Intelligent warehouse management system based on Internet of Things technology

Through IoT technology and collaborative filtering algorithms, inventory demand is intelligently predicted and warehousing layout is dynamically adjusted, solving the problems of low inventory management efficiency and high operating costs in the existing warehousing management system, and achieving more efficient and flexible inventory management and space utilization .

CN118469445BActive Publication Date: 2024-11-05深圳市光启仓储物流有限公司
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410639899.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-05-22
Publication Date
2024-11-05
Estimated Expiration
2044-05-22

AI Technical Summary

Technical Problem

The existing warehousing management system lacks effective analysis and demand forecasting methods for market trends, resulting in overstock or shortage of inventory, making it difficult to optimize warehousing layout and handling paths, resulting in low inventory management efficiency, high operating costs, and reduced customer satisfaction.

Method used

Using an intelligent warehousing management system based on IoT technology, collects inventory location and environmental data through sensor networks, combines historical sales databases for market trend analysis, uses collaborative filtering algorithm to predict future inventory demand, and dynamically adjusts storage location and handling paths to identify And optimize the bottleneck of warehousing layout.

Benefits of technology

It has achieved accurate prediction of future inventory demand, avoided inventory surplus and shortages, improved inventory turnover, maximized warehouse space utilization, optimized operational efficiency, reduced operating costs, and significantly improved the overall performance and operational efficiency of the warehousing management system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118469445B_ABST
    Figure CN118469445B_ABST
Patent Text Reader

Abstract

The present invention relates to the field of inventory management technology, specifically to an intelligent warehouse management system based on the Internet of Things technology, the system includes: a sensor network data acquisition module based on the Internet of Things sensor network, collects inventory item locations, environmental monitor parameters and operation data, sets nodes including inventory items, storage locations, and handling robots, sets edges including handling paths and storage location relationships, performs summary updates, and generates multidimensional warehouse data. In the present invention, through market trend analysis and collaborative filtering algorithms of historical sales data, it is possible to accurately predict future inventory demand, improve inventory turnover, combine environmental parameters and operation data in the warehouse, model and optimize inventory item storage locations and handling paths, identify and solve warehouse layout bottlenecks, maximize warehouse space utilization, monitor inventory dynamics in real time and dynamically adjust operation processes, reallocate storage locations and handling paths according to changes in inventory item demand, and optimize operational efficiency.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of inventory management, and in particular to an intelligent warehouse management system based on the Internet of Things technology. Background Art

[0002] The field of inventory management technology involves systems and methods for optimizing inventory levels, reducing costs and improving efficiency by using data processing, information technology and systematic methods. This field includes various technical tools and solutions for tracking, managing and optimizing inventory, covering supply chain management, warehouse management systems (WMS), enterprise resource planning (ERP) systems, automatic identification and data capture (AIDC) technologies (such as barcodes and RFID), and Internet of Things (IoT) technologies. These technologies help companies ensure inventory visibility, update inventory status in real time, predict demand, automatically replenish, reduce inventory surpluses and shortages, thereby enhancing operational efficiency and customer satisfaction.

[0003] Among them, warehouse management system is a management system used to monitor and control the inventory in the warehouse. Its main uses include inventory record maintenance, inbound and outbound process management, inventory counting, cargo tracking, and order fulfillment. These systems manage inventory through manual records or using simple information technology tools to ensure the correct storage and timely dispatch of goods. Warehouse management systems help companies maintain inventory accuracy, optimize warehouse space utilization, reduce inventory losses, and support the effective operation of the supply chain.

[0004] Existing technologies have many deficiencies in inventory management. Traditional warehouse management systems lack effective analysis of market trends and demand forecasting methods, which can easily lead to overstocking or shortages, resulting in increased inventory costs and reduced service quality. Insufficient optimization of warehouse layout and transportation routes makes it difficult to identify and solve bottlenecks in the warehouse process, resulting in low warehouse space utilization and low operating efficiency. The lack of dynamic scheduling control means makes it impossible to adjust storage locations and transportation routes in a timely manner according to changes in inventory demand, resulting in insufficient flexibility and responsiveness of inventory management. In actual operations, these problems often lead to a series of negative effects such as inefficient inventory management, high operating costs, and reduced customer satisfaction. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and propose an intelligent warehouse management system based on the Internet of Things technology.

[0006] In order to achieve the above purpose, the present invention adopts the following technical solution: The intelligent warehouse management system based on the Internet of Things technology includes:

[0007] The sensor network data acquisition module is based on the IoT sensor network, collects inventory item locations, environmental monitor parameters and operation data, sets nodes including inventory items, storage locations, and handling robots, sets edges including the handling path and storage location relationship, summarizes and updates the location, status and operation path data, and generates multi-dimensional warehouse data;

[0008] The demand forecasting and analysis module performs market trend analysis of historical sales data, inventory change data, and historical sales data based on the multi-dimensional warehouse data. It uses collaborative filtering algorithms to extract user demand fitting patterns through historical sales data analysis, calculates and predicts future inventory demand, compares future inventory demand with real-time inventory status, and generates future inventory demand forecasts.

[0009] The warehouse layout optimization module models the nodes of the inventory storage location handling robots based on the future inventory demand forecast, combined with the warehouse environment parameters and operation data, and identifies the current layout bottleneck and optimization space by analyzing the relationship between the storage location of the item handling path in the warehouse layout, and generates the optimized warehouse layout;

[0010] Based on the optimized warehouse layout, the dynamic scheduling control module monitors inventory dynamics and operation status in real time, dynamically adjusts operations, analyzes the turnover speed preservation conditions of inventory item demand changes, reallocates inventory locations, dynamically adjusts inventory item storage locations and transportation paths, and generates an inventory dynamic scheduling plan.

[0011] As a further solution of the present invention, the steps of collecting inventory item locations, environmental monitor parameters and operation data are specifically as follows:

[0012] According to the data of the IoT sensor network, the location information of inventory items at different time points is collected, and the parameters and operation data of the environmental monitor are recorded. Refer to the formula:

[0013]

[0014] Generate a comprehensive dataset D;

[0015] Among them, D i Represents a comprehensive dataset of items, S ij represents the position signal strength of item i collected by sensor j, T j represents the timestamp of sensor j, E ij represents the environmental parameters of item i collected by environmental monitor j, H j Indicates the timestamp of the environmental monitor, O ij represents the operation data of item i collected by operation data sensor j, P j represents the timestamp of the operation data sensor j, K represents the position adjustment parameter, L ikrepresents the location information of item i at time point k, M represents the normalization coefficient, and ∈ represents a small value to avoid division by zero;

[0016] The comprehensive data set D is processed to remove abnormal data and calculate the average location, environmental parameters and operation data of each inventory item, referring to the formula:

[0017]

[0018] Get the average comprehensive data set A;

[0019] Among them, A i represents the average comprehensive data of item i, D ik represents the comprehensive data of item i at time point k, w k represents the weight parameter at time point k, R ip represents the comprehensive data residual of item i in dimension p, and N represents the normalization coefficient;

[0020] The cluster analysis was performed using the average comprehensive data set A, referring to the formula:

[0021]

[0022] Determine the location area, environmental monitor parameters and operation path of the inventory items to obtain the comprehensive area R of the items i ;

[0023] Among them, R i Indicates the comprehensive area of ​​the item, A ip represents the average comprehensive data of item i on dimension p, C p represents the location of the cluster center on dimension p, V iq represents the deviation value of item i on dimension q, d represents the number of dimensions of comprehensive data, B i represents the base value of item i;

[0024] By looping through all items, the combined area R of the item i , refer to the formula:

[0025]

[0026] Determine whether there is a situation that exceeds the preset threshold. If so, adjust the position and update the storage position to obtain the inventory item location, environmental monitor parameters and operation data;

[0027] Among them, R′ i represents the comprehensive area of ​​the adjusted items, R i represents the comprehensive area of ​​the item, Threshold represents the preset threshold, Δ represents the position adjustment value, and Z itrepresents the correction factor of item i at time point t, and Z represents the total number of correction factors.

[0028] As a further solution of the present invention, the step of acquiring the multi-dimensional warehouse data is specifically as follows:

[0029] The inventory item location, environmental monitor parameters and operating data are called using the formula:

[0030]

[0031] Obtain the data of sensor nodes and perform preliminary aggregation to generate preliminary aggregation results W of inventory item locations, environmental monitor parameters and operation data;

[0032] Among them, W represents the preliminary summary results, p i Indicates the inventory item location, L i represents the environmental monitor parameter, γ i Represents the weight coefficients, α and β i To adjust the parameters;

[0033] Compare the preliminary summary result W with the historical data to determine whether the environmental monitor parameter exceeds the preset threshold T, using the formula:

[0034]

[0035] Associating the operation data with the inventory item location to generate location status data V;

[0036] Among them, V represents the position status data, H j represents the jth historical data, σ j represents the historical data adjustment coefficient, δ is the smoothing parameter, and T represents the preset threshold;

[0037] According to the position status data V, combined with the inventory item location and transportation path data, the transportation path is adjusted, the transportation path is recalculated, and the summary update result of the position, status and operation path data is generated. The reference formula is as follows:

[0038]

[0039] Among them, U represents the summary update result, P k represents the kth transport path, R k represents the adjusted k-th transport path, φ k represents the path adjustment coefficient;

[0040] Perform multi-dimensional processing of the position state data V, the operation path data and the summary update result U, referring to the formula:

[0041]

[0042] Combining the multidimensional processing coefficient K and the warehouse data integration coefficient D, multidimensional warehouse data M is formed;

[0043] Among them, M represents multidimensional warehouse data, U represents the summary update result, K represents the multidimensional processing coefficient, and D l represents the first warehouse data integration coefficient, η l is the data integration adjustment coefficient, and θ is the smoothing coefficient.

[0044] As a further solution of the present invention, the steps of analyzing the market trend of historical sales data inventory change data are specifically as follows:

[0045] Call the multi-dimensional warehouse data to obtain historical sales data using the formula:

[0046]

[0047] Summarize historical sales data and inventory change data to generate historical sales summary results S;

[0048] Among them, S represents the historical sales summary results, h i represents historical sales data, k i Indicates inventory change data;

[0049] The historical sales summary result S is subjected to time series analysis using the formula:

[0050]

[0051] Extract time series features T;

[0052] Among them, T represents the time series characteristics, S j represents the j-th historical sales summary result, t j Represents time series data, c and d j To adjust the parameters, generate time series features T;

[0053] The market trend analysis of the time series feature T is performed according to the formula:

[0054]

[0055] Combine market influencing factors to generate market trend analysis results M;

[0056] Among them, M represents the market trend analysis results, T k represents the kth time series feature, f k represents the market impact factor, e k and g are smoothing parameters;

[0057] According to the market trend analysis result M, the formula is adopted:

[0058]

[0059] Conduct market trend analysis of historical sales data, inventory change data;

[0060] Among them, A represents the historical sales data, inventory change data, market trend analysis results, n l represents the market data integration coefficient, and v is the data integration adjustment coefficient.

[0061] As a further solution of the present invention, the step of extracting the user demand fitting pattern is specifically:

[0062] From the user's time series features T i Get data, apply collaborative filtering technology, and use the formula:

[0063]

[0064] Identify and aggregate the characteristics of similar user groups to obtain similar user group data U;

[0065] Among them, U is the data of the same user group, w i is the user weight, T i is the user's time series feature, r, s i To adjust the parameters and balance the feature weights of differentiated users;

[0066] Combined with the similar user group data U and purchase behavior history, behavioral pattern analysis is used, using the formula:

[0067]

[0068] Aggregate purchase patterns and create purchase pattern data P;

[0069] Among them, P is the purchase mode data, U j is the group data of the jth user, b j Indicates the corresponding purchase behavior, h j , t is the adjustment coefficient, balancing the impact of purchasing behavior;

[0070] Through the purchase pattern data P, market changes and product trends are analyzed using the formula:

[0071]

[0072] Adjust demand forecasting parameters to form a user demand fitting model F;

[0073] Among them, F is the user demand fitting model, P k is the kth purchase pattern data, dk is the demand adjustment factor, x, y k is the smoothing parameter.

[0074] As a further solution of the present invention, the steps of obtaining the future inventory demand forecast are specifically:

[0075] Get the output from the user demand fitting model F, apply the market forecasting model, and use the formula:

[0076]

[0077] Calculate the inventory demand in the future target time period to form the preliminary future inventory demand R;

[0078] Among them, R is the initial future inventory demand, F i represents the output of the i-th user demand fitting mode, m i is the corresponding market trend impact coefficient, a i , b is the adjustment parameter to match the demand forecast under differentiated market conditions;

[0079] Analyze the difference between the preliminary future inventory demand R and the current inventory status, using the formula:

[0080]

[0081] Evaluate and quantify the inventory gap to form inventory gap data G;

[0082] Among them, G is the inventory gap data, R j is the predicted inventory demand at the jth time point, s j is the actual inventory status at the same time point, c j and d are smoothing parameters;

[0083] Combining the inventory gap data G with economic indicators and supply chain capacity parameters, the formula is adopted:

[0084]

[0085] Evaluate and adjust future inventory plans to form future inventory demand forecast Q;

[0086] Among them, Q is the future inventory demand forecast, G k is the kth inventory gap data, v k is the economic indicator adjustment coefficient, o k , u is the smoothing parameter of supply chain capability.

[0087] As a further solution of the present invention, the step of obtaining the optimized storage layout is specifically as follows:

[0088] According to the future inventory demand forecast Q, the environmental parameters and operation data in the warehouse are collected and the formula is used:

[0089]

[0090] Generate warehouse data model D;

[0091] Among them, D is the warehouse data model, Q i represents the inventory demand forecast for period i, e i is the influence coefficient of environmental parameters, f i is the influence coefficient of the operating data, g is the constant adjustment parameter;

[0092] Using the warehouse data model D, each storage location and handling robot node is modeled, and the relationship between the handling path and the storage location is analyzed using the formula:

[0093]

[0094] Construct path relationship model P;

[0095] Among them, P is the path relationship model, D j represents the jth storage element data extracted from the D model, h j is the influence coefficient of the transport path, k, l j To adjust the parameters;

[0096] By analyzing the relationship model P between the transport path and the storage location, the formula is adopted:

[0097]

[0098] Identify bottlenecks and optimization space in warehouse layout and form layout optimization plan S;

[0099] Among them, S is the layout optimization solution, P j is the output of the relationship model between the jth transport path and the storage location, v represents the transport efficiency under ideal conditions, o j , p is the efficiency adjustment parameter;

[0100] Based on the layout optimization scheme S, the warehouse layout is redesigned and adjusted using the formula:

[0101]

[0102] Form an optimized storage layout L;

[0103] Among them, L is the optimized storage layout, S n represents the nth optimization solution obtained from S, r n is the implementation weight of the suggestion.

[0104] As a further solution of the present invention, the steps of obtaining the dynamic inventory scheduling solution are specifically as follows:

[0105] Real-time monitoring of inventory dynamics and operations in the optimized warehouse layout, using the formula:

[0106]

[0107] Integrate key data and build real-time monitoring data model E;

[0108] Among them, E is the real-time monitoring data model, L i represents the real-time information of the i-th inventory location, o i is the corresponding operation data, p i , q is the adjustment parameter;

[0109] The real-time monitoring data model E is used to analyze the demand changes and turnover speed of inventory items, using the formula:

[0110]

[0111] Identify inventory adjustment requirements N;

[0112] Among them, N is the inventory adjustment demand, M j represents the monitoring data of the jth inventory item, r j is the demand change rate, s, t j is the adjustment factor;

[0113] According to the inventory adjustment demand N, the inventory location is dynamically reallocated using the formula:

[0114]

[0115] Form an inventory position adjustment plan P;

[0116] Among them, P is the inventory position adjustment plan, N k is the demand data of the kth inventory item, v k In order to adjust the coefficient angle, the trigonometric function sin(v k ) simulates the impact of cyclical demand changes, w k , x is the adjustment parameter; based on the inventory location adjustment plan P, combined with the warehouse operation strategy and resource constraints, the formula is adopted:

[0117]

[0118] Generate a dynamic inventory scheduling plan S;

[0119] Among them, S is the inventory dynamic scheduling plan, P nis the output of the nth position adjustment plan, u represents the scheduling optimization goal, y n , z is the efficiency adjustment parameter.

[0120] Compared with the prior art, the advantages and positive effects of the present invention are:

[0121] In the present invention, through market trend analysis of historical sales data and collaborative filtering algorithms, future inventory demand can be accurately predicted, overstocking and shortages can be avoided, and inventory turnover can be improved. In combination with environmental parameters and operating data in the warehouse, modeling and optimization of inventory item storage locations and transportation paths are performed to identify and resolve warehouse layout bottlenecks and maximize warehouse space utilization. Real-time monitoring of inventory dynamics and dynamic adjustment of operating processes are performed to reallocate storage locations and transportation paths according to changes in inventory item demand, optimize operational efficiency, and reduce operating costs. Overall, through these innovative measures, more efficient, accurate, and flexible inventory management is achieved, significantly improving the overall performance and operational benefits of the warehouse management system. BRIEF DESCRIPTION OF THE DRAWINGS

[0122] Figure 1 is a system flow chart of the present invention;

[0123] Figure 2 A flow chart of the steps for collecting inventory item locations, environmental monitor parameters and operating data of the present invention;

[0124] Figure 3 A flowchart of the steps for obtaining multi-dimensional warehouse data of the present invention;

[0125] Figure 4 A flowchart of the steps of analyzing the market trend of historical sales data, inventory change data, and market trends of the present invention;

[0126] Figure 5 A flowchart of the steps of extracting a user demand fitting model according to the present invention;

[0127] Figure 6 A flow chart of the steps for obtaining future inventory demand forecasts of the present invention;

[0128] Figure 7 A flow chart of the steps for obtaining the optimized warehouse layout of the present invention;

[0129] Figure 8 The flowchart is a step chart of obtaining the inventory dynamic scheduling solution of the present invention. DETAILED DESCRIPTION

[0130] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0131] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, "multiple" means two or more, unless otherwise clearly and specifically defined.

[0132] Embodiment 1

[0133] See also Figure 1 , the intelligent warehouse management system based on Internet of Things technology includes:

[0134] The sensor network data acquisition module is based on the IoT sensor network, collects inventory item locations, environmental monitor parameters and operation data, sets nodes including inventory items, storage locations, and handling robots, sets edges including the handling path and storage location relationship, summarizes and updates the location, status and operation path data, and generates multi-dimensional warehouse data;

[0135] The demand forecasting and analysis module conducts market trend analysis of historical sales data, inventory change data, and historical sales data based on multi-dimensional warehouse data. It uses collaborative filtering algorithms to extract user demand fitting patterns through historical sales data analysis, calculates and predicts future inventory demand, compares future inventory demand with real-time inventory status, and generates future inventory demand forecasts.

[0136] The warehouse layout optimization module is based on the forecast of future inventory demand, combined with the environmental parameters and operation data in the warehouse, to model the nodes of the robot handling the storage location of the inventory items. By analyzing the relationship between the storage location of the item handling path in the warehouse layout, the current layout bottleneck and optimization space are identified, and the optimized warehouse layout is generated.

[0137] The dynamic scheduling control module monitors inventory dynamics and operation status in real time based on the optimized warehouse layout, dynamically adjusts operations, and reallocates inventory locations by analyzing the turnover speed preservation conditions of inventory item demand changes. It dynamically adjusts inventory item storage locations and transportation routes to generate a dynamic inventory scheduling plan.

[0138] Multidimensional warehouse data includes location data, environmental data, and operation data. Future inventory demand forecasts include demand quantity, forecast time, and product category. The optimized warehouse layout includes storage areas, transportation routes, and optimization plans. The dynamic inventory scheduling plan includes scheduling plans, location adjustments, and optimized paths.

[0139] See also Figure 2 ,The steps for collecting inventory item locations, environmental monitor parameters and operation data are as follows:

[0140] According to the data of the IoT sensor network, the location information of inventory items at different time points is collected, and the parameters and operation data of the environmental monitor are recorded. Refer to the formula:

[0141]

[0142] Generate a comprehensive dataset D;

[0143] Among them, D i Represents a comprehensive dataset of items, S ij represents the position signal strength of item i collected by sensor j, T j represents the timestamp of sensor j, E ij represents the environmental parameters of item i collected by environmental monitor j, H j Indicates the timestamp of the environmental monitor, O ij represents the operation data of item i collected by operation data sensor j, P j represents the timestamp of the operation data sensor j, K represents the position adjustment parameter, L ik represents the location information of item i at time point k, M represents the normalization coefficient, and ∈ represents a small value to avoid division by zero;

[0144] Process the comprehensive data set D, remove abnormal data and calculate the average location, environmental parameters and operation data of each inventory item, referring to the formula:

[0145]

[0146] Get the average comprehensive data set A;

[0147] Among them, A i represents the average comprehensive data of item i, D ik represents the comprehensive data of item i at time point k, w k represents the weight parameter at time point k, R ip represents the comprehensive data residual of item i in dimension p, and N represents the normalization coefficient;

[0148] Use the average comprehensive data set A for cluster analysis, refer to the formula:

[0149]

[0150] Determine the location area, environmental monitor parameters and operation path of the inventory items to obtain the comprehensive area R of the items i ;

[0151] Among them, R iIndicates the comprehensive area of ​​the item, A ip represents the average comprehensive data of item i on dimension p, C p represents the location of the cluster center on dimension p, V iq represents the deviation value of item i on dimension q, d represents the number of dimensions of comprehensive data, B i represents the base value of item i;

[0152] By looping through the comprehensive area R of all items i , refer to the formula:

[0153]

[0154] Determine whether there is a situation that exceeds the preset threshold. If so, adjust the position and update the storage position to obtain the inventory item location, environmental monitor parameters and operation data;

[0155] Among them, R′ i represents the comprehensive area of ​​the adjusted items, R i represents the initial comprehensive area of ​​the item, Threshold represents the preset threshold, Δ represents the position adjustment value, and Z it represents the correction factor of item i at time point t, and Z represents the total number of correction factors.

[0156] Improved formula:

[0157]

[0158] Explanation and derivation:

[0159] D i represents a comprehensive dataset of item i;

[0160] S ij Represents the location information of item i at time point j;

[0161] T j A timestamp representing the location data;

[0162] E ij represents the environmental parameters of item i collected by environmental monitor j;

[0163] H j Represents the timestamp of environmental data;

[0164] O ij represents the operation data of item i collected by operation data sensor j;

[0165] P j The timestamp representing the operation data sensor j;

[0166] K represents the position adjustment parameter, which is assumed to be 1.5;

[0167] L ik Represents the location information of item i at time point k;

[0168] M represents the standardized coefficient, which is assumed to be 100;

[0169] ∈ represents a small value to avoid division by zero, assumed to be 0.01.

[0170] Calculation process:

[0171] Collect sensor signals, assuming the sensor signal strength is S ij =[5,10,15];

[0172] Record the timestamp, assuming the timestamp is T j =[1,2,3];

[0173] Collect environmental parameters, assuming that the environmental parameters are E ij =[0.1,0.2,0.3];

[0174] Collect operation data, assuming that the operation data is O ij =[10,20,30];

[0175] Record the timestamp, assuming the timestamp is P j =[1,2,3];

[0176] Calculate the position information, assuming the position information is L ik =[50,60,70];

[0177] Substitute into the formula:

[0178]

[0179] Process the comprehensive data set D, remove abnormal data and calculate the average location, environmental parameters and operation data of each inventory item to obtain the average comprehensive data set A;

[0180] Improved formula:

[0181]

[0182] Explanation and derivation:

[0183] A i Indicates the average comprehensive data of the item;

[0184] D ik Represents the comprehensive data of item i at time point k;

[0185] w k Represents the weight parameter at time point k, assuming it is w k=[1,1,1];

[0186] R ip Represents the comprehensive data residual of item i in dimension p, assuming it is R ip =[0.5,0.6,0.7]; N represents the normalization coefficient, which is assumed to be 3;

[0187] Calculation process:

[0188] Bring the comprehensive data of step into D ik =[48,50,52];

[0189] Calculate the weighted average;

[0190] Calculate the residuals of the composite data;

[0191] Substitute into the formula:

[0192]

[0193] Improved formula:

[0194]

[0195] Explanation and derivation:

[0196] R i A comprehensive area representing items;

[0197] A ip Represents the average comprehensive data of item i on dimension p;

[0198] C p Represents the location of the cluster center on dimension p, assuming it is C p =[50,51,52];

[0199] V iq Represents the deviation value of item i on dimension q, assuming it is V iq =[0.2,0.3,0.4];

[0200] B i Indicates the base value of the item, assuming it is 50;

[0201] d represents the number of dimensions of the comprehensive data, which is assumed to be 3;

[0202] Calculation process:

[0203] Substitute the average comprehensive data from step 2 into A ip =[50.6,50.6,50.6];

[0204] Calculate the Euclidean distance from the cluster center;

[0205] Calculate the absolute difference between the deviation value and the reference value;

[0206] Substitute into the formula:

[0207]

[0208] Improved formula:

[0209]

[0210] Explanation and derivation:

[0211] R′ i represents the comprehensive area of ​​the adjusted items;

[0212] R i represents the initial comprehensive area of ​​item i;

[0213] Threshold indicates the preset threshold, which is assumed to be 52;

[0214] Δ represents the position adjustment value, which is assumed to be 1;

[0215] Z it Represents the correction factor of item i at time point t, assuming it is Z it =[0.1,0.2,0.3]; Z represents the total number of correction factors, assuming it is 3;

[0216] Calculation process:

[0217] Bring the comprehensive regional data from step 3 into R i =51.27;

[0218] Determine whether it exceeds the high value;

[0219] If the threshold is exceeded, the adjusted value is calculated;

[0220] Substitute into the formula:

[0221]

[0222] See also Figure 3 ,The specific steps for obtaining multi-dimensional warehouse data are:

[0223] Call inventory item locations, environmental monitor parameters, and operational data using the formula:

[0224]

[0225] Obtain the data of sensor nodes and perform preliminary aggregation to generate preliminary aggregation results W of inventory item locations, environmental monitor parameters and operation data;

[0226] Among them, W represents the preliminary summary results, p iIndicates the inventory item location, L i represents the environmental monitor parameter, γ i Represents the weight coefficients, α and β i To adjust the parameters;

[0227] Compare the preliminary summary result W with the historical data to determine whether the environmental monitor parameter exceeds the preset threshold T, using the formula:

[0228]

[0229] Associating the operation data with the inventory item location to generate location status data V;

[0230] Among them, V represents the position status data, H j represents the jth historical data, σ j represents the historical data adjustment coefficient, δ is the smoothing parameter, and T represents the preset threshold;

[0231] According to the position status data V, combined with the inventory item location and transportation path data, the transportation path is adjusted, the transportation path is recalculated, and the summary update result of the position, status and operation path data is generated. The reference formula is as follows:

[0232]

[0233] Among them, U represents the summary update result, P k represents the kth transport path, R k represents the adjusted k-th transport path, φ k represents the path adjustment coefficient;

[0234] Perform multi-dimensional processing of the position state data V, the operation path data, and the summary update result U, referring to the formula:

[0235]

[0236] Combining the multidimensional processing coefficient K and the warehouse data integration coefficient D, multidimensional warehouse data M is formed;

[0237] Among them, M represents multidimensional warehouse data, U represents the summary update result, K represents the multidimensional processing coefficient, and D l represents the first warehouse data integration coefficient, η l is the data integration adjustment coefficient, and θ is the smoothing coefficient.

[0238] formula:

[0239]

[0240] Parameter details:

[0241] p i: The specific location value of the inventory item location, assumed to be randomly distributed in the range of 10-100.

[0242] L i : Environmental monitor parameters, such as temperature, humidity, etc., assuming the temperature range is 15-25℃.

[0243] γ i : Weight coefficient, reflecting the importance of different data, assumed to be in the range of 0.5-1.5.

[0244] α: Tuning parameter, assumed to be a constant 5.

[0245] β i : Adjust the parameter, assuming the range is 1-10.

[0246] Calculation example:

[0247] Assume there are 3 node data:

[0248] p1=20, L1=18, γ1=1.0, β1=2

[0249] p2=35, L2=22, γ2=1.2, β2=5

[0250] p3=50, L3=19, γ3=0.8, β3=3

[0251] Enter the formula to calculate:

[0252]

[0253] W≈135.85+292.41+268.91

[0254] W≈697.17

[0255] Calculation process:

[0256] Get inventory item locations, environmental monitor parameters, and weight coefficients.

[0257] The parameters α and β will be adjusted i Plug into the formula.

[0258] Calculate the value of each node and sum them up to get the preliminary summary result W.

[0259] formula:

[0260]

[0261] Parameter details:

[0262] W: Preliminary summary results, the value calculated in the previous step.

[0263] H j: Historical data, assuming the summary results of the past three days are 700, 680, and 690. j : Historical data adjustment coefficient, assumed to be in the range of 0.8-1.2.

[0264] δ: smoothing parameter, assumed to be 0.1.

[0265] T: preset threshold, assumed to be 300.

[0266] Calculation example:

[0267] Assume the historical data is:

[0268] H1=700,σ1=1.0

[0269] H2=680,σ2=0.9

[0270] H3=690,σ3=1.1

[0271] Enter the formula to calculate:

[0272]

[0273] V≈0.069

[0274] Calculation process:

[0275] Combine the preliminary summary results W with the historical data H j Make a comparison.

[0276] Adjust the coefficient σ using historical data j and smoothing parameter δ adjusts each difference.

[0277] Sum all the adjusted differences and divide them by the preset threshold T to get the position status data V. Step 3: Calculation of the summary update result U

[0278] formula:

[0279]

[0280] Parameter details:

[0281] V: Position status data, the value calculated in the previous step.

[0282] P k : Transportation path, assuming the path length is 20, 30.

[0283] R k : The adjusted transport path is assumed to have a path length of 18 or 28.

[0284] φ k : Path adjustment coefficient, assumed to be 1.1 and 1.2.

[0285] Calculation example:

[0286] Assume that the general transport path and the adjusted transport path are:

[0287] P1=20,R1=18,φ1=1.1

[0288] P2=30,R2=28,φ2=1.2

[0289] Enter the formula to calculate:

[0290]

[0291] U=|-15.11-21.61|

[0292] U=|-36.72|

[0293] U=36.72

[0294] Calculation process:

[0295] Call position status data V and transport path P k .

[0296] Combined with the adjusted transport path R k and the path adjustment factor φ k Calculate the difference of each path. Sum all the differences and take the absolute value to get the summary update result U.

[0297] formula:

[0298]

[0299] Parameter details:

[0300] U: Summary update result, the value calculated in the previous step.

[0301] K: multidimensional processing coefficient, assumed to be 2.

[0302] D l : Warehouse data integration coefficient, assumed to be 1.5 or 2.0.

[0303] η l : Data integration adjustment coefficient, assumed to be 0.9 and 1.0.

[0304] θ: smoothing coefficient, assumed to be 0.1.

[0305] Calculation example:

[0306] Assume that the warehouse data integration coefficient is:

[0307] D1=1.5,η1=0.9

[0308] D2=2.0,η2=1.0

[0309] Enter the formula to calculate:

[0310]

[0311] M≈4.32

[0312] Calculation process:

[0313] Call summary to update the result U and multi-dimensional processing coefficient K.

[0314] Combined with warehouse data integration coefficient D l and data integration adjustment coefficient η l Perform data integration.

[0315] Sum up all the integrated data and take the cube root to obtain the multidimensional warehouse data M.

[0316] See also Figure 4 The specific steps for market trend analysis of historical sales data, inventory change data are as follows:

[0317] Call multi-dimensional warehouse data to obtain historical sales data using the formula:

[0318]

[0319] Summarize historical sales data and inventory change data to generate historical sales summary results S;

[0320] Among them, S represents the historical sales summary results, h i represents historical sales data, k i Indicates inventory change data;

[0321] Perform time series analysis on the historical sales summary results S using the formula:

[0322]

[0323] Extract time series features T;

[0324] Among them, T represents the time series characteristics, S j represents the j-th historical sales summary result, t j Represents time series data, c and d j To adjust the parameters, generate time series features T;

[0325] Conduct market trend analysis on time series feature T, refer to the formula:

[0326]

[0327] Combine market influencing factors to generate market trend analysis results M;

[0328] Among them, M represents the market trend analysis results, T k represents the kth time series feature, f k represents the market impact factor, e k and g are smoothing parameters;

[0329] According to the market trend analysis result M, the formula is adopted:

[0330]

[0331] Conduct market trend analysis of historical sales data, inventory change data;

[0332] Among them, A represents the historical sales data, inventory change data, market trend analysis results, n l represents the market data integration coefficient, and v is the data integration adjustment coefficient.

[0333] formula:

[0334]

[0335] Parameter details:

[0336] h i : The historical sales volume at the ith time point, assumed to be unit sales volume.

[0337] k i : The inventory change at the ith time point, assumed to be a unit change.

[0338] n: The total number of time points.

[0339] Assumptions of the example:

[0340] Suppose there are sales and inventory change data at three time points:

[0341] h1=100,k1=50

[0342] h2=150,k2=75

[0343] h3=200,k3=100

[0344] Calculation process:

[0345] Multiply the sales and inventory change data at each point in time.

[0346] The results of all time points are accumulated to get S.

[0347] Calculation process:

[0348] S=(100·50)+(150·75)+(200·100)

[0349] S = 5000 + 11250 + 20000

[0350] S=36250

[0351] formula:

[0352]

[0353] Parameter details:

[0354] S j : The historical sales summary result at the jth time point is obtained from step 1.

[0355] t j : The time coefficient at the jth time point, assumed to represent the seasonal influence factor.

[0356] c: constant used to adjust the size of the denominator, assumed to be 1.

[0357] d j : The adjustment parameter at the jth time point, used to reflect the special situation at the time point, assumed to be fixed to 1. Example assumptions:

[0358] t1=0.9,d1=1

[0359] t2=1.0,d2=1

[0360] t3=1.1,d3=1

[0361] Calculation process:

[0362] For each time point S j and t j Perform the product.

[0363] The product results of all time points are accumulated.

[0364] Calculate the square root of the denominator and divide the accumulated result by the square root to get T.

[0365] Calculation process:

[0366]

[0367] T≈76923.77

[0368] formula:

[0369]

[0370] Parameter details:

[0371] T k: The kth time series feature, obtained from step 2.

[0372] f k : The kth market impact factor, assumed to be a fixed value.

[0373] e k : The kth adjustment coefficient, assumed to be fixed.

[0374] g: smoothing parameter, assumed to be fixed.

[0375] Assumptions of the example:

[0376] f k =1.05,e k =0.5,g=0.5

[0377] Assume there is only one set of data T k =76923.77

[0378] Calculation process:

[0379] For each time point T k and f k Perform the product.

[0380] The product results of all time points are accumulated.

[0381] Use e k Calculate the denominator and divide by that value to get M.

[0382] Calculation process:

[0383]

[0384] M=80769.96

[0385] See also Figure 5 ,The steps of extracting the user demand fitting model are as follows:

[0386] From the user's time series features T i Get data, apply collaborative filtering technology, and use the formula:

[0387]

[0388] Identify and aggregate the characteristics of similar user groups to obtain similar user group data U;

[0389] Among them, U is the data of the same user group, w i is the user weight, T i is the user's time series feature, r, s i To adjust the parameters and balance the feature weights of differentiated users;

[0390] Combined with the data of similar user groups U and the purchase behavior history, using behavioral pattern analysis, the formula is adopted:

[0391]

[0392] Aggregate purchase patterns and create purchase pattern data P;

[0393] Among them, P is the purchase mode data, U j is the group data of the jth user, b j Indicates the corresponding purchase behavior, h j , t is the adjustment coefficient, balancing the impact of purchasing behavior;

[0394] Through the purchase pattern data P, analyze the market changes and product trends, using the formula:

[0395]

[0396] Adjust demand forecasting parameters to form a user demand fitting model F;

[0397] Among them, F is the user demand fitting model, P k is the kth purchase pattern data, d k is the demand adjustment factor, x, y k is the smoothing parameter.

[0398] Detailed explanation and derivation process:

[0399] The following assumptions will be used for derivation and calculations, assuming that the parameter values ​​are within reasonable ranges:

[0400] User weight coefficient w i : The assumed value is 1.2, reflecting the user's influence relative to the group.

[0401] Time series features T i : Assume that the purchase behavior index of a user within a certain time window is 30.

[0402] Adjusting parameters r and s i :Assume r = 0.5, s i =0.8, used to adjust the feature weights of different users.

[0403] Purchase behavior j : The number of purchases for each user, assumed to be 5.

[0404] Purchase behavior adjustment coefficient h j and t: Let h j =0.7, t=0.3, adjusting the influence of purchase behavior data.

[0405] Demand adjustment factor d k: Set to 1.5 to adjust forecast demand based on market dynamics.

[0406] Smoothing parameters x and y k :Let x=1, y k =0.9, processing model for smooth transition between different user behaviors.

[0407] First formula: Similar user group data U

[0408]

[0409] Assume there are 3 users and perform the following calculation:

[0410]

[0411] U≈31.58+31.58+31.58

[0412] U≈94.74

[0413] Second formula: Purchase mode data P

[0414]

[0415] Calculate with the same three users:

[0416]

[0417] P = 473.7 + 473.7 + 473.7

[0418] P = 1421.1

[0419] The third formula: User demand fitting model F

[0420]

[0421] Using the previous P value, calculate:

[0422]

[0423] F=1544.23 2

[0424] F≈2385634.37

[0425] Calculation process

[0426] The time series features of each user are weighted and summed, with the weights based on the user influence and the adjustment parameters.

[0427] Aggregate user purchase behaviors and sum them based on purchase frequency and behavior adjustment coefficients to simulate purchase patterns.

[0428] Use purchasing pattern data, combined with demand adjustment factors and smoothing parameters, to predict user demand and enhance the model's responsiveness.

[0429] See also Figure 6 , the specific steps for obtaining future inventory demand forecasts are:

[0430] Get the output from the user demand fitting model F, apply the market forecasting model, and use the formula:

[0431]

[0432] Calculate the inventory demand in the future target time period to form the preliminary future inventory demand R;

[0433] Among them, R is the initial future inventory demand, F i represents the output of the i-th user demand fitting mode, m i is the corresponding market trend impact coefficient, a i , b is the adjustment parameter to match the demand forecast under differentiated market conditions;

[0434] Analyze the difference between the preliminary future inventory demand R and the current inventory status, using the formula:

[0435]

[0436] Evaluate and quantify the inventory gap to form inventory gap data G;

[0437] Among them, G is the inventory gap data, R j is the predicted inventory demand at the jth time point, s j is the actual inventory status at the same time point, c j and d are smoothing parameters;

[0438] Combining the inventory gap data G with economic indicators and supply chain capacity parameters, the formula is used:

[0439]

[0440] Evaluate and adjust future inventory plans to form future inventory demand forecast Q;

[0441] Among them, Q is the future inventory demand forecast, G k is the kth inventory gap data, v k is the economic indicator adjustment coefficient, o k , u is the smoothing parameter of supply chain capability.

[0442] Formula 1: Preliminary future inventory demand R

[0443]

[0444] Parameter details:

[0445] F i : User demand fitting model output, assumed to be the predicted purchase intention strength, ranging from 0-100.

[0446] m i : Market trend impact coefficient, reflecting the impact of seasonal fluctuations or market activities, is assumed to be 0.8-1.2.

[0447] a i , b: Adjust parameters to adapt to different market conditions, such as a i is the market-specific adjustment coefficient, b is the global adjustment constant, assuming a i is 0.1-0.5, b is 0.2.

[0448] Assumption example:

[0449] Three user groups n = 3

[0450] F = [50, 60, 70]

[0451] m=[1.0,1.1,1.2]

[0452] a=[0.2,0.3,0.4]

[0453] b=0.2

[0454] Calculation process:

[0455]

[0456] R≈781.25+1936.2+3848.41=6565.86

[0457] Formula 2: Inventory gap data G

[0458]

[0459] Parameter details:

[0460] R j : The predicted inventory demand at the jth time point, obtained from the previous formula.

[0461] s j : Actual inventory status, assuming the data comes from actual warehouse records, is 40, 50, 60.

[0462] c j , d: Smoothing parameter, adjusts the sensitivity of inventory difference calculation, assuming c j is 0.1, and d is 0.05.

[0463] Assumption example:

[0464] R = [781, 1936, 3848] (simplified result of previous calculation)

[0465] s = [700, 1800, 3600]

[0466] Calculation process:

[0467]

[0468] G≈38319.7+83528.5+179603.8=301452

[0469] See also Figure 7 , the specific steps for obtaining the optimized warehouse layout are:

[0470] According to the future inventory demand forecast Q, collect the environmental parameters and operation data in the warehouse, and use the formula:

[0471]

[0472] Generate warehouse data model D;

[0473] Among them, D is the warehouse data model, Q i represents the inventory demand forecast for period i, e i is the influence coefficient of environmental parameters, f i is the influence coefficient of the operating data, g is the constant adjustment parameter;

[0474] Using the warehouse data model D, each storage location and handling robot node is modeled, and the relationship between the handling path and the storage location is analyzed using the formula:

[0475]

[0476] Construct path relationship model P;

[0477] Among them, P is the path relationship model, D j represents the jth storage element data extracted from the D model, h j is the influence coefficient of the transport path, k, l j To adjust the parameters;

[0478] By analyzing the relationship model P between the transport path and the storage location, the formula is adopted:

[0479]

[0480] Identify bottlenecks and optimization space in warehouse layout and form layout optimization plan S;

[0481] Among them, S is the layout optimization solution, Pj is the output of the relationship model between the jth transport path and the storage location, v represents the transport efficiency under ideal conditions, o j , p is the efficiency adjustment parameter;

[0482] Based on the layout optimization plan S, the warehouse layout is redesigned and adjusted using the formula:

[0483]

[0484] Form an optimized storage layout L;

[0485] Among them, L is the optimized storage layout, S n represents the nth optimization solution obtained from S, r n is the implementation weight of the suggestion.

[0486] Warehouse Data Model D

[0487] formula:

[0488]

[0489] Parameter details:

[0490] Q i : Inventory demand forecast for period i, e.g. 1000 units.

[0491] e i : The impact coefficient of environmental parameters, such as 1.05, represents the increase ratio of environmental factors to demand.

[0492] f i : The impact coefficient of operating data, such as 0.95, represents the reduction ratio of operating efficiency to demand.

[0493] g: A constant adjustment parameter, such as 2, used to ensure that the denominator is non-zero and smooth the data.

[0494] Calculation example:

[0495] Assume that there are three time periods of data, and the specific data of each period is as follows:

[0496] Q1=1000,e1=1.05,f1=0.95

[0497] Q2=1500,e2=1.10,f2=0.90

[0498] Q3=1200,e3=1.08,f3=0.92

[0499] Q=2

[0500] Substitute into the formula:

[0501]

[0502] Calculation process:

[0503]

[0504] D≈(610.192) 2.5 +(968.316) 2.5 +(758.336) 2.5

[0505] D≈143873+290779+181353

[0506] D≈616005

[0507] Path Relationship Model P

[0508] formula:

[0509]

[0510] Parameter details:

[0511] D j : Data model output of the jth storage element.

[0512] h j : Transportation path influence coefficient.

[0513] k, l j : Adjust parameters.

[0514] Assumptions:

[0515] D1=616005,h1=1.2,k=1,l1=1.5

[0516] D2=606000,h2=1.1,l2=1.4

[0517] calculate:

[0518]

[0519] P≈(739206 / 1.225)+(666600 / 1.183)

[0520] P≈603433+563467

[0521] P≈1166900

[0522] Layout optimization suggestions

[0523] formula:

[0524]

[0525] Parameter details:

[0526] P j : Output of the relationship model between the j-th transport path and the storage location.

[0527] v: ideal transport efficiency, for example, set to 1000 units.

[0528] o j , p: efficiency adjustment parameter, o j is a specific parameter for the transport path, and p is a constant adjustment parameter.

[0529] Assumptions:

[0530] P1=1166900 (the result of the previous formula)

[0531] o1=200,p=50

[0532] Calculation process:

[0533]

[0534] S=min(-1165900 / 15.81)

[0535] S≈-73768

[0536] This step selects the area that needs optimization the most, that is, the largest bottleneck area, to highlight the focus of improvement.

[0537] Optimized warehouse layout

[0538] formula:

[0539]

[0540] Parameter details:

[0541] S: Minimization function result obtained from the layout optimization suggestions.

[0542] r n : The relative importance weights of the optimization implementation, assuming r1=0.7, r2=0.3.

[0543] S n : Specific values ​​in the optimization suggestion, assuming S1 = -73768 and S2 = -50000.

[0544] Calculation process:

[0545] L=-73768+(0.7·exp(-73768)+0.3·exp(-50000))

[0546] L≈-73768+(0.7·0+0.3·0)

[0547] L≈-73768

[0548] This calculation uses an exponential function to amplify the impact of each optimization suggestion, although in the case of extreme values ​​(such as very large negative numbers in the example), the exponential function approaches zero, highlighting the need to reconsider parameter settings or optimization strategies.

[0549] See also Figure 8 , the steps for obtaining the dynamic inventory scheduling plan are as follows:

[0550] Real-time monitoring of inventory dynamics and operations in the optimized warehouse layout, using the formula:

[0551]

[0552] Integrate key data and build real-time monitoring data model E;

[0553] Among them, E is the real-time monitoring data model, L i represents the real-time information of the i-th inventory location, o i is the corresponding operation data, p i , q is the adjustment parameter;

[0554] Using the real-time monitoring data model E, we analyze the changes in inventory demand and turnover speed using the formula:

[0555]

[0556] Identify inventory adjustment requirements N;

[0557] Among them, N is the inventory adjustment demand, M j represents the monitoring data of the jth inventory item, r j is the demand change rate, s, t j is the adjustment factor;

[0558] According to the inventory adjustment demand N, the inventory location is dynamically reallocated using the formula:

[0559]

[0560] Form an inventory position adjustment plan P;

[0561] Among them, P is the inventory position adjustment plan, N k is the demand data of the kth inventory item, v k In order to adjust the coefficient angle, the trigonometric function sin(v k ) simulates the impact of cyclical demand changes, w k, x is the adjustment parameter; based on the inventory location adjustment plan P, combined with the warehouse operation strategy and resource constraints, the formula is adopted:

[0562]

[0563] Generate a dynamic inventory scheduling plan S;

[0564] Among them, S is the inventory dynamic scheduling plan, P n is the output of the nth position adjustment plan, u represents the scheduling optimization goal, y n , z is the efficiency adjustment parameter.

[0565] Real-time monitoring of the calculation process of the data model M

[0566] formula:

[0567]

[0568] Parameter details:

[0569] L i : Real-time information of the i-th inventory location, such as inventory quantity.

[0570] o i : Operational data for the ith inventory location, such as visit frequency.

[0571] p i : The variation adjustment parameter of the location, taking into account specific changes in inventory locations.

[0572] q: Globally adjust parameters to ensure overall data consistency.

[0573] Assumed parameter values ​​and examples:

[0574] Assume there are three inventory locations, and the parameters of each location are as follows:

[0575] L1=100 pieces, o1=10 times / day, p1=5

[0576] L2=150 pieces, o2=15 times / day, p2=3

[0577] L3=200 pieces, o3=20 times / day, p3=2

[0578] q=2

[0579] Calculation process:

[0580]

[0581] E = 102453.4

[0582] Calculation process of demand change analysis N

[0583] formula:

[0584]

[0585] Parameter details:

[0586] M j : Monitoring data of the j-th inventory item.

[0587] r j : Demand change rate, which is the sales growth rate or decline rate. s: Global adjustment coefficient, used for data consistency.

[0588] t j : Adjustment coefficient for the j-th item.

[0589] Assumed parameter values ​​and examples:

[0590] Assume that the demand change rate r1 = 1.05, r2 = 1.10, r3 = 0.95 and the global parameter s = 0.5

[0591] t1=0.1,t2=0.2,t3=0.3

[0592] Calculation process:

[0593]

[0594] N = 6992.2

[0595] Calculation process of inventory location optimization P

[0596] formula:

[0597]

[0598] Parameter details:

[0599] N k : Demand data for the kth inventory item.

[0600] v k : Adjust the coefficient angle to simulate periodic demand changes. k : Adjustment parameters for each project.

[0601] x: Global adjustment parameter.

[0602] Assumed parameter values ​​and examples:

[0603] N1=2000, N2=3000, N3=1992.2

[0604] v1=30°,v2=45°,v3=60°

[0605] w1=0.1,w2=0.2,w3=0.1

[0606] x=0.5

[0607] Calculation process:

[0608]

[0609] P = 3884.0

[0610] The calculation process of generating S for dynamic scheduling scheme

[0611] formula:

[0612]

[0613] Parameter details:

[0614] P n : Output of the nth position adjustment plan.

[0615] u: Scheduling optimization objective, usually operating efficiency or cost.

[0616] y n : Efficiency tuning parameters for each tuning plan.

[0617] z: Global efficiency adjustment parameter.

[0618] Assumed parameter values ​​and examples:

[0619] P1=774.6, P2=1772.8, P3=1336.5

[0620] u=2000

[0621] y1=0.1,y2=0.2,y3=0.15

[0622] z=0.5

[0623] Calculation process:

[0624]

[0625] S=271.4

[0626] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. Intelligent warehouse management system based on Internet of Things technology, characterized by: The system comprises: The sensor network data acquisition module is based on the IoT sensor network, collects inventory item locations, environmental monitor parameters and operation data, sets nodes including inventory items, storage locations, and handling robots, sets edges including the handling path and storage location relationship, summarizes and updates the location, status and operation path data, and generates multi-dimensional warehouse data; The demand forecasting and analysis module performs market trend analysis of historical sales data, inventory change data, and historical sales data based on the multi-dimensional warehouse data. It uses collaborative filtering algorithms to extract user demand fitting patterns through historical sales data analysis, calculates and predicts future inventory demand, compares future inventory demand with real-time inventory status, and generates future inventory demand forecasts. The steps of market trend analysis of historical sales data inventory change data are specifically as follows: Call the multi-dimensional warehouse data to obtain historical sales data using the formula: Summarize historical sales data and inventory change data to generate historical sales summary results S; Among them, S represents the historical sales summary results, h i represents historical sales data, k i Indicates inventory change data; The historical sales summary result S is subjected to time series analysis using the formula: Extract time series features T; Among them, T represents the time series characteristics, S j represents the j-th historical sales summary result, t j Represents time series data, c and d j To adjust the parameters, generate time series features T; The market trend analysis of the time series feature T is performed according to the formula: Combine market influencing factors to generate market trend analysis results M; Among them, M represents the market trend analysis results, T k represents the kth time series feature, f k represents the market impact factor, e k and g are smoothing parameters; According to the market trend analysis result M, the formula is adopted: Conduct market trend analysis of historical sales data, inventory change data; Among them, A represents the historical sales data, inventory change data, market trend analysis results, n l represents the market data integration coefficient, v is the data integration adjustment coefficient; The step of extracting the user demand fitting mode is specifically as follows: From the user's time series features T i Get data, apply collaborative filtering technology, and use the formula: Identify and aggregate the characteristics of similar user groups to obtain similar user group data U; Among them, U is the data of the same user group, w i is the user weight, T i is the user's time series feature, r, s i To adjust the parameters and balance the feature weights of differentiated users; Combined with the similar user group data U and purchase behavior history, behavioral pattern analysis is used, using the formula: Aggregate purchase patterns and create purchase pattern data P; Among them, P is the purchase mode data, U j is the group data of the jth user, b j Indicates the corresponding purchase behavior, h j , t is the adjustment coefficient, balancing the impact of purchasing behavior; Through the purchase pattern data P, market changes and product trends are analyzed using the formula: Adjust demand forecasting parameters to form a user demand fitting model F; Among them, F is the user demand fitting model, P k is the kth purchase pattern data, d k is the demand adjustment factor, x, y k is the smoothing parameter; The steps for obtaining the future inventory demand forecast are specifically as follows: Get the output from the user demand fitting model F, apply the market forecasting model, and use the formula: Calculate the inventory demand in the future target time period to form the preliminary future inventory demand R; Among them, R is the initial future inventory demand, F i represents the output of the i-th user demand fitting mode, m i is the corresponding market trend impact coefficient, a i , b is the adjustment parameter to match the demand forecast under differentiated market conditions; Analyze the difference between the preliminary future inventory demand R and the current inventory status, using the formula: Evaluate and quantify the inventory gap to form inventory gap data G; Among them, G is the inventory gap data, R j is the predicted inventory demand at the jth time point, s j is the actual inventory status at the same time point, c j and d are smoothing parameters; Combining the inventory gap data G with economic indicators and supply chain capacity parameters, the formula is adopted: Evaluate and adjust future inventory plans to form future inventory demand forecast Q; Among them, Q is the future inventory demand forecast, G k is the kth inventory gap data, v k is the economic indicator adjustment coefficient, o k , u is the smoothing parameter of supply chain capacity; The warehouse layout optimization module models the nodes of the inventory storage location handling robots based on the future inventory demand forecast, combined with the warehouse environment parameters and operation data, and identifies the current layout bottleneck and optimization space by analyzing the relationship between the storage location of the item handling path in the warehouse layout, and generates the optimized warehouse layout; Based on the optimized warehouse layout, the dynamic scheduling control module monitors inventory dynamics and operation status in real time, dynamically adjusts operations, analyzes the turnover speed preservation conditions of inventory item demand changes, reallocates inventory locations, dynamically adjusts inventory item storage locations and transportation paths, and generates an inventory dynamic scheduling plan.

2. According to claim 1, the intelligent warehouse management system based on Internet of Things technology is characterized in that: The steps for collecting the inventory item location, environmental monitor parameters and operation data are specifically as follows: According to the data of the IoT sensor network, the location information of inventory items at different time points is collected, and the parameters and operation data of the environmental monitor are recorded. Refer to the formula: Generate a comprehensive dataset D; Among them, D i Represents a comprehensive dataset of items, S ij represents the position signal strength of item i collected by sensor j, T j represents the timestamp of sensor j, E ij represents the environmental parameters of item i collected by environmental monitor j, H j Indicates the timestamp of the environmental monitor, O ij represents the operation data of item i collected by operation data sensor j, P j represents the timestamp of the operation data sensor j, K represents the position adjustment parameter, L ik represents the location information of item i at time point k, M represents the normalization coefficient, and ∈ represents a small value to avoid division by zero; The comprehensive data set D is processed to remove abnormal data and calculate the average location, environmental parameters and operation data of each inventory item, referring to the formula: Get the average comprehensive data set A; Among them, A i represents the average comprehensive data of item i, D ik represents the comprehensive data of item i at time point k, w k represents the weight parameter at time point k, R ip represents the comprehensive data residual of item i in dimension p, and N represents the normalization coefficient; The cluster analysis was performed using the average comprehensive data set A, referring to the formula: Determine the location area, environmental monitor parameters and operation path of the inventory items to obtain the comprehensive area R of the items i ; Among them, R i Indicates the comprehensive area of ​​the item, A ip represents the average comprehensive data of item i on dimension p, C p represents the location of the cluster center on dimension p, V iq represents the deviation value of item i on dimension q, d represents the number of dimensions of comprehensive data, B i represents the base value of item i; By looping through all items, the combined area R of the item i , refer to the formula: Determine whether there is a situation that exceeds the preset threshold. If so, adjust the position and update the storage position to obtain the inventory item location, environmental monitor parameters and operation data; Among them, R′ i represents the comprehensive area of ​​the adjusted items, R i represents the comprehensive area of ​​the item, Threshold represents the preset threshold, Δ represents the position adjustment value, and Z it represents the correction factor of item i at time point t, and Z represents the total number of correction factors.

3. The intelligent warehouse management system based on Internet of Things technology according to claim 2 is characterized in that: The steps for obtaining the multidimensional warehouse data are specifically as follows: The inventory item location, environmental monitor parameters and operating data are called using the formula: Obtain the data of sensor nodes and perform preliminary aggregation to generate preliminary aggregation results W of inventory item locations, environmental monitor parameters and operation data; Among them, W represents the preliminary summary results, p i Indicates the inventory item location, L i represents the environmental monitor parameter, γ i Represents the weight coefficients, α and β i To adjust the parameters; Compare the preliminary summary result W with the historical data to determine whether the environmental monitor parameter exceeds the preset threshold T, using the formula: Associating the operation data with the inventory item location to generate location status data V; Among them, V represents the position status data, H j represents the jth historical data, σ j represents the historical data adjustment coefficient, δ is the smoothing parameter, and T represents the preset threshold; According to the position status data V, combined with the inventory item location and transportation path data, the transportation path is adjusted, the transportation path is recalculated, and the summary update result of the position, status and operation path data is generated. The reference formula is as follows: Among them, U represents the summary update result, P k represents the kth transport path, R k represents the adjusted k-th transport path, φ k represents the path adjustment coefficient; Perform multi-dimensional processing of the position state data V, the operation path data and the summary update result U, referring to the formula: Combining the multidimensional processing coefficient K and the warehouse data integration coefficient D, multidimensional warehouse data M is formed; Among them, M represents multidimensional warehouse data, U represents the summary update result, K represents the multidimensional processing coefficient, and D l represents the first warehouse data integration coefficient, η l is the data integration adjustment coefficient, and θ is the smoothing coefficient.

4. The intelligent warehouse management system based on Internet of Things technology according to claim 1 is characterized in that: The steps for obtaining the optimized warehouse layout are specifically as follows: According to the future inventory demand forecast Q, the environmental parameters and operation data in the warehouse are collected and the formula is used: Generate warehouse data model D; Among them, D is the warehouse data model, Q i represents the inventory demand forecast for period i, e i is the influence coefficient of environmental parameters, f i is the influence coefficient of the operating data, g is the constant adjustment parameter; Using the warehouse data model D, each storage location and handling robot node is modeled, and the relationship between the handling path and the storage location is analyzed using the formula: Construct path relationship model P; Among them, P is the path relationship model, D j represents the jth storage element data extracted from the D model, h j is the influence coefficient of the transport path, k, l j To adjust the parameters; By analyzing the relationship model P between the transport path and the storage location, the formula is adopted: Identify bottlenecks and optimization space in warehouse layout and form layout optimization plan S; Among them, S is the layout optimization solution, P j is the output of the relationship model between the jth transport path and the storage location, v represents the transport efficiency under ideal conditions, o j , p is the efficiency adjustment parameter; Based on the layout optimization scheme S, the warehouse layout is redesigned and adjusted using the formula: Form an optimized storage layout L; Among them, L is the optimized storage layout, S n represents the nth optimization solution obtained from S, r n is the implementation weight of the suggestion.

5. The intelligent warehouse management system based on Internet of Things technology according to claim 4 is characterized in that: The steps for obtaining the dynamic inventory scheduling plan are specifically as follows: Real-time monitoring of inventory dynamics and operations in the optimized warehouse layout, using the formula: Integrate key data and build real-time monitoring data model E; Among them, E is the real-time monitoring data model, L i represents the real-time information of the i-th inventory location, o i is the corresponding operation data, p i , q is the adjustment parameter; The real-time monitoring data model E is used to analyze the demand changes and turnover speed of inventory items, using the formula: Identify inventory adjustment requirements N; Among them, N is the inventory adjustment demand, M j represents the monitoring data of the jth inventory item, r j is the demand change rate, s, t j is the adjustment factor; According to the inventory adjustment demand N, the inventory location is dynamically reallocated using the formula: Form an inventory position adjustment plan P; Among them, P is the inventory position adjustment plan, N k is the demand data of the kth inventory item, v k In order to adjust the coefficient angle, the trigonometric function sin(v k ) simulates the impact of cyclical demand changes, w k , x is the adjustment parameter; based on the inventory location adjustment plan P, combined with the warehouse operation strategy and resource constraints, the formula is adopted: Generate a dynamic inventory scheduling plan S; Among them, S is the inventory dynamic scheduling plan, P n is the output of the nth position adjustment plan, u represents the scheduling optimization goal, y n , z is the efficiency adjustment parameter.