Dynamic scheduling optimization method for strong coupling type goods-to-person picking system

Abstract

The invention relates to a dynamic scheduling optimization method for a strong coupling type'goods-to-people 'sorting system, which comprises the steps of performing strong coupling type'goods-to-people' sorting system optimization problem modeling and a dynamic scheduling algorithm, and establishing a target function with the minimum total sum of accumulated stockout waiting time of a sorting station by using a target function with the minimum total sum of task-free waiting time of ex-warehouse equipment; and in the order dynamic sorting algorithm, order issuing sorting is carried out by taking the minimum difference value between all warehouse-out equipment and the safety task quantity value of the warehouse-out equipment after the order is issued as a target, and if there is no sorting station whose to-be-sorted task quantity is smaller than the lower limit value of the to-be-sorted task quantity, the task of the sorting station whose remaining task quantity is minimum is issued. According to the dynamic scheduling optimization method for the strong coupling type goods-to-person picking system, efficient cooperation of the ex-warehouse equipment and the picking station is achieved, the waiting time of the ex-warehouse equipment and the waiting time of the picking station are shortened, and finally the purposes of shortening the total working hours of the system and improving the overall working efficiency of the system are achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to a dynamic scheduling optimization method, more particularly, to a dynamic scheduling optimization method for a strong coupling type "goods-to-person" picking system. BACKGROUND

[0002] A "goods-to-person" picking system is usually composed of one or more picking stations and their connected outbound devices. According to the degree of association between the outbound device and the picking station, the "goods-to-person" picking system can be divided into a strong coupling type system and a weak coupling type system. When the outbound device takes out the goods to be picked only for a unique picking station, the system is a strong coupling type system; when the outbound device takes out the goods to be picked for different picking stations, the system is a weak coupling type system.

[0003] The outbound device and the picking station of the strong coupling type "goods-to-person" picking system are closely related to each other and influence each other. From the perspective of the influence of the outbound device on the picking station, the task execution order of the outbound device affects the work efficiency of each picking station. If the outbound device does not supply the boxes to a certain picking station in time, it will cause the picker in the station to wait for goods; if the outbound device supplies too many boxes to a certain picking station, it will cause the boxes on the buffer conveyor line in the station to overflow, block the outbound main line, affect the supply of boxes to the subsequent picking station, and cause the picker in the station to wait for goods. From the perspective of the influence of the picking station on the outbound device, the order issuing order of the picking station affects the utilization rate of the outbound device. Each picking station independently processes complete orders, and the number of orders processed limits the number of orders processed by the outbound device at the same time. The picking station completes an order, and the system can issue a new order again, and the number of outbound boxes between the two order issuing intervals is defined as the order issuing interval outbound quantity. Each outbound device of the system is only responsible for the inbound and outbound tasks of the boxes in the same aisle, and under the premise that the goods items and the storage locations are one-to-one, the order issuing order selection determines the order issuing interval outbound quantity of each outbound device. If the order issuing interval outbound quantity of a certain outbound device is too small, it may cause the outbound device to wait for tasks before the system issues a new order.

[0004] The current research on the overall scheduling optimization of the "goods-to-person" picking system mainly focuses on the weak coupling type, and there is little research on the scheduling optimization of the strong coupling type system, without considering the influence of the coupling relationship between the outbound device and the picking station on the work efficiency of the system. SUMMARY

[0005] The present application provides a dynamic scheduling optimization method for a strong coupling type "goods-to-person" picking system to overcome the above technical problems.

[0006] The present invention discloses a dynamic scheduling optimization method for a strongly coupled "goods-to-person" picking system. The strongly coupled "goods-to-person" picking system consists of bin shelves, outbound equipment, conveying equipment, and picking stations. The bin shelves store boxes, each containing only one type of material. Outbound / return stations are located at positions corresponding to the bin shelves on the conveying equipment. A buffer conveyor line and a hoist are connected between the picking station and the conveying equipment. The outbound equipment places the boxes corresponding to the materials to be picked on the outbound / return stations. The conveying equipment transports the boxes to be picked to the buffer conveyor line, which then transports the boxes to be picked to the picking station.

[0007] The key feature is that the dynamic scheduling optimization method for a strongly coupled "goods-to-person" picking system includes modeling the optimization problem of the strongly coupled "goods-to-person" picking system and a dynamic scheduling algorithm. The modeling of the optimization problem includes mathematical modeling of the order placement optimization problem and the task execution optimization problem. In the mathematical modeling of the order placement optimization problem, an objective function is established to minimize the total waiting time of outbound equipment without tasks. In the mathematical modeling of the task execution optimization problem, an objective function is established to minimize the total cumulative waiting time of out-of-stock at picking stations. The dynamic scheduling algorithm includes an order dynamic sorting algorithm and an outbound task dynamic scheduling algorithm. In the order dynamic sorting algorithm, the order placement is sorted with the objective of minimizing the difference between all outbound equipment and their safe task quantity after the order is placed. In the outbound task dynamic scheduling algorithm, tasks are first placed at picking stations where the number of tasks to be picked is less than the lower limit of the number of tasks to be picked at the picking station. If there is no picking station where the number of tasks to be picked is less than the lower limit of the number of tasks to be picked at the picking station, then the task of the picking station with the smallest remaining task quantity is placed.

[0008] The dynamic scheduling optimization method for a strongly coupled "goods-to-person" picking system of the present invention is implemented through the following steps in modeling the optimization problem of the strongly coupled "goods-to-person" picking system:

[0009] 1) Parameter Definition; First, define the following data parameters:

[0010] o ij Let j be task j in order i. The order set contains I orders, where order i contains J tasks. i One outbound task;

[0011] m is the outbound equipment number, and there are a total of M outbound equipment in the system;

[0012] k is the picking station number, and there are K picking stations in the system.

[0013] This represents the total waiting time for outbound equipment without any tasks.

[0014] The total out-of-stock waiting time at the picking station;

[0015] (a,b) represents the b-th task execution decision after the a-th order issuance decision;

[0016] T (a,b) Let (a, b) be the decision time corresponding to the decision point number (a, b).

[0017] A represents the total number of order issuance decision points;

[0018] B a For decision time T (a,0) To T (a+1,0) The total number of task execution decision points between;

[0019] D is the set of all decision point numbers, where {(a,b)|(a,b)∈D,b=0} is the set of order issuance decision point numbers, and {(a,b)|(a,b)∈D,b≠0} is the set of task execution decision point numbers.

[0020] S (a,b) Let (a, b) be the set of tasks to be executed for the outbound equipment after the decision is made.

[0021] After issuing order i at decision point (a,0), at decision time T (a,0) To T (a+1,0) Between these times, the no-task waiting time of the outbound device m, i.e., from the decision time T... (a,0) In the set of tasks to be executed by outbound equipment, outbound equipment m starts with no tasks and continues until decision time T. (a+1,0) The period of time up to that point;

[0022] After task j in order i is issued to decision point (a,b), at decision time T (a,b) To T (a,b+1) Between (where b≠0), the cumulative out-of-stock waiting time of picking station k, i.e., from decision time T... (a,b) From the beginning until the decision time T (a,b+1) Up to this point, the cumulative value of the waiting time for out-of-stock items at picking station k due to the lack of available boxes for picking;

[0023] The number of tasks to be executed for outbound equipment m after order i is issued at decision point (a,0) is the number of tasks for which the system has completed order issuance but has not yet executed outbound.

[0024] U represents the safety task quantity of a single outbound device. If the number of tasks to be executed for outbound device m is less than U, it indicates that there will be a long waiting time without tasks.

[0025] After issuing order i at decision point (a,0), the following conditions are met: The number of outbound devices;

[0026] R (a,0) Let (a, 0) be the set of candidate orders.

[0027] After task j in order i is issued to decision point (a,b), the number of tasks to be picked at picking station k is the number of tasks that have been sent out but not picked at picking station k in the system.

[0028] After task j in order i is issued to decision point (a,b), the remaining number of tasks at picking station k is the number of tasks that have been issued to picking station k but have not yet been shipped out.

[0029] V up This represents the upper limit of the number of tasks to be picked at the picking station.

[0030] V down This represents the lower limit of the number of tasks to be picked at the picking station.

[0031]

[0032] 2) Setting conditions; then, set the following assumptions:

[0033] Condition 1: The number of outbound devices and the number of operating picking stations in the system are fixed;

[0034] Condition 2: Each type of goods corresponds to a unique cargo box, and the cargo box has a fixed location within the outbound equipment;

[0035] Condition 3: The order set is known, and the quantity of goods in the outbound equipment meets the picking quantity requirements of the order set;

[0036] 3) Establishment of mathematical modeling for the order placement optimization problem;

[0037] The objective function, as shown in formula (1), is established to minimize the total idle time of outbound equipment:

[0038]

[0039] At the same time, establish the constraints shown in formulas (2) to (5):

[0040]

[0041] in:

[0042] Formula (2) indicates that all orders in the order set have been issued;

[0043] Formula (3) indicates that each order can only be issued once during system operation;

[0044] Formula (4) indicates that each order issuance decision point can only issue one order;

[0045] Formula (5) means that after order i is issued at decision point (a,0), all outbound tasks in order i will be added to the nearest decision point (a-1,B). a The set of tasks to be executed by the outbound equipment. Construct the set of tasks S to be executed for the outbound equipment after the decision point (a,0) is determined. (a,0) ;

[0046] 4) Establishment of mathematical modeling for the task execution optimization problem;

[0047] The objective function is established as shown in formula (6) to minimize the total cumulative out-of-stock waiting time at the picking station:

[0048]

[0049] At the same time, establish the constraints as shown in formulas (7) to (11):

[0050]

[0051] Formula (7) indicates that all tasks in the order set are executed;

[0052] Formula (8) indicates that each task can only be executed once during system operation;

[0053] Formula (9) indicates that only one task can be executed at each task execution decision point;

[0054] Formula (10) indicates that decision point (a,b) can only select one task from the set of tasks to be executed by the outbound equipment after decision point (a,b-1);

[0055] Formula (11) represents the set of tasks S to be executed from the equipment that left the warehouse after decision point (a,b) issues task j in order i. (a,b-1) Delete and construct the set of tasks S to be executed for the outbound equipment after the decision point (a,b) is made. (a,b) .

[0056] The dynamic scheduling optimization method for a strongly coupled "goods-to-person" picking system of the present invention is implemented through the following steps:

[0057] a) Add unsent orders to R (a,0) ;

[0058] b) Obtain the decision point (a-1, B) a The set of tasks to be executed for equipment to be dispatched after decision-making.

[0059] c). R (a,0) Add outbound task to order i calculate

[0060] d) From R (a,0) Delete matching Orders;

[0061] e) Determine whether, after order i is issued, the number of pending tasks for all outbound devices is not less than the safe task quantity value U for a single outbound device, i.e., determine... If the condition is true, proceed to step f); if the condition is false, proceed to step g.

[0062] f). If R (a,0) middle Then calculate To satisfy The orders are treated as a set of pending orders;

[0063] g) If calculate To satisfy The orders are treated as a set of pending orders;

[0064] h) Randomly select an order from the pending order set and send it out, thus completing the dynamic sorting of orders.

[0065] The dynamic scheduling optimization method for a strongly coupled "goods-to-person" picking system of the present invention is implemented through the following steps:

[0066] A. Obtain the set of tasks S to be executed for the outbound equipment after the decision point (a, b-1). (a,b-1) ;

[0067] B) Calculate the number of tasks to be picked at picking station k. Number of remaining tasks

[0068] C) Determine if it exists. If the picking station exists, proceed to step D); otherwise, proceed to step E.

[0069] D) will The picking station corresponding to the minimum value is determined as a candidate picking station, and step F is executed.

[0070] E). The picking station corresponding to the minimum value is determined as a candidate picking station, and step F is executed.

[0071] F). From S (a,b-1)The task of obtaining candidate picking stations is used to construct a set of tasks to be sent.

[0072] G) Randomly select a task from the set of tasks to be dispatched to complete the dynamic sorting of tasks to be executed.

[0073] The beneficial effects of this invention are as follows: The dynamic scheduling optimization method for the strongly coupled "goods-to-person" picking system of this invention first defines symbols and variables and gives assumptions based on the actual layout and operation mode of the "goods-to-person" picking system, and establishes mathematical models for the order placement optimization problem and the task execution optimization problem, respectively. The order placement optimization problem aims to minimize the total waiting time of outbound equipment without tasks; the task execution optimization problem aims to minimize the total waiting time of accumulated stockouts at picking stations, and constraints are given according to the characteristics of each problem. Then, the coupling relationship between outbound equipment and picking stations was analyzed in depth. It was clarified that the upper and lower limits of the stacker crane's safe task quantity and the picking station's pending task quantity are the core parameters of system scheduling. Based on this, a dynamic scheduling algorithm was designed to solve the problem. In the order dynamic sorting algorithm, the goal is to minimize the difference between the safe task quantity of all outbound equipment and the order after the order is issued. In the outbound task dynamic scheduling algorithm, the tasks of picking stations with a pending task quantity less than the lower limit of the picking station's pending task quantity are issued first. If there is no picking station with a pending task quantity less than the lower limit of the picking station's pending task quantity, then the task of the picking station with the smallest remaining task quantity is issued. By adjusting the order issuance order and stacker crane task execution order through the dynamic scheduling algorithm, efficient collaboration between outbound equipment and picking stations is achieved, reducing the waiting time of outbound equipment and picking stations, and ultimately shortening the total system operation time and improving the overall system operation efficiency. Attached Figure Description

[0074] Figure 1 This is a schematic diagram of the structural principle of the strongly coupled "goods-to-person" picking system of the present invention;

[0075] Figure 2 This is a schematic diagram showing the distribution of decision points and decision times in the dynamic scheduling optimization method of the strongly coupled "goods-to-person" picking system of the present invention;

[0076] Figure 3 This is a flowchart of the dynamic order sorting algorithm in this invention;

[0077] Figure 4 This is a flowchart of the dynamic scheduling algorithm for outbound tasks in this invention.

[0078] In the diagram: 1 Outbound equipment, 2 Picking station, 3 Conveying equipment, 4 Tote box rack, 5 Outbound / Return station, 6 Buffer conveyor line, 7 Elevator. Detailed Implementation

[0079] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0080] like Figure 1 The diagram shows the structural principle of the strongly coupled "goods-to-person" picking system of the present invention. It consists of bin racks 4, outbound equipment 1, conveying equipment 3, and picking stations 2. Outbound / return stations 5 are installed on the conveying equipment 3 at positions corresponding to each set of bin racks 4. A buffer conveyor line 6 and a lift 7 are connected between the conveying equipment 3 and each picking station 2. The conveying equipment 3 has two layers: the upper layer is the return conveyor line, and the lower layer is the outbound conveyor line, connected by the lift 7. The bin racks 4 are used to store bins; both material inbound and outbound operations require the bins to carry them. Each bin stores only one type of material, which may be raw materials for manufacturing or seasonal goods. The buffer conveyor line 6 buffers a certain number of bins, such as three bins.

[0081] This invention relates to a dynamic scheduling optimization method for a strongly coupled "goods-to-person" picking system. It includes modeling the optimization problem of the strongly coupled "goods-to-person" picking system and a dynamic scheduling algorithm. The modeling of the optimization problem includes mathematical modeling of order placement optimization and task execution optimization. In the mathematical modeling of the order placement optimization problem, an objective function is established to minimize the total waiting time of outbound equipment without tasks. In the mathematical modeling of the task execution optimization problem, an objective function is established to minimize the total cumulative waiting time of out-of-stock items at picking stations.

[0082] The dynamic scheduling algorithm includes an order dynamic sorting algorithm and an outbound task dynamic scheduling algorithm. In the order dynamic sorting algorithm, the goal is to sort the orders based on minimizing the difference between the total number of outbound devices and their safe task quantity after the order is issued. In the outbound task dynamic scheduling algorithm, the task is first issued to the picking station whose number of pending tasks is less than the lower limit of the number of pending tasks. If there is no picking station whose number of pending tasks is less than the lower limit of the number of pending tasks, then the task is issued to the picking station with the smallest remaining task quantity.

[0083] like Figure 2 The diagram shows the distribution of decision points and decision times in the dynamic scheduling optimization method for the strongly coupled "goods-to-person" picking system of the present invention. The modeling of the optimization problem of the strongly coupled "goods-to-person" picking system is specifically implemented through the following steps:

[0084] 1) Parameter Definition; First, define the following data parameters:

[0085] o ij Let j be task j in order i. The order set contains I orders, where order i contains J tasks. i One outbound task;

[0086] m is the outbound equipment number, and there are a total of M outbound equipment in the system;

[0087] k is the picking station number, and there are K picking stations in the system.

[0088] This represents the total waiting time for outbound equipment without any tasks.

[0089] The total out-of-stock waiting time at the picking station;

[0090] (a,b) represents the b-th task execution decision after the a-th order issuance decision;

[0091] T (a,b) Let (a, b) be the decision time corresponding to the decision point number (a, b).

[0092] A represents the total number of order issuance decision points;

[0093] B a For decision time T (a,0) To T (a+1,0) The total number of task execution decision points between;

[0094] D is the set of all decision point numbers, where {(a,b)|(a,b)∈D,b=0} is the set of order issuance decision point numbers, and {(a,b)|(a,b)∈D,b≠0} is the set of task execution decision point numbers.

[0095] S (a,b) Let (a, b) be the set of tasks to be executed for the outbound equipment after the decision is made.

[0096] After issuing order i at decision point (a,0), at decision time T (a,0) To T (a+1,0) Between these times, the no-task waiting time of the outbound device m, i.e., from the decision time T... (a,0) In the set of tasks to be executed by outbound equipment, outbound equipment m starts with no tasks and continues until decision time T. (a+1,0) The period of time up to that point;

[0097] After task j in order i is issued to decision point (a,b), at decision time T (a,b) To T (a,b+1) Between (where b≠0), the cumulative out-of-stock waiting time of picking station k, i.e., from decision time T... (a,b) From the beginning until the decision time T (a,b+1) Up to this point, the cumulative value of the waiting time for out-of-stock items at picking station k due to the lack of available boxes for picking;

[0098] The number of tasks to be executed for outbound equipment m after order i is issued at decision point (a,0) is the number of tasks for which the system has completed order issuance but has not yet executed outbound.

[0099] U represents the safety task quantity of a single outbound device. If the number of tasks to be executed for outbound device m is less than U, it indicates that there will be a long waiting time without tasks.

[0100] After issuing order i at decision point (a,0), the following conditions are met: The number of outbound devices;

[0101] R (a,0) Let (a, 0) be the set of candidate orders.

[0102] After task j in order i is issued to decision point (a,b), the number of tasks to be picked at picking station k is the number of tasks that have been sent out but not picked at picking station k in the system.

[0103] After task j in order i is issued to decision point (a,b), the remaining number of tasks at picking station k is the number of tasks that have been issued to picking station k but have not yet been shipped out.

[0104] In order to deal with the number of picking tasks and the number of remaining tasks To illustrate this further, let's take a concrete example. Suppose the picking station currently has 10 tasks (i.e., picking 10 materials), 3 tasks have been completed, and 3 tasks have been completed and not yet picked (at this time, the toy bins are located on conveyor 3 or buffer conveyor 6). but

[0105] V up This represents the upper limit of the number of tasks to be picked at the picking station.

[0106] V down This represents the lower limit of the number of tasks to be picked at the picking station.

[0107]

[0108] 2) Setting conditions; then, set the following assumptions:

[0109] Condition 1: The number of outbound devices and the number of operating picking stations in the system are fixed;

[0110] Condition 2: Each type of goods corresponds to a unique cargo box, and the cargo box has a fixed location within the outbound equipment;

[0111] Condition 3: The order set is known, and the quantity of goods in the outbound equipment meets the picking quantity requirements of the order set;

[0112] 3) Establishment of mathematical modeling for the order placement optimization problem;

[0113] The objective function, as shown in formula (1), is established to minimize the total idle time of outbound equipment:

[0114]

[0115] At the same time, establish the constraints shown in formulas (2) to (5):

[0116]

[0117] in:

[0118] Formula (2) indicates that all orders in the order set have been issued;

[0119] Formula (3) indicates that each order can only be issued once during system operation;

[0120] Formula (4) indicates that each order issuance decision point can only issue one order;

[0121] Formula (5) means that after order i is issued at decision point (a,0), all outbound tasks in order i will be added to the nearest decision point (a-1,B). a The set of tasks to be executed by the outbound equipment. Construct the set of tasks S to be executed for the outbound equipment after the decision point (a,0) is determined. (a,0) ;

[0122] In formula (5), taking outbound equipment m as an example, if order i is issued at decision point (a,0), the number of tasks to be executed by outbound equipment m is... If the value is too small, the outbound equipment m will experience a waiting state with no tasks; therefore, the order placement decision determines S. (a,0) The task distribution of each outbound device in the middle, and then the decision time T. (a,0) To T (a+1,0) Between, the no-task waiting time of outbound device m It has an impact.

[0123] 4) Establishment of mathematical modeling for the task execution optimization problem;

[0124] The objective function is established as shown in formula (6) to minimize the total cumulative out-of-stock waiting time at the picking station:

[0125]

[0126] At the same time, establish the constraints as shown in formulas (7) to (11):

[0127]

[0128] Formula (7) indicates that all tasks in the order set are executed;

[0129] Formula (8) indicates that each task can only be executed once during system operation;

[0130] Formula (9) indicates that only one task can be executed at each task execution decision point;

[0131] Formula (10) indicates that decision point (a,b) can only select one task from the set of tasks to be executed by the outbound equipment after decision point (a,b-1);

[0132] Formula (11) represents the set of tasks S to be executed from the equipment that left the warehouse after decision point (a,b) issues task j in order i. (a,b-1) Delete and construct the set of tasks S to be executed for the outbound equipment after the decision point (a,b) is made. (a,b) .

[0133] In formula (11), taking picking station k as an example, if decision point (a,b) issues task j in order i, the number of tasks to be picked at picking station k is... If the quantity is too low, picking station k may experience stockout waiting; if the quantity of tasks to be picked at the preceding picking station (let's say k') is too low... Too many items can clog the outbound mainline, causing stockouts at picking station k due to cartons queuing on the mainline and unable to arrive in time. Both of these situations increase the decision time T. (a,b) To T (a,b+1) The out-of-stock waiting time at picking station k. Therefore, the task execution decision determines S. (a,b) The number of tasks to be picked at each picking station, and thus the decision time T. (a,b) To T (a,b+1) Between (where b≠0), the out-of-stock waiting time at picking station k It has an impact.

[0134] like Figure 3 The flowchart of the dynamic order sorting algorithm in this invention is shown. The dynamic order sorting algorithm in this invention is specifically implemented through the following steps:

[0135] a) Add unsent orders to R (a,0) ;

[0136] b) Obtain the decision point (a-1, B) aThe set of tasks to be executed for equipment to be dispatched after decision-making.

[0137] c). R (a,0) Add outbound task to order i calculate

[0138] d) From R (a,0) Delete matching Orders;

[0139] e) Determine whether, after order i is issued, the number of pending tasks for all outbound devices is not less than the safe task quantity value U for a single outbound device, i.e., determine... If the condition is true, proceed to step f); if the condition is false, proceed to step g.

[0140] f). If R (a,0) middle Then calculate To satisfy The orders are treated as a set of pending orders;

[0141] g) If calculate To satisfy The orders are treated as a set of pending orders;

[0142] h) Randomly select an order from the pending order set and send it out, thus completing the dynamic sorting of orders.

[0143] In step d), if the candidate order set R (a,0) The outbound task of order i is added to the pending task set of the outbound device. After that, if (that is, satisfied) If the maximum value of the number of outbound devices is less than M, then after step d), the "delete" condition is met. After processing the order, only the one with the largest value is retained. The orders. The value can only be less than or equal to M, when When this occurs, it indicates that the number of pending tasks for all outbound devices is greater than the safe task capacity U for a single outbound device. In this case, the tasks should be distributed as evenly as possible among each outbound device. This can be achieved through step f). Filter out orders that distribute the number of tasks equally among each outbound device.

[0144] when When this occurs, it indicates that the number of pending tasks for a single outbound device is less than the safe task capacity U for that device. In this case, tasks should be allocated to each outbound device with a pending task capacity less than U, i.e., through step g). If the equipment distributes the tasks equally, then proceed through step f). Filter out orders where the number of tasks to be executed is less than U and distribute the task quantity equally among each outbound device.

[0145] like Figure 4 The flowchart shown is a presentation of the dynamic scheduling algorithm for outbound tasks in this invention. The dynamic scheduling algorithm for outbound tasks in this invention is specifically implemented through the following steps:

[0146] A. Obtain the set of tasks S to be executed for the outbound equipment after the decision point (a, b-1). (a,b-1) ;

[0147] B) Calculate the number of tasks to be picked at picking station k. Number of remaining tasks

[0148] C) Determine if it exists. If the picking station exists, proceed to step D); otherwise, proceed to step E.

[0149] D) will The picking station corresponding to the minimum value is determined as a candidate picking station, and step F is executed.

[0150] E). The picking station corresponding to the minimum value is determined as a candidate picking station, and step F is executed.

[0151] F). From S (a,b-1) The task of obtaining candidate picking stations is used to construct a set of tasks to be sent.

[0152] G) Randomly select a task from the set of tasks to be dispatched to complete the dynamic sorting of tasks to be executed.

[0153] As can be seen, the dynamic scheduling optimization method for a strongly coupled "goods-to-person" picking system of the present invention first conducts an in-depth analysis of the functional layout, equipment operation status, and on-site operation mode of the "goods-to-person" picking system. It defines symbols and variables, gives assumptions, and establishes a mathematical model for order placement optimization with the objective of minimizing the total waiting time without tasks for outbound equipment, and a mathematical model for task execution optimization with the objective of minimizing the total waiting time for accumulated stockouts at picking stations. Constraints are given according to the characteristics of each problem. Then, the coupling relationship between outbound equipment and picking stations is analyzed in depth. The upper and lower limits of the stacker crane's safe task capacity and the picking station's pending task capacity are determined as key parameters for system scheduling. Based on this, a dynamic scheduling algorithm is designed to solve the problem. The dynamic scheduling algorithm includes a dynamic order sorting algorithm and a dynamic outbound task scheduling algorithm, achieving efficient collaboration between outbound equipment and picking stations by scheduling the order placement order and the outbound task execution order.

Claims

1. A dynamic scheduling optimization method for a strongly coupled "goods-to-person" picking system. The strongly coupled "goods-to-person" picking system consists of a bin rack (4), an outbound device (1), a conveying device (3), and a picking station (2). The bin rack stores bins, and each bin stores only one type of material. An outbound / return station (5) is set up at the position corresponding to the bin rack of the conveying device. A buffer conveyor line (6) and a hoist (7) are set up between the picking station and the conveying device. The outbound device places the bins corresponding to the materials to be outbound on the outbound / return station. The conveying device transports the bins to be picked to the buffer conveyor line. The buffer conveyor line transports the bins to be picked to the picking station (2). Its features are: The dynamic scheduling optimization method for a tightly coupled "goods-to-person" picking system includes modeling the optimization problem of the tightly coupled "goods-to-person" picking system and dynamic scheduling algorithms. The modeling of the optimization problem of the tightly coupled "goods-to-person" picking system includes mathematical modeling of the order placement optimization problem and the task execution optimization problem. In the mathematical modeling of the order placement optimization problem, an objective function is established to minimize the total waiting time of outbound equipment without tasks. In the mathematical modeling of the task execution optimization problem, an objective function is established to minimize the total cumulative waiting time of out-of-stock at picking stations. The dynamic scheduling algorithm includes an order dynamic sorting algorithm and an outbound task dynamic scheduling algorithm. In the order dynamic sorting algorithm, the order placement is sorted with the objective of minimizing the difference between the total number of outbound equipment and its safe task quantity after the order is placed. In the outbound task dynamic scheduling algorithm, tasks are first placed at picking stations where the number of tasks to be picked is less than the lower limit of the number of tasks to be picked at each picking station. If there is no picking station where the number of tasks to be picked is less than the lower limit of the number of tasks to be picked at each picking station, then the task is placed at the picking station with the smallest remaining task quantity.

2. The dynamic scheduling optimization method for a strongly coupled "goods-to-person" picking system according to claim 1, characterized in that, Modeling the optimization problem of a tightly coupled "goods-to-person" picking system is achieved through the following steps: 1) Parameter Definition; First, define the following data parameters: o ij Let j be task j in order i. The order set contains I orders, where order i contains J tasks. i One outbound task; m is the outbound equipment number, and there are a total of M outbound equipment in the system; k is the picking station number, and there are K picking stations in the system. T st w represents the total waiting time without tasks for outbound devices; The total out-of-stock waiting time at the picking station; (a,b) represents the b-th task execution decision after the a-th order issuance decision; T (a,b) Let (a, b) be the decision time corresponding to the decision point number (a, b). A represents the total number of order issuance decision points; B a For decision time T (a,0) To T (a+1,0) The total number of task execution decision points between; D is the set of all decision point numbers, where {(a,b)|(a,b)∈D,b=0} is the set of order issuance decision point numbers, and {(a,b)|(a,b)∈D,b≠0} is the set of task execution decision point numbers. S (a,b) Let (a, b) be the set of tasks to be executed for the outbound equipment after the decision is made. After issuing order i at decision point (a,0), at decision time T (a,0) To T (a+1,0) Between these times, the no-task waiting time of the outbound device m, i.e., from the decision time T... (a,0) In the set of tasks to be executed by outbound equipment, outbound equipment m starts with no tasks and continues until decision time T. (a+1,0) The period of time up to that point; After task j in order i is issued to decision point (a,b), at decision time T (a,b) To T (a,b+1) Between (where b≠0), the cumulative out-of-stock waiting time of picking station k, i.e., from decision time T... (a,b) From the beginning until the decision time T (a,b+1) Up to this point, the cumulative value of the waiting time for stockouts at picking station k due to the lack of available boxes for picking; The number of tasks to be executed for outbound equipment m after order i is issued at decision point (a,0) is the number of tasks for which the system has completed order issuance but has not yet executed outbound. U represents the safety task quantity of a single outbound device. If the number of tasks to be executed for outbound device m is less than U, it indicates that there will be a long waiting time without tasks. After issuing order i at decision point (a,0), the following conditions are met: The number of outbound devices; R (a,0) Let (a, 0) be the set of candidate orders. After task j in order i is issued to decision point (a,b), the number of tasks to be picked at picking station k is the number of tasks that have been sent out but not picked at picking station k in the system. After task j in order i is issued to decision point (a,b), the remaining number of tasks at picking station k is the number of tasks that have been issued to picking station k but have not yet been shipped out. V up This represents the upper limit of the number of tasks to be picked at the picking station. V down This represents the lower limit of the number of tasks to be picked at the picking station. 2) Setting conditions; then, set the following assumptions: Condition 1: The number of outbound devices and the number of operating picking stations in the system are fixed; Condition 2: Each type of goods corresponds to a unique cargo box, and the cargo box has a fixed location within the outbound equipment; Condition 3: The order set is known, and the quantity of goods in the outbound equipment meets the picking quantity requirements of the order set; 3) Establishment of mathematical modeling for the order placement optimization problem; The objective function, as shown in formula (1), is established to minimize the total idle time of outbound equipment: At the same time, establish the constraints shown in formulas (2) to (5): in: Formula (2) indicates that all orders in the order set have been issued; Formula (3) indicates that each order can only be issued once during system operation; Formula (4) indicates that each order issuance decision point can only issue one order; Formula (5) means that after order i is issued at decision point (a,0), all outbound tasks in order i will be added to the nearest decision point (a-1,B). a The set of tasks to be executed for the outbound equipment S (a-1,Ba) Construct the set of tasks S to be executed for the outbound equipment after the decision point (a,0) is reached. (a,0) ; 4) Establishment of mathematical modeling for the task execution optimization problem; The objective function is established as shown in formula (6) to minimize the total cumulative out-of-stock waiting time at the picking station: At the same time, establish the constraints as shown in formulas (7) to (11): Formula (7) indicates that all tasks in the order set are executed; Formula (8) indicates that each task can only be executed once during system operation; Formula (9) indicates that only one task can be executed at each task execution decision point; Formula (10) indicates that decision point (a,b) can only select one task from the set of tasks to be executed by the outbound equipment after decision point (a,b-1); Formula (11) represents the set of tasks S to be executed from the equipment that left the warehouse after decision point (a,b) issues task j in order i. (a,b-1) Delete and construct the set of tasks S to be executed for the outbound equipment after the decision point (a,b) is made. (a,b) .

3. The dynamic scheduling optimization method for a strongly coupled "goods-to-person" picking system according to claim 2, characterized in that, The dynamic order sorting algorithm is implemented through the following steps: a) Add unsent orders to R (a,0) ; b) Obtain the decision point (a-1, B) a The set of tasks to be executed for the outbound equipment after the decision is made. (a-1,Ba) ; c). R (a,0) Add the outbound task of order i to S (a-1,Ba) ,calculate d) From R (a,0) Delete matching Orders; e) Determine whether, after order i is issued, the number of pending tasks for all outbound devices is not less than the safe task quantity value U for a single outbound device, i.e., determine... If the condition is true, proceed to step f); if the condition is false, proceed to step g. f). If R (a,0) middle Then calculate To satisfy The orders are treated as a set of pending orders; g) If calculate To satisfy The orders are treated as a set of pending orders; h) Randomly select an order from the pending order set and send it out, thus completing the dynamic sorting of orders.

4. The dynamic scheduling optimization method for a strongly coupled "goods-to-person" picking system according to claim 2 or 3, characterized in that, The dynamic scheduling algorithm for outbound tasks is implemented through the following steps: A. Obtain the set of tasks S to be executed for the outbound equipment after the decision point (a, b-1). (a,b-1) ; B) Calculate the number of tasks to be picked at picking station k. Number of remaining tasks C) Determine if it exists. If the picking station exists, proceed to step D); otherwise, proceed to step E. D) will The picking station corresponding to the minimum value is determined as a candidate picking station, and step F is executed. E). The picking station corresponding to the minimum value is determined as a candidate picking station, and step F is executed. F). From S (a,b-1) The task of obtaining candidate picking stations is used to construct a set of tasks to be sent. G) Randomly select a task from the set of tasks to be dispatched to complete the dynamic sorting of tasks to be executed.

Images