Flexible processing box type stored leaf logistics control method and system

Abstract

The invention discloses a box type stored leaf logistics control method and system for flexible processing. The method comprises the following steps: S1, boxing: establishing a total boxing amount estimation model, and predicting an optimal solution of the total boxing amount; then, the weight and the number of the boxed boxes are obtained in real time, a dynamic boxing algorithm is called, and the expected boxing value and the total boxing number of each box are dynamically corrected; finally, judging whether the packing data is noise data with inaccurate scene description or not, and cleaning the noise data; s2, warehousing: warehousing the boxed cigarette boxes for storage; s3, ex-warehouse: obtaining formula information in the plan according to the ex-warehouse plan; according to the formula information, obtaining a corresponding ex-warehouse mode and calling a corresponding ex-warehouse algorithm; and forming an ex-warehouse smoke box sequence according to the ex-warehouse plan and the ex-warehouse mode for ex-warehouse. The method is realized through a logistics system. Through a dynamic boxing mode, the utilization rate of the cigarette boxes is increased, and a guarantee is provided for the subsequent process; and various warehouse-out modes are achieved, the flexibility of production organizations is improved, the uniformity of raw materials of cigarette products is improved, and the overall quality of the products is further improved.

Description

Technical Field

[0001] The present invention relates to the field of warehouse management, and in particular to a flexible processing box-type leaf storage logistics control method and system. Background Art

[0002] With the development requirements of refined processing of Chinese cigarette characteristics, automated box-type leaf storage logistics systems have been widely used in tobacco industry enterprises.

[0003] In the box-type tobacco storage logistics system, wooden boxes are often used as the main containers for box storage, which are mainly divided into the packing area, storage area and box turning area. The main process is as follows: first pack at the packing station, that is, put tobacco leaves into the empty box; then the box filled with tobacco leaves is transported to the warehouse for storage and aging through the conveying equipment; when the tobacco leaves need to be used, the stored and aging tobacco leaves are transported to the box turning feeding station again through the conveying equipment, and the box turning robot grabs the tobacco box, turns the box and unloads the materials to the feeding bin for subsequent tobacco processing; the empty box after unloading is transported by the conveying equipment to the empty box buffer area or storage warehouse; the entire flow process of the tobacco box forms an automated closed-loop system.

[0004] In the current cigarette manufacturing process, batches are used as a processing unit; a batch is divided into two processing modes: full recipe and group recipe. When using a box-type leaf storage logistics system in the group recipe processing mode, each module is packed into a box and put into storage separately; when leaving the warehouse, multiple modules with different numbers of cigarette boxes in each module form a complete batch for box turnover and unloading.

[0005] Therefore, the traditional box-type leaf storage system has the following shortcomings: 1) When the quantitative packing mode is adopted for warehousing, it is often impossible to evenly distribute various categories. The excess unallocated cigarette boxes are called tail boxes. The probability of producing tail boxes will increase, resulting in increased inventory and reduced utilization rate of cigarette boxes, affecting the uniformity of material blending and the stability of material flow; 2) The outbound mode is single and cannot meet the needs of multiple outbound modes at the same time; 3) If the outbound sorting method is fixed, it cannot meet the sorting requirements of different quantities of multiple modules; if disordered outbound is adopted, the same modules in the same batch will be repeatedly stacked to a certain extent, increasing the possibility of uneven use of raw materials, which is not conducive to the addition of moisture and liquid in the subsequent feeding process; 4) When the cigarette boxes are "wrongly branded and serially numbered" when entering and leaving the warehouse, there is no complete automated processing mechanism.

[0006] Nowadays, the box storage logistics automation system is the key development direction of the technological transformation of cigarette manufacturing enterprises. It organically combines logistics with the production and operation management of the enterprise, realizes the flexibility and automation of the production, transportation, storage and other links of the enterprise, and meets the needs of reform. Summary of the invention

[0007] In order to solve the problems of the prior art, the purpose of the present invention is to provide a box-type leaf storage logistics control method and system for flexible processing. On the basis of the traditional quantitative packing mode, it is improved to a dynamic packing mode, which solves the "tail box" that appears in the packing process and eliminates manual intervention in the "packing volume", thereby improving the utilization rate of the cigarette box and providing guarantees for subsequent processes; organizing production in a variety of outbound modes according to the production plan, improving the flexibility of production organization, and fully meeting the needs of "group processing, homogenization processing, and information processing" in cigarette production; complete error-proof and abnormal equipment and procedures ensure the stability of system operation and avoid wrong brand and serial number in cigarette production. It provides reliable software and hardware support for box-type logistics, and improves the production, operation and manufacturing level and the information management level of the enterprise as a whole. The specific contents are as follows:

[0008] A flexible processing box-type leaf storage logistics control method comprises the following steps:

[0009] S1. Packing, including S1.1. Before packing begins, a packing total quantity estimation model is established based on packing data of historical batches to predict the optimal solution for the total packing quantity; S1.2. During the packing process, the weight and quantity of the packed boxes are obtained in real time, and a dynamic packing algorithm is called to dynamically correct the expected packing value and the total number of packed boxes for each box; S1.3. After packing is completed, it is determined whether the packing data is noise data that does not accurately describe the scene. If it is noise data, the data record is cleaned;

[0010] S2, storage, storing the packed cigarette boxes;

[0011] S3, outbound delivery, including steps S3.1, according to the outbound delivery plan, obtaining the recipe information in the plan; S3.2, according to the recipe information, obtaining the outbound delivery method in the recipe, the outbound delivery method includes full recipe outbound delivery mode, grouped recipe outbound delivery mode and flower arrangement sorting outbound delivery mode, and calling the corresponding outbound delivery algorithm according to different outbound delivery methods; S3.3, after forming an outbound cigarette box sequence according to the outbound delivery plan and the outbound delivery method, the cigarette boxes are transported to the unloading platform through the logistics system and then outbound, and the sorting batch sequence is confirmed before outbound delivery; S3.4, after the outbound delivery is completed, the materials in the cigarette box are dumped out by the box turning and material emptying cleaning robot, and the empty cigarette box after the box turning and material emptying is recovered; S3.5, the dumped logistics are mixed evenly and enter the downstream section.

[0012] Further, before the start of packing in step S1.1, the following steps are included: According to historical data, the material weight, incoming cabinet moisture, environmental temperature and humidity, leaf storage time in the upstream process storage cabinet are related to the final packing total. The decision tree regression algorithm is used to divide the sample space into multiple sub-spaces, and each sub-space corresponds to a decision tree node. At each node, the algorithm selects an optimal partitioning variable and partitioning point, divides the sample into two subsets, and then recursively partitions each subset until a preset stopping condition is reached. When predicting, the algorithm starts from the root node along the decision tree, continuously traverses downward according to the feature values of the test samples, and finally reaches a leaf node, and the value corresponding to this leaf node is the predicted value. After the historical data collection is completed, use the training data to build a decision tree model, determine the optimal partitioning variable and partitioning point for each node, find the most suitable decision tree parameters, perform pruning on the decision tree to avoid overfitting, use the test data to verify the decision tree, predict the results, evaluate the performance of the model, use the trained model for prediction and interpretation, predict the new data, and finally obtain the optimal solution of the predicted packing total amount.

[0013] Further, in step S1.2 during the packing process, the following steps are included: The dynamic packing mode is divided into two types. Mode 1 is applicable to the cigarette boxes containing tobacco leaves, and both the packing weight and the number of packed boxes of the cigarette box are dynamic values. Mode 2 is applicable to the cigarette boxes containing reconstituted tobacco sheets, the packing weight is a dynamic value, and the number of packed boxes is a fixed value. Before the start of packing, determine the corresponding packing method according to the material carried by the cigarette box.

[0014] When using Mode 1, both the packing weight and the number of packed boxes are dynamic values. According to the quotient of the predicted packing total amount or the remaining total amount M (KG) and the set central packing weight value N (KG) in S1.1, the preset value of the number of packed boxes i is obtained. Add 1 to the preset value of the number of packed boxes to get j, that is, i + 1, and thus obtain:

[0015] i = TRUNC(M / N), j = i + 1,

[0016] For the predicted packing total amount or the remaining total amount M (KG) and the two predicted values of the number of packed boxes i and j, the two quotient values are obtained to get two preset values of the average packing weight N 1 (KG), N 2 (KG), and thus obtain:

[0017] N 1 = [ROUND * (M / i * 100.0)] / 100, N 2 = [ROUND * (M / j * 100.0)] / 100,

[0018] For the two preset values of the average packing weight N 1 (KG), N 2(KG) is compared with the absolute value of the difference from the center value N (KG) of the packing weight, that is:

[0019] K 1 = ABS(N - N 1 ), K 2 = ABS(N - N 2 ),

[0020] If K 2 is greater than K 1 , and the preset value i of the packing quantity is closer to the center value of the packing weight, then the average preset packing weight N 1 is used for packing, and the final estimated packing quantity is i; conversely, if K 1 is greater than K 2 , and the preset value i of the packing quantity is closer to the center value of the packing weight, then the packing station uses the average preset packing weight N 2 for packing, and the final estimated packing quantity is j (j is i + 1). If the actual packing value is X (kg), then the remaining total packing quantity is M = M – X. After packing is completed, M and N are substituted into the above formula, and the process enters a loop until the total packing quantity is 0;

[0021] When using Method 2, the weight is a dynamic value and the number of packing boxes is a fixed value. The quotient value obtained from the predicted total packing quantity or the remaining total quantity M (KG) and the number of packing boxes Y according to step S1.1 is used to obtain the average preset packing weight N (KG). The packing station packs according to this value. After one box of materials is packed, based on the cumulative packed weight X (KG) of the packing station, the total weight M (KG) of the remaining materials to be packed is obtained. Using this weight and the remaining number of packing boxes Y, the average packing weight N (KG) of the remaining cigarette boxes is obtained. By analogy through the iterative algorithm, if the cumulative packing quantity of the previous cigarette boxes in the same batch is positively biased, then the average packing weight of each box is adjusted by calculation to make the subsequent packing weight negatively biased, and vice versa. The expected packing value and the total number of packings of each box are dynamically corrected to obtain

[0022] N = [ROUND * (M / Y * 100.0)] / 100,

[0023] If the actual packing value is X (kg), then the remaining total packing quantity is M = M – X. After packing is completed, M and N are substituted into the above formula, and the process enters a loop until the total packing quantity is 0.

[0024] Further, after the step S1.3, after the packing is completed, the following steps are included: judging the noise data respectively through the packing time, whether to stop the machine, and the packing parameters; setting the packing time as T (T is a system-set parameter) minutes. If the actual packing time is greater than the set packing time, the system determines that a failure occurs during the packing process, performs data cleaning, and does not record it into the packing history data; if the machine stops during the packing process, the system determines that a failure occurs during the packing process and performs data cleaning; if there is a deviation between the packing parameters and the set parameters, the system determines that the packing parameter setting is incorrect and performs data cleaning.

[0025] Further, the specific steps of the flower arrangement sorting outbound mode are as follows.

[0026] a), Suppose there are a total of N types of modules for outbound (when N ∈ (2, +∞), it is the leaf group formula outbound in order), and the outbound modules are (X 1 , X 2 , …, X N-1 , X N ), then the outbound quantity of each module is (M 1 , M 2 , …, M N-1 , M N ).

[0027] Traverse the quantity of each module. When M Y = 1, the module number and quantity are pushed onto the stack for storage;

M Y ∈ (M 1 , M 2 , …, M N-1 , M N )

[0028] b), Suppose the modules are divided into K groups, K ∈ (2, MIN(M 1 , M 2 , …, M N-1 , M N ))

When K = 1, it is the full formula outbound

[0029] Traverse K (2, MIN(M 1 , M 2 , …, M N-1 , M N ))

[0030] When K = 2, M 1 % 2 = L 21 , M 2 % 2 = L 22 , …, M N % 2 = L 2N , U 2 = MAX(L 21 , L 22 , …, L2N );

[0031] When K = 3, M 1 % 3 = L 31 , M 2 % 3 = L 32 , …, M N % 3 = L 3N , U 3 = MAX(L 31 , L 32 , …, L 3N );

[0032] …………

[0033] Then, the general formula is:

[0034] M 1 % K = L K1 , M 2 % K = L K2 , …, M N % K = L KN , U K = MAX(L K1 , L K2 , …, L KN );

[0035] c), Let J, if U J = MIN(U 2 , U 3 , …, U n-1 , U n ), then the optimal solution of K is J,

[0036] If U O = U P = … = U q

(U O , U P , …, U q ) ∈ (U 2 , U 3 , …, U n-1 , U n )

[0037] Divide the number of outgoing cigarette boxes into J groups;

[0038] Then the number of groups for the optimal solution is (H J1 , H J2 , …, H JN-1 , H JN ), with a modulus of (L J1 , L J2 , …, L JN-1 , L JN ) Then:

[0039] H J1 =FLOOR(M 1 / J),L J1 =M 1 %J;

[0040] H J2 =FLOOR(M 2 / J),L J2 =M 2 %J;

[0041] ……

[0042] H JN =FLOOR(M N / J),L JN =M N J;

[0043] Then,

[0044] If the outbound module is (X 1 ,X 2 ,…,X N-1 ,X N ), then the outbound quantity of each module in each group is (H J1 ,H J2 ,…,wH JN-1 ,H JN ), the remaining single boxes that cannot be grouped are (L J1 ,L J2 ,…,L JN-1 ,L JN ). Traverse (L J1 ,L J2 ,…,L JN-1 ,L JN ). When L JY =1, the module number and quantity are pushed onto the stack and saved;

L JY ∈(L J1 ,L J2 ,…,L JN-1 ,L JN )

[0045] Group each group and repeat iterations a) - c) until each group is a single cigarette box;

[0046] Pop the data from the stack to obtain the sorted data.

[0047] Furthermore, in the step S1, when the empty cigarette box enters the packing station during packing, the RFID reader reads the RFID electronic tag at the bottom of the empty cigarette box; after packing is completed, in step S2, when entering the warehouse, the RFID reader writes the relevant information of the material in the cigarette box into the RFID electronic tag to ensure that the information recorded in the system database, the information recorded in the RFID electronic tag of the cigarette box and the material in the cigarette box are consistent; a warehouse-out buffer area is provided, and in step S3.3, the solid boxes are transported to the warehouse-out buffer area before the sorting batch sequence is confirmed, and when the solid boxes enter the solid box warehouse-out buffer area, the RFID reader reads the RFID electronic tag at the bottom of the empty cigarette box, and the information carried in the RFID electronic tag of the cigarette box is the same as the information recorded in the system database, then the next step is entered normally; the information recorded in the RFID electronic tag of the cigarette box is different from the information recorded in the system database, the system alarms, and the RFID electronic tag of the cigarette box and the system database are both marked as unqualified cigarette boxes, and this cigarette box will not be turned over and unloaded subsequently, and the system will replenish the materials, and then the unqualified cigarette boxes will be directly transported to manual processing.

[0048] Furthermore, there are two ways to replenish materials. When the outbound delivery mode is either full formula outbound delivery or group formula outbound delivery, which are two unordered outbound delivery modes, replenishment is carried out at the end of the batch; when the outbound delivery mode is the flower arrangement outbound delivery mode, the emergency replenishment mode is used to replace and replenish materials for specific cigarette boxes.

[0049] A logistics system for a box-type leaf storage with flexible processing, including a shelf, a stacker crane, a cigarette box turnover conveyor line, a double-station straight shuttle car, a packing station, a manual processing area, an empty box buffer area, a reciprocating vertical elevator, a box-turning and material-dumping robot, a storage-type feeding elevator, a single-station straight shuttle car, a circular sorting area, and a full box outbound buffer area; the packing station is connected to the empty box buffer area and the stacker crane through the cigarette box turnover conveyor line, the stacker crane is connected to the full box outbound buffer area through the cigarette box turnover conveyor line, the full box outbound buffer area is connected to the box-turning, material-dumping and cleaning robot through the reciprocating vertical elevator, the box-turning, material-dumping and cleaning robot is connected to the storage-type feeding elevator, the double-station straight shuttle car can be connected to the stacker crane, the packing station, the empty box buffer area, the reciprocating vertical elevator, and the manual processing area through the cigarette box turnover conveyor line, and the single-station straight shuttle car can be connected to the stacker crane and the full box outbound buffer area through the cigarette box turnover conveyor line; a circular sorting area is configured at the entrance of the full box outbound buffer area. After the circular sorting area, it is divided into three outbound channels, and an RFID reader is configured on each channel to identify whether the information of the incoming cigarette box matches the database information; RFID readers are installed on the packing station, the manual processing area, the box-turning, material-dumping robot, and the full box outbound buffer area, and RFID electronic tags are installed at the bottom of the cigarette box; the RFID readers are used to read the cigarette box information in the RFID electronic tags to identify whether the RFID information is consistent with the information in the database.

[0050] Further, the system is configured with two packing stations, which can simultaneously and separately pack tobacco leaves and cut tobacco. The packing station is connected to the upstream feeding section through a belt conveyor, and the tobacco leaves enter the packing station through the belt conveyor; the packing station is connected to the empty box buffer area and the stacker crane through the cigarette box turnover conveyor line, and the empty cigarette boxes in the empty box buffer area are transported into the packing station for packing the tobacco leaves in boxes and then can be transported to the stacker crane for storage in the warehouse.

[0051] Further, the system is configured with four stacker cranes. The stacker cranes for loading and receiving goods are configured in two layers, respectively located on the ground and the steel platform; each stacker crane is provided with a receiving port and a discharging port on the first layer for outbound and inbound of empty boxes and full boxes; each stacker crane is provided with two discharging ports on the second-layer steel platform for outbound of full boxes, and the setting of the two discharging ports can support different outbound modes simultaneously.

[0052] Further, the stacker is connected to the outbound buffer area of filled cartons through the carton turnover and conveying line. The single-station straight shuttle car is connected to the stacker and the outbound buffer area of filled cartons through the carton turnover and conveying line. The outbound buffer area of filled cartons is divided into three outbound channels. Any one of the outbound channels is connected to the carton turning, dumping and cleaning robot through a reciprocating vertical elevator. The carton turning, dumping and cleaning robot is connected to the storage type feeder elevator. The storage type feeder elevator is connected to the downstream section through a belt conveyor.

[0053] Further, the reciprocating vertical elevator has a double-inlet and single-outlet structure, that is, the carton can be input from the inlet on the first or second floor and output from the outlet on the first floor. The double-station straight shuttle car is connected to the stacker, the carton packing station, the empty carton buffer area, the reciprocating vertical elevator, and the manual processing area through the carton turnover and conveying line, and can supply empty cartons in the empty carton buffer area and return the excess empty cartons to the warehouse. The defective cartons are transported to the manual processing area for manual processing and then returned to the warehouse. The filled cartons in the warehouse are transported to the inlet on the first floor of the reciprocating vertical elevator for emergency replenishment. The entire batch of cartons before being transported to the carton turning and dumping robot is returned to the warehouse uniformly.

[0054] The working principle of the present invention is as follows:

[0055] For warehousing of the present invention, a packing total amount prediction model is established by using the decision tree regression algorithm, the packing historical data, the material weight of the upstream process storage cabinet, the moisture content entering the cabinet, the leaf storage time, and the current ambient temperature and humidity to predict the optimal solution of the packing total amount. The weight and quantity of the packed cartons are obtained in real time, and the dynamic packing algorithm is called to dynamically correct the expected value of each carton's packing and the total number of packed cartons. Finally, it is judged whether the packing data this time is noise data with inaccurate scene description, and the noise data is cleaned.

[0056] When the present invention is for outbound, it has multiple outbound modes such as full formula, grouped formula, and interspersed sorting. The interspersed sorting adopts an automated sorting algorithm, which can sort according to the number of any module in the batch and the number of any carton in each module. The optimal solution is obtained by using the methods of grouping, traversing, and taking the modulus. After obtaining the optimal solution, the solution numbers are saved in a stack manner, and finally the stack is popped to obtain the best sorting method, and then the outbound is carried out in sequence according to the calculated sorting method, so that the materials of multiple modules in the batch can be fully mixed. It is equipped with a corresponding logistics system to accurately grab and store the cartons according to the calculation results, avoiding waste of storage space and reducing manual intervention.

[0057] The present invention has the following beneficial effects compared with the prior art:

[0058] (1) Based on the traditional quantitative boxing mode, the present invention is improved to a dynamic boxing mode, which solves the "last box" problem in the boxing process, eliminates manual intervention in the "boxing quantity", improves the utilization rate of cigarette boxes and provides guarantee for subsequent processes; organizes production according to the production plan in various outbound modes, improves the flexibility of production organization, and fully meets the needs of cigarette production for "group processing, homogenization processing, and informatization processing"; after using a certain order of outbound modes corresponding to the demand, the uniformity of raw materials of cigarette products can be increased, the overall quality of products can be further improved, and the occurrence of extremely unqualified products can be avoided.

[0059] (2) The complete error-proof and abnormality-proof equipment and procedures of the present invention ensure the stability of system operation and avoid misbranding and serial number mixing in cigarette production; the present invention provides reliable software and hardware support for box-type logistics, and overall improves the enterprise's production, operation and manufacturing level and the enterprise's informatization management level. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 It is the flowchart of the method described in the embodiment of the present invention;

[0061] Figure 2 It is the schematic diagram of the first-layer layout described in the embodiment of the present invention;

[0062] Figure 3 It is the schematic diagram of the second-layer layout described in the embodiment of the present invention;

[0063] Figure 4 It is the flowchart of the dynamic boxing described in the embodiment of the present invention;

[0064] Figure 5 It is the flowchart of the outbound box turning described in the embodiment of the present invention;

[0065] Figure 6 It is the flowchart of the flower arrangement sorting algorithm described in the embodiment of the present invention;

[0066] Figure 7 It is the specific outbound box sorting schematic diagram in the embodiment of the present invention.

[0067] In the figure:

[0068] 1 - Shelf; 2 - Stacker; 3 - Cigarette box turnover conveyor line; 4 - Double-station straight shuttle car;

[0069] 5 - Boxing station; 6 - Manual processing area; 7 - Empty box buffer area; 8 - Reciprocating vertical elevator;

[0070] 9 - Box turning and discharging robot; 10 - Warehouse-type feeding elevator; 11 - Single-station straight shuttle car;

[0071] 12 - Ring sorting area; 13 - Full box outbound buffer area. DETAILED DESCRIPTION OF THE INVENTION

[0072] In order to make the technical means, creative features, and objectives achieved by the present invention easy to understand, the following further describes the technical solution of the invention in conjunction with one of the embodiments and specific implementation manners of a flexible processing box-type leaf storage logistics control method and system given by the present invention.

[0073] As Figure 1-7 shown, the specific embodiments given for the invention are as follows:

[0074] A box-type leaf storage logistics system for flexible processing is mainly divided into a packing and warehousing area, a storage area, and an outbound box-turning area according to functional areas. The box storage container is a wooden box. Packing and warehousing area: The tobacco leaves are boxed through the packing station 5 and then transported to the storage area; Storage area: According to the roadway equal division strategy, the cigarette boxes of one module are placed disorderly for storage and aging; Outbound box-turning area: The cigarette boxes of different modules are unloaded from the storage area according to the quantity and then emptied. The system includes a shelf 1, a stacker 2, a cigarette box turnover conveyor line 3, a double-station straight shuttle car 4, a packing station 5, a manual processing area 6, an empty box buffer area 7, a reciprocating vertical elevator 8, a box-turning and emptying robot 9, a warehousing feeder elevator 10, a single-station straight shuttle car 11, a circular sorting area 12, a full-box outbound buffer area 13, and a control system. The packing station 5 is connected to the empty box buffer area 7 and the stacker 2 through the cigarette box turnover conveyor line 3. The stacker 2 is connected to the full-box outbound buffer area 13 through the cigarette box turnover conveyor line 3. The full-box outbound buffer area 13 is connected to the box-turning, emptying, and cleaning robot through the reciprocating vertical elevator 8. The box-turning, emptying, and cleaning robot is connected to the warehousing feeder elevator 10. The double-station straight shuttle car 4 is connected to the stacker 2, the packing station 5, the empty box buffer area 7, the reciprocating vertical elevator 8, and the manual processing area 6 through the cigarette box turnover conveyor line 3. The single-station straight shuttle car 11 is connected to the stacker 2 and the full-box outbound buffer area 13 through the cigarette box turnover conveyor line 3. The system is configured with two packing stations 5 in total, which can pack tobacco leaves and flakes simultaneously and separately. The packing station 5 is connected to the upstream feeding section through a belt conveyor, and the tobacco leaves enter the packing station 5 through the belt conveyor; The packing station 5 is connected to the empty box buffer area 7 and the stacker 2 through the cigarette box turnover conveyor line 3. The empty cigarette boxes in the empty box buffer area 7 are transported into the packing station 5 for boxing of tobacco leaves and then transported to the stacker 2 for storage in the warehouse.

[0075] The system is configured with four stackers 2 in total. The load-carrying and receiving of the stacker 2 are configured in two layers, which are located on the ground and the steel platform respectively; Each stacker 2 is provided with a receiving port and a discharging port on the first floor for the outbound and inbound of empty boxes and full boxes; Each stacker 2 is provided with two discharging ports on the second-floor steel platform for the outbound of full boxes. The setting of the two discharging ports can support two disorderly outbound modes of full-formula outbound and grouped-formula outbound simultaneously, as well as the flower-arrangement sorting outbound mode.

[0076] The stacker 2 is connected to the outbound buffer area 13 of full cases through the cigarette case turnover and conveying line 3. The single-station straight shuttle car 11 is connected to the stacker 2 and the outbound buffer area 13 of full cases through the cigarette case turnover and conveying line 3. The outbound buffer area 13 of full cases is divided into three outbound channels. Any one of the outbound channels is connected to the case turning, dumping and cleaning robot through the reciprocating vertical elevator 8. The case turning, dumping and cleaning robot is connected to the storage type feeding and lifting machine 10. The storage type feeding and lifting machine 10 is connected to the downstream process section through the belt conveyor.

[0077] The outbound case turning area is configured in the order of the stacker 2, the single-station straight shuttle car 11, and the outbound buffer area 13 of full cases. An annular sorting area 12 is configured at the entrance of the outbound buffer area 13 of full cases. After the annular sorting area 12, it is divided into three outbound channels. An RFID reader is configured on each channel to identify whether the information of the entering cigarette case matches the database information.

[0078] The materials in the cigarette case are poured into the storage type feeding and lifting machine 10 through the case turning and dumping robot 9. The storage type feeding and lifting machine 10 can store the tobacco leaves of a certain number of cigarette cases. The materials in different cigarette cases can be mixed evenly in the storage type feeder and then enter the downstream process section.

[0079] RFID radio frequency identification readers are installed on the packing station 5, the manual processing area 6, the case turning, dumping and cleaning robot 9, and the outbound buffer area 13 of full cases. An RFID electronic tag is installed at the bottom of the cigarette case. The RFID reader is used to read the cigarette case information in the RFID electronic tag to identify whether the RFID information is consistent with the information in the database.

[0080] The reciprocating vertical elevator 8 has a double-inlet and single-outlet structure, that is, the cigarette case can be input from the inlet on the first or second floor and output from the outlet on the first floor.

[0081] The double-station straight shuttle is connected to the stacker 2, the packing station 5, the empty case buffer area 7, the reciprocating vertical elevator 8, and the manual processing area 6 through the cigarette case turnover and conveying line 3, and can realize: supplying empty cigarette cases to the empty case buffer area 7 and returning the excess empty cigarette cases to the warehouse; transporting the defective cigarette cases to the manual processing area 6 for manual processing and then returning them to the warehouse; transporting the full cases in the warehouse to the first-floor inlet of the reciprocating vertical elevator 8 for emergency replenishment; returning the entire batch of cigarette cases before being transported to the case turning, dumping and cleaning robot 9 to the warehouse.

[0082] The packing station 5 uses a dynamic packing mode to pack materials such as tobacco leaves or flakes into cases. This packing mode changes the original packing mode with a fixed weight per case to a packing mode that predicts the weight per case before packing and calculates the weight per case in real time during the packing process, making the weight per case of each module batch uniform and avoiding the generation of tail cases.

[0083] Such as Figure 4As shown in the figure, the system uses a dynamic packing method for packing and warehousing, and the steps are as follows:

[0084] Step 1: Before packing starts, establish a total packing quantity prediction model through the packing data of historical batches to predict the optimal solution of the total packing quantity. Before packing starts, establish a total packing quantity prediction model through the packing data of historical batches to predict the optimal solution of the total packing quantity. The detailed steps are as follows: According to historical data, the material weight, incoming cabinet moisture, environmental temperature and humidity, leaf storage time in the upstream process storage cabinet are related to the final total packing quantity. The decision tree regression algorithm is used to divide the sample space into multiple sub-spaces, and each sub-space corresponds to a decision tree node. At each node, the algorithm selects an optimal partitioning variable and partitioning point to divide the samples into two subsets, and then recursively partitions each subset until a preset stopping condition is reached. When predicting, the algorithm starts from the root node along the decision tree and continuously traverses downward according to the characteristic values of the test samples, and finally reaches a leaf node, and the value corresponding to this leaf node is the predicted value. After the historical data collection is completed, use the training data to construct a decision tree model, determine the optimal partitioning variable and partitioning point of each node, find the most suitable decision tree parameters, prune the decision tree to avoid overfitting, use the test data to verify the decision tree, predict the results, evaluate the performance of the model, use the trained model to make predictions and interpretations, and make predictions on new data to finally obtain the predicted optimal solution of the total packing quantity.

[0085] Step 2: During the packing process, obtain the packed weight and the number of items already packed in real time, and call the dynamic packing algorithm to dynamically correct the expected packing value and the total number of packages for each box. The dynamic packing mode is divided into two types: Method 1: Both the packing weight and the number of packing boxes are dynamic values; Method 2: The packing weight is a dynamic value and the number of packing boxes is a fixed value. Before packing starts, it is necessary to determine which packing method to use.

[0086] Using Method 1, both the packing weight and the number of packing boxes are dynamic values. According to the quotient of the predicted total packing quantity (or remaining total quantity) M (KG) in Step 1 and the set central packing weight value N (KG), obtain the preset value i of the number of packing items. After incrementing the preset value i of the number of packing items by 1 to get j, that is, i + 1, the following can be obtained:

[0087] i = TRUNC(M / N), j = i + 1,

[0088] For the predicted total packing quantity (or remaining total quantity) M (KG) and the two predicted values i and j of the number of packing items, obtain the two quotient values to obtain two preset average packing weight values N 1 (KG), N 2 (KG), and the following can be obtained:

[0089] N 1= [ROUND*(M / i*100.0)] / 100, N 2 = [ROUND*(M / j*100.0)] / 100, and preset two average packing weight values N 1 (KG), N 2 Compare the absolute values of the differences between (KG) and the packing weight center value N (KG), that is:

[0090] K 1 = ABS(N - N 1 ), K 2 = ABS(N - N 2 ),

[0091] If K 2 is greater than K 1 , the preset packing quantity i is closer to the center value of the packing weight, then use the average packing weight preset value N 1 for packing, and the final estimated packing quantity is i; conversely, if K 1 is greater than K 2 , the preset packing quantity i is closer to the center value of the packing weight, then the packing station uses the average packing weight preset value N 2 for packing, and the final estimated packing quantity is j (j is i + 1). If the actual packing value is X (kg), then the remaining total packing quantity is M = M – X. After packing is completed, substitute M and N into the above formula and enter a loop until the total packing quantity is 0.

[0092] Adopt Method 2, where the weight is a dynamic value and the number of packing boxes is a fixed value. Obtain the average packing weight preset value N (KG) according to the quotient value obtained from the predicted total packing quantity (or remaining total quantity) M (KG) and the number of packing boxes Y predicted in Step 1. The packing station 5 packs according to this value. After completing the packing of one box of materials, based on the cumulative packed weight X (KG) of the packing station 5, obtain the total weight M (KG) of the remaining materials to be packed. Using this weight and the remaining number of packing boxes Y, obtain the average packing weight N (KG) of the remaining cigarette boxes. By analogy through the iterative algorithm, if the cumulative packing quantity of the same batch of cigarette boxes is positively biased in the previous packing, then adjust the average packing weight of each box through calculation to make the subsequent packing weight negatively biased, and vice versa. Dynamically correct the expected packing value and the total number of packing boxes for each box to obtain

[0093] N = [ROUND*(M / Y*100.0)] / 100,

[0094] If the actual packing value is X (kg), then the remaining total packing quantity is M = M – X. After packing is completed, substitute M and N into the above formula and enter a loop until the total packing quantity is 0.

[0095] Step 3: After the packing is completed, determine whether the data packed this time is noise data, that is, data that inaccurately describes the scenario. If it is noise data, clean the record of this data. Determine whether the packing time is much greater than the set packing time according to the packing time. If it is much greater than the packing time, which is set to 20 minutes in the system, the system determines that a failure has occurred during the packing process, cleans the data, and does not record it in the packing history data. Determine according to the number of shutdowns. If a shutdown occurs during the packing process, the system determines that a failure has occurred during the packing process and cleans the data. Determine whether the packing parameters conform to the set parameters according to the packing parameters. If there is a large deviation in the packing parameters, the system determines that the packing parameter setting is incorrect and cleans the data.

[0096] As Figure 5 shown, the system includes an outbound method, and its steps are as follows:

[0097] Step 1: According to the outbound plan, obtain the formula information in the plan;

[0098] Step 2: According to the formula information, obtain the outbound method in the formula: If it is the full formula outbound mode, call the full formula outbound algorithm and perform outbound according to the "first in, first out" principle;

[0099] If it is the grouped formula outbound mode, call the full formula outbound algorithm and perform outbound according to the "first in, first out" principle according to the quantity of each module;

[0100] If it is the flower arrangement sorting outbound mode, call the flower arrangement sorting algorithm to sort all the cigarette boxes of all modules that need to be outbound, and perform outbound according to the sorted order;

[0101] Step 3: After forming the outbound cigarette box sequence according to the outbound plan and the outbound method, if it is one of the two unordered outbound modes of full formula outbound or grouped formula outbound, the stacker 2 transports the cigarette boxes to the unloading platform of the stacker 2, and the cigarette boxes are transported from the cigarette box transfer conveyor line 3 to the outbound real box buffer area 13;

[0102] If it is the flower arrangement sorting outbound mode, the stacker 2 transports the cigarette boxes to the unloading platform of the stacker 2 according to the sorting order. After the cigarette boxes are transported to the receiving platform of the single-station straight shuttle car 11, they wait. The single-station straight shuttle car 11 loads and transports the goods to the unloading platform according to the sorting order, and the cigarette boxes are transported from the cigarette box transfer conveyor line 3 to the outbound real box buffer area 13 in sequence;

[0103] Step 4: After the cigarette boxes enter the real box buffer area, they enter the sorted and allocated real box buffer channel, and reconfirm the order of the batches that need to be sorted. If the order is incorrect, it can be adjusted in the circular sorting area 12;

[0104] Step 5: The RFID reader reads the RFID electronic tag at the bottom of the cigarette box;

[0105] Step 6: After the outbound is completed, wait in the outbound full-case buffer area 13 and then enter the case-tilting and discharging area. The materials in the cigarette cases are discharged into the storage type feeding elevator 10 by the case-tilting and discharging cleaning robot. After the case-tilting and discharging is completed, the empty cigarette cases are placed into the case return path and enter the rack 1 or the empty-case buffer area 7.

[0106] Step 7: The storage type feeding elevator 10 is connected to the downstream section through a belt conveyor. After the materials are fully mixed evenly in the storage type feeding elevator 10, they enter the downstream section.

[0107] The three channels in the full-case buffer area can perform outbound and case-tilting and discharging operations simultaneously, realizing an outbound mode that can have both full formula and grouped formula. On the basis of the grouped formula outbound mode, the flower-arranging sorting outbound mode is further realized.

[0108] When performing grouped formula outbound, the formula module performs outbound in the order required by the formula, and the formula order adopts the flower-arranging outbound mode of evenly distributing and sorting by the total number of cases.

[0109] As Figure 6 shown, the flower-arranging sorting outbound mode adopts the flower-arranging outbound algorithm, and the steps are as follows:

[0110] Step 1: Assume that there are a total of N types of modules for outbound [when N ∈ (2, +∞), it is the leaf group formula outbound in sequence]. The outbound modules are (X 1 , X 2 , …, X N-1 , X N ), then the outbound quantity of each module is (M 1 , M 2 , …, M N-1 , M N ).

[0111] Traverse the quantity of each module. When M Y = 1, the module number and quantity are pushed onto the stack for storage. [M Y ∈ (M 1 , M 2 , …, M N-1 , M N )]

[0112] Step 2: Assume that the modules are divided into K groups, K ∈ (2, MIN(M 1 , M 2 , …, M N-1 , M N )) [when K = 1, it is full formula outbound]

[0113] K traverses (2, MIN(M 1 , M 2 , …, M N-1 , M N ))

[0114] When K = 2, M 1 %2 = L 21 , M 2 %2 = L 22 , …, M N %2 = L 2N , U 2 = MAX(L 21 , L 22 , …, L 2N );

[0115] When K = 3, M 1 %3 = L 31 , M 2 %3 = L 32 , …, M N %3 = L 3N , U 3 = MAX(L 31 , L 32 , …, L 3N );

[0116] …………

[0117] Then, the general formula is:

[0118] M 1 %K = L K1 , M 2 %K = L K2 , …, M N %K = L KN , U K = MAX(L K1 , L K2 , …, L KN );

[0119] Step 3: Let J. If U J = MIN(U 2 , U 3 , …, U n-1 , U n ), then the optimal solution of K is J,

[0120] If U O = U P = … = U q

(U O , U P , …, U q ) ∈ (U 2 , U 3 , …, U n-1 , U n )

[0121] Divide the number of outbound cigarette boxes into J groups.

[0122] The number of groups in the optimal solution is (H J1 , H J2 , …, H JN-1 , H JN ), with a modulus of (L J1 , L J2 , …, L JN-1 , L JN ). Then:

[0123] H J1 = FLOOR(M 1 / J), L J1 = M 1 % J;

[0124] H J2 = FLOOR(M 2 / J), L J2 = M 2 % J;

[0125] ……

[0126] H JN = FLOOR(M N / J), L JN = M N J;

[0127] Then,

[0128] If the outbound module is (X 1 , X 2 , …, X N-1 , X N ), then the outbound quantity of each module in each group is (H J1 , H J2 , …, wH JN-1 , H JN ), and the remaining single boxes that cannot be grouped are (L J1 , L J2 , …, L JN-1 , L JN ). Traverse (L J1 , L J2 , …, L JN-1 , L JN ). When L JY = 1, the module number and quantity are pushed onto the stack for storage.

L JY ∈ (L J1 , L J2 , …, L JN-1 , L JN )

[0129] Group each group and repeat the iterative steps 1 to 3 until each group is a single cigarette box.

[0130] The following uses an example to illustrate this algorithm. There are 5 boxes of Module A, 10 of Module B, and 10 of Module C, with a total batch of 25 boxes. The specific order of box removal is Figure 7 as shown.

[0131] Let M A = 5, M B = 10, M C = 10, and the outgoing stock is

[0132] Also, assume that the outgoing quantity is divided into K groups, where K ∈ (2,..., 5). Traverse K, and

[0133] When K = 2, M A % K = 5 % 2 = 1, M B % K = 10 % 2 = 0, M C % K = 10 % 2 = 0, then U 2 = MAX(1, 0, 0) = 1;

[0134] When K = 3, M A % K = 5 % 3 = 2, M B % K = 10 % 3 = 1, M C % K = 10 % 3 = 1, then U 3 = MAX(2, 1, 1) = 2;

[0135] When K = 4, M A % K = 5 % 4 = 1, M B % K = 10 % 4 = 2, M C % K = 10 % 4 = 2, then U 4 = MAX(1, 2, 2) = 2;

[0136] When K = 5, M A % K = 5 % 5 = 0, M B % K = 10 % 5 = 0, M C % K = 10 % 5 = 0, then U 5 = MAX(0, 0, 0) = 0;

[0137] Then,

[0138] MIN(U 2 , U 3 , U 4 , U 5 ) = MIN(1, 2, 2, 0) = 0. Let the optimal solution of K be J, and J = 5. That is, it is divided into 5 groups in total. Let the number of cigarette boxes of each group of modules be (H 5A , H 5B , H 5C ), and the modulus be (L 5A , L 5B,L 5C ) Then,

[0139] H 5A = FLOOR(M A / J) = FLOOR(5 / 5) = 1;

[0140] H 5B = FLOOR(M B / J) = FLOOR(10 / 5) = 2;

[0141] H 5C = FLOOR(M c / J) = FLOOR(10 / 5) = 2;

[0142] L 5A = M A % J = 5 % 5 = 0;

[0143] L 5B = M B % J = 10 % 5 = 0;

[0144] L 5C = M B % J = 10 % 5 = 0;

[0145] Because L 5A = L 5B = L 5C = 0, there is no single remaining box after grouping.

[0146] In summary, the modules are divided into 5 groups. Each group has 1 box of module A, 2 boxes of module B, 2 boxes of module C, and module A is a single box. That is

[0147] (A, B * 2, C * 2) * 5, ①

[0148] Among them, module A has 1 box and cannot be further subdivided, so it is excluded from the grouping. The new grouping is [A * 1, (B * 2, C * 2)] * 5. Then, the grouping algorithm is iteratively called within each group for further grouping. That is, module B has 2 boxes and module C has 2 boxes, totaling 4 boxes.

[0149] Let M B = 2, M C = 2,

[0150] Also, let the total outbound quantity be divided into K groups, where K belongs to 2. Traverse K, and

[0151] When K = 2, M B % K = 2 % 2 = 0, M C % K = 2 % 2 = 0, then U 2 = MAX(0, 0) = 0;

[0152] Let the optimal solution of K be J, and J = K = 2. That is, it is divided into 2 groups in total. Let the number of cigarette boxes of the modules in each group be (H 2B ,H 2C ). Then,

[0153] H 2B = FLOOR(M B / J) = FLOOR(2 / 2) = 1;

[0154] H 2C = FLOOR(M C / J) = FLOOR(2 / 2) = 1;

[0155] L 2B = M B % J = 10 % 5 = 0;

[0156] L 2C = M B % J = 10 % 5 = 0;

[0157] Because L 2B = L 2C = 0, there are no separate remaining boxes after grouping.

[0158] To sum up, each module is divided into 2 groups. Each group has 1 box of B module and 1 box of C module, and both B and C modules are separate boxes. That is

[0159] (B,C)*2, ②

[0160] Substitute ② into ①, and it becomes

[0161] [A,(B,C)*2]*5, ③

[0162] Expand ③, and it becomes

[0163] [A,(B,C)*2],[A,(B,C)*2],[A,(B,C)*2],[A,(B,C)*2],[A,(B,C)*2]

[0164] = [A,(B,C),(B,C)],[A,(B,C),(B,C)],[A,(B,C),(B,C)],[A,(B,C),(B,C)],[A,(B,C),(B,C)]

[0165] = A,B,C,B,C,A,B,C,B,C,A,B,C,B,C,A,B,C,B,C,A,B,C,B,C ④

[0166] ④ is what we get after sorting.

[0167] The system is equipped with complete anti-error and anti-abnormality devices and programs, as detailed below:

[0168] When packing and storing in the warehouse, the empty cigarette boxes enter the packing station 5. The RFID reader reads the RFID electronic tag at the bottom of the empty cigarette box. After packing is completed, the RFID reader writes the information related to the materials in the cigarette box into the RFID electronic tag to ensure that the information recorded in the system database, the information recorded in the RFID electronic tag of the cigarette box, and the materials in the cigarette box are consistent.

[0169] When leaving the warehouse, the full boxes enter the full-box outbound buffer area 13. The RFID reader reads the RFID electronic tag at the bottom of the empty cigarette box. The information carried in the RFID electronic tag of the cigarette box is the same as the information recorded in the system database, and the cigarette box enters the full-box outbound buffer area 13; if the information recorded in the RFID electronic tag of the cigarette box is different from the information recorded in the system database, the system alarms, and both the RFID electronic tag of the cigarette box and the system database mark it as a non-conforming cigarette box. This cigarette box still enters the full-box outbound buffer area 13, but no box turning and material pouring will be performed subsequently, and the system will replenish the materials.

[0170] When turning the box and pouring the material, the qualified cigarette boxes pour the materials in the cigarette box into the storage type feeder elevator 10 through the box-turning and material-pouring cleaning robot. After the box turning and material pouring are completed, the empty cigarette boxes are placed into the return box path and enter the shelf 1 or the empty box buffer area 7; the box-turning and material-pouring robot 9 directly places the non-conforming cigarette boxes onto the return box path and then transports them to the manual processing area 6.

[0171] There are two ways to replenish materials. When the outbound modes are the two unordered outbound modes of full formula outbound and grouped formula outbound, replenishment is carried out at the end of the batch; when the outbound mode is the flower arrangement sorting outbound mode, an emergency replenishment mode is adopted; after the cigarette boxes for emergency replenishment leave the warehouse, they are transported to the input port on the first floor of the reciprocating vertical elevator 8 through the double-station straight shuttle car 4. When the non-conforming cigarette boxes enter the reciprocating vertical elevator 8 from the input port on the second floor and descend to the first floor, the qualified cigarette boxes for replenishment enter the reciprocating vertical elevator 8 from the input port on the first floor and follow behind the non-conforming cigarette boxes.

[0172] The present invention discloses a flexible processing box-type leaf storage logistics system and its control method, belonging to the logistics field. The logistics system includes: using the packing station 5 to pack and store in the warehouse in a dynamic packing mode; the outbound box turning adopts a triple sorting function for equipment configuration to achieve the unordered outbound modes of both full formula and grouped formula at the same time, and the flower arrangement sorting outbound mode; the system has an anti-error and anti-abnormality function.

[0173] Based on the traditional quantitative packing mode, the present invention is improved to a dynamic packing mode, which solves the "tail box" problem in the packing process, eliminates manual intervention in the "packing quantity", improves the utilization rate of cigarette boxes and provides guarantee for subsequent processes; organizes production in multiple outbound modes according to the production plan, improves the flexibility of production organization, and fully meets the needs of cigarette production for "group processing, homogenization processing, and informatization processing"; the complete equipment and procedures for preventing errors and abnormalities ensure the stability of system operation and avoid misbranding and serial number mixing in cigarette production; the present invention provides reliable software and hardware support for box-type logistics, and overall improves the enterprise's production, operation and manufacturing level and the enterprise's informatization management level.

[0174] It should be understood that the above specific embodiments are only used for illustrative explanation or interpretation of the principle of the present invention, and do not constitute a limitation to the present invention. Therefore, any modifications, equivalent replacements, improvements, etc. made without departing from the spirit and scope of the present invention shall be included within the protection scope of the present invention. The appended claims of the present invention are intended to cover all changes and modifications that fall within the scope and boundaries of the appended claims, or equivalent forms of such scope and boundaries.

Claims

1. A flexible processing box-type leaf storage logistics control method, characterized in that: The following steps are involved: S1. Packing, including S1.

1. Before packing begins, a packing total quantity estimation model is established based on packing data of historical batches to predict the optimal solution for the total packing quantity; S1.

2. During the packing process, the weight and quantity of the packed boxes are obtained in real time, and a dynamic packing algorithm is called to dynamically correct the expected packing value and the total number of packed boxes for each box; S1.

3. After packing is completed, it is determined whether the packing data is noise data that does not accurately describe the scene. If it is noise data, the data record is cleaned; S2, storage, storing the packed cigarette boxes; S3, delivery, including step S3.1, according to the delivery plan, obtaining the recipe information in the plan; S3.

2. According to the recipe information, the delivery method in the recipe is obtained, which includes the full recipe delivery mode, the grouped recipe delivery mode and the flower arrangement sorting delivery mode, and the corresponding delivery algorithm is called according to the different delivery modes; S3.

3. After the outbound cigarette box sequence is formed according to the outbound plan and outbound method, the cigarette boxes are transported to the unloading platform through the logistics system and then outbound, and the sorting batch sequence is confirmed before outbound; S3.

4. After outbound delivery, the materials in the cigarette boxes are dumped out by the box turning and material emptying cleaning robot, and the empty cigarette boxes after the box turning and material emptying are recycled; S3.

5. The dumped logistics are mixed evenly and then enter the downstream section.

2. A flexible processing box-type leaf storage logistics control method as claimed in claim 1, characterized in that: The step S1.1, before packing begins, includes the following steps: according to historical data, the material weight of the upstream process storage cabinet, the moisture content of the cabinet, the ambient temperature and humidity, and the leaf storage time are related to the final total packing amount. The decision tree regression algorithm is used to divide the sample space into multiple subspaces, each subspace corresponds to a decision tree node, and at each node, the algorithm selects an optimal partitioning variable and partitioning point to divide the sample into two subsets, and then recursively divides each subset until the preset stop condition is reached. When predicting, the algorithm starts from the root node along the decision tree, and continuously traverses downward according to the characteristic value of the test sample, and finally reaches a leaf node, and the value corresponding to the leaf node is the predicted value; after the historical data is collected, the decision tree model is constructed using the training data, the optimal partitioning variable and partitioning point of each node are determined, the most appropriate decision tree parameters are found, and the decision tree is pruned to avoid overfitting, and the decision tree is verified using test data, the prediction results, the performance of the model is evaluated, and the trained model is used for prediction and interpretation, and new data is predicted to finally obtain the optimal solution for the predicted total packing amount.

3. A flexible processing box-type leaf storage logistics control method as claimed in claim 1, characterized in that: The step S1.2, the packing process includes the following steps: there are two dynamic packing modes, mode 1 is applicable to cigarette boxes containing tobacco leaves, and the packing weight and the number of boxes are dynamic values; mode 2 is applicable to cigarette boxes containing tobacco sheets, and the packing weight is a dynamic value and the number of boxes is a fixed value; before packing begins, the corresponding packing method is determined according to the material contained in the cigarette box; When using method 1, the packing weight and the number of packing boxes are dynamic values. According to the quotient of the total packing quantity or the remaining total quantity M (KG) predicted by S1.1 and the set packing weight center value N (KG), the preset packing quantity value i is obtained. Then the preset packing quantity value is incremented by 1 to obtain j, that is, i+1, so as to obtain: i=TRUNC(M / N),j=i+1, The predicted total packing quantity or the remaining total quantity M (KG) and the two packing quantity prediction values ​​i and j are used to obtain the two average packing weight preset values ​​N1 (KG) and N2 (KG), thereby obtaining: N1=[ROUND*(M / i*100.0)] / 100,N2=[ROUND*(M / j*100.0)] / 100, compare the absolute value of the difference between the two average packing weight preset values ​​N1(KG) and N2(KG) and the packing weight center value N(KG), that is: K1=ABS(N-N1), K2=ABS(N-N2), If K2 is greater than K1, the preset value i of the packing quantity is closer to the central value of the packing weight, then the average packing weight preset value N1 is used for packing, and the final packing quantity is estimated to be i; conversely, if K1 is greater than K2, the preset value i of the packing quantity is closer to the central value of the packing weight, then the packing station packs the goods at the preset value N2 of the average packing weight, and the final packing quantity is estimated to be j (j is i+1). If the actual packing value is X (kg), the remaining total amount of packing is M=M-X. After the packing is completed, M and N are substituted into the above formula to enter the loop until the total amount of packing is 0; When the second method is adopted, the weight is a dynamic value, and the number of boxes is a fixed value. The quotient obtained by the total packing amount or the remaining total packing amount M (KG) predicted in step S1.1 and the number of boxes Y is used to obtain the average packing weight preset value N (KG). The packing station packs according to this value. After completing the packing of a box of materials, the total weight M (KG) of the remaining materials is obtained according to the cumulative packed weight X (KG) of the packing station. The weight and the remaining number of packings Y are used to obtain the average packing weight N (KG) of the remaining cigarette boxes. By analogy through the iterative algorithm, if the cumulative packing amount of the previous cigarette boxes in the same batch is positively biased, then the average packing weight of each box is adjusted by calculation to make the subsequent packing weight negatively biased, and vice versa. The expected packing value of each box and the total number of packings are dynamically corrected to obtain, N=[ROUND*(M / Y*100.0)] / 100, If the actual packing value is X (kg), the remaining total amount of packing is M = M – X. After packing is completed, substitute M and N into the above formula and enter the loop until the total amount of packing is 0.

4. A flexible processing box-type leaf storage logistics control method as claimed in claim 1, characterized in that: The step S1.3, after packing is completed, includes the following steps: judging noise data respectively by packing time, whether the machine is shut down, and packing parameters; setting the packing time to T (T is a system setting parameter) minutes. If the actual packing time is greater than the set packing time, the system determines that a fault occurred during the packing process, performs data cleaning, and does not record the packing history data; if a shutdown occurs during the packing process, the system determines that a fault occurred during the packing process, performs data cleaning; if the packing parameters deviate from the set parameters, the system determines that the packing parameters are set incorrectly, and performs data cleaning.

5. A flexible processing box-type leaf storage logistics control method as claimed in claim 1, characterized in that: The specific steps of the flower arrangement sorting and outbound mode are as follows: a) Assume that there are N modules for outbound delivery [when N∈(2,+∞), it is the leaf group formula order outbound delivery], and the outbound delivery modules are (X1,X2,…,X N-1 ,X N ), then the number of modules shipped out is (M1,M2,…,M N-1 ,M N ), Traverse each module number, when M Y =1, the module number and quantity are pushed into the stack for storage; [M Y ∈(M1,M2,…,M N-1 ,M N )】 b) Assume that each module is divided into K groups, K∈(2,MIN(M1,M2,…,M N-1 ,M N ))【When K=1, the full formula is shipped】 K traverses (2, MIN (M1, M2, ..., M N-1 ,M N )) When K = 2, M1%2 = L 21 ,M2%2=L 22 ,…,M N %2=L 2N ,U2=MAX(L 21 ,L 22 ,…,L 2N ); When K = 3, M1%3 = L 31 ,M2%3=L 32 ,…,M N %3=L 3N ,U3=MAX(L 31 ,L 32 ,…,L 3N ); ………… Then, the general formula is: M1%K=L K1 ,M2%K=L K2 ,…,M N %K=L KN ,U K =MAX(L K1 ,L K2 ,…,L KN ); c) Let J, if U J =MIN(U2,U3,…,U n-1 ,U n ), then the optimal solution of K is J, If U O = U P = … = U q 【(U O , U P , …, U q ) ∈ (U2, U3, …, U n-1 , U n )】, then J = MAX(O, P, … Q) Divide the number of cigarette boxes leaving the warehouse into J groups; Then the number of optimal solution groups is (H J1 ,H J2 ,…,H JN-1 ,H JN ), modulo (L J1 ,L J2 ,…,L JN-1 ,L JN )but: H J1 =FLOOR(M1 / J),L J1 =M1%J; H J2 =FLOOR(M2 / J),L J2 =M2%J; …… H JN =FLOOR(M N / J),L JN =M N J; but, If the outbound module is (X1, X2,…, X N-1 ,X N ), then the number of modules shipped out of each group is (H J1 ,H J2 ,…,w H JN-1 ,H JN ), the single remaining box that cannot be grouped is (L J1 ,L J2 ,…,L JN-1 ,L JN ), traverse (L J1 ,L J2 ,…,L JN-1 ,L JN ), when L JY =1, the module number and quantity are pushed into the stack for storage; [L JY ∈(L J1 ,L J2 ,…,L JN-1 ,L JN )】; Divide each group into groups and repeat a) to c) until each group is a single smoke box; Pop the data from the stack to get the sorted data.

6. A flexible processing box-type leaf storage logistics control method as claimed in claim 1, characterized in that: In the step S1, when the empty cigarette box enters the packing station (5), the RFID reader reads the RFID electronic tag at the bottom of the empty cigarette box; After the packing is completed, in step S2, when entering the warehouse, the RFID reader writes the relevant information of the materials in the cigarette box into the RFID electronic tag to ensure that the information recorded in the system database, the information recorded in the RFID electronic tag of the cigarette box and the materials in the cigarette box are consistent; a warehouse-out buffer area is set, and in step S3.3, the solid box is transported to the warehouse-out buffer area before the sorting batch sequence is confirmed. When the solid box enters the solid box warehouse-out buffer area (13), the RFID reader reads the RFID electronic tag at the bottom of the empty cigarette box. If the information carried in the RFID electronic tag of the cigarette box is the same as the information recorded in the system database, then it is normal. Go to the next step; the information recorded in the RFID electronic tag of the cigarette box is different from the information recorded in the system database, the system alarms, and the cigarette box is marked as an unqualified cigarette box in both the RFID electronic tag of the cigarette box and the system database. This cigarette box will not be turned over and unloaded subsequently, and the system will refill the material, and then directly transport the unqualified cigarette box to manual processing; there are two ways to refill the material. When the outbound mode is the full formula outbound or group formula outbound, which are two disordered outbound modes, refilling is carried out at the end of the batch; when the outbound mode is the flower arrangement sorting outbound mode, the emergency refilling mode is used to replace and refill specific cigarette boxes.

7. A logistics system used in a flexible processing box-type leaf storage logistics control method, comprising a shelf (1), a stacker (2), a cigarette box turnover conveyor line (3), a double-station straight shuttle vehicle (4), a packing station (5), a manual processing area (6), an empty box buffer area (7), a reciprocating vertical lift (8), a box turning and unloading robot (9), a warehouse-type feeding elevator (10), a single-station straight shuttle vehicle (11), a circular sorting area (12), and a full box outbound buffer area (13); characterized in that: The packing station (5) is connected to the empty box buffer area (7) and the stacker (2) through the cigarette box turnover conveyor line (3); the stacker (2) is connected to the full box outbound buffer area (13) through the cigarette box turnover conveyor line (3); the full box outbound buffer area (13) is connected to the box turning over and unloading cleaning robot through the reciprocating vertical lift (8); the box turning over and unloading cleaning robot is connected to the warehouse-type feeding elevator (10); the double-station straight shuttle car (4) can be connected to the stacker (2), the packing station (5), the empty box buffer area (7), the reciprocating vertical lift (8), and the manual processing area (6) through the cigarette box turnover conveyor line (3); the single-station straight shuttle car (4) is connected to the empty box buffer area (7), the reciprocating vertical lift (8), and the manual processing area (6) through the cigarette box turnover conveyor line (3); The shuttle vehicle (11) can be connected to the stacker (2) and the solid box outbound buffer area (13) through the cigarette box turnover conveyor line (3); a circular sorting area (12) is configured at the entrance of the solid box outbound buffer area (13), and the circular sorting area (12) is divided into three outbound channels, and each channel is configured with an RFID reader to identify whether the information of the incoming cigarette box is consistent with the database information; RFID readers are installed on the packing station (5), the manual processing area (6), the box turning and unloading robot (9), and the solid box outbound buffer area (13), and an RFID electronic tag is installed at the bottom of the cigarette box; the RFID reader is used to read the cigarette box information in the RFID electronic tag to distinguish whether the RFID information is consistent with the information in the database.

8. The logistics system used in the flexible processing box-type leaf storage logistics control method as claimed in claim 7 is characterized in that The system is equipped with two packing stations (5) which can pack tobacco leaves and slices simultaneously and separately. The packing station (5) is connected to the upstream feeding section via a belt conveyor, and tobacco leaves enter the packing station (5) via the belt conveyor; the packing station (5) is connected to the empty box buffer area (7) and the stacker (2) via the cigarette box turnover conveyor line (3), and the empty cigarette boxes in the empty box buffer area (7) are transported to the packing station (5) for packing of tobacco leaves. After loading, the goods can be transported to the stacker (2) and then stored in the warehouse; the system is equipped with four stackers (2), and the stackers (2) are configured in two layers for loading and receiving goods, which are respectively located on the ground and the steel platform; each stacker (2) is provided with a receiving port and a discharge port on the first layer, which are used for empty boxes and full boxes to be taken out of the warehouse and put into the warehouse; each stacker (2) is provided with two discharge ports on the second layer steel platform, which are used for the full boxes to be taken out of the warehouse, and the setting of the two discharge ports can support different outbound modes at the same time.

9. The logistics system used in the flexible processing box-type leaf storage logistics control method as claimed in claim 8, characterized in that The stacker (2) is connected to the solid box outbound buffer zone (13) through the cigarette box turnover conveyor line (3), the single-station straight shuttle vehicle (11) is connected to the stacker (2) and the solid box outbound buffer zone (13) through the cigarette box turnover conveyor line (3), the solid box outbound buffer zone (13) is divided into three outbound channels, any of the outbound channels is connected to the box turning over and emptying cleaning robot through a reciprocating vertical lift (8), the box turning over and emptying cleaning robot is connected to a storage-type feeding elevator (10), and the storage-type feeding elevator (10) is connected to a downstream section through a belt conveyor.

10. The logistics system used in the flexible processing box-type leaf storage logistics control method as claimed in claim 9, characterized in that The reciprocating vertical lift (8) is of a double-entry single-exit structure, that is, the cigarette boxes can be input from the entrance of the first floor or the second floor and then output from the exit of the first floor; the double-station straight shuttle vehicle (4) is connected to the stacker (2), the packing station (5), the empty box buffer area (7), the reciprocating vertical lift (8), and the manual processing area (6) through the cigarette box turnover conveyor line (3), so as to realize the supply of empty cigarette boxes to the empty box buffer area (7) and the return of excess empty cigarette boxes to the warehouse; the faulty cigarette boxes are transported to the manual processing area (6) for manual processing and then returned to the warehouse; the solid boxes in the warehouse are transported to the first floor entrance of the reciprocating vertical lift (8) for emergency refilling; the entire batch of cigarette boxes transported to the box turning and unloading robot (9) are uniformly returned to the warehouse.

Images