Supply chain warehouse management system and method based on AI digital intelligence

Abstract

The invention discloses a supply chain warehouse management system and method based on AI digital intelligence, and relates to the technical field of logistics management, comprising the combination of the Internet of Things, the virtual reality technology and the digital twinning technology, and used for realizing the mapping of an entity warehouse full life cycle process; and constructing a virtual production board, analyzing warehouse-in and warehouse-out probabilities, optimizing job scheduling and carrying out production line deduction planning. Aiming at the problem that the operation efficiency is greatly reduced due to equipment fault troubleshooting errors or failures, the equipment fault can be rapidly positioned and solved through an accurate fault monitoring and intelligent troubleshooting mechanism, the influence of local abnormity on the overall operation efficiency is reduced to the maximum extent, efficient and stable operation of warehousing operation is ensured, and the working efficiency is improved. The problem that the management cost is increased due to the fact that existing stereoscopic warehouse management, production, warehouse and other data are scattered and lack of a unified supervision system is solved, centralized and standardized overall management is achieved by integrating and uniformly managing various data, and the management cost is remarkably reduced.

Description

Technical Field

[0001] The present invention relates to the field of logistics management technology, and in particular to a supply chain warehousing management system and method based on AI digitization. Background Art

[0002] In today's warehouse management field, with the widespread application of automated warehouses, a series of problems have gradually emerged. In modern automated warehouses, a large number of sorting, transportation, warehousing and other operations are carried out every day. However, once the equipment troubleshooting is wrong or fails, the local abnormal situation will quickly affect the whole, resulting in a significant reduction in operating efficiency.

[0003] At present, many high-rise warehouses have defects in management. Their management, production, warehouse and other related data are controlled by independent operating systems respectively, and there is a lack of a unified supervision system. This undoubtedly increases management costs and restricts the improvement of management efficiency.

[0004] The traditional production resource arrangement, verification, adjustment and control are obviously not intelligent enough and rely heavily on manual operation. This makes human factors have a great impact on the production efficiency of the production line, making it difficult to achieve efficient and accurate production management.

[0005] In the actual production process, due to the combined effects of multiple factors such as unreasonable warehouse design, abnormal equipment conditions, frequent switching of multiple tasks, high-frequency entry and exit of multiple tools, and frequent rotation of storage locations, each sub-module often fails to achieve the expected production efficiency. This makes it extremely difficult to accurately analyze the efficiency performance of each sub-module in the actual production process and accurately locate inefficient modules.

[0006] In addition, the traditional marketing service model mainly focuses on the supervision of the quality of its own products, which cannot achieve effective linkage between the supply side and the demand side and lacks systematic management. This model leads to narrow product sales channels and high after-sales costs, making it difficult to meet the diversified needs of the market and the requirements of efficient operation. Summary of the invention

[0007] The purpose of the present invention is to provide a supply chain warehouse management system and method based on AI digitization to solve at least any one of the problems raised in the above background technology.

[0008] To achieve the above-mentioned purpose, the present invention provides the following technical solutions: a supply chain warehouse management system and method based on AI digitization, including a combination of the Internet of Things, virtual reality technology and digital twin technology, for realizing the mapping of the entire life cycle process of physical warehousing;

[0009] 3D computer vision technology, deployed on high-computing edge computing platforms, for cargo detection and spatial perception;

[0010] The application of deep learning and other machine learning techniques to analyze people, objects, and the interactions between people and objects;

[0011] Multi-dimensional sensing technology and AI digital supply chain technology are used to restore offline physical spaces to cloud systems in real time;

[0012] Match the best algorithm to automatically perform data preparation, analysis, and fusion, and conduct in-depth mining of twin data.

[0013] Preferably, it also includes an ARIMA model for real-time monitoring and prediction of cargo flow, and a linear programming method for optimizing warehouse layout and inventory management.

[0014] Preferably, multi-class logistic regression and K-means clustering algorithms are also included to improve the accuracy and efficiency of classification and clustering tasks in warehousing operations.

[0015] Preferably, the digital twin technology is used to improve the value of data and the responsiveness and accuracy of various services.

[0016] Preferably, it includes deploying a system that combines the Internet of Things, virtual reality technology and digital twin technology to utilize data such as physical models, sensors and operation history to achieve full life cycle process mapping of physical warehousing.

[0017] Preferably, a vertical warehouse business production dashboard is constructed in the virtual space, integrating the vertical warehouse area business data, equipment operation and maintenance information, ledger information, three-dimensional models and other data. Through comprehensive management dashboards such as precise equipment positioning, real-time data viewing, panoramic detail display, and abnormal alarm prompts, the vertical warehouse line status is analyzed and displayed in the twin space, and differentiated and refined production planning strategies are compared to improve the efficiency of enterprise decision-making.

[0018] Preferably, based on the existing zoning of the three-dimensional warehouse, the probability map of inbound and outbound operations is analyzed and displayed, the distribution and frequency of inbound and outbound tasks are analyzed, and the available time, online, offline, fault and other operating states of the stacker are statistically analyzed, and early warnings are issued for the line and inventory, so as to further optimize the storage strategy of the three-dimensional warehouse, improve storage efficiency, and reduce inventory costs and operational risks.

[0019] Preferably, in the virtual digital twin system, the warehousing and outbound processes are analyzed, the stacker priority scheduling strategy and warehousing and outbound constraint principles are designed, and with the help of technologies such as genetic algorithms and neural network optimization algorithms, reasonable warehouse positions are automatically derived and the stacker is guided to move along the optimal path, thereby achieving optimal scheduling of multiple operation tasks of warehousing and outbound, and at the same time recording and tracing data from every link from the warehousing of raw materials to the delivery of products to control product quality.

[0020] Preferably, the warehousing in and out process of the vertical warehouse is simulated according to the design scale of the production line, pre-set production and key operating parameters, etc., and the production efficiency, equipment status and whether the operation process is abnormal are evaluated in real time during the entire simulation process, so as to further improve and re-verify the design plan, evaluate the best design plan, and dynamically plan the scale of the incremental vertical warehouse production line, so as to provide decision-level strategic support for the future addition of vertical warehouse production lines or planning.

[0021] The beneficial effects of the present invention are as follows:

[0022] In the present invention, firstly, in view of the problem that the operation efficiency is greatly reduced due to errors or failures in equipment fault troubleshooting, the present invention can quickly locate and solve equipment faults through accurate fault monitoring and intelligent troubleshooting mechanisms, minimize the impact of local abnormalities on overall operation efficiency, and ensure efficient and stable operation of warehousing operations;

[0023] Secondly, it solves the problem of scattered data on existing warehouse management, production, warehouses, etc. and lack of a unified supervision system, which leads to increased management costs. By integrating and unifying various types of data, it realizes centralized and standardized overall management, significantly reduces management costs, improves management efficiency and the scientific nature of decision-making;

[0024] Furthermore, it overcomes the defects of low intelligence level and great influence of human factors in traditional production resource arrangement, verification and adjustment control. By using advanced AI technology and intelligent algorithms, it realizes accurate arrangement, verification and automatic adjustment of production resources, reduces human interference and greatly improves the production efficiency of the production line.

[0025] At the same time, it can accurately analyze the efficiency performance of each sub-module in the actual production process, accurately identify and locate inefficient modules, and use multi-dimensional sensing technology and digital twin technology to comprehensively simulate and monitor the actual operation status of the warehouse in real time, providing strong support for optimizing the production process and improving overall efficiency.

[0026] In addition, it breaks the limitations of the traditional marketing service model in which the supply side and the demand side cannot be linked and lack systematic management, and realizes the effective coordination of product quality supervision, supply-side resource allocation and demand-side market dynamics, broadens product sales channels, reduces after-sales costs, and improves customer satisfaction and market competitiveness;

[0027] In summary, the present invention effectively solves many existing problems in the current warehouse management field, significantly improves the efficiency, accuracy and intelligence level of warehouse management, and brings significant economic benefits and competitive advantages to enterprises. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 It is a schematic diagram of the supply chain warehousing management method of the present invention;

[0029] Figure 2 This is a schematic diagram of the steps of the supply chain warehousing management method of the present invention;

[0030] Figure 3 It is a partial schematic diagram of the supply chain warehouse management system of the present invention. DETAILED DESCRIPTION

[0031] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0032] The present invention provides Figure 1-3 The following is a supply chain warehouse management system and method based on AI digitalization, where the system part:

[0033] It includes the combination of the Internet of Things, virtual reality technology and digital twin technology to achieve mapping of the entire life cycle of physical warehousing; three-dimensional computer vision technology, deployed on a high-computing-powerful edge computing platform for cargo detection and spatial perception; the application of deep learning and other machine learning technologies to analyze the interaction between people, objects and between people and objects; multi-dimensional perception technology and AI digital supply chain technology to restore offline physical spaces to cloud systems in real time; matching the best algorithm to automatically perform data preparation, analysis and fusion, and conduct in-depth mining of twin data. It also includes the ARIMA model for real-time monitoring and prediction of cargo flow, and a linear programming method for optimizing warehouse layout and inventory management. It also includes multi-category logistic regression and K-means clustering algorithms to improve the accuracy and efficiency of classification and clustering tasks in warehousing operations. The digital twin technology is used to improve the value of data and the responsiveness and accuracy of various services.

[0034] The following embodiments are included:

[0035] Innovatively integrate digitalization and hardware to create a new digital warehouse management system, which aims to effectively address the many pain points faced by enterprises in traditional warehouse management when it comes to warehouse management, procurement and inventory management;

[0036] In the actual operation of enterprise supply chain warehousing management, there are a lot of complex tasks that are difficult to perceive intuitively, such as procurement processes, goods in and out of the warehouse, storage space utilization, and procurement timeliness; this solution cleverly introduces digital twin technology that deeply integrates the Internet of Things and virtual reality technology into the warehousing scene. With the full use of rich information such as physical models, data collected by sensors, and operation history records, it integrates simulation processes covering multiple disciplines, multiple physical quantities, multiple scales, and multiple probabilities, and successfully builds an efficient interactive mapping between real space and virtual space, thereby accurately and comprehensively reflecting the corresponding physical warehouse from the beginning to the end of the entire life cycle process mapping;

[0037] The specific implementation of this system adopts cutting-edge 3D computer vision technology and has the powerful ability to perform accurate cargo detection on a high-computing edge computing platform. This visual system is equipped with a high-resolution imaging sensor and a unique and carefully optimized combination of hardware and software. Its AI algorithm is deeply optimized for NVID IA GPU and is tailored for edge computer vision reasoning. Through advanced 2D image processing, accurate object recognition and efficient object detection technology, it achieves excellent spatial perception capabilities and can accurately judge the user's full-process movement trajectory and collection behavior from entering the warehouse to leaving the warehouse in the application scenario;

[0038] By using various machine learning technologies, represented by deep learning, and restoring the physical world to the computer system using 3D reconstruction technology, we can deeply analyze people, objects, and the interaction between people and objects. With the help of the powerful computing power of computers, we can deeply understand this behavior and accurately grasp the key elements such as the category, quantity, and time dimension of the goods.

[0039] Through the exquisite combination of this multi-dimensional perception technology and AI digital supply chain technology, the offline physical space is fully, accurately and in real time restored and perfectly presented in the cloud system, namely the multi-dimensional perception digital twin technology;

[0040] The ARI MA model includes the following formula:

[0041]

[0042] Y t : represents the cargo flow value at time t, which is the target value to be predicted or analyzed;

[0043] μ: mean term, which is the constant term of the model, indicating that when all other terms are 0, Y t average level;

[0044] Autoregressive coefficient, i represents the order of lag, Indicates the current value Y t and the value Y at the past p moments t-i The linear relationship between each Y t-i Multiply by the corresponding autoregressive coefficient And sum; Y t-i : The cargo flow value at the i-th moment in the past;

[0045] p: autoregressive order, which determines how many past values ​​the model considers to influence the current value;

[0046] θ j : Moving average coefficient, j represents the order of lag, Represents the current random interference (error term)∈ t Linear relationship with random interference at the past q moments;

[0047] ∈ t-j : Random interference (error term) at the jth moment in the past;

[0048] q: Moving average order, which determines how many past moments of random interference the model considers to affect the current random interference;

[0049] ∈ t : White noise sequence, that is, a random variable sequence with a mean of 0 and a constant variance, represents random interference or new information at time t.

[0050] In summary, this formula comprehensively considers historical data, random interference and their lagged effects to predict the cargo flow value at the current moment by continuously updating data and reestimating model parameters to improve the accuracy and adaptability of the prediction.

[0051] Multi-class logistic regression and K-means clustering algorithms include:

[0052] Multi-class logistic regression is a statistical model used to handle classification problems with multiple categories. In warehousing operations, it can be used for:

[0053] Cargo classification: Cargo is accurately classified into different categories based on its various characteristics, such as size, weight, value, shelf life, etc., so as to enable targeted management and operation.

[0054] Predict demand categories: Based on historical data and related factors, predict the demand categories for various types of goods in different time periods to help optimize inventory allocation.

[0055] Its mathematical principle is based on establishing a linear relationship between each category and the input feature, and converting the linear output into a probability distribution through a specific function to determine the probability that the sample belongs to each category; K is the number of categories. For a given input feature X (which can be a vector of multiple features), the model's predicted probability P(Y=k|X) can be calculated by the following formula where k = 1, 2, ..., K, β k is the parameter vector associated with class k.

[0056] K-means clustering is an unsupervised learning algorithm used to group data points into K different clusters; similar to: Warehouse partitioning: divide the warehouse space into different areas according to the flow frequency of goods, storage requirements, etc. to improve storage and retrieval efficiency;

[0057] Customer grouping: Divide customers into different groups based on their purchasing behavior, order patterns, etc., so as to formulate personalized service strategies;

[0058] The working principle is as follows: first, randomly select K initial center points and then assign them to the nearest cluster according to the distance between the data points and the center points. Then recalculate the center point of each cluster and repeat this process until the center point no longer changes significantly or reaches the specified number of iterations.

[0059] Including the following Euclidean distance formula Where x and y are two data points, and n is the number of features. By integrating multi-class logistic regression and K-means clustering algorithms into the system, we can give full play to their advantages in classification and clustering tasks, significantly improve the accuracy and efficiency of warehousing operations, and achieve more optimized warehouse management.

[0060] In summary, the combination of multi-class logistic regression and K-means clustering algorithms can give full play to their respective advantages. Multi-class logistic regression accurately classifies goods based on multiple characteristics, while K-means clustering groups goods according to their intrinsic similarities. The combination of the two can achieve a more detailed and accurate division of goods, which is helpful to formulate targeted warehousing strategies, such as optimizing the storage layout of goods, accurately predicting demand categories, and reasonably arranging inventory configurations.

[0061] Methods:

[0062] This includes deploying a system that combines the Internet of Things, virtual reality technology, and digital twin technology to map the entire life cycle of physical warehousing using data such as physical models, sensors, and operational history;

[0063] Construct a production dashboard for vertical warehouse business in the virtual space, integrating vertical warehouse area business data, equipment operation and maintenance information, ledger information, 3D models and other data. Through comprehensive management dashboards such as precise equipment positioning, real-time data viewing, panoramic detail display, and abnormal alarm prompts, analyze and display the status of the vertical warehouse line in the twin space, compare differentiated and refined production planning strategies, and improve enterprise decision-making efficiency.

[0064] Based on the existing zoning of the stereoscopic warehouse, the probability map of inbound and outbound operations is analyzed and displayed, the distribution and frequency of inbound and outbound tasks are analyzed, and the operation status of the stacker, such as available time, online, offline, and faults, is statistically analyzed. Early warnings are issued for the line and inventory, and the storage strategy of the stereoscopic warehouse is further optimized to improve storage efficiency and reduce inventory costs and operational risks.

[0065] In the virtual digital twin system, the warehousing and outbound processes are analyzed, and the priority scheduling strategy and warehousing and outbound constraints of the stacker are designed. With the help of technologies such as genetic algorithms and neural network optimization algorithms, the reasonable warehouse positions are automatically obtained and the stacker is guided to move along the optimal path, realizing the optimal scheduling of multiple tasks of warehousing and inbound. At the same time, the data of every link from the warehousing of raw materials to the outbound of products is recorded and traced to control product quality.

[0066] The vertical warehouse in and out process is simulated according to the production line design scale, pre-set production and key operating parameters, and the production efficiency, equipment status and whether the operation process is abnormal are evaluated in real time during the entire simulation process, so as to further improve and re-verify the design plan, evaluate the best design plan, and dynamically plan the scale of the incremental vertical warehouse production line, providing decision-level strategic support for future new vertical warehouse production lines or planning.

[0067] This AI-based supply chain warehouse management method combines the above-mentioned AI-based supply chain warehouse management system with a system that combines the Internet of Things, virtual reality technology, and digital twin technology, bringing significant beneficial effects;

[0068] The application of digital twin technology has achieved accurate mapping of the entire life cycle of physical warehousing, greatly improving the scientific nature and foresight of warehousing management. For example, in the warehousing planning stage, digital twin technology can simulate the warehousing operation effects under different layouts and equipment configurations in advance, so as to select the optimal solution and avoid resource waste and inefficiency caused by planning errors.

[0069] In daily operations, by building a production dashboard for the vertical warehouse business in the virtual space and integrating multi-dimensional data, managers can intuitively and comprehensively understand the status of the vertical warehouse line, quickly compare differentiated and refined production planning strategies, and significantly improve the efficiency of corporate decision-making. For example, a large manufacturing company can use this function to quickly adjust production plans to respond to changes in market demand and shorten production cycles;

[0070] The use of digital twin technology to analyze and display the probability map of inbound and outbound goods can accurately grasp the flow of goods and optimize storage strategies;

[0071] In terms of job scheduling, the algorithms in the digital twin system can independently derive reasonable storage locations and optimal paths, achieving efficient and optimized scheduling of multiple job tasks; and for production line simulation, digital twin technology can evaluate and predict production conditions in real time, providing strong support for design improvements and production line scale planning. In summary, the application of digital twin technology in this supply chain warehousing management method has comprehensively improved the efficiency, quality, and benefits of warehousing management.

[0072] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A supply chain warehouse management system based on AI digitalization, characterized by: It includes the combination of the Internet of Things, virtual reality technology and digital twin technology to achieve the mapping of the entire life cycle process of physical warehousing; 3D computer vision technology, deployed on high-computing edge computing platforms, for cargo detection and spatial perception; The application of deep learning and other machine learning techniques to analyze people, objects, and the interactions between people and objects; Multi-dimensional sensing technology and AI digital supply chain technology are used to restore offline physical spaces to cloud systems in real time; Match the best algorithm to automatically perform data preparation, analysis, and fusion, and conduct in-depth mining of twin data.

2. The AI-based supply chain warehouse management system according to claim 1, characterized in that: Also included are ARIMA models for real-time monitoring and forecasting of goods flows, and linear programming methods for optimizing warehouse layout and inventory management.

3. The supply chain warehouse management system based on AI digitalization according to claim 1 is characterized by: It also includes multi-class logistic regression and K-means clustering algorithms to improve the accuracy and efficiency of classification and clustering tasks in warehousing operations.

4. The AI-based supply chain warehouse management system according to claim 1, characterized in that: The digital twin technology is used to improve the value of data and the responsiveness and accuracy of various services.

5. A supply chain warehousing management method based on AI digitalization, characterized by: This includes deploying a system that combines the Internet of Things, virtual reality technology and digital twin technology to use data such as physical models, sensors and operation history to achieve full life cycle process mapping of physical warehousing.

6. The supply chain warehousing management method based on AI digitalization according to claim 5 is characterized by: Construct a vertical warehouse business production dashboard in the virtual space, integrate the vertical warehouse area business data, equipment operation and maintenance information, ledger information, three-dimensional models and other data, and use comprehensive management dashboards such as precise equipment positioning, real-time data viewing, panoramic detail display, and abnormal alarm prompts to analyze and display the status of the vertical warehouse line in the twin space, compare differentiated and refined production planning strategies, and improve the efficiency of enterprise decision-making.

7. The supply chain warehousing management method based on AI digitalization according to claim 5 is characterized by: Based on the existing zoning of the three-dimensional warehouse, the probability map of inbound and outbound operations is analyzed and displayed, the distribution and frequency of inbound and outbound tasks are analyzed, and the operating status of the stacker, such as available time, online, offline, and faults, is statistically analyzed. Early warnings are issued for the line and inventory, and the storage strategy of the three-dimensional warehouse is further optimized to improve storage efficiency and reduce inventory costs and operational risks.

8. The supply chain warehousing management method based on AI digitalization according to claim 5 is characterized by: In the virtual digital twin system, the warehousing and outbound processes are analyzed, and the stacker priority scheduling strategy and warehousing and outbound constraint principles are designed. With the help of technologies such as genetic algorithms and neural network optimization algorithms, reasonable storage positions are automatically derived and the stacker is guided to move along the optimal path, realizing optimized scheduling of multiple warehousing and outbound tasks. At the same time, data from every link from the warehousing of raw materials to the delivery of products is recorded and traced to control product quality.

9. The supply chain warehousing management method based on AI digitalization according to claim 5 is characterized by: The vertical warehouse in and out process is simulated according to the production line design scale, pre-set production and key operating parameters, and the production efficiency, equipment status and whether the operation process is abnormal are evaluated in real time during the entire simulation process, so as to further improve and re-verify the design plan, evaluate the best design plan, and dynamically plan the scale of the incremental vertical warehouse production line, providing decision-level strategic support for future new vertical warehouse production lines or planning.

Images