact that the operation efficiency is pursued once, the customer satisfaction is lowered, and the enterprise competitiveness is reduced.
Disclosure of Invention
The invention aims to: the invention aims to provide an ASRS task scheduling and goods allocation method under classified storage for reducing task completion time and delay time; the second purpose of the invention is to provide an ASRS task scheduling and cargo space allocation system under the classification storage which can reduce the task completion time and the delay time.
The technical scheme is as follows: according to the ASRS task scheduling and goods location allocation method under classified storage, a goods shelf is partitioned according to goods characteristics, and the goods characteristics comprise goods in-out frequency and goods attributes; the model of the warehouse-in and warehouse-out task scheduling and the goods location distribution is a dynamic planning model comprising warehouse-in and warehouse-out task stacker distribution, warehouse-in task sequencing, warehouse-out task sequencing and state transfer of a goods location set; empty goods positions generated by the ex-warehouse task can be used by subsequent warehousing tasks; each stroke of the stacker is a warehouse entering and exiting stroke which comprises a warehouse entering task and a warehouse exiting task; the method for scheduling the warehouse entry and exit tasks and allocating the goods space comprises the following steps:
aiming at minimizing the completion time of the warehouse-in and warehouse-out process and the delay time of the warehouse-out task, solving the dynamic planning model by adopting a cultural genetic algorithm, and outputting the warehouse-out and warehouse-in tasks and the processing sequence which are executed by each stacker and the optimal solution of the goods space corresponding to each warehouse-in and warehouse-out task when the termination condition is met;
the cultural genetic algorithm obtains the warehousing task sequence and the allocation of the pilers by solving an assignment problem model taking a minimized warehousing-in-and-out task matching metric index as a target in global optimization according to the initial warehousing-out task sequence and the allocation of the pilers;
the cultural genetic algorithm further optimizes the warehousing task sequence and the ex-warehouse task sequence by utilizing local search; the local search comprises two goods space exchange operators which respectively represent the exchange of the warehousing goods space and the ex-warehouse goods space of two ex-warehouse strokes on a stacker, and during each local search, one of the two goods space exchange operators is selected to perform local search according to the historical expression of the two goods space exchange operators to form a new ex-warehouse stroke.
Further, the matching measurement index of the ex-warehouse task and the in-warehouse task is a novel hamming distance, and the formula of the novel hamming distance is as follows:
wherein
In order to make the delivery task goods area sequence,
for warehousing task cargo area sequences, a i ∈Ω，b i E is omega, omega represents a cargo area sequence, the cargo area is sequentially divided into p types of cargo areas from near to far according to the distance from an I/O port, and omega = (omega is used for expressing the cargo area sequence) 1 ，Ω2...Ω p ). By using the novel Hamming distance as a matching measurement index of the ex-warehouse task and the in-warehouse task, the problem of sorting the in-warehouse tasks is converted into an assignment problem to be solved.
Furthermore, the dynamic planning model divides stages according to the number of the warehouse-out tasks, and each stage is based on the current goods position set state and selects the goods position for the warehouse-in and warehouse-out tasks by solving an integer planning model taking the completion time of the minimized warehouse-in and warehouse-out journey and the delay time of the warehouse-out tasks as targets; and after the single stage is finished, updating the goods position set state, and transferring to the next stage until all the warehouse-in and warehouse-out tasks are executed.
Further, the cultural gene algorithm adopts a genetic algorithm to perform global search, and the method comprises the following steps: carrying out chromosome coding on the ex-warehouse task sequence and the distribution of the stacking machine to generate an initial ex-warehouse task sequence and the distribution of the stacking machine;
selecting goods positions for the warehouse-in and warehouse-out tasks according to the chromosome coding and the warehouse-in task sequence and the allocation of a stacker; selecting an ex-warehouse goods position which is closest to the I/O port and contains ex-warehouse goods in a goods area to which the ex-warehouse task belongs, and selecting an empty goods position which can minimize the completion time of an in-warehouse and out-warehouse stroke in the goods area to which the in-warehouse task belongs; and after the distribution of the goods positions for one warehouse-in and warehouse-out stroke is finished, updating the states of the goods positions until the distribution of all the goods positions is finished.
Further, the method for further optimizing the warehousing task sequence and the ex-warehouse task sequence by using local search comprises the following steps:
performing local search on the initial population in the cultural genetic algorithm and the crossed and mutated individuals, and using P (N) = c in each local search N /∑ N∈{1，2} c N Selecting a goods space interchange operator to perform local search, wherein N belongs to {1,2}, represents two goods space interchange operators, P (N) is the probability of selecting the goods space interchange operator, and c (N) is the probability of selecting the goods space interchange operator N The counter is used for recording the historical expression of the goods space exchange operator, and when the solution after local search is performed by using a certain goods space exchange operator is superior to the current solution, the counter is increased by one.
Further, in the cultural genetic algorithm, the inverse number of a target function of minimizing the completion time of the warehouse-in and warehouse-out journey and the delay time of the warehouse-out task is used as a fitness function, and the optimal solution of the cultural genetic algorithm is the solution which enables the fitness function value to be maximum when the termination condition is met.
The ASRS task scheduling and goods space allocation system under classified storage comprises:
the model establishing unit is used for establishing a dynamic planning model of the distribution of the warehouse entry and exit task stacker, the sorting of the warehouse entry tasks, the sorting of the warehouse exit tasks and the state transfer of the goods location set, and empty goods locations generated by the warehouse exit tasks can be used by subsequent warehouse entry operation; each stroke of the stacker is a warehouse-in and warehouse-out stroke which comprises a warehouse-in task and a warehouse-out task; the goods shelves are partitioned according to goods characteristics, wherein the goods characteristics comprise goods in-out frequency and goods attributes;
the model solving unit is used for solving the dynamic planning model by adopting a cultural genetic algorithm with the aim of minimizing the completion time of the warehouse-in and warehouse-out process and the delay time of the warehouse-out task, and outputting the warehouse-out and warehouse-in tasks and the processing sequence which are executed by each stacker and the optimal solution of the goods space corresponding to each warehouse-in and warehouse-out task when the termination condition is met;
the cultural genetic algorithm determines an initial ex-warehouse task sequence and stacker allocation through population initialization in global optimization, and obtains an in-warehouse task sequence and stacker allocation through solving an assignment problem model taking a minimized ex-warehouse task matching metric index as a target;
the cultural genetic algorithm further optimizes the warehousing task sequence and the ex-warehouse task sequence by utilizing local search; the local search comprises two goods position exchange operators which respectively represent the exchange of the warehousing goods position and the ex-warehouse goods position of two ex-warehouse strokes on a stacker, and during each local search, one of the two goods position exchange operators is selected to perform local search according to the historical performance of the two goods position exchange operators, so that a new in-warehouse stroke is formed.
The electronic device comprises a memory, a processor and a computer program which is stored on the memory and can run on the processor, wherein when the computer program is loaded to the processor, the ASRS task scheduling and goods space allocation method under the classified storage is realized.
The computer-readable storage medium of the present invention stores a computer program, wherein the computer program is executed by a processor to implement the ASRS task scheduling and cargo space allocation method under any one of the classified storage.
Has the beneficial effects that: compared with the prior art, the invention has the advantages that: (1) Considering dynamic allocation and classified storage of goods positions, proposing a dynamic planning model combined with an integer planning model to describe problems, and enabling empty goods positions generated by the ex-warehouse task to be used by subsequent warehousing operation; (2) By combining global search with a local search strategy based on an ex-warehouse goods space exchange operator, the algorithm can search in a neighborhood space generating a better solution, and the completion time of all tasks and the delay time of ex-warehouse tasks are reduced on the whole; (3) According to the shelf partition characteristics, the novel Hamming distance is used as a matching measurement index of the warehousing and ex-warehousing tasks, the warehousing task ordering subproblems are converted into classical assignment problems to be solved, the complexity of solving the task ordering problem is reduced, poor stroke is avoided, and quick calculation is realized; (4) And integrating the warehouse-in and warehouse-out task scheduling problem and the goods allocation problem in a dynamic planning model to realize the integrated optimization of the two problems.
Drawings
Fig. 1 is a top view of an automated stereoscopic warehouse in an embodiment of the present invention.
Fig. 2 is a front view of a shelf in an embodiment of the invention.
FIG. 3 is a flowchart of the warehouse entry task scheduling and cargo space allocation method of the present invention.
FIG. 4 is a schematic diagram of the warehousing cargo space exchange operator according to the present invention.
FIG. 5 is a diagram illustrating the ex-warehouse goods location swap operator according to the present invention.
FIG. 6 is a graph comparing the solution quality of the method of the present invention with discrete ICA and FCFS in an embodiment of the present invention.
FIG. 7 is a graph comparing the average computation time of the method of the present invention and discrete ICA in accordance with an embodiment of the present invention.
Detailed Description
The technical scheme of the invention is further explained by combining the attached drawings.
As shown in fig. 1, the research object of the ASRS task scheduling and cargo space allocation method and system under classified storage according to the present invention is an ASRS using a classified storage strategy, a tunnel is provided between every two rows of shelves in the ars, and a stacker operates on each tunnel. The stacker can only carry one unit of goods at a time and its entrance and exit (I/O) are located at the front end of each lane. In the embodiment, each shelf is divided into four regions, and the regions can also be divided according to actual requirements, as shown in fig. 2. The stacker adopts a cross access mode to execute the warehouse entry and exit tasks, namely, one stroke is used for completing warehouse entry and warehouse exit, and each stroke is a warehouse entry and exit stroke. For warehousing operation, unloading transported warehoused goods and temporarily storing the warehoused goods in a warehousing buffer area, transporting the goods to an assigned tunnel conveyor belt station by a forklift, placing a transport robot in charge of the tunnel on an automatic conveyor belt to be transported to an I/O port, and then executing internal warehousing operation by a corresponding stacker. Otherwise, for the warehouse-out operation, the stacker takes the warehouse-out goods out of the corresponding goods position, places the warehouse-out goods on the I/O port conveyor belt, and conveys the warehouse-out goods to the general warehouse-out port through the conveyor belt to finish warehouse-out. The ex-warehouse tasks are driven by customer orders and must typically be completed before a given cutoff time to ensure that the truck can depart within a specified time window during subsequent shipping delivery, etc. The problem of the present invention can be represented by a triplet symbol as: [ F, para | IO 2 ，zone|∑C i ，∑T i ]Where para denotes a plurality of parallel stackers.
In the ASRS adopting the classified storage strategy, (1) as the warehousing and ex-warehouse tasks have the goods area constraint, the sequencing of the warehousing and ex-warehouse tasks plays a key role in improving the operation efficiency of the stacker and reducing the task delay time. When the execution sequence of the warehouse-in and warehouse-out tasks is unreasonable, the operation time of the stacker is prolonged, and the task completion time and the delay time are prolonged. In order to improve the warehouse-in and warehouse-out efficiency and the customer satisfaction, the invention not only considers the sorting of the warehouse-out tasks, but also optimizes the sequence of the warehouse-in tasks. (2) In the invention, the ASRS of multiple lanes is considered, goods to be delivered out of the warehouse and empty goods positions are arranged on each shelf, namely, delivery tasks can be distributed to any stacker for processing, so that the optimal stacker distribution of delivery tasks is considered. (3) The goods in and out warehouse have a plurality of optional goods positions in the area to which the goods belong, so that the proper goods positions need to be allocated for the goods in and out warehouse task. To better utilize storage space, the present invention allows for dynamic allocation of cargo space, allowing for reuse of empty locations created by ex-warehouse operations.
In summary, the present invention decomposes the problem into the following four sub-problems: the method comprises the steps of sorting the warehousing tasks, sorting the ex-warehouse tasks, distributing stackers for the ex-warehouse tasks and selecting the ex-warehouse goods space, wherein the first two subproblems realize the pairing of the warehousing tasks. The four sub-problems are associated and influence each other, and the problem solving is more complex. To clarify the problem, the present invention makes the following assumptions:
● The warehouse has no stock shortage condition, and in order to ensure the formation of warehouse-in and warehouse-out journey, the warehouse-in and warehouse-out tasks are equal in quantity.
● The same type of goods do not exist in the warehousing task queue, and if the goods exist, the goods can be directly taken out from the warehousing queue.
● The initial state of the partition and the goods position of the goods shelf is known in advance, and the deadline of the goods area of the warehousing and warehousing task and the deadline of the warehousing and warehousing task are known in advance.
● The stacker can move vertically and horizontally at the same time, and the speed is constant. The time for executing the warehouse-in and warehouse-out journey is calculated by using the Chebyshev metric method commonly used in the literature, and the time for executing the access operation at the goods position does not influence the optimization result,
neglected.
As shown in fig. 3, the ASRS task scheduling and cargo space allocation method under classified storage includes the following steps:
(1) Establishing a mathematical model for scheduling of warehouse-in and warehouse-out tasks and allocating goods space
The invention considers the dynamic allocation of goods space, namely the empty goods space generated by the ex-warehouse task can be used by the subsequent warehousing operation, the allocation mode of the goods space can lead the empty goods space set which is taken as the input parameter of the model and the goods space set containing the ex-warehouse goods to change along with the decision of each goods space selection, and the problem of the invention is difficult to be completely carved by a single integer programming model. The dynamic planning model divides the phases according to the number of the ex-warehouse tasks, namely, each time the out-warehouse and in-warehouse journey operation is executed, the operation can be regarded as one phase. In each stage, based on the current goods space set state, the goods space is selected for the in-out task and the in-out travel completion time and the out-of-warehouse task delay time are calculated by solving an integer programming model taking the minimized current task completion and delay time as targets. And after the single stage is finished, updating the goods position set state, and then transferring to the next stage until all tasks are executed.
The stages in the dynamic planning model are divided according to the number of the outbound tasks, the n outbound tasks represent n stages, and the state variable of the dynamic planning model is
C mb The decision variable is
The state transition equation is:
the recursion formula of the optimal value function of the dynamic programming model is as follows:
wherein f is b (r, S) represents a minimum target value when the r-th ex-warehouse task is arranged after including the b task sets S; r =1,2,. N; b =1,2,. N-1;
as a boundary condition, the r-th out-warehouse task is arranged in an empty set
A minimum target value at a later time; r =1,2.
In formula (4)
Is shown in a state of
C mb The completion time and delay time of the time-th warehouse-in and warehouse-out journey are calculated by the following integer programming model, wherein
C mb Input parameters for the integer programming model.
The integer programming model is:
Minimize Obj r ＝ω·C m(b+1) +(1-ω)·TT r (6)
equation (6) is an objective function, i.e. minimizing the completion time and delay time of the r-th in-out warehouse trip; constraint (7) ensures that only one warehousing and ex-warehouse goods space is selected to execute warehousing and ex-warehouse travel operation; constraints (8) and (9) ensure that goods to be warehoused are stored in empty positions in the area to which the task belongs, and the warehouse-out goods position is a goods position containing warehouse-out goods in the area to which the task belongs; constraint (10) determines the time of the warehousing journey, the value of which is calculated by the Chebyshev formula; constraint (11) determines the completion time of the warehousing journey; the constraint (12) represents the delay time of the outbound task.
(2) Cultural gene algorithm solving model
For the warehousing task sequence optimization subproblems, on the premise that the ex-warehouse task sequence is known, a novel ex-warehouse matching index based on Hamming distance is constructed, the subproblems are converted into assignment problems, and the warehousing task sequence optimization is realized by solving the assignment problems. And (5) for the sub-problems of sorting and distribution of ex-warehouse tasks and goods space selection, optimizing and solving by adopting a cultural gene algorithm.
(2.1) warehousing task ordering
On the premise of giving the ex-warehouse task sequence, the invention takes the matching degree of the ex-warehouse and in-warehouse tasks into consideration, and optimizes the in-warehouse task sequence so as to reduce the walking distance of the stacker. According to the method, the initial ex-warehouse task sequence and the stacker allocation are determined through population initialization, the goods area characteristics of classified storage are considered, a novel ex-warehouse task matching measurement index based on the Hamming distance is introduced, and the matching efficiency can be effectively improved.
Hamming Distance (Hamming Distance) is a Distance measurement, i.e. if there are two character strings with equal length, the number of different characters at the same position is the Hamming Distance. In the ASRS adopting a classified storage strategy, the time required by the stacker to execute the warehouse entry and warehouse exit journey of the same or similar cargo area is shorter, so that the matching degree of tasks in the same or similar cargo area is high, and the matching degree of tasks in the cargo areas far away from each other is low. The traditional Hamming distance adopts exclusive OR operation, and can not reflect the matching degree of the classified storage and delivery interval. Therefore, the invention adaptively adapts the Hamming distance calculation mode, and the distance between the cargo region and the I/O port is defined by the rank of the cargo region, thereby converting the cargo region type into sequential data capable of reflecting the difference of the degree. Accordingly, the novel hamming distance for in-out task matching based on cargo zone rank is defined as follows:
defining (a novel Hamming distance H matched with warehouse-in and warehouse-out tasks based on the rank of a goods area under classified storage) and sequentially dividing the goods area of a goods shelf into p types of goods areas from near to far according to the distance I/O port, wherein omega represents the sequence of the goods area, and omega = (omega) 1 ，Ω 2 ...Ω p ) And the sequence of the distance between the cargo area and the I/O port is defined as the Rank of the cargo area, namely Rank (omega) p ) = p, the farther the cargo area is from the I/O port, the larger the rank. Order of goods area of outbound taskIs listed as
The warehousing task cargo area sequence is
Then the hamming distance H for the two outbound and inbound task matches is calculated as equation (13):
the Hamming distance measurement index under the classified storage can quickly evaluate the matching degree of the warehousing and ex-warehouse tasks, so that the warehousing task ordering subproblems can be converted into assignment problems. Wherein the weight degree assigned between tasks is measured based on H index, thereby constructing to minimize
An assignment problem model for the target, the assignment problem model being a warehousing ordering integer programming model:
x ro ∈{0，1} r，o＝1，2，...，n (17)
the formula (14) is an objective function, the constraint (15) shows that one warehousing task can only be matched with one ex-warehouse task, the constraint (16) shows that one ex-warehouse task can only be matched with one warehousing task, and the constraint(17) Is a binary decision variable, x ro And the result is 1, the r warehouse-out task and the o warehouse-in task are paired to form a warehouse-in and warehouse-out stroke, and otherwise, the result is 0. And the model solving result can be converted into a warehousing task sequence with the optimal warehousing-in and warehousing-out stroke matching degree.
Taking 6 ex-warehouse and 6 in-warehouse tasks and 4 cargo areas as examples, the cargo area sequence Ω = (a, B, C, D), then Rank (Ω) = (1, 2,3, 4). Cargo area sequence of ex-warehouse task
Initial sequence of cargo area to which warehousing task belongs
At this time, the hamming distance matched with the warehouse-in and warehouse-out journey of the warehouse-in and warehouse-out task is as follows:
optimizing the warehousing task sequence by solving the warehousing sorting integer programming model to obtain the warehousing task cargo area sequence matched with the optimal warehousing and ex-warehouse journey
The Hamming distance is:
(2.2) solving model combining global search and local search
A genetic algorithm is adopted as a global search strategy, and a local search strategy based on warehouse-in and warehouse-out goods space exchange is provided according to problem characteristics.
The global search strategy comprises:
(1) chromosome coding and population initialization
A complete coding scheme includes the stacker allocation of ex-warehouse tasks and the processing order thereof. The invention numbers the ex-warehouse tasks as 1,2.. N according to the original sequence, sets the chromosome code as an integer sequence consisting of 1-n without repeated numbers, and then evenly inserts z-1 numbers which are more than n to represent the distribution of the stacker (z is the number of the stacker). The coding strategy ensures the balance of the number of the tasks processed by each stacker, and improves the algorithm efficiency. For example, the encoding of one solution for 6 ex-warehouse tasks and 2 stackers can be expressed as: [2 4 5 7 16 ], this scheme shows that the ex-warehouse tasks numbered 2,4,5 are assigned to the first stacker process in this order, and the ex-warehouse tasks numbered 1,6,3 are assigned to the second stacker process in this order. In order to ensure the diversity of the population, the algorithm initial population is generated in a random mode. Meanwhile, to give a good initial solution to the algorithm, a solution generated by EDD (Early Due Date) rules is inserted in the population.
(2) Dynamic allocation of cargo space under sorted storage
The invention provides an effective warehouse-in and warehouse-out goods space allocation strategy, which selects goods spaces for warehouse-in and warehouse-out tasks according to chromosome codes and warehouse-in task sequences. The method comprises the steps of firstly selecting an ex-warehouse goods position which is closest to an I/O port and contains ex-warehouse goods in a goods area of an ex-warehouse task, and then selecting an empty goods position which can enable the time of an in-warehouse trip to be minimum in the goods area of an in-warehouse task. Once the warehousing goods location is determined, the warehousing travel time and the delay of the ex-warehouse task can be calculated. And after the goods positions are distributed to a group of warehouse-in and warehouse-out strokes, updating the goods position state, and continuously selecting the goods positions for the subsequent warehouse-in and warehouse-out strokes according to the strategy.
(3) Fitness evaluation and population management strategy
The objective function of the invention is to minimize the completion time of the warehouse-in and warehouse-out task and the delay time of the warehouse-out task, therefore, the reciprocal of the objective function is taken as the fitness function, and is defined as a formula (18), wherein pop is the population size, C is the population size m For all task completion times, TT, on stacker m r Is the delay time of the r-th ex-warehouse task. The invention adopts an elite reservation strategy to reserve part of excellent chromosomes in the male parent and combine with part of excellent chromosomes in the offspring to form a new population.
(4) Crossover and mutation operations
The invention adopts a roulette strategy to select proper individuals from the population for cross and variation operation, and the probability Prob of individual selection p Calculated by equation (19), and the crossover and mutation operations employ a partial mapping crossover operator and an interpolation operator, respectively.
The local search strategy comprises the following steps:
the invention provides two goods space exchange operators to respectively carry out local search on the initial population, the crossed and mutated individuals so as to further optimize the solution. At each local search, an adaptive selection mechanism is adopted to select two operators LO N And N is equal to {1,2}, one of the N is selected for local search. Design counter c N The method is used for recording the historical performances of two operators, the initial value of each counter is set to be 1, and when the solution obtained after local search is carried out by using one operator is superior to the current solution, the counter of the operator is increased by one. The invention uses the roulette selection strategy to select operators, namely, the probability of selecting the operator with better historical performance is higher, and the operator selection formula is as follows: p (N) = c N /∑ N∈{1，2} c N 。
The two operators are described in detail as follows:
(1) Warehouse storage position exchange operator LO 1 : the warehousing goods positions of two warehousing-in and warehousing-out strokes are exchanged on one stacker, as shown in fig. 4. The exchange can change the sequence of the warehousing tasks and the sequence of the ex-warehouse tasks is not changed, so that a new ex-warehouse stroke is formed, and the sequence of the warehousing tasks can be further optimized.
(2) Warehouse-out goods position exchange operator LO 2 : the delivery positions of two delivery routes are exchanged on a stacker, as shown in fig. 5. The exchange can change the sequence of the ex-warehouse tasks, and the sequence of the in-warehouse tasks is not changed, so that a new in-warehouse and out-warehouse process is formed, and the ex-warehouse task sequence can be further optimized.
Because the invention adopts dynamic goods position distribution, a priority relationship may exist between two goods positions in the solution, namely, the delivery goods position of one delivery journey may be the storage goods position of the subsequent delivery journey, and if the sequence of the two goods positions is changed, the solution is not feasible. To improve algorithm efficiency, the local search of the present invention is performed only within the feasible domain of the solution.
(3) And outputting the warehouse-out and warehouse-in tasks and the processing sequence thereof which are executed by each stacker and the optimal solution of the goods space corresponding to each warehouse-in and warehouse-out task after meeting the termination condition, wherein the termination condition is that the maximum iteration times are reached.
The symbol descriptions are shown in the following table:
the method of the present invention is verified by experiments below.
In this embodiment, the shelf size is set to 10 × 10, and the shelf section is as shown in fig. 2. The number of the lanes in the warehouse is set to be 3, and the total number of the lanes is 100 multiplied by 6=600 cargo spaces; meanwhile, the utilization rate of each goods area on each shelf is 80%, each goods area contains 5 types of goods, and the state of each goods area is randomly generated; the total quantity N of the warehouse-in and warehouse-out tasks is set to be 40, 80 and 120, wherein 40% of the warehouse-out tasks are goods in a goods area A, 30% of the warehouse-out tasks are goods in a goods area B, 20% of the warehouse-out tasks are goods in a goods area C, and 10% of the warehouse-out tasks are goods in a goods area D. Deadline of ex-warehouse task is
[min{p r |r∈R}，(2·(1-γ)·∑ r∈R p r +min{p r |r∈R})/z]Is uniformly generated within the interval, wherein p r The travel time of the warehouse-in and warehouse-out for the task is calculated by adopting a goods space selection strategy according to the initial sorting of the warehouse-in and warehouse-out, gamma describes the tightness degree of the deadline, the smaller the tightness degree of the deadline is, the looser the deadline is, and the gamma is set to be {0.6,0.7 and 0.8} in the embodiment.
The termination conditions are two, (1) the maximum iteration number is set as 100 in the experiment; (2) the algorithm has no improvement on the maximum number of iterations, which is set to 50 in this experiment.
In the embodiment, an FCFS algorithm commonly used by an enterprise and a discrete ICA algorithm (discrete imperial competition algorithm) for solving the task scheduling problem of the automated stereoscopic warehouse in the prior art are used as a comparison scheme, and an FCFS policy is used for averagely allocating tasks to stackers according to the arrival sequence of the tasks, namely, the ordering of the tasks entering and exiting the warehouse and the allocation of the stackers are not considered; the results of the mass-contrast algorithm under each combination of examples (γ _ N) are shown in fig. 6, where CT is the task completion time, tar is the task delay time, obj is the target value,
GAP1＝100％*(Alg_CT-Matheuristic_CT)/Matheuristic_CT
GAP2＝100％*(Alg_Tar-Matheuristic_Tar)/Matheuristic_Tar
GAP3＝100％*(Alg_Obj-Matheuristic_Obj)/Matheuristic_Obj
matheuristic refers to the method of the present invention, and Alg stands for discrete ICA or FCFS algorithm.
The average calculation time of the method of the present invention combined with the discrete ICA for each algorithm (γ _ N) is shown in fig. 7.
From fig. 6 and 7, it can be seen that: (1) The invention is superior to the discrete ICA in terms of solving quality, and the improvement amount is increased along with the increase of the number of tasks, and the average improvement amount is 16.63%. The method is superior to discrete ICA and FCFS in terms of task completion time and task delay. The invention realizes the optimization of the warehousing task sequence by introducing the Hamming distance measurement index matched with the warehousing tasks based on the goods area rank, realizes the pairing of the warehousing tasks by combining a method combining global search and local search, and can avoid the formation of poor warehousing and ex-warehouse strokes by combining the characteristics of classified storage compared with the method adopting a 'hitching windmill' strategy for optimizing the warehousing task sequence in the ICA, thereby reducing the execution time and delay time of the tasks. Meanwhile, the method combining global search and local search can search in the neighborhood space generating better solution by combining the global search strategy based on genetic algorithm and the local search strategy based on the warehouse-in and warehouse-out goods space exchange operator.
(2) The calculation time of the invention is significantly lower than that of the discrete ICA, because compared with the discrete ICA, the invention uses the distance between specific goods positions as the matching index, and the invention can realize rapid calculation by sequencing the warehousing tasks by considering the warehousing and ex-warehouse task matching index based on the goods region rank. Meanwhile, in the embodiment, although Gurobi is adopted to solve the integer programming model, the matching index is adopted for pre-calculation in the algorithm implementation process, so that multiple times of cyclic calculation of the integer programming model are avoided, and the overall time complexity and the calculation time of the algorithm are reduced.
Based on the same inventive concept, the ASRS task scheduling and cargo space allocation system under classified storage comprises:
the model establishing unit is used for establishing a dynamic planning model for the distribution of the warehouse entry task stacker, the sorting of the warehouse entry tasks, the sorting of the warehouse exit tasks and the state transfer of the goods space set, and empty goods spaces generated by the warehouse exit tasks can be used by subsequent warehouse entry operation; each stroke of the stacker is a warehouse entering and exiting stroke which comprises a warehouse entering task and a warehouse exiting task; the goods shelf is partitioned according to goods characteristics, wherein the goods characteristics comprise goods in-out frequency and goods attributes;
the model solving unit is used for solving the dynamic planning model by adopting a cultural genetic algorithm with the aim of minimizing the completion time of the warehouse-in and warehouse-out process and the delay time of the warehouse-out task, and outputting warehouse-out and warehouse-in tasks and the processing sequence which are executed by each stacker in charge and the optimal solution of the goods space corresponding to each warehouse-in and warehouse-out task when the termination condition is met;
the cultural genetic algorithm determines an initial ex-warehouse task sequence and stacker allocation through population initialization in global optimization, and obtains an in-warehouse task sequence and stacker allocation through solving an assignment problem model taking a minimized ex-warehouse task matching metric index as a target;
the cultural genetic algorithm further optimizes the warehousing task sequence and the ex-warehouse task sequence by utilizing local search; the local search comprises two goods space exchange operators which respectively represent the exchange of the warehousing goods space and the ex-warehouse goods space of two ex-warehouse strokes on a stacker, and during each local search, one of the two goods space exchange operators is selected to perform local search according to the historical expression of the two goods space exchange operators to form a new ex-warehouse stroke.
Based on the same inventive concept, the electronic device of the present invention includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program, when loaded into the processor, implements the ASRS task scheduling and cargo space allocation method under the classified storage.
Based on the same inventive concept, the computer-readable storage medium of the present invention stores a computer program, and is characterized in that the computer program, when executed by a processor, implements the ASRS task scheduling and cargo space allocation method under the classified storage.
The computer-readable storage medium can include RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, flash memory, or any other medium that can be used to store desired program code in the form of instructions or data structures and that can be accessed by a computer.
The processor is used for executing the computer program stored in the memory to realize the steps of the method related to the embodiment.
Claims (9)
1. An ASRS task scheduling and goods space allocation method under classified storage is characterized in that a goods shelf is partitioned according to goods characteristics, wherein the goods characteristics comprise goods in-out frequency and goods attributes; the model of the warehouse-in and warehouse-out task scheduling and the goods location distribution is a dynamic planning model comprising warehouse-in and warehouse-out task stacker distribution, warehouse-in task sequencing, warehouse-out task sequencing and state transfer of a goods location set; empty goods positions generated by the ex-warehouse task can be used by the subsequent in-warehouse task; each stroke of the stacker is a warehouse-in and warehouse-out stroke which comprises a warehouse-in task and a warehouse-out task; the method for scheduling the warehouse entry and exit tasks and allocating the goods space comprises the following steps:
aiming at minimizing the completion time of the warehouse-in and warehouse-out process and the delay time of the warehouse-out task, solving the dynamic planning model by adopting a cultural genetic algorithm, and outputting the warehouse-out and warehouse-in tasks and the processing sequence which are executed by each stacker and the optimal solution of the goods space corresponding to each warehouse-in and warehouse-out task when the termination condition is met;
the cultural genetic algorithm obtains the warehousing task sequence and the stacker allocation by solving an assigned problem model taking a minimized warehousing and ex-warehousing task matching measurement index as a target in global optimization according to the initial warehousing and ex-warehousing task sequence and the stacker allocation;
the cultural genetic algorithm further optimizes the warehousing task sequence and the ex-warehouse task sequence by utilizing local search; the local search comprises two goods position exchange operators which respectively represent the exchange of the warehousing goods position and the ex-warehouse goods position of two ex-warehouse strokes on a stacker, and during each local search, one of the two goods position exchange operators is selected to perform local search according to the historical performance of the two goods position exchange operators, so that a new in-warehouse stroke is formed.
2. The ASRS task scheduling and cargo space allocation method under classified storage according to claim 1, wherein the matching metric index of the outbound task and the inbound task is a novel Hamming distance, and the formula of the novel Hamming distance is as follows:
wherein
In order to carry out the warehouse task goods area sequence,
for warehousing task cargo area sequences, a i ∈Ω，b i E is omega, omega represents a cargo area sequence, the cargo area is sequentially divided into p types of cargo areas from near to far according to the distance from an I/O port, and omega = (omega is used for expressing the cargo area sequence) 1 ，Ω 2 ...Ω p )。
3. The ASRS task scheduling and cargo space allocation method under classified storage according to claim 1, wherein the dynamic planning model divides stages according to the number of ex-warehouse tasks, each stage based on the current cargo space set state, selects a cargo space for the ex-warehouse tasks by solving an integer planning model aiming at minimizing the completion time of the ex-warehouse journey and the delay time of the ex-warehouse tasks; and after the single stage is finished, updating the goods position set state, and transferring to the next stage until all the warehouse-in and warehouse-out tasks are executed.
4. The ASRS task scheduling and cargo space allocation method under classified storage according to claim 1, wherein the cultural genetic algorithm adopts a genetic algorithm to perform global search, and the method comprises the following steps: carrying out chromosome coding on the ex-warehouse task sequence and the distribution of the stacking machine to generate an initial ex-warehouse task sequence and the distribution of the stacking machine;
selecting goods positions for the warehouse-in and warehouse-out tasks according to the chromosome coding and the warehouse-in task sequence and the distribution of the stacking machine; selecting an ex-warehouse goods position which is closest to the I/O port and contains ex-warehouse goods in a goods area to which the ex-warehouse task belongs, and selecting an empty goods position which can minimize the completion time of an in-warehouse and out-warehouse stroke in the goods area to which the in-warehouse task belongs; and after the distribution of the goods positions for one warehouse-in and warehouse-out stroke is finished, updating the states of the goods positions until the distribution of all the goods positions is finished.
5. The ASRS task scheduling and cargo space allocation method under classified storage according to claim 1, wherein the method for further optimizing the warehousing task sequence and the ex-warehouse task sequence by using local search comprises:
performing local search on the initial population in the cultural genetic algorithm and the crossed and mutated individuals, and using P (N) = c in each local search N /∑ N∈{1，2} c N Selecting a goods space exchange operator to carry out local search, wherein N is epsilon {1,2}, represents two goods space exchange operators, P (N) is the probability of selecting the goods space exchange operator, and c is N The counter is used for recording the historical performance of the goods position exchange operator, and when the solution after local search is performed by using a certain goods position exchange operator is superior to the current solution, the counter countsAdding one to the device.
6. The ASRS task scheduling and cargo space allocation method under classified storage according to claim 4, wherein in the cultural genetic algorithm, the inverse of an objective function that minimizes the completion time of the warehouse-in and warehouse-out journey and the delay time of the warehouse-out task is used as a fitness function, and the optimal solution of the cultural genetic algorithm is a solution that maximizes the fitness function value when a termination condition is satisfied.
7. An ASRS task scheduling and cargo space allocation system under classified storage, comprising:
the model establishing unit is used for establishing a dynamic planning model for the distribution of the warehouse entry task stacker, the sorting of the warehouse entry tasks, the sorting of the warehouse exit tasks and the state transfer of the goods space set, and empty goods spaces generated by the warehouse exit tasks can be used by subsequent warehouse entry operation; each stroke of the stacker is a warehouse entering and exiting stroke which comprises a warehouse entering task and a warehouse exiting task; the goods shelf is partitioned according to goods characteristics, wherein the goods characteristics comprise goods in-out frequency and goods attributes;
the model solving unit is used for solving the dynamic planning model by adopting a cultural genetic algorithm with the aim of minimizing the completion time of the warehouse-in and warehouse-out process and the delay time of the warehouse-out task, and outputting warehouse-out and warehouse-in tasks and the processing sequence which are executed by each stacker in charge and the optimal solution of the goods space corresponding to each warehouse-in and warehouse-out task when the termination condition is met;
the cultural genetic algorithm determines an initial ex-warehouse task sequence and stacker allocation through population initialization in global optimization, and obtains an in-warehouse task sequence and stacker allocation through solving an assignment problem model taking a minimized ex-warehouse task matching metric index as a target;
local search is utilized in the cultural gene algorithm to further optimize the sequence of warehousing tasks and the sequence of ex-warehouse tasks; the local search comprises two goods space exchange operators which respectively represent the exchange of the warehousing goods space and the ex-warehouse goods space of two ex-warehouse strokes on a stacker, and during each local search, one of the two goods space exchange operators is selected to perform local search according to the historical expression of the two goods space exchange operators to form a new ex-warehouse stroke.
8. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the computer program, when loaded into the processor, implements the method of ASRS task scheduling and cargo space allocation under sorted storage according to any of claims 1-7.
9. A computer-readable storage medium, in which a computer program is stored, which, when being executed by a processor, implements the ASRS task scheduling and cargo space allocation method under sorted storage according to any of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211422100.4A CN115730789A (en) | 2022-11-14 | 2022-11-14 | ASRS task scheduling and goods allocation method and system under classified storage |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211422100.4A CN115730789A (en) | 2022-11-14 | 2022-11-14 | ASRS task scheduling and goods allocation method and system under classified storage |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115730789A true CN115730789A (en) | 2023-03-03 |
Family
ID=85295633
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211422100.4A Pending CN115730789A (en) | 2022-11-14 | 2022-11-14 | ASRS task scheduling and goods allocation method and system under classified storage |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115730789A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116523445A (en) * | 2023-06-30 | 2023-08-01 | 江西启烨物联技术有限公司 | Warehouse bin warehouse-in and warehouse-out scheduling method |
| CN117314316A (en) * | 2023-10-17 | 2023-12-29 | 湘南学院 | Warehouse-in and warehouse-out scheduling method of customized furniture plate automatic sorting system |
-
2022
- 2022-11-14 CN CN202211422100.4A patent/CN115730789A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116523445A (en) * | 2023-06-30 | 2023-08-01 | 江西启烨物联技术有限公司 | Warehouse bin warehouse-in and warehouse-out scheduling method |
| CN116523445B (en) * | 2023-06-30 | 2023-08-25 | 江西启烨物联技术有限公司 | Warehouse bin warehouse-in and warehouse-out scheduling method |
| CN117314316A (en) * | 2023-10-17 | 2023-12-29 | 湘南学院 | Warehouse-in and warehouse-out scheduling method of customized furniture plate automatic sorting system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115730789A (en) | ASRS task scheduling and goods allocation method and system under classified storage | |
| CN108550007B (en) | Goods space optimization method and system for automatic stereoscopic warehouse of pharmaceutical enterprise | |
| CN106934580B (en) | Inventory control method and device | |
| CN110245890B (en) | Goods sorting method and goods sorting system | |
| Ahmadizar et al. | Two-level vehicle routing with cross-docking in a three-echelon supply chain: A genetic algorithm approach | |
| CN107480922B (en) | Method for establishing goods position distribution scheduling model under two-end type same-rail double-vehicle running mode | |
| CN111144568B (en) | Multi-target city logistics distribution path planning method | |
| CN109764882B (en) | Multi-target vehicle path planning method based on self-adaptive local search chain | |
| CN105469201A (en) | Method for logistics dispensing center work task processing and scheduling | |
| CN109919541A (en) | A kind of model solution method of multistage positioning inventory routing problem | |
| CN112183812B (en) | Finished cigarette logistics vehicle scheduling method considering short-time and low-cost | |
| Jiang et al. | Picking-replenishment synchronization for robotic forward-reserve warehouses | |
| CN112884257A (en) | Goods taking path optimization method, device and system based on genetic algorithm | |
| CN115345549B (en) | Vehicle path adjustment method and system combined with loading scheme | |
| CN111860957B (en) | Multi-vehicle-type vehicle path planning method considering secondary distribution and balancing | |
| Ko et al. | A rollout heuristic algorithm for order sequencing in robotic compact storage and retrieval systems | |
| CN110942251B (en) | Batch scheduling method based on joint ant colony algorithm | |
| Guo et al. | Storage assignment for newly arrived items in forward picking areas with limited open locations | |
| CN111626516B (en) | Order ordering optimization method of double-deep four-way shuttle system considering cargo pouring strategy | |
| CN116933939A (en) | Flexible workshop collaborative production method and system based on improved raccoon optimization algorithm | |
| CN112016750A (en) | Improved method for solving problem of vehicle path with constraint | |
| CN115345550B (en) | Vehicle path planning method and system | |
| CN116542365A (en) | Order allocation and AGV scheduling combined optimization method in mobile robot fulfillment system | |
| CN116342039A (en) | Optimizing method for goods distribution and sorting of stereoscopic warehouse | |
| CN112862212B (en) | Multi-AGV (automatic guided vehicle) scheduling method, device and equipment based on improved sparrow 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 |