Cargo loading information management method and system based on intelligent shelf

Abstract

The invention discloses a cargo loading information management method and system based on an intelligent shelf, and relates to the technical field of Internet of Things information management, and the method comprises the steps: collecting cargo warehousing information; uWB anchor points are deployed on the intelligent goods shelf, sensor signal features are collected, a support vector machine classification algorithm is used to identify NLoS conditions, and error correction is carried out; determining the distance between the cargo and the anchor point by using a three-dimensional positioning formula to obtain the three-dimensional coordinate of the cargo, and monitoring the safety state of the cargo in real time; historical demand data is collected, the cargo demand at the future moment is simulated and predicted by using Monte Carlo, and a replenishment plan is generated in combination with the Bayesian law to replenish the warehouse cargo inventory; and after the goods are delivered, a business report is generated. According to the invention, accurate HE efficient management of cargo information is realized, the safety, transparency and intelligent degree of the information management system are improved, optimization of warehouse management is realized, and the efficiency and safety of cargo management and the comprehensive efficiency of warehouse management are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of Internet of Things information management, and in particular to a method and system for managing cargo loading information based on smart shelves. Background Art

[0002] With the rapid development of intelligent warehousing technology, cargo information management has gradually shifted from traditional manual operations to automated and intelligent management modes. In the current market, warehousing management technologies based on the Internet of Things, blockchain, and big data analysis have not only promoted the improvement of cargo management efficiency, but also gradually improved the reliability and real-time nature of data. For example, the use of technologies such as RFID tags, sensor networks, and real-time positioning systems (RTLS) can realize the automatic collection and update of cargo in and out of the warehouse information. Especially with the introduction of blockchain technology, the transparency and security of cargo information have been effectively enhanced.

[0003] Although the existing smart shelf technology has improved the intelligence level of cargo management to a certain extent, there is still much room for improvement in precise positioning and inventory optimization. Traditional warehouse management systems mostly rely on fixed sensors and hardware equipment, and fail to fully utilize real-time positioning data for high-precision error correction. Especially in the real-time positioning and monitoring of cargo, the current solution is difficult to achieve accurate measurement of three-dimensional space, which in turn affects the real-time monitoring of the safety status of cargo. In addition, the existing system also has shortcomings in predicting cargo demand, and usually cannot effectively combine historical data for accurate prediction and inventory adjustment, which affects warehouse management efficiency and resource optimization allocation. Summary of the invention

[0004] In view of the problems existing in the above-mentioned existing method and system for managing cargo loading information based on smart shelves, the present invention is proposed.

[0005] Therefore, the problem to be solved by the present invention is that the existing cargo loading information management has deficiencies in positioning accuracy and cargo demand prediction.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: a method for managing loading cargo information based on smart shelves, comprising collecting cargo warehousing information, building a blockchain platform, and storing the collected cargo warehousing information in the blockchain platform;

[0007] Deploy UWB anchors on smart shelves, collect sensor signal features based on cargo RFID tags and smart shelf UWB anchors, use support vector machine classification algorithm to identify NLoS conditions, and perform error correction;

[0008] Based on the error correction results, the ToA algorithm is used to measure the signal flight time, calculate the distance between the goods and the anchor point, and use the three-dimensional positioning formula to determine the three-dimensional coordinates of the goods. The sensor group deployed on the smart shelf and the three-dimensional coordinate data of the goods can be used to monitor the safety status of the goods in real time.

[0009] Collect historical demand data, use Monte Carlo simulation to predict future demand for goods, and combine Bayesian law to calculate the safety stock and inventory order point of goods, and generate a replenishment plan to replenish warehouse goods inventory;

[0010] After the goods leave the warehouse, inventory analysis is performed and business reports are generated.

[0011] As a preferred solution of the method for managing cargo loading information based on smart shelves of the present invention, wherein: the collecting cargo warehousing information, building a blockchain platform, and storing the collected cargo warehousing information in the blockchain platform include:

[0012] Before the goods are put into storage, the operator uses an RFID writing device to set an RFID tag for the goods and uses an RFID card reader to activate the RFID tag, scan its stored information, read the RFID tag information and transmit it to the warehouse central control system through a wireless sensor network. The RFID tag information refers to the goods ID, category, weight, volume and expiration date;

[0013] Based on blockchain technology, a blockchain platform is built to upload the cargo RFID tag information to the blockchain platform, and a hash algorithm is used to generate a hash value of the RFID tag information, which is then stored on the blockchain platform;

[0014] According to the cargo RFID tag information, the cargo information is automatically matched to the preset shelf space and the shelf location is marked.

[0015] As a preferred solution of the method for managing cargo loading information based on smart shelves of the present invention, the UWB anchor point is deployed on the smart shelf, sensor signal features are collected based on cargo RFID tags and smart shelf UWB anchor points, and NLoS conditions are identified using a support vector machine classification algorithm, and error correction is performed, including:

[0016] UMB fixed anchor points are deployed at the four corners of the smart shelf. The cargo RFID tags send UMB signals to the anchor points, record the signal flight time, and use the ToA ranging method to calculate the distance d from the cargo to each tag based on the signal flight time. ToA ;

[0017] During the UWB ranging process, the UWB ranging data is recorded, and RSSI and CIR feature extraction is performed to generate a feature vector D, wherein the RSSI feature refers to the extracted UMB received signal strength data, and the CIR feature refers to the number and strength of signal delay peaks on multiple paths during the entire response time from the transmission to the reception of the signal pulse;

[0018] Based on the support vector machine classification algorithm, an SVM model is constructed. Historical RSSI and CIR data are collected, feature data is extracted to generate historical feature vectors Di, and the historical feature vectors Di are input into the SVM model. Through the feature vectors Di in the training data, the SVM algorithm searches for the best hyperplane to maximize the classification interval and constructs the target loss function:

[0019]

[0020] Where w is the hyperplane weight vector, b is the hyperplane offset, ζ i is a slack variable used to tolerate misclassification of training samples, C is a penalty parameter that controls the model's tolerance to misclassification and balances classification accuracy and generalization ability, and B is the total number of samples;

[0021] During the SVM model training process, the SVM model automatically adjusts the position and direction of the hyperplane and adjusts the value of parameter C through cross-validation to achieve the best classification result;

[0022] The feature vector D generated in real time is input into the trained SVM model, and the data is classified by the SVM decision function f(x);

[0023] If f(x) is a positive number, it is determined to be LoS, and the current ranging data is directly used for subsequent positioning calculations;

[0024] If f(x) is a negative number, it is determined to be NLoS, and error correction is performed to adjust the signal distance data to improve the ranging accuracy;

[0025] Based on the identified NLoS signal, the corresponding RSSI and CIR data are extracted, and the historical RSSI and CIR data and their corresponding known errors are used to construct a training set. The training set data is input into the K nearest neighbor regression model, and the generalization ability of the model is verified by the cross-validation method. The model parameters are adjusted according to the verification results, and the RSSI and CIR features of the newly collected NLoS signal are input into the trained error estimation model to output the NLoS error value.

[0026] Use the estimated NLoS error value to measure the ranging data d ToA To make a correction:

[0027] d e =d ToA -εNLoS

[0028] Among them, d e is the corrected distance measurement value, ε NLoS is the NLoS error value obtained by the error estimation model.

[0029] As a preferred solution of the method for managing cargo loading information based on smart shelves of the present invention, wherein: based on the error correction result, the ToA algorithm is used to measure the signal flight time, the distance between the cargo and the anchor point is calculated, and the three-dimensional coordinates of the cargo are determined using the three-dimensional positioning formula, including:

[0030] Use the revised distance measurement value and substitute it into the triangulation formula to solve the formula to obtain the three-dimensional coordinates (x, y, z) of the goods, and store the calculated three-dimensional coordinate data of the goods on the blockchain platform.

[0031] As a preferred solution of the method for managing cargo loading information based on smart shelves of the present invention, the real-time monitoring of cargo safety status through the sensor group deployed on the smart shelves and the cargo three-dimensional coordinate data includes:

[0032] According to the layout of smart shelves, temperature sensors, humidity sensors and light sensors are deployed in different areas. The sensor data sampling period is set to once per minute. Environmental data is collected in real time according to the preset sampling period, and the collected environmental data, sensor area cargo ID, sensor coordinates and corresponding timestamps are transmitted to the warehouse central control system.

[0033] Based on the cargo ID, the corresponding cargo 3D coordinate data is read from the blockchain platform, the distance between the cargo location and the sensor location is calculated using the Euclidean distance formula, and the sensor data with the smallest Euclidean distance is selected as the cargo safety status data;

[0034] Safety thresholds are set for different goods. Each time the sensor data is updated, the collected safety status data of the goods is compared with the safety threshold. If the safety status data of the goods is less than the safety threshold, the data is recorded and continuously monitored. If the safety status data of the goods is greater than or equal to the safety threshold, the goods are marked as abnormal goods, triggering the alarm mechanism, generating alarm information and notifying the warehouse management personnel through the wireless data network, and synchronously adjusting the warehouse environmental parameters.

[0035] As a preferred solution of the method for managing cargo loading information based on smart shelves of the present invention, the method of collecting historical demand data, using Monte Carlo simulation to predict cargo demand at future times, and calculating the safety inventory and inventory order point of the cargo in combination with Bayes' law, and generating a replenishment plan to replenish the warehouse cargo inventory includes:

[0036] Extract the historical demand data of goods from the warehouse database, and calculate the mean μ of the historical demand data based on the historical demand data of goods 0 and variance

[0037] Set the number of Monte Carlo simulations to N, and the future demand follows a normal distribution, and set the mean μ of the historical demand data to 0 and variance As a parameter of normal distribution, Monte Carlo simulation is performed to generate N simulated demands every day, read and calculate the number of times each demand occurs, calculate the frequency of occurrence of the demand, and obtain the probability distribution of the demand, and the sum of the frequencies of the predicted demand is 1;

[0038] Observe the daily demand for goods in real time and record the actual demand for goods on day t D obs，t , calculate the observation demand D obs，t The difference between the predicted demand and i :

[0039] ε i =D obs，t -d i

[0040] According to the density function of the normal distribution, the predicted demand is calculated as d i When the observed demand is D obs，t The probability of occurrence P(D obs，t ∣g t ):

[0041]

[0042] Among them, g t is the predicted demand for goods on day t, d i is the i-th cargo demand among the N cargo demand predicted on the t-th day;

[0043] According to Bayes' theorem, the posterior probability of each predicted demand is updated and normalized:

[0044]

[0045] Use the posterior probability of each predicted demand to perform weighted averaging and calculate the expected value of the predicted demand on day t:

[0046]

[0047] Calculate the variance of the forecast demand on day t:

[0048]

[0049] Set the service level to Z, and according to the standard deviation of the forecast demand σ b and service level Z, calculate the safety stock S:

[0050] S=Z·σ b

[0051] Add the expected value of the safety stock quantity S and the forecast demand quantity to get the goods inventory replenishment threshold R p , and generate a replenishment plan. When the inventory of goods is less than or equal to the threshold, replenishment will be carried out, and the replenishment plan will be notified to suppliers and warehouse managers.

[0052] As a preferred solution of the method for managing cargo loading information based on smart shelves of the present invention, after the cargo is shipped out of the warehouse, inventory analysis is performed and business reports are generated, including:

[0053] When the goods are shipped out, the RFID tag recognition technology is used to record the shipment information of each piece of goods, including the goods ID, shipment time, quantity, destination and shipment personnel, and the shipment information is associated with the inventory data to update the quantity of goods in the inventory management system;

[0054] Regularly analyze inventory status, calculate inventory turnover and identify slow-moving inventory, generate business reports and store them on the blockchain platform, and send business reports to designated managers.

[0055] Another object of the present invention is to provide a system for managing cargo loading information based on smart shelves, which comprises:

[0056] Data collection module, used to collect the incoming information and store it on the blockchain platform;

[0057] Cargo positioning module, used to locate the three-dimensional position of cargo;

[0058] Safety monitoring module, used to monitor the safety status of goods in real time and issue early warnings;

[0059] Intelligent replenishment module, used to predict future demand for goods and perform intelligent replenishment;

[0060] The inventory analysis module is used to record the outbound information and generate business reports when the goods are shipped out.

[0061] A computer device comprises: a memory and a processor; the memory stores a computer program, and the processor implements the steps of a method for managing loading cargo information based on smart shelves when executing the computer program.

[0062] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of a method for managing loading cargo information based on smart shelves.

[0063] The beneficial effects of the present invention are as follows: the present invention realizes accurate, real-time and efficient management of cargo information, which not only improves the security, transparency and intelligence of the information management system, but also optimizes warehouse management through high-precision positioning and effective inventory forecasting, and improves the efficiency and security of cargo management. It effectively addresses the deficiencies of existing technologies in positioning, information storage and demand forecasting, and significantly improves the comprehensive efficiency of warehouse management. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0065] Figure 1 The figure is a flow chart of a method for managing cargo loading information based on smart shelves.

[0066] Figure 2 This is a structural diagram of a system for managing cargo loading information based on smart shelves. DETAILED DESCRIPTION

[0067] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings.

[0068] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0069] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor is it a separate or selective embodiment that is mutually exclusive with other embodiments.

[0070] Example 1

[0071] Reference Figure 1 , which is the first embodiment of the present invention, and which provides a method for managing cargo loading information based on a smart shelf. The method for managing cargo loading information based on a smart shelf includes:

[0072] S1. Collect the goods warehousing information, build a blockchain platform, and store the collected goods warehousing information in the blockchain platform;

[0073] Furthermore, the goods warehousing information is collected, a blockchain platform is constructed, and the collected goods warehousing information is stored in the blockchain platform, including:

[0074] Before the goods are put into storage, the operator uses an RFID writing device to set an RFID tag for the goods and uses an RFID card reader to activate the RFID tag, scan its stored information, read the RFID tag information and transmit it to the warehouse central control system through a wireless sensor network. The RFID tag information refers to the goods ID, category, weight, volume and expiration date;

[0075] Based on blockchain technology, a blockchain platform is built to upload the cargo RFID tag information to the blockchain platform, and a hash algorithm is used to generate a hash value of the RFID tag information, which is then stored on the blockchain platform;

[0076] According to the cargo RFID tag information, the cargo information is automatically matched to the preset shelf space and the shelf location is marked.

[0077] By storing the goods entry information on the blockchain platform and utilizing the decentralized and encrypted storage characteristics of the blockchain, data security is ensured, which can effectively prevent data tampering or loss caused by human errors or malicious attacks in traditional centralized database systems, thereby improving the security and credibility of the warehouse management information system. The blockchain platform is transparent and open, and the entry information of all goods can be traced back to the original records in chronological order, ensuring that the source and modification traces of each piece of data can be tracked. Warehouse management personnel can easily check the entry records of any goods and achieve full traceability of information, which is crucial for the logistics and warehousing fields, and helps to enhance the transparency of the supply chain and ensure that data in the process of goods circulation is open and transparent. In addition, the blockchain platform can ensure data consistency between all nodes, even when multiple warehouses or storage nodes in different geographical locations access data at the same time, it can ensure that the acquired data is synchronized and consistent in real time. In this way, management errors caused by data delays or inconsistencies can be avoided, ensuring that the goods entry information is always kept up to date in the warehouse management system, and improving the accuracy and real-time nature of information management.

[0078] S2, deploy UWB anchor points on smart shelves, collect sensor signal features based on cargo RFID tags and smart shelf UWB anchor points, use support vector machine classification algorithm to identify NLoS conditions, and perform error correction;

[0079] Furthermore, UWB anchors are deployed on smart shelves. Based on the cargo RFID tags and smart shelf UWB anchors, sensor signal features are collected and the support vector machine classification algorithm is used to identify the NLoS conditions and perform error correction including,

[0080] UMB fixed anchor points are deployed at the four corners of the smart shelf. The cargo RFID tags send UMB signals to the anchor points, record the signal flight time, and use the ToA ranging method to calculate the distance d from the cargo to each tag based on the signal flight time. ToA :

[0081] d ToA =c×t i

[0082] Where c is the speed of light, t i is the signal flight time;

[0083] During the UWB ranging process, the UWB ranging data is recorded, and RSSI and CIR feature extraction is performed to generate a feature vector D, wherein the RSSI feature refers to the extracted UMB received signal strength data, and the CIR feature refers to the number and strength of signal delay peaks on multiple paths during the entire response time from the transmission to the reception of the signal pulse;

[0084] Based on the support vector machine classification algorithm, an SVM model is constructed. Historical RSSI and CIR data are collected, feature data is extracted to generate historical feature vectors Di, and the historical feature vectors Di are input into the SVM model. Through the feature vectors Di in the training data, the SVM algorithm searches for the best hyperplane to maximize the classification interval and constructs the target loss function:

[0085]

[0086] Where w is the hyperplane weight vector, b is the hyperplane offset, ζ i is a slack variable used to tolerate misclassification of training samples, C is a penalty parameter that controls the model's tolerance to misclassification and balances classification accuracy and generalization ability, and B is the total number of samples;

[0087] During the SVM model training process, the SVM model automatically adjusts the position and direction of the hyperplane and adjusts the value of parameter C through cross-validation to achieve the best classification result;

[0088] The feature vector D generated in real time is input into the trained SVM model, and the data is classified by the SVM decision function. The SVM decision function is:

[0089]

[0090] Among them, x is the feature vector of the test sample, x iis the support vector, which indicates the specific sample points selected by the model during the training process, y i is the category label corresponding to the support vector, α i is the Lagrange multiplier coefficient, K(x i , x) is the kernel function used to calculate the relationship between sample x and support vector x i The similarity between them, b is the bias term of the decision function, and M is the total number of support vectors;

[0091] If f(x) is a positive number, it is determined to be LoS, and the current ranging data is directly used for subsequent positioning calculations;

[0092] If f(x) is a negative number, it is determined to be NLoS, and error correction is performed to adjust the signal distance data to improve the ranging accuracy;

[0093] Based on the identified NLoS signal, the corresponding RSSI and CIR data are extracted, and the historical RSSI and CIR data and their corresponding known errors are used to construct a training set. The training set data is input into the K nearest neighbor regression model, and the generalization ability of the model is verified by the cross-validation method. The model parameters are adjusted according to the verification results to improve the accuracy and stability of the model prediction error. The RSSI and CIR features of the newly collected NLoS signal are input into the trained error estimation model to output the NLoS error value.

[0094] Use the estimated NLoS error value to measure the ranging data d ToA To make a correction:

[0095] d e =d ToA -ε NLoS

[0096] Among them, d e is the corrected distance measurement value, ε NLoS is the NLoS error value obtained by the error estimation model.

[0097] By deploying UWB anchor points and using the support vector machine (SVM) classification algorithm to identify NLoS conditions, it is possible to effectively distinguish line-of-sight LoS and non-line-of-sight NLoS signals, thereby determining whether the signal is blocked. Error correction under NLoS conditions can significantly reduce positioning errors and improve positioning accuracy. This approach effectively avoids the problem of reduced positioning accuracy due to complex environments, ensuring that location data in cargo information management is more accurate and reliable. The support vector machine classification algorithm can quickly classify and process sensor signals, identify NLoS conditions, and then perform timely error correction during the positioning process. Compared with traditional positioning error correction methods, the SVM algorithm has fast processing speed and low computational complexity, which facilitates rapid analysis and feedback of real-time positioning data, ensuring data processing efficiency in real-time positioning. With the improvement of NLoS conditions, the cargo positioning information can be quickly synchronized to the warehouse information management system, providing timely and accurate location information support for warehouse decisions, and optimizing the positioning accuracy without human intervention. This automated processing method not only reduces the maintenance workload, but also reduces the errors and operating costs that may be caused by human intervention, further improving the autonomous operation capability and economic benefits of the entire warehouse system. Finally, by identifying NLoS conditions and performing error correction, the integrity and accuracy of the positioning data are ensured, providing high-quality basic data for the warehouse management system, further supporting the cargo tracking, safety monitoring and inventory optimization functions in the intelligent warehouse system, and effectively improving the intelligence level of the entire warehouse management system, so that it can better meet the needs of modern warehousing for efficient, safe and accurate positioning.

[0098] S3. Based on the error correction result, the ToA algorithm is used to measure the signal flight time, calculate the distance between the goods and the anchor point, and use the three-dimensional positioning formula to determine the three-dimensional coordinates of the goods. The sensor group deployed on the smart shelf and the three-dimensional coordinate data of the goods are used to monitor the safety status of the goods in real time;

[0099] Furthermore, based on the error correction result, the ToA algorithm is used to measure the signal flight time, calculate the distance between the cargo and the anchor point, and use the three-dimensional positioning formula to determine the three-dimensional coordinates of the cargo, including:

[0100] Use the revised distance measurement value and substitute it into the triangulation formula to solve the formula to obtain the three-dimensional coordinates (x, y, z) of the cargo. The triangulation formula is:

[0101]

[0102] Among them, (x i ,y i , z i ) is the fixed anchor point coordinate of the smart shelf UMB, is the measured distance from the cargo tag to the i-th fixed coordinate;

[0103] The calculated three-dimensional coordinate data of the goods is stored on the blockchain platform.

[0104] By using the arrival time ToA algorithm to measure the signal flight time and calculate the distance between the goods and the anchor point, the accuracy of the distance measurement is ensured. Based on the error-corrected data, the three-dimensional coordinates are solved by the triangulation formula to obtain the accurate spatial position (x, y, z) of the goods. This three-dimensional positioning method can not only provide the plane position of the goods, but also accurately obtain its height information, which is particularly important for the storage management of multi-layer shelves. The accuracy of three-dimensional positioning directly determines the effectiveness of cargo tracking, ensuring that the storage system can grasp the accurate position of the goods in real time, thereby improving the efficiency of cargo management and scheduling. After realizing three-dimensional positioning on the smart shelf, the storage system can analyze the placement status of the goods based on the three-dimensional position data of the goods, and detect whether there is any displacement or tilt beyond the specified range. This is crucial for the safe management of goods, especially when storing valuables or fragile items. Real-time three-dimensional monitoring can promptly detect abnormal conditions and prevent potential damage risks in advance. The introduction of three-dimensional positioning data provides more reliable data support for warehouse safety management and significantly improves the monitoring and early warning capabilities of the system.

[0105] Furthermore, through the sensor group deployed on the smart shelf and the three-dimensional coordinate data of the goods, the safety status of the goods can be monitored in real time, including:

[0106] According to the layout of smart shelves, temperature sensors, humidity sensors and light sensors are deployed in different areas. The sensor data sampling period is set to once per minute. Environmental data is collected in real time according to the preset sampling period, and the collected environmental data, sensor area cargo ID, sensor coordinates and corresponding timestamps are transmitted to the warehouse central control system.

[0107] Based on the cargo ID, the corresponding cargo 3D coordinate data is read from the blockchain platform, the distance between the cargo location and the sensor location is calculated using the Euclidean distance formula, and the sensor data with the smallest Euclidean distance is selected as the cargo safety status data;

[0108] According to the manufacturer's specifications and national standards and systems, safety thresholds are set for different goods. Each time the sensor data is updated, the collected cargo safety status data is compared with the safety threshold. If the cargo safety status data is less than the safety threshold, the data is recorded and continuously monitored. If the cargo safety status data is greater than or equal to the safety threshold, the cargo is marked as abnormal, triggering the alarm mechanism, generating alarm information and notifying the warehouse management personnel through the wireless data network, and synchronously adjusting the warehouse environmental parameters to keep the cargo in a safe storage environment.

[0109] By deploying a variety of sensors such as temperature, humidity, and light on smart shelves, key parameters of the cargo storage environment can be collected and monitored in real time. A data sampling cycle of once a minute can be set to ensure that environmental data is always up to date. The real-time monitoring mechanism can not only continuously detect abnormal conditions in the warehouse environment, but also effectively prevent quality changes or damage to cargo caused by factors such as temperature, humidity, and light. This real-time data collection and monitoring function realizes full-process visual management of the cargo safety status, greatly improving the warehousing system's ability to respond to cargo protection. The three-dimensional coordinates of the cargo are extracted from the blockchain in combination with the cargo ID, and the distance between the cargo and the sensor is calculated using the Euclidean distance formula, so that the closest sensor data is selected as the cargo safety status data. It ensures that each piece of monitoring data is most relevant to a specific product, avoiding monitoring errors caused by inaccurate matching of environmental data. Through this precise positioning calculation method, it can ensure the accuracy and reliability of cargo monitoring data, and provide targeted guidance for the adjustment of the storage environment. In addition, by combining manufacturer specifications and national standards to set safety thresholds, it ensures that the storage conditions of various types of goods meet industry and legal requirements, and provides a scientific and reasonable safety benchmark for different types of goods to avoid quality degradation or safety hazards caused by environmental abnormalities. This not only helps the standardized operation of warehouse management, but also effectively prevents product quality problems or legal liabilities caused by environmental non-compliance, increases compliance protection for warehouse operation processes, and by monitoring the cargo environment,

[0110] Combined with the alarm mechanism, managers can deal with abnormal goods in the first time and adjust the environmental parameters of the warehouse simultaneously to ensure that the goods storage environment meets safety requirements. This greatly improves the response speed of warehouse management to emergencies and helps reduce the risk of goods damage and property losses.

[0111] S4. Collect historical demand data, use Monte Carlo simulation to predict future demand for goods, and calculate the safety stock and inventory order point of goods in combination with Bayes' theorem, and generate a replenishment plan to replenish the warehouse inventory;

[0112] Furthermore, historical demand data is collected, and Monte Carlo simulation is used to predict future demand for goods. In combination with Bayesian law, the safety stock and inventory order point of goods are calculated, and a replenishment plan is generated to replenish the warehouse inventory, including:

[0113] Extract the historical demand data of goods from the warehouse database, and calculate the mean μ of the historical demand data based on the historical demand data of goods 0 and variance

[0114] Set the number of Monte Carlo simulations to N, and the future demand follows a normal distribution, and set the mean μ of the historical demand data to0 and variance As a parameter of normal distribution, Monte Carlo simulation is performed to generate N simulated demands every day, read and calculate the number of times each demand occurs, calculate the frequency of occurrence of the demand, and obtain the probability distribution of the demand, and the sum of the frequencies of the predicted demand is 1;

[0115] Observe the daily demand for goods in real time and record the actual demand for goods on day t D obs，t , calculate the observation demand D obs，t The difference between the predicted demand and i :

[0116] ε i =D obs，t -d i

[0117] According to the density function of the normal distribution, the predicted demand is calculated as d i When the observed demand is D obs，t The probability of occurrence P(D obs，t ∣g t ):

[0118]

[0119] Among them, g t is the predicted demand for goods on day t, d i is the i-th cargo demand among the N cargo demand predicted on the t-th day;

[0120] According to Bayes' theorem, the posterior probability of each predicted demand is updated and normalized:

[0121]

[0122] Use the posterior probability of each predicted demand to perform weighted averaging and calculate the expected value of the predicted demand on day t:

[0123]

[0124] Calculate the variance of the forecast demand on day t:

[0125]

[0126] According to industry standards, the service level is set to Z, and the standard deviation of the predicted demand is σ b and service level Z, calculate the safety stock S:

[0127] S=Z·σ b

[0128] Add the expected value of the safety stock quantity S and the forecast demand quantity to get the goods inventory replenishment threshold R p , and generate a replenishment plan. When the inventory of goods is less than or equal to the threshold, replenishment will be carried out, and the replenishment plan will be notified to suppliers and warehouse managers.

[0129] By collecting historical demand data and calculating the mean and variance, a normal distribution can be constructed based on the historical demand data of goods, thus providing reliable basic parameters for Monte Carlo simulation. Using Monte Carlo simulation to repeatedly generate daily demand for goods in the future forecast period can effectively take into account the volatility and uncertainty of demand. Multiple simulations can provide N sets of forecast data. Through frequency statistics and analysis of these simulation results, the demand in the future can be predicted more accurately. This probability-based forecasting method effectively improves the accuracy of demand forecasting, helps companies better grasp inventory dynamics, and provides scientific data support for subsequent inventory management decisions. Combined with Bayes' theorem, the safety inventory and inventory replenishment points of goods are calculated and generated. Bayes' theorem provides more accurate inventory demand data by observing the actual demand for goods in real time and dynamically correcting it with the preliminary forecast results. The calculation of safety stock not only takes into account historical demand fluctuations, but also makes adjustments based on real-time data to ensure that while maintaining service levels, excessive inventory backlogs are avoided and inventory costs are reduced. This method of dynamically adjusting inventory levels effectively improves the turnover efficiency and economy of warehouse inventory. In addition, the system can initiate replenishment requests when the inventory of goods approaches the critical point, promptly notify suppliers and warehouse managers, reduce the lag of replenishment decisions, and combine the safety stock S of the goods to provide early warning of replenishment needs, reduce the risks caused by supply chain fluctuations, thereby optimizing inventory management and further reducing the risks of excess and out-of-stock.

[0130] S5. After the goods leave the warehouse, inventory analysis is performed and business reports are generated;

[0131] Furthermore, after the goods leave the warehouse, inventory analysis is performed and business reports are generated, including:

[0132] When the goods are shipped out, the RFID tag recognition technology is used to record the shipment information of each piece of goods, including the goods ID, shipment time, quantity, destination and shipment personnel, and the shipment information is associated with the inventory data to update the quantity of goods in the inventory management system;

[0133] Regularly analyze inventory status, calculate inventory turnover and identify slow-moving inventory, generate business reports and store them on the blockchain platform, and send business reports to designated managers for inventory optimization.

[0134] By using RFID tag identification technology to record the outbound information of goods when they leave the warehouse, and associating this information with the existing inventory data in the inventory management system, real-time updating of inventory data is achieved, and information such as cargo ID, outbound time, quantity, destination, and outbound personnel are automatically recorded, ensuring the accuracy and timeliness of inventory data, reducing the possibility of manual entry errors, improving the data management efficiency of the warehousing system, and ensuring real-time updating of inventory quantities, so that managers can timely grasp the dynamic changes of goods in the warehouse, and by conducting inventory analysis, calculating inventory turnover rate and identifying unsalable inventory, accurate data support is provided for decision-making in warehousing management. Analyzing inventory turnover rate can understand the average residence time of goods in the warehouse, evaluate inventory liquidity and management efficiency, and identifying unsalable inventory helps managers take strategies such as cleaning or promotion for backlogged goods, optimize the utilization of storage space, and help enterprises better allocate resources and control inventory costs, ultimately improving the economic benefits of warehousing management.

[0135] Example 2

[0136] Reference Figure 2 , which is the second embodiment of the present invention, is different from the previous embodiment and provides a system for managing cargo loading information based on smart shelves, including:

[0137] Data collection module, used to collect the incoming information and store it on the blockchain platform;

[0138] Cargo positioning module, used to locate the three-dimensional position of cargo;

[0139] Safety monitoring module, used to monitor the safety status of goods in real time and issue early warnings;

[0140] Intelligent replenishment module, used to predict future demand for goods and perform intelligent replenishment;

[0141] The inventory analysis module is used to record the outbound information and generate business reports when the goods are shipped out.

[0142] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program codes.

[0143] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0144] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0145] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used to implement: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0146] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for managing cargo loading information based on smart shelves, characterized in that: include, Collect goods warehousing information, build a blockchain platform, and store the collected goods warehousing information in the blockchain platform; Deploy UWB anchors on smart shelves, collect sensor signal features based on cargo RFID tags and smart shelf UWB anchors, use support vector machine classification algorithm to identify NLoS conditions, and perform error correction; Based on the error correction results, the ToA algorithm is used to measure the signal flight time, calculate the distance between the goods and the anchor point, and use the three-dimensional positioning formula to determine the three-dimensional coordinates of the goods. The sensor group deployed on the smart shelf and the three-dimensional coordinate data of the goods can be used to monitor the safety status of the goods in real time. Collect historical demand data, use Monte Carlo simulation to predict future demand for goods, and combine Bayesian law to calculate the safety stock and inventory order point of goods, and generate a replenishment plan to replenish warehouse goods inventory; After the goods leave the warehouse, inventory analysis is performed and business reports are generated.

2. The method for managing cargo loading information based on the smart shelf according to claim 1, characterized in that: The collecting of goods warehousing information, building a blockchain platform, and storing the collected goods warehousing information in the blockchain platform include: Before the goods are put into storage, the operator uses an RFID writing device to set an RFID tag for the goods and uses an RFID card reader to activate the RFID tag, scan its stored information, read the RFID tag information and transmit it to the warehouse central control system through a wireless sensor network. The RFID tag information refers to the goods ID, category, weight, volume and expiration date; Based on blockchain technology, a blockchain platform is built to upload the cargo RFID tag information to the blockchain platform, and a hash algorithm is used to generate a hash value of the RFID tag information, which is then stored on the blockchain platform; According to the cargo RFID tag information, the cargo information is automatically matched to the preset shelf space and the shelf location is marked.

3. The method for managing cargo loading information based on the smart shelf according to claim 2, characterized in that: The method of deploying UWB anchor points on smart shelves, collecting sensor signal features based on cargo RFID tags and smart shelf UWB anchor points, using support vector machine classification algorithms to identify NLoS conditions, and performing error correction includes: UMB fixed anchor points are deployed at the four corners of the smart shelf. The cargo RFID tags send UMB signals to the anchor points, record the signal flight time, and use the ToA ranging method to calculate the distance d from the cargo to each tag based on the signal flight time. ToA ; During the UWB ranging process, the UWB ranging data is recorded, and RSSI and CIR feature extraction is performed to generate a feature vector D, wherein the RSSI feature refers to the extracted UMB received signal strength data, and the CIR feature refers to the number and strength of signal delay peaks on multiple paths during the entire response time from the transmission to the reception of the signal pulse; Based on the support vector machine classification algorithm, an SVM model is constructed. Historical RSSI and CIR data are collected, feature data is extracted to generate historical feature vectors Di, and the historical feature vectors Di are input into the SVM model. Through the feature vectors Di in the training data, the SVM algorithm searches for the best hyperplane to maximize the classification interval and constructs the target loss function: Where w is the hyperplane weight vector, b is the hyperplane offset, ζ i is a slack variable used to tolerate misclassification of training samples, C is a penalty parameter that controls the model's tolerance to misclassification and balances classification accuracy and generalization ability, and B is the total number of samples; During the SVM model training process, the SVM model automatically adjusts the position and direction of the hyperplane and adjusts the value of parameter C through cross-validation to achieve the best classification result; The feature vector D generated in real time is input into the trained SVM model, and the data is classified by the SVM decision function f(x); If f(x) is a positive number, it is determined to be LoS, and the current ranging data is directly used for subsequent positioning calculations; If f(x) is a negative number, it is determined to be NLoS, and error correction is performed to adjust the signal distance data to improve the ranging accuracy; Based on the identified NLoS signal, the corresponding RSSI and CIR data are extracted, and the historical RSSI and CIR data and their corresponding known errors are used to construct a training set. The training set data is input into the K nearest neighbor regression model, and the generalization ability of the model is verified by the cross-validation method. The model parameters are adjusted according to the verification results, and the RSSI and CIR features of the newly collected NLoS signal are input into the trained error estimation model to output the NLoS error value. Use the estimated NLoS error value to measure the ranging data d ToA To make a correction: d e =d ToA -e NLoS Among them, d e is the corrected distance measurement value, ε NLo S is the NLoS error value obtained by the error estimation model.

4. The method for managing cargo loading information based on the smart shelf according to claim 3, characterized in that: Based on the error correction result, the ToA algorithm is used to measure the signal flight time, calculate the distance between the cargo and the anchor point, and use the three-dimensional positioning formula to determine the three-dimensional coordinates of the cargo, including: Use the revised distance measurement value and substitute it into the triangulation formula to solve the formula to obtain the three-dimensional coordinates (x, y, z) of the goods, and store the calculated three-dimensional coordinate data of the goods on the blockchain platform.

5. The method for managing cargo loading information based on smart shelves according to claim 4, characterized in that: The real-time monitoring of the safety status of goods by using the sensor group deployed on the smart shelf and the three-dimensional coordinate data of the goods includes: According to the layout of smart shelves, temperature sensors, humidity sensors and light sensors are deployed in different areas. The sensor data sampling period is set to once per minute. Environmental data is collected in real time according to the preset sampling period, and the collected environmental data, sensor area cargo ID, sensor coordinates and corresponding timestamps are transmitted to the warehouse central control system. Based on the cargo ID, the corresponding cargo 3D coordinate data is read from the blockchain platform, the distance between the cargo location and the sensor location is calculated using the Euclidean distance formula, and the sensor data with the smallest Euclidean distance is selected as the cargo safety status data; Safety thresholds are set for different goods. Each time the sensor data is updated, the collected safety status data of the goods is compared with the safety threshold. If the safety status data of the goods is less than the safety threshold, the data is recorded and continuously monitored. If the safety status data of the goods is greater than or equal to the safety threshold, the goods are marked as abnormal goods, triggering the alarm mechanism, generating alarm information and notifying the warehouse management personnel through the wireless data network, and synchronously adjusting the warehouse environmental parameters.

6. The method for managing cargo loading information based on the smart shelf according to claim 5, characterized in that: The above-mentioned collection of historical demand data, the use of Monte Carlo simulation to predict future cargo demand, and the calculation of the safety stock and inventory order point of the cargo in combination with Bayes' law, and the generation of a replenishment plan to replenish the warehouse cargo inventory include: Extract the historical demand data of goods from the warehouse database, and calculate the mean μ0 and variance of the historical demand data based on the historical demand data of goods. Set the number of Monte Carlo simulations to N, and the future demand follows a normal distribution, and set the mean μ0 and variance of the historical demand data to As a parameter of normal distribution, Monte Carlo simulation is performed to generate N simulated demands every day, read and calculate the number of times each demand occurs, calculate the frequency of occurrence of the demand, and obtain the probability distribution of the demand, and the sum of the frequencies of the predicted demand is 1; Observe the daily demand for goods in real time and record the actual demand for goods on day t D obs，t , calculate the observation demand D obs，t The difference between the predicted demand and i : e i =D obs，t -d i According to the density function of the normal distribution, the predicted demand is calculated as d i When the observed demand D obs，t The probability of occurrence P(D obs，t ∣g t ): Among them, g t is the predicted demand for goods on day t, d i is the i-th cargo demand among the N cargo demand predicted on the t-th day; According to Bayes' theorem, the posterior probability of each predicted demand is updated and normalized: Use the posterior probability of each predicted demand to perform weighted averaging and calculate the expected value of the predicted demand on day t: Calculate the variance of the forecast demand on day t: Set the service level to Z, and according to the standard deviation of the forecast demand σ b and service level Z, calculate the safety stock S: S=Z·σ b Add the expected value of the safety stock quantity S and the forecast demand quantity to get the goods inventory replenishment threshold R p , and generate a replenishment plan. When the inventory of goods is less than or equal to the threshold, replenishment will be carried out, and the replenishment plan will be notified to suppliers and warehouse managers.

7. The method for managing cargo loading information based on smart shelves according to claim 6, characterized in that: After the goods are shipped out, inventory analysis is performed and business reports are generated, including: When the goods are shipped out, the RFID tag recognition technology is used to record the shipment information of each piece of goods, including the goods ID, shipment time, quantity, destination and shipment personnel, and the shipment information is associated with the inventory data to update the quantity of goods in the inventory management system; Regularly analyze inventory status, calculate inventory turnover and identify slow-moving inventory, generate business reports and store them on the blockchain platform, and send business reports to designated managers.

8. A system for managing cargo loading information based on a smart shelf based on the method for managing cargo loading information based on a smart shelf according to any one of claims 1 to 7, characterized in that: include, Data collection module, used to collect the incoming information and store it on the blockchain platform; Cargo positioning module, used to locate the three-dimensional position of cargo; Safety monitoring module, used to monitor the safety status of goods in real time and issue early warnings; Intelligent replenishment module, used to predict future demand for goods and perform intelligent replenishment; The inventory analysis module is used to record the outbound information and generate business reports when the goods are shipped out.

9. A computer device comprising: Memory and processor; The memory stores a computer program, characterized in that when the processor executes the computer program, the steps of the method for managing loading cargo information based on smart shelves according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for managing loading cargo information based on a smart shelf according to any one of claims 1 to 7 are implemented.

Images