A steel storage layout method combining EIQ-ABC and FP-growth
By combining EIQ-ABC and FP-growth algorithms to analyze and mine steel outbound data, the problem of unreasonable warehouse layout in the steel logistics park was solved, the steel storage layout was optimized, and the outbound efficiency and operation efficiency were improved.
Patent Information
- Application Number
- CN202111262142.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-10-27
- Publication Date
- 2024-11-01
- Estimated Expiration
- 2041-10-27
AI Technical Summary
Due to unreasonable warehouse settings and yard layout in the steel logistics park, the cargo yard is queuing long queues, affecting the operating efficiency of the logistics park. It is difficult for the existing technology to effectively optimize the storage and outbound large steel.
Combining the EIQ-ABC and FP-growth algorithms, the steel outbound data is analyzed, the steel categories are divided through the ABC classification method, and the steel combinations are mined with FP-growth, the steel storage layout is optimized, and specific stacking solutions are provided.
It improves the outbound efficiency and operation efficiency of the steel logistics park, provides reasonable steel placement and layout through data analysis and algorithm mining, and reduces the flow time of freight yard vehicles.
Smart Images

Figure CN113971542B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of warehouse storage optimization, and in particular to a steel storage layout method combining EIQ-ABC and FP-growth. Background Art
[0002] In recent years, with the substantial increase in steel trading volume and the diversification of steel varieties in trading, the problems of insufficient storage and chaotic layout in steel logistics parks have become increasingly serious. The warehousing and distribution system is an important link in the logistics supply chain that cannot be ignored.
[0003] Steel logistics products are large in size, heavy in weight, and bulky to move, so special methods and technologies are required for storage, handling, distribution, and information processing. In addition, there are many types of steel products, and the stacking rules are also different. These problems are likely to increase the amount of picking operations, resulting in a long circulation time of outbound vehicles in the cargo yard, affecting the operating efficiency of the logistics park.
[0004] However, the problem of long queues of vehicles in the cargo yard is common in steel logistics parks due to unreasonable warehouse layout and yard layout. Therefore, how to scientifically and rationally optimize the warehouse layout of steel logistics parks and promote the efficient operation of various logistics activities in the park has become an urgent problem to be solved.
[0005] The Chinese patent application number is: CN201810930069.2, and its name is: A dynamic cargo location optimization method for plate parts for warehousing. The invention discloses a dynamic cargo location optimization method for plate parts for warehousing in the technical field of dynamic cargo location optimization methods. It mainly considers the two factors of the stored goods' own attributes and turnover rate, sets the optimization goals of the simulation mathematical model, and establishes a multi-objective optimization model. The ultimate goal is to reduce the time for goods to enter and leave the warehouse. According to the analysis of the current status of warehouse management, reasonable constraint variables are set first, and then the final optimization goal is quantified, and finally the optimized storage location design scheme is obtained to improve the efficiency of warehousing operations. Since steel items are heavy and inconvenient to move stacks, this invention, as a dynamic cargo location optimization method for plate parts, is not suitable for the optimization of steel warehousing.
[0006] The Chinese patent application number is: CN201911332599.8, and its name is: Automated warehouse cargo space allocation optimization method based on NSGA-II. The invention first combines cargo information and adopts ABC classification analysis method to classify cargo; secondly, it determines the optimization objectives and constraints of the automated warehouse cargo space allocation method, establishes a constrained multi-objective optimization problem mathematical model, and finally adopts the optimization algorithm based on NSGA-II to solve the model, obtain the optimal Pareto solution set, and allocate optimization weights according to the actual specific conditions of the automated warehouse space to obtain the only non-dominated solution as the optimal solution. However, the invention is suitable for small and medium-sized automated warehouses such as smart express cabinets, smart vending machines, and self-service storage cabinets, and is not suitable for storage optimization of large goods such as steel. Summary of the invention
[0007] The purpose of the present invention is to solve the problem of unreasonable storage layout in steel logistics parks, and to provide a steel storage layout method combining EIQ-ABC and FP-growth. By combining EIQ-ABC and FP-growth analysis methods, the outbound data of the steel logistics park is analyzed from the two indicators of IQ and IK, and combined with the ABC analysis method, it is divided into three categories of ABC according to the importance of shipment of different types of steel. Then, the FP-growth algorithm is used to mine the steel combinations with medium and high shipment frequencies of various types of ABC, and a reference basis is provided for the specific stacking layout of each type of ABC steel, so as to achieve the purpose of optimizing the cargo space of the steel logistics park.
[0008] To achieve the above object, the technical solution of the present invention is: a steel storage layout method combining EIQ-ABC and FP-growth, comprising the following steps:
[0009] First, implement the EIQ-ABC analysis of steel:
[0010] 1) Calculate the IQ values of various steels and arrange them in descending order to obtain the steel IQ set {IQ1, IQ2, ..., IQ n}, IQ represents the order quality of each steel material, and n is the number of steel types;
[0011] 2) According to the IQ set, the mass percentage of the corresponding single steel material in the total mass of the steel materials is calculated in turn to form a mass percentage set, and the mass percentage of each steel material is accumulated in turn to form a mass cumulative percentage set IQ-rate;
[0012] 3) Realize IQ-ABC steel classification: According to the ABC classification principle, the steel corresponding to the IQ-rate interval of [0,60%] is defined as QA, the steel corresponding to the interval of (60%,85%] is defined as QB, and the steel corresponding to the interval of (85%,100%] is defined as QC;
[0013] 4) Count the IK values of various types of steel and arrange them in descending order to obtain the IK set of steel {IK1, IK2, ..., IK n}, IK represents the number of orders for each item;
[0014] 5) According to the sorted IK set, calculate the percentage of the shipment frequency of the corresponding single steel to the total shipment frequency of all steels in turn to form a frequency percentage set, and add up the frequency percentages of each steel in turn to form a frequency cumulative percentage set IK-rate;
[0015] 6) Implement IK-ABC steel classification: According to the ABC classification principle, the products corresponding to the cumulative shipment frequency percentage interval of [0,60%] are defined as KA, the products corresponding to the interval of (60%,85%] are defined as KB, and the products corresponding to the interval of (85%,100%] are defined as KC;
[0016] 7) Conduct IQ-IK cross analysis: Analyze the IQ and IK of each steel to determine whether the IQ and IK changes of each steel are basically consistent;
[0017] 8) Determine the EIQ-ABC classification results: Based on the IQ-IK cross analysis, the classification results of IQ-ABC and IK-ABC are combined to determine the final ABC classification results of each steel;
[0018] Then use the FP-growth algorithm to mine various types of steel combinations with medium and high frequency shipments in ABC:
[0019] 9) Establish a database of steel shipment records, with the order number as the identifier TID, one sales order number as a transaction T, different types of steel under the same order number as different items, and the set of all transactions as D;
[0020] 10) Set the minimum support min_sup, where min_sup is a threshold. Scan the transaction set D for the first time and count the support count_x of each steel item. The support refers to the number of times the steel appears in D. Filter the transaction set D, record the steel items with count_x ≥ min_sup as frequent items and retain them, otherwise remove them. Then sort the frequent items in transaction T in descending order of support to get D. * ;
[0021] 11) Build FP_tree: First create the root node null, then * Perform the second scan. Every time a transaction T is read in, create a node marked with the corresponding type of steel, and then form a path from the root node null to each steel node until each transaction is mapped to a path in the FP_tree. After reading in all transactions, a complete FP_tree is formed.
[0022] 12) Construct the conditional pattern base of steel items: For the i-th steel, find its node position in FP_tree. The combination of items traversed along the direction of the root node is defined as a conditional pattern base of the steel item. Summarize all paths to obtain the conditional pattern base combination of each steel item, which is recorded as set U i ;
[0023] 13) Construct steel item condition FP_tree: For steel item i, use the conditional pattern base as input data, repeat steps 10)-11), and read in the conditional pattern base U ij , then form a path from the root node null to each steel node until each conditional pattern base is mapped to a path of the conditional FP_tree, and form a complete conditional FP_tree after reading all transactions;
[0024] 14) Mining frequent item sets: Repeat steps 12)-13) to obtain the conditional FP_tree of each steel item, and mine the combination of each conditional FP_tree in order to obtain the final frequent item set P that meets the conditions, and count its frequent item sets in the transaction set Da * The number of times it appears in the list is sorted in descending order, and finally the frequent item set result is obtained;
[0025] 15) Similarly, repeat steps 9)-14) for Class B and Class C steels, and mine out the combinations of steel items that appear frequently in the transaction set for each category, and finally obtain the combination information of each frequently sold steel category, thereby providing a basis for the specific placement of each category of steel.
[0026] Compared with the prior art, the present invention has the following beneficial effects:
[0027] 1) The present invention provides a complete process for optimizing the layout of steel materials in a steel logistics park through a data analysis method, which helps to improve the operational efficiency of the logistics park.
[0028] 2) Through the EIQ-ABC analysis method, steel types are divided into three categories, ABC, according to their importance in warehouse delivery at the macro level. Through FP-growth and algorithms, various types of ABC steel combinations with high frequency of shipment are mined at the micro level. The adjacent stacking of these combinations of steel can improve the warehouse delivery efficiency of the logistics park. BRIEF DESCRIPTION OF THE DRAWINGS
[0029] Figure 1 This is a flow chart of the EIQ-ABC analysis method of the present invention;
[0030] Figure 2 The flowchart of mining frequent itemsets by the FP-growth algorithm of the present invention is as follows;
[0031] Figure 3This is the overall flow chart of storage optimization of the steel logistics park described in the present invention. DETAILED DESCRIPTION
[0032] The technical solution of the present invention is described in detail below in conjunction with the accompanying drawings.
[0033] like Figure 3 As shown, a flow chart of a steel storage layout method combining EIQ-ABC and FP-growth of the present invention is provided. For the convenience of description, the steel information is first encoded: 18 different types of steel are obtained by statistics. Since each type of steel carries information on steel type, diameter and length, for example, [HRB400E, 20, 12000] represents one type of steel. For the convenience of the following description, different types of steel are encoded using different Arabic numerals. Here, 1, 2, ..., 18 represent the 18 types of steel in the logistics park, as shown in Table 1. Scan historical orders to construct a steel database D = [[9, 7, 6, 4, 1, 11, 8], [10, 16, 7, 4, 1, 11, 8], ..., [2, 3, 4]], where each bracket represents a shipping order number, and the different numbers inside represent the different commodities contained in the order number;
[0034] Table 1
[0035] model diameter length coding HRB400E 12 12000 1 HRB400E 22 9000 2 HRB400E 22 12000 3 HRB400E 12 9000 4 HRB400E 25 12000 5 HRB400E 18 12000 6 HRB400E 14 12000 7 HRB400E 25 9000 8 HRB400E 18 9000 9 HRB400E 16 9000 10 HRB400E 16 12000 11 HRB400E 28 9000 12 HRB400E 20 9000 13 HRB400E 20 12000 14 HRB400E 32 9000 15 HRB400E 14 9000 16 HRB400E 32 12000 17 HRB400E 28 12000 18
[0036] First, implement the EIQ-ABC analysis of steel. See the flow chart for details. Figure 1 , the specific steps are as follows:
[0037] 1) Calculate the IQ values of various steels (IQ represents the order quality of each steel) and arrange them in descending order to obtain the steel IQ set {IQ1, IQ4, ..., IQ 15};
[0038] 2) According to the IQ set, the mass percentage of the corresponding single steel material in the mass of the whole steel material is calculated in sequence to form a mass percentage set, and the mass percentage of each steel material is accumulated in sequence to form a mass cumulative percentage set IQ-rate = [12.81%, 25.12%, ..., 100%];
[0039] 3) Realize IQ-ABC steel classification: According to the ABC classification principle, the steel numbers [5,4,3,8,1] with an IQ-rate interval of [0,60%] are classified as QA, the steel numbers [16,7,2,11] corresponding to the interval of (60%,85%] are classified as QB, and the steel numbers [10,18,14,13,6,12,9,17,15] with an IQ-rate interval of (85%,100%] are classified as QC;
[0040] 4) Count the IK values of various types of steel (IK represents the number of orders for each single product) and arrange them in descending order to obtain the IK set of steel {IK1, IK5, ..., IK 15};
[0041] 5) According to the sorted IK set, calculate the percentage of the shipment frequency of the corresponding single steel to the total shipment frequency of all steels in turn to form a frequency percentage set, and add up the frequency percentage of each steel in turn to form a frequency cumulative percentage set IK-rate = [12.34%, 24.37%, ..., 100%];
[0042] 6) Implement IK-ABC steel classification. According to the ABC classification principle, the steel numbers [4,5,3,8,1] with a cumulative shipment frequency percentage in the interval of [0,60%] are classified as KA, the steel numbers [2,16,7,11] with a frequency percentage in the interval of (60%,85%] are classified as KB, and the steel numbers [10,18,13,14,6,12,9,17,15] with a frequency percentage in the interval of (85%,100%] are classified as KC;
[0043] 7) Carry out IQ-IK cross analysis, analyze the IQ and IK of each steel, and find that the changes of IQ and IK of each steel are basically consistent, indicating that compared with varieties with small shipments, steel varieties with large shipments are also ordered more frequently. Therefore, it can be considered that the results obtained by IQ and IK analysis are basically consistent. Finally, the results obtained by IQ-ABC analysis method are selected to determine the ABC classification of various steels;
[0044] 8) Determine the EIQ-ABC classification results: Based on the IQ-IK cross analysis, the classification results of IQ-ABC and IK-ABC are combined to determine the final classification results of each steel: steel number [4,5,3,8,1] is Class A, steel number [2,16,7,11] is Class B, and steel number [10,18,13,14,6,9,12,17,15] is Class C;
[0045] Then, the FP-growth algorithm is used to mine various types of steel combinations with medium and high frequency shipments in ABC. The process is shown in Figure 2 , the specific steps are as follows:
[0046] 9) Filter database D to obtain data information related only to type A steel, denoted as Da.
[0047] Da=[[4,8,1],[4,8,1],...,[4,3]], similarly, classes B and C are denoted as Db and Dc respectively;
[0048] 10) Taking category A as an example, set the minimum support of this category min_sup = 100, scan the transaction set Da for the first time, and count the corresponding support count_x of steels numbered 1, 3, 4, 5, 8 = [302, 341, 404, 423, 363]. Filter the transaction set Da, record the steel items [1, 3, 4, 5, 8] with count_x ≥ 100 as frequent items and retain them, then sort the frequent items in transaction Da in descending order of support to obtain Da * =[[4,8,1],[4,8,1],...,[4,3]];
[0049] 11) Build FP_tree, first create the root node null, then the transaction set Da * Perform a second scan, read the first transaction T1 = [4, 8, 1], create a node marked with the corresponding type of steel, and then form a path with the root node null-4-8-1. Read each transaction set in turn until each transaction is mapped to a path in the FP_tree. After reading all transactions, a complete FP_tree is formed.
[0050] 12) Construct the conditional pattern base of steel items: For steel item 1, find its node position in FP_tree. The combination of items traversed along the direction of the root node is defined as a conditional pattern base of the steel item. Summarize all paths to obtain the conditional pattern base combination of each steel item to form a set U1 = [[4, 8], [3, 5], ..., [3]];
[0051] 13) Construct steel item condition FP_tree: For steel item 1, use the conditional pattern base as input data, repeat steps 10)-11), and read in the first conditional pattern base U 11 = [4, 8], create a node marked with the corresponding type of steel, then form a path with the root node null-4-8, read each conditional pattern base in turn until each conditional pattern base is mapped to a path of the conditional FP_tree, and form a complete conditional FP_tree after reading all transactions;
[0052] 14) Mining frequent item sets: Repeat steps 12)-13) to obtain the conditional FP_tree of each steel item, and mine the combination of each conditional FP_tree in order to obtain the final frequent item set P = [[1,3],...,[4,5]] that meets the conditions, and count its frequent item sets in the transaction set Da * The number of times they appear in the table is sorted in descending order, and the final result is shown in Table 2. Observe the mining results. For example, in category A, steel 5 and steel 8 are sold together the most, so these two steels should be placed next to each other;
[0053] Table 2
[0054] Steel combination Frequency Steel combination Frequency Steel combination Frequency 5,8 223 1,4 162 1,5 125 3,5 174 4,5 150 1,3 107 4,8 163 3,4 142 3,8 103
[0055] 15) Similarly, repeat steps 9)-14) for Class B and Class C steels, and mine out the combinations of steel items that appear frequently in the transaction set for each category, and finally obtain the combination information of each frequently sold steel category, thereby providing a basis for the specific placement of each category of steel.
[0056] The above are preferred embodiments of the present invention. Any changes made according to the technical solution of the present invention, as long as the resulting functions do not exceed the scope of the technical solution of the present invention, belong to the protection scope of the present invention.
Claims
1. A steel storage layout method combining EIQ-ABC and FP-growth, characterized in that: The steps include: First, implement the EIQ-ABC analysis of steel: 1) Calculate the IQ values of various steels and arrange them in descending order to obtain the steel IQ set {IQ1, IQ2, ..., IQ n }, IQ represents the order quality of each steel material, and n is the number of steel types; 2) According to the IQ set, the mass percentage of the corresponding single steel material in the total mass of the steel materials is calculated in turn to form a mass percentage set, and the mass percentage of each steel material is accumulated in turn to form a mass cumulative percentage set IQ-rate; 3) Realize IQ-ABC steel classification: According to the ABC classification principle, the steel corresponding to the IQ-rate interval of [0,60%] is defined as QA, the steel corresponding to the interval of (60%,85%] is defined as QB, and the steel corresponding to the interval of (85%,100%] is defined as QC; 4) Count the IK values of various types of steel and arrange them in descending order to obtain the IK set of steel {IK1, IK2, ..., IK n }, IK represents the number of orders for each item; 5) According to the sorted IK set, calculate the percentage of the shipment frequency of the corresponding single steel to the total shipment frequency of all steels in turn to form a frequency percentage set, and add up the frequency percentages of each steel in turn to form a frequency cumulative percentage set IK-rate; 6) Implement IK-ABC steel classification: According to the ABC classification principle, the products corresponding to the cumulative shipment frequency percentage interval of [0,60%] are defined as KA, the products corresponding to the interval of (60%,85%] are defined as KB, and the products corresponding to the interval of (85%,100%] are defined as KC; 7) Conduct IQ-IK cross analysis: Analyze the IQ and IK of each steel to determine whether the IQ and IK changes of each steel are consistent; 8) Determine the EIQ-ABC classification results: Based on the IQ-IK cross analysis, the classification results of IQ-ABC and IK-ABC are combined to determine the final ABC classification results of each steel; Then use the FP-growth algorithm to mine various types of steel combinations with medium and high frequency shipments in ABC: 9) Establish a database of steel shipment records, with the order number as the identifier TID, one sales order number as a transaction T, different types of steel under the same order number as different items, and the set of all transactions as D; 10) Set the minimum support min_sup, where min_sup is a threshold. Scan the transaction set D for the first time and count the support count_x of each steel item. The support refers to the number of times the steel appears in D. Filter the transaction set D, record the steel items with count_x ≥ min_sup as frequent items and retain them, otherwise remove them. Then sort the frequent items in transaction T in descending order of support to get D. * ; 11) Build FP_tree: First create the root node null, then * Perform the second scan. Every time a transaction T is read in, create a node marked with the corresponding type of steel, and then form a path from the root node null to each steel node until each transaction is mapped to a path in the FP_tree. After reading in all transactions, a complete FP_tree is formed. 12) Construct the conditional pattern base of steel items: For the i-th steel, find its node position in FP_tree. The combination of items traversed along the direction of the root node is defined as a conditional pattern base of the steel item. Summarize all paths to obtain the conditional pattern base combination of each steel item, which is recorded as set U i ; 13) Construct steel item condition FP_tree: For steel item i, use the conditional pattern base as input data, repeat steps 10)-11), and read in the conditional pattern base U ij , then form a path from the root node null to each steel node until each conditional pattern base is mapped to a path of the conditional FP_tree, and form a complete conditional FP_tree after reading all transactions; 14) Mining frequent item sets: Repeat steps 12)-13) to obtain the conditional FP_tree of each steel item, and mine the combination of each conditional FP_tree in order to obtain the final frequent item set P that meets the conditions, and count its frequent item sets in the transaction set Da * The number of times it appears in the list is sorted in descending order, and finally the frequent item set result is obtained; 15) Similarly, repeat steps 9)-14) for Class B and Class C steels, and mine out the combinations of steel items that appear frequently in the transaction set for each category, and finally obtain the combination information of each frequently sold steel category, thereby providing a basis for the specific placement of each category of steel.
