Intelligent warehousing system and method based on supply chain management

Abstract

The present invention relates to the field of data analysis and processing technology, and in particular to an intelligent warehousing system and method based on supply chain management. First, the original data is obtained and transformed to obtain market trend data and extract market trend data features, and a time series decomposition method is introduced to perform trend prediction through a deep learning model. Seasonal adjustments are performed through high-order Fourier transforms, and an integrated learning model is introduced to generate predictions for future inventory demand; secondly, a real-time data processing framework and an adaptive data analysis model based on edge computing are introduced, compatibility and conflict resolution mechanisms are designed, modular design and plug-in architecture are adopted, and a comprehensive maintenance tool is introduced. This solves the problem that inventory prediction in the prior art relies on traditional prediction methods, cannot accurately process complex market data, resulting in inaccurate inventory predictions and excess or shortages of inventory; and limits the ability to extract useful information from complex data, resulting in poor system performance.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis and processing, and in particular to an intelligent warehousing system and method based on supply chain management. Background Art

[0002] In modern supply chain management, with the increasing diversification of market demand and the acceleration of globalization, the complexity of the supply chain has increased significantly, among which inventory management has become a core challenge. Proper inventory management is not only related to cost control, but also directly affects customer satisfaction and the continuity of the supply chain. Excessive inventory may lead to waste of resources and backlog of funds, while insufficient inventory may cause supply chain disruptions and affect customer service quality.

[0003] In addition, the rapid development of technology, especially the advancement of big data, artificial intelligence, the Internet of Things and cloud computing, has provided new optimization methods for supply chain management. These technologies make supply chain management more intelligent, can effectively process and analyze massive data, and improve the accuracy of decision-making and the efficiency of operations.

[0004] my country's patent application number: CN202310971577.6, publication date: 2023.11.03, discloses a management method and system for supply chain warehousing goods supply, including the following steps: receiving an uploaded purchase order, analyzing the purchase information of the purchase order, the purchase information including the purchase target name and purchase target quantity; matching the purchase target quantity of the goods supply demand according to the expected inventory of the goods in the current period to obtain the goods surplus information, and generating a goods scheduling plan; according to the goods scheduling plan, managing the supply and transportation of goods through the supply chain; according to the exponential smoothing algorithm module, performing an exponential smoothing forecast for the next period of time on the obtained scheduling quantity and forecast quantity of the purchase target of the current period, obtaining the forecasted expected inventory demand information of the goods, and generating an inventory scheduling plan. This invention collects the goods demand required by the purchaser and the expected inventory of the supply chain warehousing, specifies the goods transportation plan according to the goods demand, and sends the goods directly to the purchaser.

[0005] However, the above technology has at least the following technical problems: inventory forecasting in existing technologies usually relies on traditional forecasting methods, which cannot accurately process complex market data, resulting in inaccurate inventory forecasting and may cause problems such as excess or shortage of inventory; it limits the ability to extract useful information from complex data, thereby affecting the accuracy and efficiency of decision-making, resulting in poor overall system performance and difficulty in achieving efficient data processing and information sharing. Summary of the invention

[0006] The embodiments of the present invention provide an intelligent warehousing system and method based on supply chain management, which solves the problem that inventory forecasting in the prior art usually relies on traditional forecasting methods and cannot accurately process complex market data, resulting in inaccurate inventory forecasting, which may cause excess or shortage of inventory; it limits the ability to extract useful information from complex data, thereby affecting the accuracy and efficiency of decision-making, resulting in poor overall system performance and difficulty in achieving efficient data processing and information sharing. An efficient, accurate and flexible supply chain management intelligent warehousing system is implemented, which significantly improves the accuracy and response speed of inventory forecasting through advanced data processing and analysis technology, while reducing the complexity and cost of system maintenance.

[0007] The present invention provides an intelligent warehousing system and method based on supply chain management, which specifically includes the following technical solutions:

[0008] The intelligent warehousing system based on supply chain management includes the following parts:

[0009] Data acquisition module, data transformation module, feature extraction module, decomposition module, trend prediction module, seasonal adjustment module, prediction module;

[0010] The data acquisition module is used to collect raw data from the supply chain management system and market research, including inventory levels, transportation and distribution, order processing, supplier performance, customer needs and feedback, and external environment; the data acquisition module is connected to the data transformation module through data transmission;

[0011] The data transformation module is used to transform the original data to obtain market trend data; the data transformation module is connected to the feature extraction module by means of data transmission;

[0012] The feature extraction module is used to extract features from market trend data based on natural language processing, principal component analysis or cluster analysis technology; the feature extraction module is connected to the decomposition module and the trend prediction module by means of data transmission;

[0013] The decomposition module is used to distinguish trends, seasonal patterns and random fluctuations in the data by using a time series decomposition method; the decomposition module is connected to the trend prediction module and the seasonal adjustment module by means of data transmission;

[0014] The trend prediction module is used to use a deep learning model, including a long short-term memory network, to perform trend prediction based on the output of the decomposition module and the input of the feature extraction module; the trend prediction module is connected to the prediction module by means of data transmission;

[0015] The seasonal adjustment module is used to perform seasonal adjustment through high-order Fourier transform; the seasonal adjustment module is connected to the prediction module through data transmission;

[0016] The forecasting module is used to generate a forecast of future inventory demand by integrating the influence of trend forecasting, seasonal adjustment and residuals.

[0017] The intelligent warehousing method based on supply chain management includes the following steps:

[0018] S100: Obtain and transform the original data, obtain market trend data and extract market trend data features, introduce time series decomposition method, perform trend forecasting through deep learning model, perform seasonal adjustment through high-order Fourier transform, introduce integrated learning model, and generate forecasts for future inventory demand;

[0019] S200: Introduce a real-time data processing framework and adaptive data analysis model based on edge computing, design compatibility and conflict resolution mechanisms, discover and resolve conflicts between data processing logics in real time, and ensure the stable operation of the system.

[0020] Preferably, the S100 specifically includes:

[0021] Obtain raw data from supply chain management systems and market research, including inventory level data, transportation and distribution data, order processing data, supplier performance data, customer demand and feedback data, and external environment data; transform the raw data, the specific formula is:

[0022]

[0023] Among them, X′ represents the transformed data, that is, the market trend data, X represents the original data, including historical transportation volume, cargo arrival time or cargo circulation speed; μ and σ are the mean and variance of the logarithmic transformed data; λ is the adjustment parameter; ∈ is a very small number used to avoid taking the logarithm of zero; tan -1 (X) is the inverse tangent transform.

[0024] Preferably, the S100 further includes:

[0025] Use principal component analysis or cluster analysis to extract features from market trend data; the extracted features include the type of goods, destination, seasonal demand fluctuations, and output a transformed feature matrix.

[0026] Preferably, the S100 further includes:

[0027] Time series decomposition is used to distinguish trends, seasonal patterns and random fluctuations in the data; the output of time series decomposition will reflect the information of the feature matrix.

[0028] Preferably, the S100 further includes:

[0029] In the trend prediction stage, the deep learning model uses the features in the feature matrix to predict future logistics trends; in the seasonal adjustment stage, the high-order Fourier transform uses the features in the feature matrix to adjust the seasonal pattern.

[0030] Preferably, the S200 specifically includes:

[0031] A real-time data processing framework based on edge computing is introduced to perform data processing close to the data source; an adaptive data analysis model is designed, which combines historical and real-time data to automatically adjust analysis parameters.

[0032] Preferably, the S200 further includes:

[0033] Design compatibility and conflict resolution mechanisms to adjust and optimize data flows; increase modular design and plug-in architecture to adapt to different business needs and market changes; compatibility and conflict resolution mechanisms define a set of unified interface standards and data exchange formats, and all modules and plug-ins must follow this set of standards at the beginning of design to ensure the consistency and accuracy of data flowing within the system; compatibility and conflict resolution mechanisms reduce the degree of coupling between various parts of the system to achieve true modularization. In addition, by introducing a central control unit to continuously monitor data flows and processing processes, conflicts between data processing logics can be discovered and resolved in real time to ensure the stable operation of the system.

[0034] Beneficial effects:

[0035] The multiple technical solutions provided in the embodiments of the present invention have at least the following technical effects or advantages:

[0036] 1. By using prediction algorithms, combined with time series analysis, deep learning and natural language processing techniques, the present invention can accurately predict inventory demand and optimize inventory levels; this not only reduces the risk of inventory backlogs, but also ensures the timely supply of goods in the supply chain, improving the overall efficiency and accuracy of inventory management;

[0037] 2. By introducing a real-time data processing framework based on edge computing, the present invention can quickly respond to changes in the market and supply chain and reduce data processing delays; enabling the warehousing system to more agilely adapt to fluctuations in market demand, thereby improving the overall supply chain response speed;

[0038] 3. The application of adaptive data analysis models enables the system to combine historical and real-time data and automatically adjust analysis parameters, thus improving the accuracy and efficiency of data analysis. This precise data analysis is essential for developing effective inventory strategies.

[0039] 4. By designing compatibility and conflict resolution mechanisms, the present invention ensures that different technical features and system components can work together effectively, reduces the risk of system failure, and enhances the stability and reliability of the system; this design strategy provides long-term sustainable development capabilities and ensures the long-term competitiveness of the system.

[0040] 5. The technical solution of the present invention can effectively solve the problem that inventory forecasting in the prior art usually relies on traditional forecasting methods and cannot accurately process complex market data, resulting in inaccurate inventory forecasting, which may cause problems such as excess or shortage of inventory; it limits the ability to extract useful information from complex data, thereby affecting the accuracy and efficiency of decision-making, resulting in poor overall system performance and difficulty in achieving efficient data processing and information sharing. It realizes an efficient, accurate and flexible supply chain management intelligent warehousing system, which significantly improves the accuracy and response speed of inventory forecasting through advanced data processing and analysis technology, while reducing the complexity and cost of system maintenance. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 This is a structural diagram of the intelligent warehousing system based on supply chain management described in the present invention;

[0042] Figure 2 This is a flow chart of the intelligent warehousing method based on supply chain management described in the present invention. DETAILED DESCRIPTION

[0043] The embodiments of the present invention provide an intelligent warehousing system and method based on supply chain management, thereby solving the problem that inventory forecasting in the prior art usually relies on traditional forecasting methods and cannot accurately process complex market data, resulting in inaccurate inventory forecasting, which may cause problems such as excess or shortage of inventory; it limits the ability to extract useful information from complex data, thereby affecting the accuracy and efficiency of decision-making, resulting in poor overall system performance, and making it difficult to achieve efficient data processing and information sharing.

[0044] The technical solution in the embodiment of the present invention is to solve the above problems, and the overall idea is as follows:

[0045] By using prediction algorithms combined with time series analysis, deep learning and natural language processing technologies, we can accurately predict inventory demand and optimize inventory levels. This not only reduces the risk of inventory backlogs, but also ensures the timely supply of goods in the supply chain, improving the overall efficiency and accuracy of inventory management. By introducing a real-time data processing framework based on edge computing, we can quickly respond to changes in the market and supply chain and reduce data processing delays. This enables the warehousing system to adapt to fluctuations in market demand more agilely, thereby improving the overall supply chain response speed. The application of adaptive data analysis models enables the system to combine historical and real-time data and automatically adjust analysis parameters, improving the accuracy and efficiency of data analysis. This precise data analysis is essential for formulating effective inventory strategies. By designing compatibility and conflict resolution mechanisms, it ensures that different technical features and system components can work together effectively, reducing the risk of system failures and enhancing the stability and reliability of the system. This design strategy provides long-term sustainable development capabilities and ensures the long-term competitiveness of the system.

[0046] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0047] Refer to the attached Figure 1 The intelligent warehousing system based on supply chain management described in the present invention includes the following parts:

[0048] Data acquisition module, data transformation module, feature extraction module, decomposition module, trend prediction module, seasonal adjustment module, prediction module;

[0049] The data acquisition module is used to collect raw data from the supply chain management system and market research, including inventory levels, transportation and distribution, order processing, supplier performance, customer needs and feedback, external environment, etc. The data acquisition module is connected to the data transformation module by data transmission;

[0050] The data transformation module is used to transform the original data for further analysis; the data transformation module is connected to the feature extraction module by means of data transmission;

[0051] The feature extraction module is used to extract key features from the transformed market trend data using natural language processing, principal component analysis or cluster analysis. The feature extraction module is connected to the decomposition module and the trend prediction module by means of data transmission.

[0052] The decomposition module is used to distinguish the main trends, seasonal patterns and random fluctuations in the data by using the time series decomposition method; the decomposition module is connected to the trend prediction module and the seasonal adjustment module by means of data transmission;

[0053] The trend prediction module is used to use a deep learning model, such as a long short-term memory network, to perform trend prediction based on the output of the decomposition module and the input of the feature extraction module; the trend prediction module is connected to the prediction module by means of data transmission;

[0054] The seasonal adjustment module is used to perform seasonal adjustment through high-order Fourier transform; the seasonal adjustment module is connected to the prediction module through data transmission;

[0055] The forecasting module is used to integrate the impact of trend forecasting, seasonal adjustment and residuals to generate accurate forecasts of future inventory demand.

[0056] Refer to the attached Figure 2 The intelligent warehousing method based on supply chain management described in the present invention comprises the following steps:

[0057] S100: Obtain and transform the original data, obtain market trend data and extract market trend data features, introduce time series decomposition method, perform trend forecasting through deep learning model, perform seasonal adjustment through high-order Fourier transform, introduce integrated learning model, and generate forecasts for future inventory demand;

[0058] In order to solve the problem of inventory optimization in the warehousing system, accurately predict product demand and effectively adjust inventory, a prediction algorithm is designed to provide accurate inventory demand forecasts through in-depth analysis of raw data. Combined with data processing technology, machine learning models and mathematical analysis methods, it can achieve efficient prediction of inventory demand dynamics in the supply chain. In this way, the intelligent warehousing system can not only cope with the rapid changes in the market, but also prepare inventory in advance, thereby avoiding excess or shortage situations and optimizing the overall supply chain management process.

[0059] The data acquisition module obtains raw data from the supply chain management system and market research, including inventory level data, transportation and distribution data, order processing data, supplier performance data, customer demand and feedback data, and external environment data. Among them, the supplier performance data records the supplier's delivery time, delivery reliability, quality control, etc., which helps to evaluate the stability and risk of the supply chain; the customer demand and feedback data involves customer order preferences, complaint records, service evaluations, etc.; the external environment data refers to information such as economic indicators, traffic conditions, and weather conditions.

[0060] The data transformation module transforms the original data. The specific formula is:

[0061]

[0062] Among them, X′ represents the transformed data, that is, the market trend data, X represents the original data, such as historical transportation volume, cargo arrival time or cargo circulation speed, etc.; μ and σ are the mean and variance of the data after logarithmic transformation, which are used to standardize the data; λ is an adjustment parameter used to adjust the impact of the inverse tangent transformation; ∈ is a constant used to avoid taking the logarithm of zero, which increases the stability of the formula; tan -1 (X) is the inverse tangent transformation, which provides nonlinear processing capabilities. The above formula converts the original data into a form more suitable for complex analysis, handling extreme values ​​and skewed distributions.

[0063] Use natural language processing technology to deeply analyze market trend data. The feature extraction module uses principal component analysis or cluster analysis to extract key features that have a significant impact on inventory demand from market trend data. In the scenario of supply chain data, principal component analysis helps to reduce the dimension of the data and highlight the main factors affecting transportation efficiency, while cluster analysis can group similar transportation modes or types of goods, making data processing more targeted. The extracted features include the type of goods, destination, seasonal demand fluctuations, etc., and the transformed feature matrix F is output. The feature calculation formula is:

[0064]

[0065] Among them, M i is the i-th feature of the market trend data, ω i is the feature weight, which is used to indicate the importance of different features in prediction.

[0066] The decomposition module uses time series decomposition to distinguish the main trends, seasonal patterns and random fluctuations in the data, laying the foundation for more accurate predictions. The specific formula is:

[0067]

[0068] Among them, Y t Represents time series data, such as cargo transportation volume or transportation time; T t , S t , R t Represent the trend component, seasonal component and residual component respectively. The feature matrix F is used to assist in identifying and explaining the trend of the data T t and seasonal S t Components. For example, a feature such as the degree of congestion on a particular route (with a high weight in F) is significantly correlated with seasonal fluctuations in freight volume, and this association can be identified and exploited in time series analysis. The output of time series decomposition (trend and seasonal components) will reflect the key information in the feature matrix F, making trend and seasonal analysis more targeted to the actual logistics situation.

[0069] The trend prediction module uses deep learning models, such as long short-term memory networks, to perform trend prediction. The specific formula is:

[0070]

[0071] Among them, h t-1 is the hidden state of the previous time step, representing historical information; x t is the input at the current time t; W h is the weight of the hidden state, W x is the weight of the input data, b is the bias term; σ is the activation function, which is used to introduce nonlinearity. The LSTM network model uses its long-term dependencies on time series data to model and improve the accuracy of trend forecasting. In the trend forecasting stage, the deep learning model can use the features in F to more accurately predict future logistics trends. For example, if the diversity of cargo types (a feature) is important in predicting future demand, this can be used to adjust the parameters of the deep learning model to take this factor into account when predicting future demand.

[0072] The seasonal adjustment module uses high-order Fourier transform to perform seasonal adjustment. The specific formula is:

[0073]

[0074] Among them, S t represents the seasonally adjusted output, k is the index, A k and B k are the amplitudes, representing the strength of the cosine and sine terms respectively; f k is the frequency, which determines the period of the seasonal pattern; φ k and ψ k is the phase, which determines the starting point of the cyclical pattern. The purpose of seasonal adjustment is to capture seasonal fluctuations in sales data, such as the impact of holidays or seasonal changes. During the seasonal adjustment stage, the Fourier transform can use the features in the feature matrix F to accurately adjust the seasonal pattern. This means that the model is not only based on seasonal fluctuations in historical data, but also considers the impact of features such as market trends or special events on seasonal patterns.

[0075] The forecasting module combines the impact of trend forecasting, seasonal adjustment and residuals to generate accurate forecasts of future inventory demand. An integrated learning model is introduced with the formula:

[0076]

[0077] in, is the final predicted inventory demand, α, γ, and δ are weights determined through training data, representing the relative importance of different components in the final prediction.

[0078] When the above method is applied to an intelligent warehousing system based on supply chain management, it can accurately predict transportation demand in different periods, helping logistics companies optimize transportation planning and inventory management, reduce transportation costs, and improve efficiency. For example, it can predict transportation peaks before and after holidays, allowing companies to allocate resources in advance, or identify low-demand periods to reduce unnecessary transportation and storage costs. Through accurate predictions, logistics companies can better manage vehicle scheduling, reduce empty trips, and optimize route planning, thereby improving overall logistics efficiency and customer satisfaction.

[0079] S200: Introduce a real-time data processing framework and adaptive data analysis model based on edge computing, design compatibility and conflict resolution mechanisms, discover and resolve conflicts between data processing logics in real time, and ensure the stable operation of the system.

[0080] In order to avoid the problem of response delay when processing real-time data, the system introduces a real-time data processing framework based on edge computing. This framework can process data closer to the data source, reducing the time for data to be transmitted between the central server and the warehousing system, thereby speeding up the system's response. For example, if a sensor in the warehouse detects a sudden drop in the inventory of a certain product, the framework can immediately process this information and update the system without waiting for the data to be transmitted to the remote server.

[0081] First, determine the location to deploy the edge computing node, which can be inside the warehouse, distribution center, or other logistics nodes. The principle of selection is to be close to the data source to reduce data transmission time. Select appropriate hardware according to data processing requirements. Edge computing nodes are dedicated servers, industrial personal computers, or high-performance embedded devices. The operating system and necessary data processing software are deployed on the edge nodes. In addition, define data transmission and processing protocols to ensure that data is transmitted safely and efficiently between nodes. Run the prediction algorithm in real time on the edge node, deeply analyze the original data, and provide accurate inventory demand forecasts.

[0082] However, the real-time data processing framework introduces problems with the accuracy and efficiency of data analysis. To address this problem, the system designs an adaptive data analysis model. This model is able to combine historical and real-time data and automatically adjust its analysis parameters to improve the accuracy of data analysis. For example, if the system detects a sudden increase in the immediate demand for a certain commodity, the model will automatically adjust its forecast parameters to more accurately reflect this change.

[0083] Integrate real-time data from edge computing nodes and data from historical data storage systems, and select machine learning models that are suitable for time series data and have adaptive capabilities as adaptive data analysis models, such as recurrent neural networks (RNNs) or long short-term memory networks (LSTMs). Use historical data to train the model. Use adaptive gradient descent methods to dynamically adjust model parameters, such as learning rate, model weights, etc., according to real-time data streams. Implement online learning or incremental learning mechanisms so that the model can continuously learn from new data and adapt to changes in data patterns in a timely manner.

[0084] There are compatibility and collaboration issues between the adaptive data analysis model and the prediction algorithm and real-time data processing framework. To resolve this conflict, a compatibility and conflict resolution mechanism is designed in the system. The compatibility and conflict resolution mechanism adjusts and optimizes the data flow to achieve coordination and consistency between different features. For example, when the prediction algorithm and the real-time data processing framework conflict, the compatibility and conflict resolution mechanism will intervene to adjust their data flow and processing logic to ensure that the data processing of the entire system is both accurate and efficient.

[0085] Compatibility and conflict resolution mechanisms define a set of unified interface standards and data exchange formats. All modules and plug-ins must follow this set of standards at the beginning of their design to ensure the consistency and accuracy of data flowing within the system. Compatibility and conflict resolution mechanisms reduce the degree of coupling between the various parts of the system and achieve true modularity. In addition, by introducing a central control unit, the data flow and processing process are constantly monitored, and conflicts between data processing logics are discovered and resolved in real time to ensure the stable operation of the system. The central control unit uses advanced algorithms, including machine learning (ML) algorithms, genetic algorithms (GA), and reinforcement learning (RL) algorithms, to dynamically adjust conflict resolution strategies, enhance the system's adaptability and ability to flexibly respond to market changes.

[0086] As a specific example, taking the Q learning algorithm as an example, assume that the intelligent warehousing system needs to dynamically adjust the inventory level to optimize inventory costs and meet market demand. In this scenario, the definitions are as follows:

[0087] Status: can be composed of multiple factors, such as current inventory levels (low, medium, high), recent order trends (increasing, stable, decreasing), seasonal factors (peak season, off-season), etc.

[0088] Action A: includes increasing inventory, maintaining current levels, or reducing inventory.

[0089] Rewards: Depending on the goals of inventory management, rewards can be designed based on avoiding out-of-stocks, reducing excess inventory, and reducing inventory costs.

[0090] At the beginning, for each possible state of the warehousing system and each possible action in that state, a Q value is initialized, which is set to zero at the beginning, indicating that there is no prior knowledge about the utility of taking a specific action in a specific state. The current inventory status, order trends, and other relevant factors are monitored and evaluated in real time to determine which state it is currently in. Based on the current Q table, the system takes an action and selects the optimal action based on the ε-greedy strategy. In order to encourage exploration, the system sometimes randomly selects an action instead of always selecting the action with the highest Q value. Adjust the inventory level based on the selected action, and then observe the actual market response and inventory changes. This includes monitoring the completion of orders, changes in inventory, and changes in costs.

[0091] Based on the result of the action, the system receives a reward value. This reward value reflects the contribution of the action to the goal of inventory optimization. Then, the system uses the Q learning formula to update the Q table to improve future action decisions.

[0092]

[0093] Where Q(s,a) is the Q value of the current state-action pair, α is the learning rate, which controls how quickly new learning information overwrites old information; γ is the discount factor, which determines the current value of future rewards, and s ′ is the new state after s executes the action, a ′ is in the new state s ′ One of all possible actions that could be taken.

[0094] Repeat the above process, and adjust the action strategy based on the previous learning results in each iteration. Over time, the Q table will gradually stabilize, reflecting the optimal expected utility of taking each action in a given state.

[0095] Assume that at a certain moment, the system determines that the current state is "high inventory, stable order trend, peak season". In this state, the possible actions are "increase inventory", "maintain current level" and "reduce inventory". According to the Q table, the system decides that "reduce inventory" is the best choice because it can reduce the cost caused by excess inventory. After executing this action, the system will monitor the actual effect, including cost changes and the ability to meet orders, and then update the Q table based on the observed results.

[0096] In summary, the intelligent warehousing system and method based on supply chain management described in the present invention are completed.

[0097] The technical solutions in the above embodiments of the present invention have at least the following technical effects or advantages:

[0098] 1. By using prediction algorithms, combined with time series analysis, deep learning and natural language processing techniques, the present invention can accurately predict inventory demand and optimize inventory levels; this not only reduces the risk of inventory backlogs, but also ensures the timely supply of goods in the supply chain, improving the overall efficiency and accuracy of inventory management;

[0099] 2. By introducing a real-time data processing framework based on edge computing, the present invention can quickly respond to changes in the market and supply chain and reduce data processing delays; enabling the warehousing system to more agilely adapt to fluctuations in market demand, thereby improving the overall supply chain response speed;

[0100] 3. The application of adaptive data analysis models enables the system to combine historical and real-time data and automatically adjust analysis parameters, thus improving the accuracy and efficiency of data analysis. This precise data analysis is essential for developing effective inventory strategies.

[0101] 4. By designing compatibility and conflict resolution mechanisms, the present invention ensures that different technical features and system components can work together effectively, reduces the risk of system failure, and enhances the stability and reliability of the system; this design strategy provides long-term sustainable development capabilities and ensures the long-term competitiveness of the system.

[0102] Effect research:

[0103] The technical solution of the present invention can effectively solve the problem that inventory forecasting in the prior art usually relies on traditional forecasting methods and cannot accurately process complex market data, resulting in inaccurate inventory forecasting, which may cause problems such as excess or shortage of inventory; it limits the ability to extract useful information from complex data, thereby affecting the accuracy and efficiency of decision-making, resulting in poor overall system performance and difficulty in achieving efficient data processing and information sharing. In addition, the above-mentioned system or method has undergone a series of effect surveys and verifications, and finally realized an efficient, accurate and flexible supply chain management intelligent warehousing system. Through advanced data processing and analysis technology, the accuracy and response speed of inventory forecasting have been significantly improved, while reducing the complexity and cost of system maintenance.

[0104] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0105] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0106] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0107] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. An intelligent warehousing method based on supply chain management, characterized in that: The following steps are involved: S100: Obtain the original data and transform the original data. The specific formula is: ， in, Represents the transformed data, i.e., market trend data, Represents raw data, including historical shipping volume, cargo arrival time, or cargo circulation speed; and is the mean and variance of the log-transformed data; is the adjustment parameter; is a very small number used to avoid taking the logarithm of zero; is the inverse tangent transformation; Obtain market trend data and extract market trend data features, introduce time series decomposition method, use deep learning model to make trend forecasts, and use high-order Fourier transform to make seasonal adjustments. The specific formula is: ， in, represents seasonally adjusted output; is the index, and are the amplitudes, representing the strength of the cosine and sine terms, respectively; is the frequency; and It is the phase; Indicates the current time point; An integrated learning model is introduced to generate a forecast of future inventory demand. The specific formula is: ， in, is the final forecasted inventory requirement; , and are weights determined by training data; and represent the trend component and the residual component respectively; S200: Introduce a real-time data processing framework and adaptive data analysis model based on edge computing, design compatibility and conflict resolution mechanisms, discover and resolve conflicts between data processing logics in real time, and ensure the stable operation of the system.

2. The intelligent warehousing method based on supply chain management according to claim 1 is characterized in that: The S100 further includes: Use principal component analysis or cluster analysis to extract features from market trend data; the extracted features include the type of goods, destination, seasonal demand fluctuations, and output a transformed feature matrix.

3. The intelligent warehousing method based on supply chain management according to claim 1 is characterized in that: The S100 further includes: Time series decomposition method is introduced to distinguish trends, seasonal patterns and random fluctuations in data; the output of time series decomposition will reflect the information of feature matrix.

4. The intelligent warehousing method based on supply chain management according to claim 1 is characterized in that: The S100 further includes: In the trend prediction stage, the deep learning model uses the features in the feature matrix to predict future logistics trends; in the seasonal adjustment stage, the high-order Fourier transform uses the features in the feature matrix to adjust the seasonal pattern.

5. The intelligent warehousing method based on supply chain management according to claim 1 is characterized in that: The S200 specifically includes: A real-time data processing framework based on edge computing is introduced to perform data processing close to the data source; an adaptive data analysis model is designed, which combines historical and real-time data to automatically adjust analysis parameters.

6. The intelligent warehousing method based on supply chain management according to claim 1 is characterized in that: The S200 further includes: Design compatibility and conflict resolution mechanisms to adjust and optimize data flows; increase modular design and plug-in architecture to adapt to different business needs and market changes; compatibility and conflict resolution mechanisms define a unified set of interface standards and data exchange formats, and all modules and plug-ins must follow this set of standards at the beginning of design to ensure the consistency and accuracy of data flowing within the system; compatibility and conflict resolution mechanisms reduce the degree of coupling between various parts of the system to achieve true modularity; in addition, by introducing a central control unit to continuously monitor data flows and processing processes, conflicts between data processing logics can be discovered and resolved in real time to ensure stable operation of the system.

7. An intelligent warehousing system based on supply chain management, applied to the intelligent warehousing method based on supply chain management as claimed in claim 1, characterized in that: Includes the following sections: Data acquisition module, data transformation module, feature extraction module, decomposition module, trend prediction module, seasonal adjustment module, prediction module; The data acquisition module is used to collect raw data from the supply chain management system and market research, including inventory levels, transportation and distribution, order processing, supplier performance, customer needs and feedback, and external environment; the data acquisition module is connected to the data transformation module through data transmission; The data transformation module is used to transform the original data to obtain market trend data; the data transformation module is connected to the feature extraction module by means of data transmission; The feature extraction module is used to extract features from market trend data based on natural language processing, principal component analysis or cluster analysis technology; feature The extraction module is connected with the decomposition module and the trend prediction module by means of data transmission; The decomposition module is used to distinguish trends, seasonal patterns and random fluctuations in the data using a time series decomposition method; The decomposition module is connected with the trend prediction module and the seasonal adjustment module through data transmission; The trend prediction module is used to use a deep learning model, including a long short-term memory network, to perform trend prediction based on the output of the decomposition module and the input of the feature extraction module; The trend prediction module is connected to the prediction module by means of data transmission; The seasonal adjustment module is used to perform seasonal adjustment through high-order Fourier transform; the seasonal adjustment module is connected to the prediction module through data transmission; The forecasting module is used to generate a forecast of future inventory demand by integrating the influence of trend forecasting, seasonal adjustment and residuals.

Images