Four-way shuttle vehicle positioning system based on bar code recognition

Abstract

The invention discloses a four-direction shuttle vehicle positioning system based on bar code recognition, belongs to the technical field, and remarkably improves the efficiency and accuracy of warehouse operation through the high-precision bar code recognition and intelligent scheduling functions of the four-direction shuttle vehicle positioning system. The system can automatically and rapidly identify cargo bar codes and command the shuttle vehicle to accurately execute carrying tasks, manual intervention and operation time are greatly reduced, the operation error rate is reduced, the smoothness and reliability of overall warehousing operation are improved, meanwhile, the bar codes are used for replacing traditional counting pieces and two-dimensional codes to serve as positioning reference, and the positioning accuracy is improved. The one-dimensional bar code with the gum is pasted on the side edge of a track where a four-way vehicle runs, a counting piece and a two-dimensional code are replaced, a support is saved, construction links are reduced, maintenance is convenient, meanwhile, the defect of false induction of a photoelectric sensor is overcome, the defect of misreading and misreading caused by dust accumulation of the two-dimensional code which is placed horizontally is overcome, and the four-way shuttle vehicle can run more stably.

Description

Technical Field

[0001] The invention belongs to the technical field of shuttle vehicle positioning, and in particular relates to a four-way shuttle vehicle positioning system based on barcode recognition. Background Art

[0002] With the rapid development of the modern logistics industry, warehouse automation and intelligence have become key factors in improving logistics efficiency and reducing operating costs. In automated warehouses, the four-way shuttle is an efficient cargo handling device, and the accuracy and intelligence level of its positioning system directly affect the overall operating efficiency of the warehouse.

[0003] There are two ways to locate the traditional four-way shuttle:

[0004] 1. Based on RFID + counting sheet combination positioning, the RFID module identifies the label on the counting sheet, and the photoelectric switch signal identifies the counting sheet positioning. However, this positioning method requires writing the RFID tag data on the counting sheet in advance, which is labor-intensive and costly. The photoelectric switch signal positioning accuracy is poor and it is very easy to be mis-sensed, causing the four-way shuttle vehicle to malfunction, and has high requirements for the shelf;

[0005] 2. Based on QR code positioning, this positioning method generally places the QR code horizontally on the shelf. The QR code is prone to dust accumulation, causing the sensor to miss or misread the code. Manufacturers generally provide a sweeper to regularly clean the dust on the QR code, which increases customer costs. At the same time, multiple sweeps can easily damage the QR code. Summary of the invention

[0006] The purpose of the present invention is to provide a four-way shuttle positioning system based on barcode recognition in order to solve the above-mentioned problems.

[0007] The technical solution adopted by the present invention is as follows: a four-way shuttle positioning system based on barcode recognition, the system comprising: a barcode recognition module, a communication module, a control module, an artificial intelligence module, a cloud platform module, a scheduling module and an auxiliary module;

[0008] The barcode recognition module is connected to the communication module to send the read barcode information to the control module;

[0009] The communication module is connected to the control module to send the barcode information and the motor encoder data to the control module;

[0010] The control module is connected to the artificial intelligence module, sends real-time coordinate data to the artificial intelligence module, and receives path optimization and task prediction instructions;

[0011] The control module is connected to the cloud platform module, sends real-time coordinate data to the cloud platform module, and receives remote monitoring and management instructions;

[0012] The control module is connected to the scheduling module, sends real-time coordinate data to the scheduling module, and receives task scheduling instructions;

[0013] The artificial intelligence module is connected to the cloud platform module, sends the prediction results to the cloud platform module, and receives cloud data for training and learning;

[0014] The artificial intelligence module is internally provided with a data preprocessing module, a feature engineering module, a model training module, a model evaluation module, a model deployment module, a model updating module and a visualization module.

[0015] In a preferred embodiment, the feature scaling of the data preprocessing module uses a standardized method to scale the coordinate data features to a range of a mean of 0 and a standard deviation of 1, and the calculation formula is:

[0016] x' = (x-min) / (max-min);

[0017] x' represents the normalized eigenvalue; x represents the original eigenvalue; min represents the minimum value of all the values ​​of the feature; max represents the maximum value of all the values ​​of the feature;

[0018] The data preprocessing module uses the mean filling method to replace the missing values ​​with the average value of all the values ​​of the feature, and uses the two valid coordinate data before and after to estimate the missing coordinate data through linear interpolation. The calculation formula is:

[0019] x=x0+(x1-x0)*(t-t0) / (t1-t0);

[0020] Where x represents the estimated coordinate data;

[0021] x0, x1 represent the two valid coordinate data before and after;

[0022] t0, t1 represent the timestamps corresponding to the two valid coordinate data before and after;

[0023] t: The timestamp corresponding to the missing coordinate data.

[0024] In a preferred embodiment, the feature engineering module evaluates the information content of the feature to the target variable through information gain, and selects the feature with the largest information gain, and the calculation formula is:

[0025] IG(T,A)=Entropy(T)-Σ|T_v| / |T|*Entropy(T_v);

[0026] Where: IG(T,A) represents the information gain of feature A to target variable T

[0027] Entropy(T) represents the entropy of the target variable T

[0028] |T_v| represents the number of samples where feature A takes the value v

[0029] |T| represents the total number of samples

[0030] Entropy(T_v) represents the entropy of samples with feature A value v

[0031] The feature engineering module uses the principal component analysis method to convert the original features into new features through linear transformation, reducing the feature dimension while retaining most of the information. The formula is:

[0032] Z=W*X; wherein Z represents the new feature matrix, W represents the orthogonal matrix including the feature vector, and X represents the original feature matrix.

[0033] In a preferred embodiment, the algorithm flow of the model training module using the linear regression model for training includes:

[0034] Step S1: Initialize model parameters, set initial weight w and bias b;

[0035] Step S2: Forward propagation, calculate the predicted value y^, the calculation formula is:

[0036] y^(x)=w·x+b;

[0037] Where y^(x) represents the predicted output for input x; w represents the weight vector; x represents the input feature vector; b represents the bias term;

[0038] Step S3: Calculate the loss function and use the mean square error (MSE) as the loss function;

[0039] Formula: L(y,y)=1 / nΣ(yi-y^_i)2, where L(y,y^) represents the loss function; y represents the true value vector; y^ represents the predicted value vector; n represents the number of samples;

[0040] Step S4: Back propagation, calculating the gradient of the loss function with respect to the model parameters;

[0041] The gradient calculation formula for weight w is:

[0042] The gradient calculation formula for bias b is:

[0043] Step S5: Update model parameters, using gradient descent method to update weights and biases;

[0044] The updated weight calculation formula is:

[0045] The update bias calculation formula is:

[0046] Where α represents the learning rate;

[0047] Step S6: Repeat steps S2-5 until the model converges. The criterion for model convergence is that the loss function is lower than a certain threshold, or the number of iterations reaches a preset value.

[0048] In a preferred embodiment, the model evaluation module is a key link in the artificial intelligence system for measuring model performance and accuracy; it includes a series of tests and verifications on the trained model to evaluate its performance on unknown data; the evaluation indicators include accuracy, recall, F1 score, mean square error, depending on the type of model and the application scenario; through cross-validation, holdout test, A / B test methods, the model evaluation module can provide quantitative indicators of model reliability to help researchers or engineers decide whether the model is ready for practical application or whether it needs further adjustment and optimization;

[0049] The model deployment module is responsible for integrating the trained and evaluated model into the actual working environment; the model deployment module converts the model into a format that runs in the production environment, encapsulates the model into an API service, and deploys it to a cloud server or edge device.

[0050] In a preferred embodiment, the model updating module maintains its accuracy and relevance over time and as data accumulates; this module involves periodically retraining the model to incorporate new data points and features to adapt to data distribution changes or concept drift;

[0051] The visualization module provides an intuitive model representation and data analysis tool for the artificial intelligence system; it displays the model's input, output, performance indicators and prediction results through charts, dashboards and interactive interfaces, so that non-technical users can also understand and analyze the model's behavior; the visualization module includes data exploration tools, monitoring of the model training process, interpretation of prediction results and visualization of the model's decision path.

[0052] In a preferred embodiment, the barcode recognition module is the front-end sensing unit of the four-way shuttle positioning system, which is responsible for collecting and processing barcode information in the warehouse; the barcode recognition module includes a barcode scanner or camera sensor, and matching image recognition and processing software; it recognizes and extracts barcode information from a complex warehouse environment, and then converts this information into a usable data format.

[0053] In a preferred embodiment, the communication module is a bridge for information exchange between various components in the four-way shuttle positioning system; the communication module is responsible for transmitting and exchanging data between the barcode recognition module, the control module, and the artificial intelligence module unit;

[0054] The control module is the execution unit of the four-way shuttle positioning system. The control module receives instructions from the artificial intelligence module and directly controls the movement of the shuttle. The control module includes a motion controller, a driver and an actuator, which are responsible for converting instructions into specific actions of the shuttle, including forward, backward and turning.

[0055] In a preferred embodiment, the cloud platform module provides a remote data storage, processing and analysis platform for the four-way shuttle positioning system; the cloud platform module allows the system to upload data to the cloud for large-scale data processing and analysis, while supporting remote monitoring and fault diagnosis; the cloud platform module also enables data sharing and collaborative operations between multiple warehouses;

[0056] The auxiliary module is a support unit of the four-way shuttle positioning system, which includes power management, safety monitoring, and alarm system auxiliary functions; the power management module ensures that the system can have stable power supply in normal and emergency situations.

[0057] In a preferred embodiment, the scheduling module is the task allocation center in the four-way shuttle positioning system, which intelligently plans the shuttle's driving path and operation sequence according to the warehouse's operational requirements, the shuttle's status and location information; the core algorithms of the scheduling module include path planning, task scheduling and resource optimization.

[0058] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0059] 1. In the present invention, the efficiency and accuracy of warehouse operations are significantly improved through its high-precision barcode recognition and intelligent scheduling functions. The system can automatically and quickly identify the barcode of the goods, and command the shuttle car to accurately perform the handling task, which greatly reduces manual intervention and operation time, and reduces the error rate of operation, thereby improving the fluency and reliability of the overall warehousing operation. At the same time, the barcode is used to replace the traditional counting sheet and QR code as a positioning reference. The one-dimensional barcode with adhesive backing is attached to the side of the track where the four-way car travels, replacing the counting sheet and QR code, saving the bracket, reducing the construction links, and facilitating maintenance. At the same time, it solves the disadvantages of false induction of photoelectric sensors, overcomes the disadvantages of missed reading and wrong reading caused by the easy accumulation of dust on the QR code when it is placed flat, and can make the four-way shuttle car run more stably.

[0060] 2. In the present invention, the artificial intelligence module learns a large amount of historical data through machine learning algorithms, which can accurately predict and optimize the travel path of the shuttle vehicle, and significantly improve the efficiency of cargo handling in the warehouse; secondly, the module can analyze barcode recognition data in real time, automatically identify and correct errors, ensure the accuracy of cargo information, and reduce losses caused by human errors; thirdly, the artificial intelligence module continuously optimizes scheduling strategies, improves resource utilization, and reduces energy consumption through continuous learning of warehouse operations; finally, the module can also provide intelligent decision-making support, and through in-depth analysis of operational data, help enterprises discover potential problems and improvement points, thereby achieving more refined management and improving the intelligence level of the overall warehousing and logistics system. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 is the overall system block diagram of the present invention;

[0062] Figure 2 It is a system block diagram of the present invention;

[0063] Figure 3 It is a schematic diagram of the position information on the barcode in the present invention. DETAILED DESCRIPTION

[0064] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0065] Reference Figure 1-3 ,

[0066] A four-way shuttle positioning system based on barcode recognition, the system comprising: a barcode recognition module, a communication module, a control module, an artificial intelligence module, a cloud platform module, a scheduling module and an auxiliary module;

[0067] The barcode recognition module is connected to the communication module and sends the read barcode information to the control module;

[0068] The communication module is connected to the control module to send the barcode information and motor encoder data to the control module;

[0069] The control module is connected to the artificial intelligence module, sends real-time coordinate data to the artificial intelligence module, and receives path optimization and task prediction instructions;

[0070] The control module is connected to the cloud platform module, sends real-time coordinate data to the cloud platform module, and receives remote monitoring and management instructions;

[0071] The control module is connected to the scheduling module, sends real-time coordinate data to the scheduling module, and receives task scheduling instructions;

[0072] The artificial intelligence module is connected to the cloud platform module, sends the prediction results to the cloud platform module, and receives cloud data for training and learning;

[0073] The artificial intelligence module is internally configured with a data preprocessing module, a feature engineering module, a model training module, a model evaluation module, a model deployment module, a model update module, and a visualization module;

[0074] The feature scaling of the data preprocessing module uses a standardized method to scale the coordinate data features to a range of a mean of 0 and a standard deviation of 1. The calculation formula is:

[0075] x' = (x-min) / (max-min);

[0076] x' represents the normalized eigenvalue; x represents the original eigenvalue; min represents the minimum value of all the values ​​of the feature; max represents the maximum value of all the values ​​of the feature;

[0077] The data preprocessing module uses the mean filling method to replace the missing values ​​with the average value of all the values ​​of the feature, and uses the two valid coordinate data before and after to estimate the missing coordinate data through linear interpolation. The calculation formula is:

[0078] x=x0+(x1-x0)*(t-t0) / (t1-t0);

[0079] Where x represents the estimated coordinate data;

[0080] x0, x1 represent the two valid coordinate data before and after;

[0081] t0, t1 represent the timestamps corresponding to the two valid coordinate data before and after;

[0082] t: The timestamp corresponding to the missing coordinate data.

[0083] The feature engineering module evaluates the amount of information a feature provides to the target variable through information gain and selects the feature with the largest information gain. The calculation formula is:

[0084] IG(T,A)=Entropy(T)-Σ|T_v| / |T|*Entropy(T_v);

[0085] Where: IG(T,A) represents the information gain of feature A to target variable T

[0086] Entropy(T) represents the entropy of the target variable T

[0087] |T_v| represents the number of samples where feature A takes the value v

[0088] |T| represents the total number of samples

[0089] Entropy(T_v) represents the entropy of samples with feature A value v

[0090] The feature engineering module uses the principal component analysis method to convert the original features into new features through linear transformation, reducing the feature dimension while retaining most of the information. The formula is:

[0091] Z = W*X; where Z represents the new feature matrix, W represents the orthogonal matrix including the feature vector, and X represents the original feature matrix;

[0092] Suppose you need to perform feature engineering on the barcode data in a four-way shuttle positioning system to improve the performance of the path optimization model.

[0093] 1. Feature Selection:

[0094] The information gain method is used to select the features that have the greatest impact on the path optimization target variable, such as barcode recognition time and barcode recognition distance.

[0095] 2. Feature extraction:

[0096] Principal component analysis is used to extract new features and reduce feature dimensions while retaining most of the information.

[0097] 3. Feature Dimensionality Reduction:

[0098] Select some important features, such as retaining the first three principal components, to reduce the feature dimension.

[0099] The algorithm flow of the model training module using the linear regression model for training includes:

[0100] Step S1: Initialize model parameters and set initial weights w and bias b.

[0101] Step S2: Forward propagation, calculate the predicted value y^, the calculation formula is:

[0102] y^(x)=w·x+b;

[0103] Where y^(x) represents the predicted output for input x; w represents the weight vector; x represents the input feature vector; b represents the bias term;

[0104] Step S3: Calculate the loss function and use the mean square error (MSE) as the loss function.

[0105] Formula: L(y,y)=1 / nΣ(yi-y^_i)2, where L(y,y^) represents the loss function; y represents the true value vector; y^ represents the predicted value vector; n represents the number of samples;

[0106] Step S4: Back propagation, calculate the gradient of the loss function with respect to the model parameters.

[0107] The gradient calculation formula for weight w is:

[0108] The gradient calculation formula for bias b is:

[0109] Step S5: Update model parameters and update weights and biases using gradient descent.

[0110] The updated weight calculation formula is:

[0111] The update bias calculation formula is:

[0112] Where α represents the learning rate;

[0113] Step S6: Repeat steps S2-5 until the model converges. The criterion for model convergence may be that the loss function is lower than a certain threshold, or the number of iterations reaches a preset value.

[0114] Suppose a linear regression model is used to predict the X coordinate position of a four-way shuttle.

[0115] Step 1: Initialize model parameters

[0116] w = [0, 0, ..., 0] (the dimension is the same as the number of features);

[0117] b = 0;

[0118] Step 2-6: Use the above steps to perform forward propagation, calculate loss, back propagation, and parameter update until the model converges. The model training module selects a suitable algorithm, initializes parameters, performs forward propagation and back propagation, and updates parameters to ultimately train a model that can accurately predict or classify. In practical applications, it may be necessary to adjust parameters such as the learning rate and number of iterations, and use regularization techniques to prevent overfitting.

[0119] The model evaluation module is a key link in the artificial intelligence system for measuring model performance and accuracy. It includes a series of tests and verifications on the trained model to evaluate its performance on unknown data. Evaluation indicators may include accuracy, recall, F1 score, and mean square error, depending on the type of model and the application scenario. Through cross-validation, holdout testing, and A / B testing methods, the model evaluation module can provide quantitative indicators of model reliability to help researchers or engineers decide whether the model is ready for practical application or whether further adjustments and optimizations are needed;

[0120] The model deployment module is responsible for integrating the trained and evaluated model into the actual working environment. This process includes converting the model into a format that can run in a production environment, such as encapsulating the model into an API service and deploying it to a cloud server or edge device. Model deployment also needs to consider the model's operating efficiency, scalability, security, and monitoring and maintenance factors. In addition, the deployment module may also include model version control to ensure that it can quickly roll back to a stable version when problems occur, as well as implementing the model's continuous integration and continuous deployment (CI / CD) process.

[0121] The model update module is a dynamic part of the AI ​​system that ensures that the model maintains its accuracy and relevance over time and as data accumulates. This module involves periodically retraining the model to incorporate new data points and features to adapt to possible changes in data distribution or concept drift. Model updates may require automated processes to detect degradation in model performance and trigger the retraining process. In addition, this module also includes the evaluation and validation of the new model to ensure that the updated model does not introduce new errors or reduce overall performance;

[0122] The visualization module provides an intuitive model representation and data analysis tool for the AI ​​system. It displays the model's input, output, performance indicators, and prediction results through charts, dashboards, and interactive interfaces, allowing non-technical users to understand and analyze the model's behavior. The visualization module can include data exploration tools, monitoring of the model training process, explanation of prediction results, and visualization of the model's decision path. With these tools, researchers and decision makers can gain a deeper understanding of how the model works, discover patterns in the data, and identify possible deviations or anomalies, thereby improving the transparency and trust of the model.

[0123] The barcode recognition module is the front-end sensing unit of the four-way shuttle positioning system, responsible for collecting and processing barcode information in the warehouse. The barcode recognition module includes a barcode scanner or camera sensor, and the corresponding image recognition and processing software. Its main task is to recognize and extract barcode information from the complex warehouse environment, and then convert this information into a usable data format, such as barcode number and location coordinates.

[0124] The communication module is the bridge for information exchange between various components in the four-way shuttle positioning system. The communication module is responsible for transmitting and exchanging data between the barcode recognition module, control module, and artificial intelligence module;

[0125] The control module is the execution unit of the four-way shuttle positioning system. The control module receives instructions from the artificial intelligence module and directly controls the movement of the shuttle. The control module includes a motion controller, a driver, and an actuator, which is responsible for converting instructions into specific actions of the shuttle, including forward, backward, and turn.

[0126] The cloud platform module provides a remote data storage, processing and analysis platform for the four-way shuttle positioning system. The cloud platform module allows the system to upload data to the cloud for large-scale data processing and analysis, while supporting remote monitoring and fault diagnosis. The cloud platform module can also realize data sharing and collaborative operations among multiple warehouses to improve resource utilization. In addition, the cloud platform module has powerful computing and storage capabilities, which can support the operation of complex algorithms and the storage of large amounts of historical data;

[0127] The auxiliary module is a support unit for the four-way shuttle positioning system. It includes power management, safety monitoring, and alarm system auxiliary functions. The power management module ensures that the system can be stably powered in both normal and emergency situations.

[0128] The scheduling module is the task allocation center in the four-way shuttle positioning system. It intelligently plans the shuttle's travel path and operation sequence based on the warehouse's operational requirements, the shuttle's status and location information. The scheduling module's core algorithms include path planning, task scheduling, and resource optimization, with the goal of improving warehouse operation efficiency and reducing waiting time and transportation distance. The scheduling module also needs to consider real-time warehouse changes, such as cargo dynamics and equipment failures, to achieve dynamic scheduling.

[0129] A method for using a four-way shuttle positioning system based on barcode recognition specifically includes:

[0130] 1. Use barcodes to replace traditional counting sheets and QR codes as positioning references. One-dimensional barcodes with adhesive backing are affixed to the side of the track where the four-way shuttle travels, replacing counting sheets and QR codes. This saves brackets, reduces construction links, and facilitates maintenance. At the same time, it solves the drawbacks of false sensing of photoelectric sensors and overcomes the shortcomings of QR codes being easily accumulated with dust when laid flat, resulting in missed readings and wrong readings, making the four-way shuttle run more stable.

[0131] 2. The barcode recognition sensor is installed on the four-way car, and its position corresponds to the barcode on the track;

[0132] 3. Compared with the two-dimensional code recognition sensor, the barcode recognition sensor has the advantage of fast recognition speed, which improves the real-time performance of code reading when the four-way vehicle is running;

[0133] 4. The number on the barcode contains the location information of the three-dimensional library. For example, the barcode number is 01201301, where the first 3 digits "012" represent the 12th grid position in the X direction of the three-dimensional library, the middle 3 digits "013" represent the 13th grid position in the Y direction, 01 represents the first floor of the three-dimensional library, and the middle black dot is the center point of the barcode on the left and right, which is convenient for installation and labeling, as shown below Figure 3 As shown;

[0134] 5. After the barcode sensor recognizes the barcode, it uploads the barcode number and the offset between the sensor center and the barcode center to the four-way vehicle control system through CAN communication. The four-way vehicle control system can obtain the current point coordinates based on the barcode number. For example, if the data "01201301+10" is obtained, it means that the current point coordinates are X012, Y013, Z01; the current offset center of the four-way vehicle is +1mm;

[0135] 6. When the four-way vehicle enters the automatic state manually, the current absolute coordinates are calculated according to the current point coordinate position (such as X012, Y013, Z1), the X-axis storage location interval, the Y-axis storage location interval, and the starting coordinate point. The calculation method is as follows:

[0136] 6.1 Assume the absolute coordinates of the starting point X010 Y010 Z01 are x10000, y10000, z01 in mm;

[0137] 6.2 The current warehouse X-axis storage location interval is Xm=1500mm, the Y-axis storage location interval is 1200mm; the layer interval is 1;

[0138] 6.3 Absolute coordinate calculation of the current coordinate point (such as X012, Y013, Z1)

[0139] X=(12-10)*Xm+x10000=13000;

[0140] Y=(13-10)*Ym+y10000=13600;

[0141] Z = 1;

[0142] 7. After the four-way shuttle car is in automatic operation (assuming it is reset to the automatic state at X012 Y013 Z01), the motor system data is read in real time, and the coordinates of the four-way car are calculated in real time according to the encoder data of the four-way car motor;

[0143] The calculation method is as follows (assuming the current running direction is X):

[0144] 7.1 The difference between the last and current values ​​of the motor encoder δp, the number of encoder pulses n (how many pulses per revolution);

[0145] 7.2 The reduction ratio b of the motor reducer;

[0146] 7.3 Four-way wheel diameter d;

[0147] From the above, we can get that the running distance of the four-way vehicle is δd = δp*Π*d / b / n;

[0148] The coordinate difference between the two motor data obtained by the four-way car is δd. The current X coordinate is 13000+δd, and the Y and Z coordinates remain unchanged, Y=13600Z=1;

[0149] By repeating this cycle, the coordinates of the four-way vehicle can be calculated in real time;

[0150] 8. When the four-way vehicle receives the barcode information uploaded by the barcode sensor during operation, it immediately updates the current coordinates to eliminate the accumulated error calculated in step 7;

[0151] For example: When the four-way car obtains the barcode information "01301301+10" uploaded by the barcode sensor, its coordinates can be obtained according to step 6 as X=14500, Y=13600Z=1. At this time, the coordinates calculated in step 7 are assumed to be X=14460, Y=13600Z=1. The current coordinates are updated to eliminate the accumulated error of the calculation and updated to X=14500+10=14510, Y=13600Z=1.

[0152] 9. Through the above, the four-way vehicle can obtain coordinate information in real time when it is running, which can be used for scheduling by the (WCS) four-way vehicle scheduling system.

[0153] From the above, it can be seen that the present invention significantly improves the efficiency and accuracy of warehouse operations through its high-precision barcode recognition and intelligent scheduling functions. The system can automatically and quickly identify the barcode of the goods, and command the shuttle car to accurately perform the handling task, which greatly reduces manual intervention and operation time, while reducing the error rate of operation, thereby improving the fluency and reliability of the overall warehousing operation. At the same time, the barcode is used to replace the traditional counting sheet and QR code as a positioning reference. The one-dimensional barcode with adhesive backing is attached to the side of the track where the four-way car travels, replacing the counting sheet and QR code, saving the bracket, reducing the construction links, and facilitating maintenance. At the same time, it solves the disadvantages of false induction of photoelectric sensors, overcomes the disadvantages of missed reading and wrong reading caused by the easy accumulation of dust on the QR code when it is placed flat, and can make the operation of the four-way shuttle car more stable.

[0154] In the present invention, the artificial intelligence module learns a large amount of historical data through machine learning algorithms, which can accurately predict and optimize the travel path of the shuttle vehicle, and significantly improve the cargo handling efficiency in the warehouse; secondly, the module can analyze barcode recognition data in real time, automatically identify and correct errors, ensure the accuracy of cargo information, and reduce losses caused by human errors; thirdly, the artificial intelligence module continuously optimizes scheduling strategies, improves resource utilization, and reduces energy consumption through continuous learning of warehouse operations; finally, the module can also provide intelligent decision-making support, and through in-depth analysis of operational data, help enterprises discover potential problems and improvement points, thereby achieving more refined management and improving the intelligence level of the overall warehousing and logistics system.

[0155] It should be noted that, in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the term "comprise" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0156] The above description enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A four-way shuttle positioning system based on barcode recognition, characterized in that: The system includes: a barcode recognition module, a communication module, a control module, an artificial intelligence module, a cloud platform module, a scheduling module and an auxiliary module; The barcode recognition module is connected to the communication module to send the read barcode information to the control module; The communication module is connected to the control module to send the barcode information and the motor encoder data to the control module; The control module is connected to the artificial intelligence module, sends real-time coordinate data to the artificial intelligence module, and receives path optimization and task prediction instructions; The control module is connected to the cloud platform module, sends real-time coordinate data to the cloud platform module, and receives remote monitoring and management instructions; The control module is connected to the scheduling module, sends real-time coordinate data to the scheduling module, and receives task scheduling instructions; The artificial intelligence module is connected to the cloud platform module, sends the prediction results to the cloud platform module, and receives cloud data for training and learning; The artificial intelligence module is internally provided with a data preprocessing module, a feature engineering module, a model training module, a model evaluation module, a model deployment module, a model updating module and a visualization module.

2. A four-way shuttle positioning system based on barcode recognition as claimed in claim 1, characterized in that: The feature scaling of the data preprocessing module uses a standardized method to scale the coordinate data features to a range of a mean of 0 and a standard deviation of 1. The calculation formula is: x' = (x-min) / (max-min); x' represents the normalized eigenvalue; x represents the original eigenvalue; min represents the minimum value of all the values ​​of the feature; max represents the maximum value of all the values ​​of the feature; The data preprocessing module uses the mean filling method to replace the missing values ​​with the average value of all the values ​​of the feature, and uses the two valid coordinate data before and after to estimate the missing coordinate data through linear interpolation. The calculation formula is: x=x0+(x1-x0)*(t-t0) / (t1-t0); Where x represents the estimated coordinate data; x0, x1 represent the two valid coordinate data before and after; t0, t1 represent the timestamps corresponding to the two valid coordinate data before and after; t: The timestamp corresponding to the missing coordinate data.

3. A four-way shuttle positioning system based on barcode recognition as claimed in claim 1, characterized in that: The feature engineering module evaluates the amount of information a feature provides to the target variable through information gain, and selects the feature with the largest information gain, and the calculation formula is: IG(T,A)=Entropy(T)-Σ|T_v| / |T|*Entropy(T_v); Where: IG(T,A) represents the information gain of feature A to target variable T Entropy(T) represents the entropy of the target variable T |T_v| represents the number of samples where feature A takes the value v |T| represents the total number of samples Entropy(T_v) represents the entropy of samples with feature A value v The feature engineering module uses the principal component analysis method to convert the original features into new features through linear transformation, reducing the feature dimension while retaining most of the information. The formula is: Z=W*X; wherein Z represents the new feature matrix, W represents the orthogonal matrix including the feature vector, and X represents the original feature matrix.

4. A four-way shuttle positioning system based on barcode recognition as claimed in claim 1, characterized in that: The algorithm flow of the model training module using the linear regression model for training includes: Step S1: Initialize model parameters, set initial weight w and bias b; Step S2: Forward propagation, calculate the predicted value y^, the calculation formula is: y^(x)=w·x+b; Where y^(x) represents the predicted output for input x; w represents the weight vector; x represents the input feature vector; b represents the bias term; Step S3: Calculate the loss function and use the mean square error (MSE) as the loss function; Formula: L(y,y)=1 / nΣ(yi-y^_i)2, where L(y,y^) represents the loss function; y represents the true value vector; y^ represents the predicted value vector; n represents the number of samples; Step S4: Back propagation, calculating the gradient of the loss function with respect to the model parameters; The gradient calculation formula for weight w is: The gradient calculation formula for bias b is: Step S5: Update model parameters, using gradient descent method to update weights and biases; The updated weight calculation formula is: The update bias calculation formula is: Where α represents the learning rate; Step S6: Repeat steps S2-5 until the model converges. The criterion for model convergence is that the loss function is lower than a certain threshold, or the number of iterations reaches a preset value.

5. A four-way shuttle positioning system based on barcode recognition as claimed in claim 1, characterized in that: The model evaluation module is a key link in the artificial intelligence system for measuring model performance and accuracy. It includes a series of tests and verifications on the trained model to evaluate its performance on unknown data. The evaluation indicators include accuracy, recall, F1 score, and mean square error, depending on the type of model and the application scenario. The model deployment module is responsible for integrating the trained and evaluated model into the actual working environment; The model deployment module converts the model into a format that runs in a production environment, encapsulates the model into an API service, and deploys it to a cloud server or edge device.

6. A four-way shuttle positioning system based on barcode recognition as claimed in claim 1, characterized in that: The model updating module maintains its accuracy and relevance over time and as data accumulates; this module involves periodically retraining the model to incorporate new data points and features to adapt to changes in data distribution or concept drift; The visualization module provides an intuitive model representation and data analysis tool for the AI ​​system; it displays the model’s input, output, performance indicators, and prediction results through charts, dashboards, and interactive interfaces, allowing non-technical users to understand and analyze the model’s behavior; The visualization module includes data exploration tools, monitoring of the model training process, interpretation of prediction results, and visualization of the model decision path.

7. A four-way shuttle positioning system based on barcode recognition as claimed in claim 1, characterized in that: The barcode recognition module is the front-end sensing unit of the four-way shuttle positioning system, responsible for collecting and processing barcode information in the warehouse; the barcode recognition module includes a barcode scanner or camera sensor, and supporting image recognition and processing software; it recognizes and extracts barcode information from complex warehouse environments, and then converts this information into a usable data format.

8. A four-way shuttle positioning system based on barcode recognition as claimed in claim 1, characterized in that: The communication module is a bridge for information exchange between various components in the four-way shuttle positioning system; the communication module is responsible for transmitting and exchanging data between the barcode recognition module, the control module, and the artificial intelligence module unit; The control module is the execution unit of the four-way shuttle positioning system. The control module receives instructions from the artificial intelligence module and directly controls the movement of the shuttle. The control module includes a motion controller, a driver and an actuator, which are responsible for converting instructions into specific actions of the shuttle, including forward, backward and turning.

9. A four-way shuttle positioning system based on barcode recognition as claimed in claim 1, characterized in that: The cloud platform module provides a remote data storage, processing and analysis platform for the four-way shuttle positioning system; the cloud platform module allows the system to upload data to the cloud for large-scale data processing and analysis, while supporting remote monitoring and fault diagnosis; the cloud platform module also enables data sharing and collaborative operations between multiple warehouses; The auxiliary module is a support unit of the four-way shuttle positioning system, which includes power management, safety monitoring, and alarm system auxiliary functions; the power management module ensures that the system can have stable power supply in normal and emergency situations.

10. A four-way shuttle positioning system based on barcode recognition as claimed in claim 1, characterized in that: The scheduling module is the task allocation center in the four-way shuttle positioning system. It intelligently plans the shuttle's driving path and operation sequence according to the warehouse's operational requirements, the shuttle's status and location information; the core algorithms of the scheduling module include path planning, task scheduling and resource optimization.

Images