CN105858044A

CN105858044A - Optimal dispatching method for warehousing systems combining rail guided vehicles and lifts

Info

Publication number: CN105858044A
Application number: CN201610365624.2A
Authority: CN
Inventors: 杨玮; 岳婷; 杜雨潇; 罗洋洋; 刘江; 杨甜; 王婷; 王晓雅
Original assignee: Shaanxi University of Science and Technology
Current assignee: Zhejiang Jingxing Logistics Equipment Co ltd
Priority date: 2016-05-27
Filing date: 2016-05-27
Publication date: 2016-08-17
Anticipated expiration: 2036-05-27
Also published as: CN105858044B

Abstract

本发明公开了一种穿梭车与升降机结合的仓储系统优化调度方法，以调度行程时间最小化为目标，采用遗传算法对调度的出入库模型进行优化，使存取行程时间大大缩短，可以明显节约时间成本，节约能源，使穿梭子母车全自动立体仓库的高效率、高密度，以及高利用率等优点得到充分发挥，实现自动化立体仓库实时、在线的优化调度，具有较大的实际应用意义。The invention discloses a storage system optimization scheduling method combining a shuttle car and an elevator, aiming at minimizing the scheduling travel time, and adopting a genetic algorithm to optimize the scheduling in-out storage model, so that the access travel time is greatly shortened, and significant savings can be achieved Time cost, saving energy, making full use of the advantages of high efficiency, high density, and high utilization rate of the automatic three-dimensional warehouse of the shuttle bus, realizing real-time and online optimal scheduling of the automatic three-dimensional warehouse, which has great practical application significance .

Description

A storage system optimization scheduling method combining shuttle cars and elevators

技术领域technical field

本发明属于密集仓储进出库调度技术领域，具体涉及一种穿梭子母车与升降机(1:1型)混合优化调度方法。The invention belongs to the technical field of dispatching in and out of intensive storage, and in particular relates to a hybrid optimal dispatching method of a shuttle bus and an elevator (1:1 type).

背景技术Background technique

随着科学技术和工业生产的快速发展，另基于“密集存储”概念的兴起，现代企业对于生产、仓储和配送要求的不断提高，促使仓储方式从最初通过人力手工作业的简单堆积到通过叉车等简单设备的仓库式存储改进为目前采用高位叉车、无人导引小车AGV、穿梭车等自动化设备的立体仓库存储。轨道式穿梭车(RGV)更是以其速度快、成本低和稳定性好在现代制造业、物流等行业内扮演着越来越重要的角色。With the rapid development of science and technology and industrial production, and based on the rise of the concept of "intensive storage", the requirements of modern enterprises for production, warehousing and distribution have been continuously improved, and the storage method has changed from the simple accumulation of manual work at the beginning to the use of forklifts, etc. The warehouse-style storage of simple equipment is improved to the three-dimensional warehouse storage that currently uses automation equipment such as high-position forklifts, unmanned guided vehicles AGV, and shuttle vehicles. Rail shuttle vehicle (RGV) is playing an increasingly important role in modern manufacturing, logistics and other industries because of its high speed, low cost and good stability.

传统立体仓库中多采用堆垛机实现货物的进出库调度，堆垛机的优点是实现了仓库作业的机械化与自动化，大大提高工作效率，同时，利用计算机进行控制和管理，作业过程和信息处理迅速、准确、及时，可加速物资周转，降低储存费用。然而，堆垛机需占用相应巷道进行作业，立体仓库有效存储面积无法充分利用，并且，单台堆垛机的竖直与水平作业无法同时进行，货物进出库效率较低。In traditional three-dimensional warehouses, stackers are mostly used to realize the scheduling of goods in and out of the warehouse. The advantages of stackers are that they realize the mechanization and automation of warehouse operations and greatly improve work efficiency. At the same time, they use computers for control and management. The operation process and information The processing is rapid, accurate and timely, which can speed up the turnover of materials and reduce storage costs. However, the stacker needs to occupy the corresponding roadway for operation, and the effective storage area of the three-dimensional warehouse cannot be fully utilized. Moreover, the vertical and horizontal operations of a single stacker cannot be carried out at the same time, and the efficiency of goods entering and leaving the warehouse is low.

目前国内外有少数立体仓库使用穿梭车与堆垛机结合进行作业，充分利用了仓库的有效面积和储存空间，使货物储存集中化、立体化，减少占地面积，降低土地购置费用。然而，国内外学者对穿梭车的研究大多基于静态调度，很少涉及动态混合调度，同时，企业定制化服务不断深入，小批量、多批次、高时效特点的订单不断增多，传统堆垛机式立体仓库与目前未完善的穿梭车式货架已不能满足如今市场需求。At present, there are a few three-dimensional warehouses at home and abroad that use the combination of shuttle cars and stackers to operate, making full use of the effective area and storage space of the warehouse, making the storage of goods centralized and three-dimensional, reducing the area occupied, and reducing land purchase costs. However, domestic and foreign scholars' research on shuttle vehicles is mostly based on static scheduling, and rarely involves dynamic mixed scheduling. The three-dimensional warehouse and the current imperfect shuttle racks can no longer meet the needs of today's market.

发明内容Contents of the invention

本发明的目的在于提供一种穿梭车与升降机结合的仓储系统优化调度方法。The object of the present invention is to provide a storage system optimization dispatching method combining a shuttle car and an elevator.

为达到上述目的，本发明采用了以下技术方案：To achieve the above object, the present invention adopts the following technical solutions:

1)根据全自动立体仓库中一台升降机配一套穿梭子母车的调度情况建立用于计算调度行程时间的数学模型；1) Establish a mathematical model for calculating the dispatching travel time according to the dispatching situation of an elevator with a set of shuttle bus in the fully automatic three-dimensional warehouse;

2)以调度行程时间最小化为目标，利用步骤1)所建立的数学模型对调度任务中各个待调度货物的调度顺序进行优化，确定该调度任务在利用一台升降机配一套穿梭子母车完成调度情况下的最优调度顺序。2) With the goal of minimizing the scheduling travel time, use the mathematical model established in step 1) to optimize the scheduling order of each cargo to be dispatched in the dispatching task, and determine that the dispatching task is using a lift with a set of shuttle bus The optimal scheduling order in case of complete scheduling.

所述调度任务包括存取货复合作业。The dispatching task includes the compound job of depositing and picking up goods.

在存取货复合作业情况下，计算调度行程时间的数学模型表示为：In the case of a compound operation of storage and retrieval, the mathematical model for calculating the scheduling travel time is expressed as:

T_d＝∑[m_ij×T_t+(1-m_ij)×T_b]T _d ＝∑[m _ij ×T _t +(1-m _ij )×T _b ]

其中，T_d为存取货复合作业调度行程时间，T_t为存取货位在同一层时完成一个存取对的调度行程时间，T_b为存取货位不在同一层时完成一个存取对的调度行程时间，存取对表示先存后取的连续两次货物调度；m_ij＝0或1，取值为0表示存取对的存取货不在同一层，取值为1表示存取对的存取货在同一层，i为存货，j为取货。Among them, T _d is the dispatching time of the compound operation of storage and retrieval, T _t is the scheduling travel time of completing an access pair when the storage and retrieval locations are on the same floor, and T _b is the completion of an access when the storage and retrieval locations are not on the same floor. The scheduling travel time of the pair, the access pair means two consecutive cargo dispatches that are deposited first and then _fetched ; The correct storage and withdrawal are on the same floor, i is the stock, and j is the pick-up.

所述T_t和T_b分别表示为：The T _t and T _b are expressed as:

${T T}_{t t} = = \frac{22 {z z}_{i i}}{{v v}_{s the s}} + + \frac{{x x}_{i i} + + {x x}_{j j}}{{v v}_{m m}^{' '}} + + \frac{{y the y}_{i i} + + {y the y}_{j j}}{{v v}_{z z}^{' '}} + + \frac{{y the y}_{i i} + + {y the y}_{j j}}{{v v}_{z z}} + + \frac{| | {x x}_{i i} - - {x x}_{j j} | |}{{v v}_{m m}} + + 44 {t t}_{z z} + + 22 {t t}_{m m} + + 22 {t t}_{m m}^{' '} + + 22 {t t}_{s the s}^{' '}$

${T T}_{b b} = = \frac{{z z}_{i i} + + | | {z z}_{i i} - - {z z}_{j j} | | + + {z z}_{j j}}{{v v}_{s the s}} + + \frac{{x x}_{i i} + + {x x}_{j j}}{{v v}_{m m}^{' '}} + + \frac{{y the y}_{i i} + + {y the y}_{j j}}{{v v}_{z z}^{' '}} + + \frac{{y the y}_{i i} + + {y the y}_{j j}}{{v v}_{z z}} + + \frac{{x x}_{i i} + + {x x}_{j j}}{{v v}_{m m}} + + 44 {t t}_{z z} + + 22 {t t}_{m m} + + 22 {t t}_{m m}^{' '} + + 22 {t t}_{s the s} + + 22 {t t}_{s the s}^{' '}$

其中，x、y以及z为货物对应货位在OXYZ坐标系下坐标，所述OXYZ坐标系中，原点O对应全自动立体仓库的I/O位置，X轴对应母车移动方向，Y轴对应子车移动方向，Z轴对应升降机移动方向；t_z为子车取货或卸货时间；t_m为空载情况下母车取子车或卸子车的时间；t_m'为满载情况下母车取子车或卸子车的时间；t_s为空载情况下升降机取母车或卸母车的时间；t_s'为满载情况下升降机取母车或卸母车的时间；v_s为升降机速度；v_m为母车空载速度，v_m'为母车满载速度；v_z为子车空载速度，v_z'为子车满载速度。Among them, x, y, and z are the coordinates of the corresponding cargo location of the goods in the OXYZ coordinate system. In the OXYZ coordinate system, the origin O corresponds to the I/O position of the fully automatic three-dimensional warehouse, the X axis corresponds to the moving direction of the mother car, and the Y axis corresponds to The moving direction of the child car, the Z axis corresponds to the moving direction of the elevator; t _z is the time for picking up or unloading the child car; _t _m is the time for the mother car to pick up or unload the child car under no-load conditions; t _{s is the time for the elevator to pick up or unload the mother car when it is unloaded; t s} _' is the time for the elevator to take or unload the mother car when it is fully loaded; v _s is Lift speed; v _m is the no-load speed of the mother car, v _m ' is the full-load speed of the mother car; v _z is the no-load speed of the sub-car, and v _z ' is the full-load speed of the sub-car.

所述优化的过程中，若调度任务内包括的存货任务量大于取货任务量，则用相应数量的待取货物坐标为(0,0,0)的取货任务使存取货任务量相等。In the optimization process, if the amount of inventory tasks included in the scheduling task is greater than the amount of pick-up tasks, use the corresponding number of pick-up tasks whose coordinates are (0,0,0) to equalize the amount of pick-up tasks .

所述优化的方法为遗传算法。The optimization method is genetic algorithm.

所述遗传算法中交叉概率P_c为0.7～0.9，变异概率P_m为0.1～0.2。In the genetic algorithm, the crossover probability P _c is 0.7-0.9, and the mutation probability P _m is 0.1-0.2.

本发明的有益效果体现在：The beneficial effects of the present invention are reflected in:

本发明对一台升降机配一套穿梭子母车的调度情况建立数学模型，然后建立相应出入库模型，采用遗传算法对出入库模型优化后，存取行程时间大大缩短，可以明显节约时间成本，节约能源，使穿梭子母车全自动立体仓库的高效率、高密度，以及高利用率等优点得到充分发挥，实现自动化立体仓库实时、在线的优化调度，具有较大的实际应用意义。The present invention establishes a mathematical model for the dispatching situation of a lift with a set of shuttle-cars, and then establishes a corresponding storage model. After optimizing the storage model by using a genetic algorithm, the storage and retrieval travel time is greatly shortened, which can significantly save time and cost. Saving energy can give full play to the advantages of high efficiency, high density, and high utilization rate of the automatic three-dimensional warehouse of the shuttle car, and realize real-time and online optimal scheduling of the automatic three-dimensional warehouse, which has great practical application significance.

附图说明Description of drawings

图1为穿梭子母车全自动立体仓库示意图；Figure 1 is a schematic diagram of a fully automatic three-dimensional warehouse with a shuttle bus;

图2为穿梭子母车全自动立体仓库模型图；Fig. 2 is a model diagram of the fully automatic three-dimensional warehouse of the shuttle bus;

图中：1为母车轨道，2为子车轨道，3为升降机轨道。Among the figure: 1 is the parent car track, 2 is the child car track, and 3 is the elevator track.

具体实施方式detailed description

下面结合附图和实施例对本发明进行详细说明。The present invention will be described in detail below in conjunction with the accompanying drawings and embodiments.

本发明提供一种穿梭子母车与升降机混合优化调度方法，具体按照以下步骤实施：The present invention provides a mixed optimal dispatching method for a shuttle car and an elevator, which is specifically implemented according to the following steps:

步骤1对一台升降机配一套穿梭子母车(1:1型)的调度情况建立合理的数学模型；Step 1 establishes a reasonable mathematical model for the dispatching situation of an elevator with a set of shuttle car (1:1 type);

1.1对穿梭子母车全自动立体仓库进行分析研究，制定货物存取规则1.1 Analyze and study the fully automatic three-dimensional warehouse of the shuttle bus, and formulate the rules for cargo storage and withdrawal

参见图1，穿梭子母车全自动立体仓库包含：穿梭子母车(由子车和母车组成)、母车行走轨道(巷道)、子车行走轨道(货架列)、穿梭式货架、托盘垂直提升系统(升降机)、托盘输送系统、CMS设备控制系统、WMS仓库管理系统。将实际穿梭子母车全自动立体仓库转化为OXYZ坐标系下的模型，如图2所示，包括升降机轨道3、母车轨道1和子车轨道2。其中，升降机将装有货物的子母车进行Z轴方向的运送，母车将子车与货物进行X轴方向运送，子车将货物进行Y轴方向运送。单个货位坐标为[x,y,z]，这个坐标是指货位中心的坐标，以I/O点(出入库站)为上述坐标系原点O。Referring to Figure 1, the fully automatic three-dimensional warehouse of the shuttle car includes: the shuttle car (composed of the child car and the mother car), the mother car running track (roadway), the child car walking track (shelf column), the shuttle rack, and the pallet vertical Lifting system (elevator), pallet conveying system, CMS equipment control system, WMS warehouse management system. Transform the actual fully automatic three-dimensional warehouse that shuttles mother and child cars into a model in the OXYZ coordinate system, as shown in Figure 2, including elevator track 3, mother car track 1, and child car track 2. Among them, the elevator transports the sub-cart with the goods in the direction of the Z axis, the mother car transports the sub-car and the goods in the direction of the X-axis, and the sub-car transports the goods in the direction of the Y-axis. The coordinates of a single cargo location are [x, y, z], which refer to the coordinates of the center of the cargo location, with the I/O point (inbound and outbound station) as the origin O of the above coordinate system.

本发明还设定如下货物存取规则：The present invention also sets the following cargo access rules:

(1)使用一套穿梭子母车，一次运送仅能携带一个货物。单独存货(或取货)情况下，穿梭子母车存货(取货)结束后回到出入库站(I/O点)；存取货复合作业情况下，穿梭子母车存货结束后，从该点出发进行接下来的取货任务，最终任务结束后，升降机和穿梭子母车均回到出入库站。(1) Using a set of shuttle car, only one cargo can be carried at a time. In the case of separate inventory (or pick-up), return to the inbound and outbound station (I/O point) after the inventory (pickup) of the shuttle vehicle is completed; From this point, the next pick-up task is carried out. After the final task is over, the elevator and the shuttle car return to the warehouse station.

(2)升降机和穿梭子母车的初始位置在出入库站(即I/O点)。(2) The initial position of the elevator and the shuttle bus is at the in-out station (i.e. the I/O point).

出入库订单(即调度任务)下达后，仓库管理系统(WMS)对订单进行分析，按照单纯存货作业、单纯取货作业、存取货复合作业三种情况，设备控制系统(CMS)调控升降机将子母车提升到相应的层，母车在巷道运动，停留在货物所在的货物列处，母车释放子车，子车完成存货或者取货作业后返回。After the inbound and outbound order (that is, the scheduling task) is issued, the warehouse management system (WMS) analyzes the order, and the equipment control system (CMS) regulates the lift to The mother car is lifted to the corresponding floor, the mother car moves in the roadway, stays at the cargo row where the goods are located, the mother car releases the child car, and the child car returns after completing the inventory or picking up the goods.

1.2建立一台升降机配一套穿梭子母车(1:1型)存取货物数学模型1.2 Establish an elevator with a set of shuttle car (1:1 type) to access the mathematical model of goods

定义：子车装、卸货物时间为t_z；母车装、卸空载子车时间为t_m，母车装、卸满载子车时间为t_m'，升降机装、卸空载母车时间为t_s，升降机装、卸满载母车时间为t_s'，装和卸时间一样；升降机速度为v_s；母车空载速度为v_m，母车满载速度为v_m'；子车空载速度为v_z，子车满载速度为v_z'。Definition: the loading and unloading time of sub-cars is t _z ; the time of loading and unloading empty-loaded sub-cars on the mother car is t _m , the time of loading and unloading full-loaded sub-cars on the mother car is t _m ', the time of elevator loading and unloading empty-loaded mother cars is t _s , the time for the elevator to load and unload the full-loaded mother car is _t _s ', and the time for loading and unloading is the same; the speed of the elevator is _v _s ; The loaded speed is v _z , and the fully loaded speed of the sub-car is v _z '.

(1)单纯存货作业模型(1) Simple inventory operation model

因升降机和子母车的初始位置在O点，且每次存货完成后子母车又回到O点，所以单纯存货的总调度行程时间T_c如下：Since the initial positions of the elevator and the master car are at point O, and the mother car returns to point O after each inventory is completed, the total scheduling travel time T _c of pure inventory is as follows:

T_c＝∑T_c1 T _c ＝∑T _c1

${T T}_{c c 11} = = ((\frac{22 {z z}_{i i}}{{v v}_{s the s}} + + \frac{{x x}_{i i}}{{v v}_{m m}} + + \frac{{x x}_{i i}}{{v v}_{m m}^{' '}} + + \frac{{y the y}_{i i}}{{v v}_{z z}} + + \frac{{y the y}_{i i}}{{v v}_{z z}^{' '}} + + {t t}_{s the s} + + {t t}_{s the s}^{' '} + + {t t}_{m m} + + {t t}_{m m}^{' '} + + 22 {t t}_{z z})) - - - - - - ((11))$

(2)单纯取货作业模型(2) Simple pick-up operation model

因升降机和子母车的初始位置在O点，且每次取货完成后子母车又回到O点，所以单纯取货的总调度行程时间T_q如下：Since the initial positions of the elevator and the mother car are at point O, and the mother car returns to point O after each pick-up, the total scheduling travel time T _q for simple pick-up is as follows:

T_q＝∑T_q1 T _q =∑T _q1

${T T}_{q q 11} = = ((\frac{22 {z z}_{j j}}{{v v}_{s the s}} + + \frac{{x x}_{j j}}{{v v}_{m m}} + + \frac{{x x}_{j j}}{{v v}_{m m}^{' '}} + + \frac{{y the y}_{j j}}{{v v}_{z z}} + + \frac{{y the y}_{j j}}{{v v}_{z z}^{' '}} + + {t t}_{s the s} + + {t t}_{s the s}^{' '} + + {t t}_{m m} + + {t t}_{m m}^{' '} + + 22 {t t}_{z z})) - - - - - - ((22))$

(3)存取货复合作业模型(3) Composite operation model for storage and retrieval

将货物的存取进行配对，称为存取对，即一存一取为一个单元。对于每个存取对，分两种情况：Pairing the access of goods is called an access pair, that is, one deposit and one withdrawal are a unit. For each access pair, there are two cases:

(a)取货货位和存货货位在同一层；(a) The pick-up location and the storage location are on the same floor;

(b)取货货位和存货货位不在同一层；(b) The pick-up location and the storage location are not on the same floor;

同时，将存取对的存取货是否在同一层设为一个0-1变量m_ij，At the same time, set whether the storage and withdrawal of the access pair are on the same layer as a 0-1 variable m _ij ,

则每一个存取对(i为存货，j为取货)，存取货位在同一层(z_i＝z_j)的调度行程时间T_t、存取货位不在同一层的调度行程时间T_b如下：Then for each access pair (i is stock, j is pick-up), the scheduling travel time T _t when the access location is on the same floor ( _zi = z _j ), and the scheduling travel time T when the access location is not on the same floor _b as follows:

${T T}_{t t} = = \frac{22 {z z}_{i i}}{{v v}_{s the s}} + + \frac{{x x}_{i i} + + {x x}_{j j}}{{v v}_{m m}^{' '}} + + \frac{{y the y}_{i i} + + {y the y}_{j j}}{{v v}_{z z}^{' '}} + + \frac{{y the y}_{i i} + + {y the y}_{j j}}{{v v}_{z z}} + + \frac{| | {x x}_{i i} - - {x x}_{j j} | |}{{v v}_{m m}} + + 44 {t t}_{z z} + + 22 {t t}_{m m} + + 22 {t t}_{m m}^{' '} + + 22 {t t}_{s the s}^{' '} - - - - - - ((33))$

${T T}_{b b} = = \frac{{z z}_{i i} + + | | {z z}_{i i} - - {z z}_{j j} | | + + {z z}_{j j}}{{v v}_{s the s}} + + \frac{{x x}_{i i} + + {x x}_{j j}}{{v v}_{m m}^{' '}} + + \frac{{y the y}_{i i} + + {y the y}_{j j}}{{v v}_{z z}^{' '}} + + \frac{{y the y}_{i i} + + {y the y}_{j j}}{{v v}_{z z}} + + \frac{{x x}_{i i} + + {x x}_{j j}}{{v v}_{m m}} + + 44 {t t}_{z z} + + 22 {t t}_{m m} + + 22 {t t}_{m m}^{' '} + + 22 {t t}_{s the s} + + 22 {t t}_{s the s}^{' '} - - - - - - ((44))$

存取货复合作业存取对的总调度行程时间T_d如下：The total scheduling travel time T _d of the access pair of the composite job of depositing and withdrawing goods is as follows:

T_d＝∑[m_ij×T_t+(1-m_ij)×T_b] (5)T _d ＝∑[m _ij ×T _t +(1-m _ij )×T _b ] (5)

综上，对于一台升降机配一套穿梭子母车的立体仓库的存取模型如下：To sum up, the access model for a three-dimensional warehouse with one elevator and one set of shuttle car is as follows:

步骤2对存取模型进行优化仿真，确定出入库作业最佳调度顺序与最短调度时间。Step 2 conducts optimization simulation on the access model, and determines the optimal scheduling sequence and the shortest scheduling time of the inbound and outbound operations.

本发明基于遗传算法运用MATLAB进行仿真分析，针对一个升降机配一套穿梭子母车的调度情况，从两个方面进行研究，分别是：存货数量等于取货数量，以及存货数量大于取货数量。存货数量小于取货数量的情况不予考虑。在这里详细介绍存取货数量相等时30个任务量的研究结果，其中参数值见表1。具体步骤如下：The present invention uses MATLAB to carry out simulation analysis based on the genetic algorithm, and studies the dispatching situation of an elevator equipped with a set of shuttle busses from two aspects, namely: the quantity in stock is equal to the quantity to be picked up, and the quantity in stock is greater than the quantity to be picked up. The situation where the stock quantity is less than the pick-up quantity will not be considered. Here is a detailed introduction of the research results of 30 tasks when the number of deposits and withdrawals is equal, and the parameter values are shown in Table 1. Specific steps are as follows:

给所有要存取货的货位分别编号，以货位的遍历次序作为遗传算法的编码。Number all the cargo locations to be stored and retrieved separately, and use the traversal order of the cargo locations as the code of the genetic algorithm.

在MATLAB中产生由一百个(种群数n＝100)随机遍历次序构成的初始群体。货物的位置是编译前指定的，也可以使用随机生成的坐标参数。An initial population consisting of one hundred (population number n=100) random traversal order was generated in MATLAB. The location of the cargo is specified before compilation, and a randomly generated coordinate parameter can also be used.

采用基本遗传算法中的选择、交叉、变异操作对货位的遍历次序进行优化，选取迭代次数c＝50，交叉概率P_c＝0.9，变异概率P_m＝0.2，适配值淘汰加速指数m＝2。本发明中使用的适配值函数如下：Use the selection, crossover and mutation operations in the basic genetic algorithm to optimize the traversal order of the cargo location, select the number of iterations c = 50, the crossover probability P _c = 0.9, the mutation probability P _m = 0.2, and the adaptation value elimination acceleration index m = 2. The adaptation value function used in the present invention is as follows:

fitness(i,1)＝(1-((len(i,1)-minlen)/(maxlen-minlen+0.0001)))^mfitness(i,1)=(1-((len(i,1)-minlen)/(maxlen-minlen+0.0001)))^m

其中，len(i,1)表示任意个体i对应的调度行程时间，maxlen和minlen分别为群体中调度行程时间最长和最短路径所用时间。Among them, len(i,1) represents the scheduling travel time corresponding to any individual i, and maxlen and minlen are the longest and shortest scheduling travel time in the group respectively.

利用fitness>rand选择个体，将时间较小(适应度较大)个体选择并保留下来。Use fitness>rand to select individuals, and select and retain individuals with smaller time (larger fitness).

当存货数量恰好等于取货数量时：When the inventory quantity is exactly equal to the pickup quantity:

30个货物对应货位坐标分别表示为N(x y z)及N’(x y z)，其中，15个为存货(矩阵A)，N代表存货货位编号(N＝1～15)，15个为取货(矩阵B)，N’代表取货货位编号(N’＝1’～15’)。为寻求一条路径使存取货完成的时间最短，货物对应的货位坐标如下：The location coordinates corresponding to the 30 goods are represented as N(x y z) and N'(x y z), among which, 15 are inventory (matrix A), N represents the number of inventory location (N=1~15), and 15 are Goods (matrix B), N' represents the number of the pick-up location (N'=1'～15'). In order to seek a route to minimize the time to complete the storage and retrieval of goods, the coordinates of the corresponding goods are as follows:

A＝[1(13 10 3)；2(7 6 4)；3(15 11 2)；4(3 5 2)；5(14 9 3)；6(8 2 1)；7(12 12 2)；8(16 3 4)；9(20 92)；10(2 4 3)；11(3 8 2)；12(3 10 3)；13(13 9 1)；14(5 8 2)；15(14 4 3)]；A＝[1(13 10 3); 2(7 6 4); 3(15 11 2); 4(3 5 2); 5(14 9 3); 6(8 2 1); 7(12 12 2 );8(16 3 4);9(20 92);10(2 4 3);11(3 8 2);12(3 10 3);13(13 9 1);14(5 8 2); 15(14 4 3)];

B＝[1’(5 7 4)；2’(16 4 1)；3’(17 6 2)；4’(4 11 2)；5’(10 11 1)；6’(9 2 3)；7’(8 4 2)；8’(16 53)；9’(19 5 1)；10’(13 6 1)；11’(17 9 3)；12’(7 3 2)；13’(2 2 4)；14’(12 1 2)；15’(18 7 2)]；B＝[1'(5 7 4); 2'(16 4 1); 3'(17 6 2); 4'(4 11 2); 5'(10 11 1); 6'(9 2 3) ;7'(8 4 2);8'(16 53);9'(19 5 1);10'(13 6 1);11'(17 9 3);12'(7 3 2);13' (2 2 4); 14'(12 1 2); 15'(18 7 2)];

随机生成一百个关于A中元素的排列顺序(遍历次序数组a)，随机生成一百个关于B中元素的排列顺序(遍历次序数组b)，然后确定数组a中元素(代表一个遍历次序)与数组b中元素(代表一个遍历次序)的一一对应映射关系，由一百个映射关系从而形成一百个关于30个货物量的存取货复合作业的遍历次序。对数组a和数组b中元素分别利用遗传算法进行优化，改变数组中元素的取值(即遍历次序)，并按照存取货复合作业的遍历次序计算适配值(既是数组a中相应元素的适配值，也是数组b中相应元素的适配值，所谓相应即构成一个映射关系的两个元素)。Randomly generate one hundred sorting orders of elements in A (traversal order array a), randomly generate one hundred sorting orders about elements in B (traversal order array b), and then determine the elements in array a (representing a traversal order) One-to-one mapping relationship with the elements in the array b (representing a traversal order), one hundred mapping relationships form one hundred traversal orders for the composite operation of depositing and withdrawing 30 goods. The genetic algorithm is used to optimize the elements in array a and array b respectively, changing the value of the elements in the array (that is, the traversal order), and calculating the adaptation value according to the traversal order of the compound operation of accessing goods (that is, the corresponding element in array a The adaptation value is also the adaptation value of the corresponding element in the array b, the so-called correspondence refers to two elements forming a mapping relationship).

根据程序随机所得一组存取货复合作业顺序为：(8→3’)→(12→14’)→(3→12’)→(13→10’)→(9→15’)→(1→11’)→(2→1’)→(5→5’)→(10→6’)→(7→4’)→(11→7’)→(6→13’)→(14→8’)→(15→9’)→(4→2’)。行程时间为：RTime＝1015.2(s)。According to the program, a group of composite operation sequence of accessing goods randomly obtained is: (8→3')→(12→14')→(3→12')→(13→10')→(9→15')→( 1→11')→(2→1')→(5→5')→(10→6')→(7→4')→(11→7')→(6→13')→(14 →8')→(15→9')→(4→2'). The travel time is: RTime=1015.2(s).

经过优化后所得一组存取货复合作业顺序为：(13→9’)→(14→4’)→(6→10’)→(7→14’)→(9→2’)→(8→1’)→(5→8’)→(12→13’)→(11→7’)→(15→11’)→(4→12’)→(2→15’)→(10→6’)→(3→3’)→(1→5’)。行程时间为：RTime＝977.75(s)。优化效率：η＝(优化前行程时间-优化后行程时间)/优化前行程时间＝(1015.2-977.75)/1015.2＝3.7％。After optimization, the order of a group of composite operations of storage and retrieval is: (13→9')→(14→4')→(6→10')→(7→14')→(9→2')→( 8→1')→(5→8')→(12→13')→(11→7')→(15→11')→(4→12')→(2→15')→(10 →6')→(3→3')→(1→5'). The travel time is: RTime=977.75(s). Optimization efficiency: η=(travel time before optimization-travel time after optimization)/travel time before optimization=(1015.2-977.75)/1015.2=3.7%.

其他情况方法同上，将最终得到的结果归纳入表2。其中，当存货数量＞取货数量时，可引入相应数量的(0 0 0)点作为取货货位坐标，使得存货数量＝取货数量。同时，由于一套穿梭子母车时，对于单纯存货或者单纯取货的订单，其行程时间和存货或者取货的顺序无关，所以表2中并未进行记录。In other cases, the method is the same as above, and the final results are included in Table 2. Wherein, when the quantity in stock > the quantity to pick up, a corresponding number of (0 0 0) points can be introduced as the coordinates of the pick-up position, so that the quantity in stock = the quantity to pick up. At the same time, when a set of shuttle busses is used, the travel time for orders that are purely stocked or picked up has nothing to do with the order of stocking or picking up, so it is not recorded in Table 2.

由表2得知通过本发明所建立的出入库模型，经过算法优化后货位的取存操作排序得到优化，穿梭车和升降机的行程时间和明显缩短，可以大大节约时间成本，提高了效率，进一步证明了该模型的可行性，对于目前市面上的自动化立体仓库，经济效益可观。It is known from Table 2 that through the warehouse-in and out-of-warehouse model established by the present invention, after the algorithm is optimized, the sorting of the storage and retrieval operations of the cargo space is optimized, and the travel time of the shuttle car and the elevator is significantly shortened, which can greatly save time and cost and improve efficiency. It further proves the feasibility of the model. For the automated warehouses currently on the market, the economic benefits are considerable.

表1.模型编程参数Table 1. Model programming parameters

表2.仿真分析结果Table 2. Simulation analysis results

Claims

1. A storage system optimization scheduling method combining a shuttle car and a lift, characterized in that: comprising the following steps:

1) Establish a mathematical model for calculating the dispatching travel time according to the dispatching situation of an elevator with a set of shuttle bus in the fully automatic three-dimensional warehouse;

2) With the goal of minimizing the scheduling travel time, use the mathematical model established in step 1) to optimize the scheduling order of each cargo to be dispatched in the dispatching task, and determine that the dispatching task is using a lift with a set of shuttle bus The optimal scheduling order in case of complete scheduling.

2. According to claim 1, a storage system optimization scheduling method combining shuttle vehicles and elevators, characterized in that: said scheduling tasks include combined operations of storage and retrieval.

3. According to claim 2, a storage system optimization dispatching method combining a shuttle car and an elevator is characterized in that: in the case of combined operation of storage and retrieval, the mathematical model for calculating the dispatching travel time is expressed as:

T _d ＝∑[m _ij ×T _t +(1-m _ij )×T _b ]

Among them, T _d is the dispatching time of the compound operation of storage and retrieval, T _t is the scheduling travel time of completing an access pair when the storage and retrieval locations are on the same floor, and T _b is the completion of an access when the storage and retrieval locations are not on the same floor. The scheduling travel time of the pair, the access pair means two consecutive cargo dispatches that are _deposited first and then retrieved; The correct storage and withdrawal are on the same floor, i is the stock, and j is the pick-up.

4. According to claim 3, a storage system optimization scheduling method combining a shuttle car and an elevator is characterized in that: said T _t and T _b are respectively expressed as:

{T T}_{t t} = = \frac{22 {z z}_{i i}}{{v v}_{s the s}} + + \frac{{x x}_{i i} + + {x x}_{j j}}{{v v}_{m m}^{' '}} + + \frac{{y the y}_{i i} + + {y the y}_{j j}}{{v v}_{z z}^{' '}} + + \frac{{y the y}_{i i} + + {y the y}_{j j}}{{v v}_{z z}} + + \frac{| | {x x}_{i i} - - {x x}_{j j} | |}{{v v}_{m m}} + + 44 {t t}_{z z} + + 22 {t t}_{m m} + + 22 {t t}_{m m}^{' '} + + 22 {t t}_{s the s}^{' '}

{T T}_{b b} = = \frac{{z z}_{i i} + + | | {z z}_{i i} - - {z z}_{j j} | | + + {z z}_{j j}}{{v v}_{s the s}} + + \frac{{x x}_{i i} + + {x x}_{j j}}{{v v}_{m m}^{' '}} + + \frac{{y the y}_{i i} + + {y the y}_{j j}}{{v v}_{z z}^{' '}} + + \frac{{y the y}_{i i} + + {y the y}_{j j}}{{v v}_{z z}} + + \frac{{x x}_{i i} + + {x x}_{j j}}{{v v}_{m m}} + + 44 {t t}_{z z} + + 22 {t t}_{m m} + + 22 {t t}_{m m}^{' '} + + 22 {t t}_{s the s} + + 22 {t t}_{s the s}^{' '}

Among them, x, y, and z are the coordinates of the corresponding cargo location of the goods in the OXYZ coordinate system. In the OXYZ coordinate system, the origin O corresponds to the I/O position of the fully automatic three-dimensional warehouse, the X axis corresponds to the moving direction of the mother car, and the Y axis corresponds to The moving direction of the child car, the Z axis corresponds to the moving direction of the elevator; t _z is the time for picking up or unloading the child car; _t _m is the time for the mother car to pick up or unload the child car under no-load conditions; t _{s is the time for the elevator to pick up or unload the mother car when it is unloaded; t s} _' is the time for the elevator to take or unload the mother car when it is fully loaded; v _s is Lift speed; v _m is the no-load speed of the mother car, v _m ' is the full-load speed of the mother car; v _z is the no-load speed of the sub-car, and v _z ' is the full-load speed of the sub-car.

5. According to claim 4, a storage system optimization scheduling method combining a shuttle car and an elevator is characterized in that: during the optimization process, if the amount of inventory tasks included in the scheduling task is greater than the amount of picking tasks, then use The corresponding quantity of pick-up tasks whose coordinates are (0,0,0) make the deposit and pick-up tasks equal.

6. According to claim 1, a storage system optimization scheduling method combining shuttles and elevators, characterized in that: the optimization method is a genetic algorithm.

7. The storage system optimization scheduling method combining shuttles and elevators according to claim 6, characterized in that: in the genetic algorithm, the crossover probability P _c is 0.7-0.9, and the mutation probability P _m is 0.1-0.2.