Warehouse management system and method based on dynamic inventory prediction

Abstract

The invention relates to the technical field of warehouse management, and discloses a warehouse management system and method based on dynamic inventory prediction. According to the invention, market, sales, supply chain and warehouse internal data are collected in real time, a dynamic inventory prediction model is constructed in combination with a long short-term memory (LSTM) network, the future inventory demand can be accurately predicted, and the prediction precision is significantly improved; an inventory optimization strategy is generated based on a multi-objective optimization algorithm, so that the inventory holding cost, the stockout cost and the transportation cost are effectively reduced; and intelligent warehouse operation instruction generation and real-time monitoring are realized through the execution control module, and the response capability of an enterprise to market fluctuation and emergencies is enhanced. According to the system and the method, the warehouse management efficiency is improved, the operation risk is reduced, and an efficient and reliable solution is provided for modern logistics and supply chain management.

Description

Technical Field

[0001] The present invention relates to the technical field of warehouse management, and specifically to a warehouse management system and method based on dynamic inventory forecasting. Background Art

[0002] In modern logistics and supply chain management, the efficiency and accuracy of the warehouse management system directly affect the operating costs of enterprises and customer satisfaction. Traditional warehouse management systems usually rely on static inventory data and historical sales records for inventory forecasting and management. This method often shows great limitations when facing market demand fluctuations, seasonal changes, and emergencies. Static forecasting models cannot respond to market changes in real time, resulting in inventory overstock or shortage, thereby increasing the operating costs and risks of enterprises.

[0003] In recent years, with the development of big data and artificial intelligence technologies, dynamic inventory forecasting has gradually become a research hotspot in the field of warehouse management. By collecting and analyzing multi-source information such as market data, sales data, and supply chain data in real time, dynamic inventory forecasting can more accurately predict future inventory demands, thereby optimizing inventory management strategies. However, existing dynamic inventory forecasting methods still face many challenges in data processing, model construction, and real-time performance. Especially when facing a complex and changing market environment, the forecasting accuracy and response speed often fail to meet actual requirements.

[0004] Therefore, developing a warehouse management system and method based on dynamic inventory forecasting that can respond to market changes in real time and improve the accuracy of inventory forecasting and management efficiency has become an urgent problem to be solved in the current field of logistics and supply chain management. Summary of the Invention

[0005] Aiming at the deficiencies of the existing technology, the present invention provides a warehouse management system and method based on dynamic inventory forecasting to solve the problems in the background art.

[0006] In a first aspect, to achieve the above object, the present invention provides the following technical solution: A warehouse management system based on dynamic inventory forecasting, including:

[0007] A data collection module for collecting market data, sales data, supply chain data, and warehouse internal operation data in real time;

[0008] A data preprocessing module for cleaning, normalizing, and feature extracting the collected data to generate a standardized data set for prediction;

[0009] A dynamic forecasting model module for constructing and training a dynamic inventory forecasting model based on the standardized data set. The dynamic inventory forecasting model uses time series analysis and machine learning algorithms, and its forecasting formula is:

[0010] It+1 = f(S t , D t , C t , t )

[0011] where I t+1 represents the inventory demand at the next moment, S t represents the sales data at the current moment, D t represents the market demand data at the current moment, C t represents the supply chain data at the current moment, t and represents the random error term;

[0012] The inventory optimization module is used to generate an inventory optimization strategy according to the prediction result output by the dynamic prediction model, including a procurement plan, inventory allocation, and replenishment suggestions;

[0013] The execution control module is used to convert the inventory optimization strategy into specific warehouse operation instructions and monitor the execution situation in real time.

[0014] Preferably, the dynamic prediction model module uses the long short-term memory network (LSTM) as the core algorithm, and its formula is:

[0015] f t = σ(W f · [h t-1 , x t + b f )

[0016] i t = σ(W i · [h t-1 , x t + b i )

[0017]

[0018] o t = σ(W o · [h t-1 , x t + b o )

[0019] h t = o t · tanh(C t )

[0020] where f t represents the forget gate, i t represents the input gate, represents the candidate memory cell, C t represents the current memory cell, o tRepresents the output gate, h t Represents the current hidden state, W and b are the weight matrix and bias term respectively, and σ represents the sigmoid activation function.

[0021] Preferably, the inventory optimization module adopts a multi-objective optimization algorithm, and its objective function is:

[0022] min(α·C 库存 +β·C 缺货 +γ·C 运输 )

[0023] Where C 库存 Represents the inventory holding cost, C 缺货 Represents the shortage cost, C 运输 Represents the transportation cost, and α, β, and γ are the weight coefficients of each cost respectively.

[0024] Second, a warehouse management method based on dynamic inventory prediction, which is implemented based on the warehouse management system based on dynamic inventory prediction described in the first aspect, and includes the following steps:

[0025] Real-time collect market data, sales data, supply chain data, and warehouse internal operation data;

[0026] Clean, normalize, and extract features from the collected data to generate a standardized data set;

[0027] Based on the standardized data set, build and train a dynamic inventory prediction model, and the dynamic inventory prediction model adopts time series analysis and machine learning algorithms;

[0028] According to the prediction results output by the dynamic prediction model, generate inventory optimization strategies, including procurement plans, inventory allocation, and replenishment suggestions;

[0029] Convert the inventory optimization strategy into specific warehouse operation instructions and monitor the execution situation in real time.

[0030] Preferably, the training process of the dynamic inventory prediction model includes the following steps:

[0031] Divide the standardized data set into a training set and a test set;

[0032] Use the training set to train the dynamic inventory prediction model and verify the prediction accuracy of the model through the test set;

[0033] According to the verification results, adjust the model parameters until the preset prediction accuracy requirement is met.

[0034] Preferably, the generation process of the inventory optimization strategy includes the following steps:

[0035] Calculate the inventory demand for a period of time in the future according to the prediction results output by the dynamic prediction model;

[0036] Combine the current inventory level, supply chain capabilities, and transportation costs to generate an optimal procurement plan, inventory allocation, and replenishment recommendations;

[0037] Convert the generated inventory optimization strategy into specific warehouse operation instructions and monitor the execution status in real time.

[0038] Preferably, the execution process of the warehouse operation instructions includes the following steps:

[0039] Convert the inventory optimization strategy into specific warehouse operation instructions, including inbound, outbound, transfer, and replenishment of goods;

[0040] Monitor the execution status of the warehouse operation instructions in real time and dynamically adjust the inventory optimization strategy according to the actual execution results.

[0041] Preferably, the prediction accuracy of the dynamic inventory prediction model is evaluated by the mean square error (MSE), and its formula is:

[0042]

[0043] Where, I i represents the actual inventory demand, represents the predicted inventory demand, and n represents the number of samples.

[0044] Preferably, the generation process of the inventory optimization strategy also includes considering the influence of seasonal factors and emergencies, and its formula is:

[0045] I t+1 = f(S t , D t , C t , t ) + δ·S 季节性 + η·E 突发事件

[0046] Where, represents the seasonal factor, E 突发事件 represents the influence of emergencies, and δ and η are the weight coefficients of the seasonal factor and emergencies respectively.

[0047] Preferably, the execution process of the warehouse operation instructions also includes a real-time feedback mechanism, and its formula is:

[0048]

[0049] Where, ΔI t represents the deviation between the actual inventory and the predicted inventory, I t represents the actual inventory, Indicates the predicted inventory. According to the deviation results, the inventory optimization strategy is dynamically adjusted.

[0050] Compared with the prior art, the present invention has the following beneficial effects:

[0051] By collecting market, sales, supply chain, and warehouse internal data in real time and combining with the long short-term memory network (LSTM) to construct a dynamic inventory prediction model, the present invention can accurately predict future inventory demands, significantly improving the prediction accuracy; generating an inventory optimization strategy based on a multi-objective optimization algorithm, effectively reducing inventory holding costs, out-of-stock costs, and transportation costs; realizing the generation and real-time monitoring of intelligent warehouse operation instructions through an execution control module, enhancing the enterprise's response ability to market fluctuations and emergencies. The system and method improve warehouse management efficiency, reduce operational risks, and provide an efficient and reliable solution for modern logistics and supply chain management.

[0052] Other features and advantages of the present invention will be described in the subsequent specification, and, in part, will be obvious from the specification, or will be understood by implementing the present invention. The objectives and other advantages of the present invention can be realized and obtained through the structures pointed out in the specification, claims, and drawings. Brief Description of the Drawings

[0053] Figure 1 It is a flowchart of the warehouse management method based on dynamic inventory prediction of the present invention;

[0054] Figure 2 It is a block diagram of the warehouse management system based on dynamic inventory prediction of the present invention. Detailed Embodiments

[0055] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of the present invention.

[0056] 1. System Architecture and Module Functions

[0057] Please refer to Figure 1 and Figure 2 , this embodiment provides a warehouse management system and method based on dynamic inventory prediction, aiming to improve the accuracy and efficiency of inventory management through real-time data collection, dynamic prediction models, and intelligent optimization strategies. The following is the detailed implementation process of the system.

[0058] 2. Data Collection Module

[0059] The data acquisition module is responsible for real-time data acquisition from multiple data sources, including:

[0060] Market data: such as market demand trends, competitor inventory status, price fluctuations, etc.;

[0061] Sales data: such as historical sales records, real-time order data, customer return data, etc.;

[0062] Supply chain data: such as supplier delivery time, transportation cost, raw material inventory, etc.;

[0063] Warehouse internal operation data: such as inventory level, goods turnover rate, warehouse capacity utilization rate, etc.

[0064] The data acquisition module realizes real-time synchronization of multi-source data through API interfaces, Internet of Things sensors (such as RFID, barcode scanners), and enterprise resource planning (ERP) systems. For example, when a customer places an order, the sales data will be transmitted to the data acquisition module in real time through the ERP system; at the same time, the RFID sensors in the warehouse will update the inventory level in real time.

[0065] 3. Data preprocessing module

[0066] The data preprocessing module cleans, normalizes, and extracts features from the collected raw data to generate a standardized data set. The specific steps are as follows:

[0067] Data cleaning: Remove duplicate data, fill in missing values, and correct outliers. For example, for missing sales data, interpolation is used for filling.

[0068] Data normalization: Unify data with different dimensions to the same scale. For example, normalize sales data and market demand data to the [0, 1] interval.

[0069] Feature extraction: Extract key features from the raw data. For example, extract seasonal features from sales data and trend features from market demand data.

[0070] The preprocessed data will be used to train the dynamic prediction model.

[0071] 4. Dynamic prediction model module

[0072] The dynamic prediction model module uses the long short-term memory network (LSTM) as the core algorithm, and its formula is as follows:

[0073] f t =σ(W f ·[h t-1 ,x t +b f )

[0074] it = σ(W i · [h t-1 , x t + b i )

[0075]

[0076] o t = σ(W o · [h t-1 , x t + b o )

[0077] h t = o t · tanh(C t )

[0078] where f t represents the forget gate, i t represents the input gate, represents the candidate memory cell, C t represents the current memory cell, o t represents the output gate, h t represents the current hidden state, W and b are the weight matrix and bias term respectively, and σ represents the sigmoid activation function.

[0079] Training process:

[0080] 1. Divide the standardized dataset into a training set and a test set (e.g., 80% training set, 20% test set);

[0081] 2. Use the training set to train the LSTM model and verify the prediction accuracy of the model through the test set;

[0082] 3. According to the verification results, adjust the model parameters (such as the learning rate, the number of hidden layer nodes) until the preset prediction accuracy requirement is met (e.g., the mean squared error MSE < 0.01).

[0083] Prediction process:

[0084] The model predicts the inventory demand I t at the next moment based on the sales data S t , the market demand data D t , the supply chain data C t and the random error term t+1 , and its formula is:

[0085] I t+1 = f(S t , D t , C t , t )

[0086] 5. Inventory Optimization Module

[0087] The inventory optimization module generates an inventory optimization strategy based on the prediction results output by the dynamic prediction model. The optimization goal is to minimize inventory holding costs, stock-out costs, and transportation costs, and its objective function is:

[0088] min(α·C 库存 +β·C 缺货 +γ·C 运输 )

[0089] Where, C 库存 represents the inventory holding cost, C 缺货 represents the stock-out cost, C 运输 represents the transportation cost, and α, β, and γ are the weight coefficients of each cost respectively.

[0090] Optimization strategy generation process:

[0091] 1. Based on the prediction result I t+1 , calculate the inventory demand within a future period of time;

[0092] 2. Combine the current inventory level, supply chain capacity, and transportation costs to generate an optimal procurement plan, inventory allocation, and replenishment suggestions;

[0093] 3. Consider seasonal factors and unexpected events E 突发事件 , and adjust the optimization strategy. Its formula is:

[0094] I t+1 =f(S t ,D t ,C t , t )+δ·S 季节性 +η·E 突发事件

[0095] Where, δ and η are the weight coefficients of seasonal factors and unexpected events respectively.

[0096] 6. Execution Control Module

[0097] The execution control module converts the inventory optimization strategy into specific warehouse operation instructions and monitors the execution situation in real time. The specific steps are as follows:

[0098] 1. Instruction generation: Generate instructions for goods receipt, shipment, transfer, and replenishment according to the optimization strategy;

[0099] 2. Instruction execution: Execute the instructions through the warehouse management system (WMS) and automated equipment (such as AGV, robotic arm);

[0100] 3. Real-time monitoring: The execution status of instructions is monitored in real time through sensors and the WMS system, and the actual inventory I is calculated. t The deviation ΔI from the predicted inventory t :

[0101]

[0102] where represents the predicted inventory.

[0103] 4. Dynamic adjustment: Based on the deviation results, the inventory optimization strategy is dynamically adjusted. For example, when the actual inventory is lower than the predicted inventory, an emergency replenishment instruction is triggered.

[0104] Example:

[0105] Suppose an e-commerce enterprise faces a sharp increase in market demand during the "Double Eleven" promotion period. The traditional static prediction model fails to accurately predict the demand, resulting in inventory shortages and order delays. Through the system of this embodiment:

[0106] The data acquisition module collects the sales data and market demand data during the promotion in real time;

[0107] The dynamic prediction model module predicts the inventory demand for the next week, taking into account the influence of seasonal factors and promotional activities;

[0108] The inventory optimization module generates a replenishment plan and an inventory allocation strategy;

[0109] The execution control module monitors the inventory level in real time and dynamically adjusts the replenishment instruction to ensure sufficient inventory.

[0110] Finally, the enterprise achieves a zero out-of-stock rate during the promotion period, and the customer satisfaction is significantly improved.

[0111] This embodiment provides a warehouse management system and method based on dynamic inventory prediction. By collecting market, sales, supply chain, and warehouse internal data in real time, using a long short-term memory network (LSTM) to build a dynamic prediction model, accurately predicting future inventory demand, and combining a multi-objective optimization algorithm to generate an inventory optimization strategy, finally, the intelligent generation and real-time monitoring of warehouse operation instructions are realized through the execution control module. This system significantly improves the inventory prediction accuracy and management efficiency, reduces the operation cost, enhances the enterprise's response ability to market fluctuations and emergencies, and provides an efficient and reliable solution for modern logistics and supply chain management.

Claims

1. A warehouse management system based on dynamic inventory forecasting, characterized in that, Including: A data collection module for real-time collection of market data, sales data, supply chain data, and internal warehouse operation data; A data preprocessing module for cleaning, normalizing, and feature extraction of the collected data to generate a standardized data set for prediction; A dynamic prediction model module for constructing and training a dynamic inventory prediction model based on the standardized data set. The dynamic inventory prediction model uses time series analysis and machine learning algorithms, and its prediction formula is: I t+1 = f(S t , D t , C t , t ) Among them, I t+1 represents the inventory demand at the next moment, S t represents the sales data at the current moment, D t represents the market demand data at the current moment, C t represents the supply chain data at the current moment, t represents the random error term; An inventory optimization module for generating inventory optimization strategies according to the prediction results output by the dynamic prediction model, including procurement plans, inventory allocation, and replenishment suggestions; An execution control module for converting the inventory optimization strategy into specific warehouse operation instructions and monitoring the execution status in real time.

2. The warehouse management system based on dynamic inventory prediction according to claim 1, wherein The dynamic prediction model module uses a long short-term memory network (LSTM) as the core algorithm, and its formula is: f t = σ(W f · [h t-1 , x t + b f ) i t = σ(W i · [h t-1 , x t + b i ) o t = σ(W o · [h t-1 , x t + b o ) h t = o t ·tanh(C t ) Among them, f t represents the forget gate, i t represents the input gate, represents the candidate memory cell, C t represents the current memory cell, o t represents the output gate, h t represents the current hidden state, W and b are the weight matrix and bias term respectively, and σ represents the sigmoid activation function.

3. The warehouse management system based on dynamic inventory prediction according to claim 1, characterized in that, The inventory optimization module uses a multi-objective optimization algorithm, and its objective function is: min(α·C 库存 +β·C 缺货 +γ·C 运输 ) Among them, C 库存 represents the inventory holding cost, C 缺货 represents the stock-out cost, C 运输 represents the transportation cost, and α, β, and γ are the weight coefficients of each cost respectively.

4. Warehouse management method based on dynamic inventory forecasting, characterized in that, This method is implemented based on the warehouse management system for dynamic inventory prediction described in any one of claims 1-3, and the following steps are carried out: Real-time collection of market data, sales data, supply chain data, and internal warehouse operation data; Cleaning, normalizing, and feature extraction of the collected data to generate a standardized data set; Based on the standardized data set, constructing and training a dynamic inventory prediction model. The dynamic inventory prediction model uses time series analysis and machine learning algorithms; Generating inventory optimization strategies according to the prediction results output by the dynamic prediction model, including procurement plans, inventory allocation, and replenishment suggestions; Converting the inventory optimization strategy into specific warehouse operation instructions and monitoring the execution status in real time.

5. The warehouse management method based on dynamic inventory prediction according to claim 4, wherein The training process of the dynamic inventory prediction model includes the following steps: Dividing the standardized data set into a training set and a test set; Using the training set to train the dynamic inventory prediction model and verifying the prediction accuracy of the model through the test set; Adjusting the model parameters according to the verification results until the preset prediction accuracy requirement is met.

6. The warehouse management method based on dynamic inventory prediction according to claim 4, wherein The generation process of the inventory optimization strategy includes the following steps: Calculating the inventory demand in a future period according to the prediction results output by the dynamic prediction model; Combining the current inventory level, supply chain capacity, and transportation cost to generate optimal procurement plans, inventory allocation, and replenishment suggestions; Converting the generated inventory optimization strategy into specific warehouse operation instructions and monitoring the execution status in real time.

7. The warehouse management method based on dynamic inventory prediction according to claim 4, characterized in that The execution process of the warehouse operation instructions includes the following steps: Converting the inventory optimization strategy into specific warehouse operation instructions, including goods receipt, goods issue, transfer, and replenishment; Real-time monitoring of the execution status of the warehouse operation instructions and dynamically adjusting the inventory optimization strategy according to the actual execution results.

8. The warehouse management method based on dynamic inventory prediction according to claim 4, wherein The prediction accuracy of the dynamic inventory prediction model is evaluated by the mean square error (MSE), and its formula is: Among them, I i represents the actual inventory demand, represents the predicted inventory demand, and n represents the number of samples.

9. The warehouse management method based on dynamic inventory prediction according to claim 4, wherein The generation process of the inventory optimization strategy also includes considering the impact of seasonal factors and emergencies, and its formula is: I t+1 = f(S t , D t , C t , t ) + δ·S 季节性 + η·E 突发事件 Among them, represents the seasonal factor, E 突发事件 represents the impact of unexpected events, and δ and η are the weight coefficients of the seasonal factor and the unexpected event respectively.

10. The warehouse management method based on dynamic inventory prediction according to claim 4, wherein The execution process of the warehouse operation instructions also includes a real-time feedback mechanism, and its formula is: Among them, ΔI t represents the deviation between the actual inventory and the predicted inventory, and I t represents the actual inventory, represents the predicted inventory. According to the deviation result, the inventory optimization strategy is dynamically adjusted.

Images