Solder paste intelligent storage method and solder paste intelligent storage device based on image recognition

Abstract

The invention relates to the technical field of image recognition, in particular to an intelligent solder paste storage method and an intelligent solder paste storage device based on image recognition, and realizes intelligent management of solder paste storage by introducing computer vision, deep learning, an artificial intelligence optimization algorithm and an environment monitoring technology. OCR, infrared ray and ultraviolet ray imaging technologies are utilized, solder paste packaging information is automatically identified, and leakage, pollution and deterioration conditions are detected; the optimal storage position is calculated based on a reinforcement learning algorithm, and accurate carrying and storage of the solder paste are executed through an intelligent device; a built-in sensor monitors temperature, humidity and air quality in real time, and a storage environment is automatically regulated and controlled in combination with a prediction model; the manufacturing execution system and the digital twin virtual warehouse are integrated, full-process tracking, inventory prediction and intelligent replenishment early warning are achieved, and the safety, the efficiency and the management level are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of image recognition, and in particular to an intelligent storage method and device for solder paste based on image recognition. Background Art

[0002] In the prior art, as a key material in the surface mount technology (SMT) of the electronics manufacturing industry, the storage management of solder paste has an important impact on production efficiency and product quality. However, traditional solder paste storage methods mainly rely on manual management or simple automated equipment, suffering from problems such as low efficiency, high error rate, and lack of intelligence. In traditional methods, information such as the batch number, model, and production date of solder paste is usually recorded manually or identified through simple barcode / RFID tags. The storage and retrieval processes mostly rely on manual operations, easily leading to low storage efficiency, high error rate, and even the problem of deterioration and failure due to improper storage of solder paste. In addition, most existing systems are unable to achieve real-time monitoring and automatic regulation of the solder paste storage environment (such as temperature and humidity), making it difficult to ensure the quality and use safety of solder paste.

[0003] Existing automated warehousing equipment (such as automated storage and retrieval systems AS / RS and automated guided vehicles AGV) mainly relies on simple label scanning technologies (such as barcodes or RFID tags), and has problems such as limited recognition accuracy, large environmental influence, and inability to judge the state of solder paste when identifying solder paste packaging information. For example, barcodes and RFID tags may cause recognition failures in the case of dust, damage, or signal interference, and can only obtain static information on the package, unable to judge whether there is leakage, contamination, or deterioration of the solder paste through image analysis. In addition, these systems lack intelligent dynamic storage decision-making capabilities and are unable to perform intelligent scheduling and management based on factors such as the batch number, usage frequency, and storage environment of the solder paste.

[0004] To solve the above problems, there is an urgent need for a smart storage method and device for solder paste based on image recognition. By introducing computer vision, deep learning, Internet of Things (IoT), and automation control technologies, it realizes the full-process management of intelligent recognition, dynamic storage, automated handling, environmental monitoring, and data feedback of solder paste. This method can not only automatically extract label information (such as batch number, model number, production date) on the solder paste packaging through image recognition technology, but also determine whether there are abnormalities such as leakage, contamination, or deterioration of the solder paste packaging through image analysis. At the same time, the system can use AI algorithms to analyze multi-dimensional data such as the usage frequency, batch requirements, and storage environment of the solder paste, realize dynamic optimization of the storage location, ensure the efficient use of warehouse space and the improvement of access efficiency. In addition, through IoT technology, the storage environment is monitored and automatically regulated in real time to ensure that the solder paste is stored under suitable conditions and extend its service life. Through intelligent and closed-loop management, this method can significantly improve the safety, reliability, and management efficiency of solder paste storage, and provide a more efficient and secure intelligent storage solution for electronic manufacturing enterprises. Summary of the Invention

[0005] In view of the problems existing in the above-mentioned prior art, the present invention is proposed.

[0006] Therefore, the technical problems to be solved by the present invention are: aiming to solve the technical problems such as low efficiency, inaccurate recognition, insufficient environmental regulation, incomplete tracking, and rigid storage planning existing in traditional solder paste storage management. Traditional methods rely on manual records and simple label scanning, making it difficult to detect packaging leakage, contamination, or deterioration in a timely manner, and at the same time unable to monitor temperature, humidity, and air quality in real time, resulting in a decline in solder paste quality and production interruption. The present invention realizes the closed-loop management of solder paste from automatic recognition, intelligent handling, precise storage to full-process tracking and self-learning feedback by introducing computer vision, deep learning, artificial intelligence optimization algorithms, environmental monitoring and regulation, and digital twin technology, significantly improving the safety, accuracy, and efficiency of solder paste storage, and meeting the strict requirements of the electronic manufacturing industry.

[0007] To solve the above technical problems, the present invention provides the following technical solutions. A smart storage method for solder paste based on image recognition includes: automatically identifying and classifying the solder paste; after automatic identification and classification, calculating the optimal storage location through an artificial intelligence optimization algorithm, and performing the handling and storage of the solder paste through a smart storage device for solder paste; monitoring based on a prediction model and the built-in environmental monitoring and regulation function of the smart storage device for solder paste; integrating manufacturing execution and storage to predict the consumption trend of the solder paste; adopting a self-learning feedback mechanism to improve the storage plan, establishing a digital twin virtual warehouse, and performing anomaly detection.

[0008] As a preferred embodiment of the intelligent solder paste storage method based on image recognition according to the present invention, wherein: the automatic recognition utilizes OCR and image recognition technologies to automatically recognize and extract information from the labels, barcodes, QR codes, and surface states of the solder paste packages, and analyzes key information such as batch numbers, brands, models, and production dates of the solder paste through computer vision technology;

[0009] The classification includes detecting the solder paste by combining infrared imaging and ultraviolet imaging for anomaly classification.

[0010] As a preferred embodiment of the intelligent solder paste storage method based on image recognition according to the present invention, wherein: the artificial intelligence optimization algorithm includes comprehensive calculations based on the current storage state, storage operation selection, reward function, discount factor, state transition rate, and future state to obtain the optimal storage location.

[0011] As a preferred embodiment of the intelligent solder paste storage method based on image recognition according to the present invention, wherein: the prediction model includes a model function for predicting the storage state, considering the state transition rate, perturbation coefficient, and environmental variables collected in real time under the current storage state; wherein, the environmental variables are input into the prediction model for dynamic adjustment. When the storage environment temperature exceeds the safety threshold, the refrigeration device is automatically activated to lower the temperature; when the humidity exceeds the safe range, the adjustment device is activated to maintain an appropriate humidity; when the concentration of volatile organic compounds in the air exceeds the standard, the ventilation device is activated to restore the air quality;

[0012] The environmental monitoring and regulation function includes collecting temperature and humidity data in the storage environment in real time and detecting the air quality in the storage area.

[0013] As a preferred embodiment of the intelligent solder paste storage method based on image recognition according to the present invention, wherein: the execution and storage integration includes connecting the intelligent solder paste storage device with a standard interface to achieve data intercommunication;

[0014] Predicting the solder paste consumption trend includes predicting the future consumption trend based on historical production data and solder paste usage records, including dynamic inventory prediction and inventory warning and replenishment suggestions; wherein, the dynamic inventory prediction combines the production line work order plan and historical production data to predict the future solder paste consumption, and allocates the inventory according to the prediction results; when it is recognized that the solder paste inventory is lower than the safety inventory, an early warning notice is automatically sent to provide intelligent replenishment suggestions.

[0015] As a preferred solution of the solder paste intelligent storage method based on image recognition according to the present invention, wherein: the self-learning feedback mechanism includes adaptive update of the value function and optimization of intelligent storage planning; among them, the adaptive update of the value function includes, after performing the storage operation, collecting feedback data during the storage process in real time and updating the value function. When the storage operation has a positive impact, a high value is assigned to the storage operation to complete the feedback mechanism; based on the updated value function, reselect the best storage location; after the new storage operation is completed, monitor the storage result in real time and continuously optimize.

[0016] As a preferred solution of the solder paste intelligent storage method based on image recognition according to the present invention, wherein: the digital twin virtual warehouse includes real-time mirroring of the warehouse state, and collects operation state data in the warehouse in real time through sensors, including the position of the solder paste intelligent storage device, the device operation state, environmental monitoring data, and inventory status. Input the operation state data into the digital twin virtual warehouse, realize the linkage between the virtual and the real through digital twin technology, perform anomaly detection, and complete actual detection and optimization.

[0017] To solve the above technical problems, the present invention provides the following technical solutions: a solder paste intelligent storage method and a solder paste intelligent storage device based on image recognition, including: a control interface and a data intercommunication unit, provided with a standard interface for data intercommunication with an artificial intelligence optimization algorithm and a digital twin virtual warehouse;

[0018] An automatic handling and storage execution mechanism, equipped with an automatic handling device and an in-built navigation and positioning system, to achieve precise handling and storage of the solder paste from the warehousing position to the best storage position;

[0019] An environment monitoring and regulation module, with built-in temperature, humidity, and air quality sensors, collects storage environment data in real time, and is linked with an environment regulation device to automatically adjust the environmental parameters in the warehouse according to the feedback of the prediction model;

[0020] A real-time feedback and self-learning support mechanism, equipped with a feedback data collection unit, monitors handling, storage operations, and environmental changes in real time, and transmits the feedback data to the adaptive update of the value function; supports the self-learning feedback mechanism, supports the optimization of intelligent storage planning, and interacts and feedbacks the operation results with the updated storage strategy to continuously optimize the storage planning.

[0021] A computer device, including a memory and a processor, the memory stores a computer program, and is characterized in that when the processor executes the computer program, the steps of the solder paste intelligent storage method based on image recognition as described above are implemented.

[0022] A computer-readable storage medium has a computer program stored thereon, characterized in that when the computer program is executed by a processor, it implements the steps of the above-mentioned intelligent solder paste storage method based on image recognition.

[0023] Advantages of the present invention: By introducing computer vision, deep learning, artificial intelligence optimization algorithms, and environmental monitoring and control technologies, the present invention realizes the intelligent management of the entire process of solder paste storage. Using OCR, infrared, and ultraviolet imaging technologies to automatically extract solder paste packaging information, and real-time detecting abnormal situations such as packaging leakage, contamination, and deterioration, significantly improving the recognition accuracy and quality control level. At the same time, the system dynamically calculates the optimal storage location according to real-time feedback through optimization algorithms such as reinforcement learning, significantly improving the utilization rate of warehouse space and handling efficiency. The built-in temperature, humidity, and air quality sensors continuously monitor the storage environment, and actively regulate the environmental parameters in combination with the prediction model to ensure that the solder paste is stored under safe and stable conditions and extend its service life. The integration of the manufacturing execution system and the digital twin virtual warehouse realizes full-process tracking, dynamic inventory prediction, and intelligent replenishment warning, reducing the risks of manual operation and management costs. Generally speaking, the present invention significantly improves the safety, stability, and efficiency of solder paste storage, providing an efficient and reliable intelligent storage solution for the electronic manufacturing industry. Description of the Drawings

[0024] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0025] Figure 1 It is a flowchart of an intelligent solder paste storage method based on image recognition provided by an embodiment of the present invention.

[0026] Figure 2 It is a device structure diagram of an intelligent solder paste storage device provided by an embodiment of the present invention. Detailed Embodiments

[0027] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following will make a detailed description of the specific embodiments of the present invention in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all embodiments. All other embodiments obtained by those of ordinary skill in the art without creative efforts based on the embodiments of the present invention should fall within the scope of protection of the present invention.

[0028] Example 1, referring to Figure 1, which is an embodiment of the present invention. This embodiment provides a smart solder paste storage method based on image recognition, including:

[0029] S1: Automatically identify and classify the solder paste.

[0030] It should be noted that, as Figure 1 shown in S1, automatic identification uses OCR and image recognition technologies to automatically identify and extract information from the labels, barcodes, QR codes, and surface states of the solder paste packaging, and analyzes key information such as the batch number, brand, model, and production date of the solder paste through computer vision technology.

[0031] Furthermore, classification includes detecting the solder paste by combining infrared imaging and ultraviolet imaging for anomaly classification;

[0032] Among them, infrared imaging detection uses a high-resolution infrared camera to monitor the temperature distribution of the solder paste packaging in real time, detecting signs of poor packaging sealing, leakage, and contamination; when there is a small leak in the packaging, infrared imaging can identify abnormal changes in the surface temperature of the packaging, such as cold and hot spots caused by gas leakage, or uneven temperature caused by poor sealing; in addition, further analyze internal quality problems of the solder paste through temperature distribution characteristics, such as identifying internal temperature differences caused by flux separation or solder paste component sedimentation, providing a new evaluation dimension for solder paste quality classification.

[0033] Ultraviolet imaging detection irradiates the surface of the solder paste packaging and uses the fluorescence reaction of specific organic substances under ultraviolet light to accurately identify organic pollutants on the surface of the solder paste; for example, by analyzing the fluorescence intensity and spectral characteristics, the system can distinguish flux residues, oil stains, and other pollutants outside the solder paste; on the solder paste packaging returned to the warehouse after being used on the production line, ultraviolet imaging technology can also detect chemical substances or contaminants during the operation remaining on the packaging surface, realizing dynamic monitoring of the storage status of the solder paste.

[0034] Furthermore, the classification process does not rely solely on a single imaging result, but comprehensively analyzes infrared, ultraviolet, and visible light image data through multimodal data fusion; for example, when the system detects abnormal temperature distribution of the packaging through infrared imaging, it will further call the ultraviolet imaging module for verification to determine whether the abnormality is caused by contamination or leakage.

[0035] S2: After automatic identification and classification, calculate the optimal storage location through an artificial intelligence optimization algorithm, and execute the handling and storage of the solder paste through a smart solder paste storage device.

[0036] It should be noted that, as Figure 1 shown in S2, the artificial intelligence optimization algorithm is expressed as:

[0037]

[0038] Among them, V(s) represents the value function in state s, s represents the current storage state, a represents the storage operation selected in the current state, R(s, a) represents the reward function, γ represents the discount factor (0 < γ < 1), P(s'|s, a) represents the state transition rate, and s′ represents the next state;

[0039] Among them, the reward function is expressed as:

[0040] R(s, a) = αU + βQ + θE

[0041] Among them, U represents the storage space utilization rate, by calculating the impact of the current storage operation on the warehouse space utilization, for example, giving priority to placing frequently used solder pastes near the exit. Q represents the solder paste quality retention, based on the storage environment prediction model, evaluating the impact of environmental conditions at a specific location on the solder paste quality, for example, preferentially storing highly sensitive solder pastes in low-temperature stable areas. E represents the operation efficiency, evaluating the efficiency of access operations, for example, minimizing the handling path and the operation time of the robotic arm. α, β, and θ represent weight parameters.

[0042] Furthermore, the optimal action is obtained by calculating V(s) That is, the best storage location;

[0043] After selecting the best storage location, connect to the intelligent solder paste storage device through the control interface and issue storage instructions, including detailed operation information such as the target storage location, handling path, and environmental parameter settings; the intelligent solder paste storage device receives the storage instructions, starts the built-in automatic handling device, and precisely executes the handling and storage operations of the solder paste using the built-in navigation, and real-time feedback on the execution situation to update the state s; during the handling process, the intelligent solder paste storage device will monitor the operation status in real time, including handling speed, path deviation, robotic arm grasping force, etc., and compare with the target path through sensor data to ensure the safety and accuracy of the handling process.

[0044] Even further, continuously optimize the weight parameters in the value function and the reward function through the feedback data to make the storage decision more accurate and flexible.

[0045] S3: Based on the prediction model and the built-in environment monitoring and regulation function of the intelligent solder paste storage device, conduct monitoring.

[0046] It should be noted that, as Figure 1 shown in S3, the prediction model is expressed as:

[0047]

[0048] Among them, Indicates the predicted next storage state, including information such as the location of the solder paste in the warehouse, the temperature, humidity, air quality of the storage environment, and the operating status of the equipment. f(s, a, e) represents the prediction model function, which combines the current state s, the storage operation a, and the environmental variable e to predict the state change of the warehouse after the operation is executed. P(s′|s, a) represents the state transition rate, δ represents the perturbation coefficient, and e represents the environmental variable.

[0049] Furthermore, the environmental monitoring and regulation function collects the environmental data in the warehouse in real time through temperature, humidity, and air quality sensors to monitor the temperature e of the storage environment in real time T and humidity data e H , detects the air quality in the storage area e A ; inputs the environmental variable e = (e T , e H , e A ) into the prediction model for adjustment. When the temperature of the storage environment exceeds the safety threshold and rises above 15°C (exceeding the suitable storage temperature of the solder paste), the refrigeration device is automatically started to lower the temperature of the warehouse; when the humidity of the storage environment exceeds the safety range and exceeds 60% (industry environmental standard), the dehumidification equipment is automatically started to keep the air dry and prevent the solder paste from getting damp; when the concentration of volatile organic compounds in the air quality of the storage environment exceeds the safety standard (500 ppm), the automatic ventilation equipment is started to reduce the pollutant concentration and ensure the safety of the storage environment.

[0050] Even further, compare the next state output by the prediction model with the actual monitoring data. For example, when the model predicts that the humidity will drop below 50% within 10 minutes after the dehumidification equipment is started, but the actual humidity only drops to 55%, the system will adjust the perturbation coefficient δ or the state transition rate P(s′|s, a) in the prediction model; and when the prediction model predicts that an environmental anomaly is about to occur (such as excessive temperature and humidity, non-compliant air quality), it will issue a warning before the actual situation occurs.

[0051] S4: Manufacturing execution and storage integration, predicting the consumption trend of solder paste.

[0052] It should be noted that, as Figure 1 shown in S4, the execution and storage integration includes connecting the intelligent solder paste storage device with a standard interface to achieve data intercommunication;

[0053] Specifically, the data interconnection method uses a standard communication protocol to achieve real-time synchronization of data, including information such as the incoming and outgoing of solder paste, inventory status, storage location, and environmental monitoring data; the system interaction logic includes that when the production line issues a demand for solder paste, the MES system will automatically send an instruction to the intelligent solder paste storage device to retrieve the corresponding model and batch of solder paste; after the storage device executes the handling operation, it will transmit the operation result and the solder paste outgoing information back to the MES system to ensure the consistency of data between the production line and the warehousing system; the full-process tracking includes that the whole process of each batch of solder paste from incoming, storage, monitoring to outgoing is recorded, generating a unique tracking ID to achieve forward tracking (from raw materials to finished products) and reverse tracking (from finished products to raw materials) of product batches, and being able to quickly locate the problem batch in case of abnormal situations to ensure the traceability of production quality.

[0054] Furthermore, predicting the solder paste consumption trend includes predicting the future consumption trend based on historical production data and solder paste usage records, including dynamic inventory prediction and inventory warning and replenishment suggestions;

[0055] Among them, dynamic inventory prediction combines the production line work order plan and historical production data to predict the future solder paste consumption, and actively allocates inventory according to the prediction results; combines the production line work order plan, analyzes the solder paste models and usage amounts required for future production tasks, and predicts the dynamic consumption of different solder paste models through time series analysis; dynamically calculates the required solder paste inventory within the next week or a production cycle by analyzing data such as the running speed, production shifts, and equipment startup rate of the production line; simulates the inventory changes under different production plans, for example, automatically increases the safety inventory level during the production peak period to prevent production interruptions due to insufficient inventory; the prediction model prepares sufficient solder paste inventory in advance by learning the seasonal trends in historical data (such as production demand fluctuations in specific months) to cope with sudden changes in production demand; when the dynamic inventory prediction model identifies that the current inventory level is close to or lower than the safety inventory level, the system will automatically generate an inventory warning notice and push it to the warehouse management personnel or the procurement department through the MES or ERP system; provides inventory warnings and gives intelligent replenishment suggestions based on historical procurement and supply chain data; for example, recommends purchasing a specific batch or quantity of solder paste, or provides comparison information of multiple suppliers to help decision-makers select the optimal procurement plan; supports an automated replenishment process. For example, after being integrated with the supply chain management system (SCM), it can automatically generate a purchase order (PO) and send it to the supplier through the electronic data interchange (EDI) system to achieve a replenishment process without manual intervention; optimizes the replenishment strategy based on information such as inventory turnover rate, procurement lead time, and supplier supply capacity to avoid inventory backlogs caused by over-purchasing or stock-out risks caused by untimely replenishment.

[0056] Furthermore, actively allocate inventory according to the prediction model, and monitor the difference between the prediction result and the actual consumption data; if there is a deviation between the prediction result and the actual situation, the system will automatically adjust the parameters of the prediction model (such as the weight coefficient in the prediction algorithm) to continuously optimize the prediction accuracy; when the prediction model detects that the consumption of a certain type of solder paste is significantly higher than other types, the system will automatically adjust the inventory layout. For example, store the frequently consumed solder paste in a more accessible location to reduce the outbound time and improve storage and production efficiency; when the MES system detects a temporary adjustment of the production plan (such as an increase in production orders or a product model switch), the system will update the inventory prediction model in real time and quickly adjust the solder paste storage strategy according to the new production plan to ensure the efficient operation of the production line.

[0057] S5: Adopt a self-learning feedback mechanism to improve the storage plan, establish a digital twin virtual warehouse, and perform anomaly detection.

[0058] It should be noted that, as Figure 1 shown in S5, the self-learning feedback mechanism includes adaptive update of the value function and optimization of intelligent storage planning;

[0059] Among them, the adaptive update of the value function includes, after performing the storage operation a, collecting the feedback data during the storage process in real time, and updating the value function, which is expressed as:

[0060] V(s)←V(s)+α[R(s,a)+γV(s′)-V(s)]

[0061] If the operation a has a positive impact, a high value will be assigned to the operation a to complete the feedback mechanism; after the solder paste intelligent storage device completes the storage operation, the system collects real-time feedback data through sensors, navigation devices, and environmental monitoring modules, including information such as the utilization rate of the storage location, the quality retention of the solder paste, the operation efficiency, and the change of environmental parameters; for example, after performing the operation of moving the frequently used solder paste closer to the outbound area, the feedback data may include specific values such as the shortened handling time and the improved access efficiency; if the operation has a positive impact (such as an increase in operation efficiency or storage space utilization rate), the system will assign a higher value to the operation and enhance the priority of the operation in subsequent decision-making through the positive feedback mechanism; apply the real-time updated value function to the selection of the next optimal action to enable the model to continuously learn and adapt to the changes in the actual operation environment.

[0062] Furthermore, based on the updated value function, reselect the optimal storage location; by selecting the operation with the highest value function, the system can preferentially execute storage decisions that are most beneficial to the overall storage efficiency, solder paste quality maintenance, and operation safety; automatically correct the state transition probability P(s′|s,a) in the prediction model after each operation to make the prediction model more accurate and continuously optimize the storage plan; after the new storage operation is completed, monitor the storage results in real time and continuously optimize; after the new storage operation is executed, the system will start the environmental monitoring module to continuously monitor the storage results. For example, monitor the temperature and humidity changes in the storage area, air quality data, solder paste packaging status, etc.; compare and analyze the real-time monitored data with the expected results output by the prediction model. For example, after the expected temperature adjustment, it should drop to 20°C, but actually only drops to 22°C, automatically adjust the operation mode of the refrigeration equipment, and store the feedback data in the historical database as the data source for the next model training. Through the continuously accumulated feedback information, improve the accuracy of model prediction and the reliability of operation decisions.

[0063] Furthermore, the digital twin virtual warehouse includes a real-time mirror of the warehouse state. The operating state data in the warehouse is collected in real time through sensors and input into the digital twin virtual warehouse. Through digital twin technology, the virtual and the real are linked together to facilitate anomaly detection and assist in actual detection and optimization.

[0064] Specifically, the digital twin virtual warehouse collects the operation status data in the warehouse in real time through a sensor network, including the position of the solder paste intelligent storage device, the device operation status, the environmental monitoring data (such as temperature, humidity, air quality), and the inventory status; these data are synchronously transmitted to the virtual warehouse system through IoT (Internet of Things) technology, and the operation status of the physical warehouse is presented in real time on the virtual interface, forming an accurate mapping between the virtual and physical environments; the system realizes the linkage between the virtual and physical through digital twin technology. When the sensors in the physical warehouse detect abnormal environmental parameters (such as excessive humidity, decreased air quality), the virtual warehouse will synchronously display the abnormal status, and the system simulates various countermeasures through the virtual model; for example, when the prediction model shows that the VOC concentration in the air quality exceeds the standard, the virtual warehouse will simulate operations such as starting the ventilation system, adjusting the air volume, and calculating the pollutant dissipation time, and select the optimal solution through the simulation results to be applied to the actual warehouse; compare the simulation data of the virtual operation with the feedback data after the actual execution, and automatically adjust the perturbation coefficient δ and the state transition rate P(s'|s,a) in the prediction model through comparative analysis to improve the system's adaptability to environmental changes; the digital twin virtual warehouse can not only provide emergency handling suggestions when an anomaly occurs, but also perform predictive maintenance and optimization operations during daily operation; for example, when the system predicts that the temperature in a certain storage area may exceed the safety threshold in the future, the virtual warehouse will simulate the effects of different operations such as starting the refrigeration equipment and adjusting the storage location, and select the best operation suggestion to be pushed to the operator or automatically executed by analyzing the cost and benefit of each operation;

[0065] Through the closed-loop mechanism of "prediction - execution - feedback - optimization", the whole process from data collection, operation execution, feedback analysis to model optimization is realized automatically; after each operation is completed, compare the actual execution data with the expected data output by the prediction model, and adjust the model parameters through machine learning algorithms to make the model more accurate in prediction and decision-making; in the long-term operation, the system can identify long-term effective storage strategies, for example, certain specific types of solder paste have the best storage effect in a low-humidity area, and automatically solidify this experience into storage rules to improve the system's automatic decision-making ability; through the combination of the self-learning feedback mechanism and the digital twin virtual warehouse, the present invention realizes the continuous optimization of the solder paste intelligent storage system, enabling the system to always maintain an efficient, stable, and safe operation state in a dynamically changing production environment.

[0066] Embodiment 2, refer to Figure 2 , which is an embodiment of the present invention, provides a solder paste intelligent storage device, including:

[0067] The control interface and data interconnection unit is equipped with standard interfaces for data interconnection with artificial intelligence optimization algorithms and digital twin virtual warehouses; it supports mainstream industrial protocols (such as MQTT, HTTP) and realizes seamless data communication with the outside through two-way data streams. Among them, the two-way data streams include input data (receiving storage operation instructions sent by the intelligent storage decision module, including the best storage location, handling path planning, environmental control parameter settings, etc.) and output data (transmitting the operating status of the solder paste intelligent storage device, including handling progress, environmental monitoring data, and operation feedback, to the upper-level system to provide basic data for the self-learning feedback mechanism).

[0068] The automatic handling and storage execution mechanism is equipped with an automatic handling device with a built-in navigation and positioning system to achieve precise handling and storage of solder paste from the incoming location to the best storage location; it uses a robotic arm to achieve automatic handling of solder paste, with 6-axis multi-degree-of-freedom operation to accurately complete operations such as grasping, moving, and placing of solder paste, and is equipped with a vision sensor to achieve intelligent obstacle avoidance; the built-in navigation and positioning system achieves high-precision positioning in the warehouse through inertial navigation; through the force feedback sensor and displacement sensor installed on the handling device, real-time monitoring of the grasping force and moving accuracy during the operation is achieved.

[0069] The environmental monitoring and control module is built with temperature, humidity, and air quality sensors to collect storage environment data in real time (the sensor data sampling frequency can be flexibly adjusted according to environmental changes, supporting an adjustable sampling period from 1 second to 60 minutes to ensure the real-time and accuracy of data), and is linked with the environmental control device to automatically adjust the environmental parameters in the warehouse according to the feedback of the prediction model. Among them, the environmental control device includes refrigeration / heating equipment, humidification / dehumidification equipment, and a ventilation system.

[0070] The real-time feedback and self-learning support mechanism is equipped with a feedback data collection unit to monitor handling, storage operations, and environmental changes in real time, and transmit the feedback data to the value function for adaptive update; it supports the self-learning feedback mechanism and intelligent storage planning optimization, and interacts and feedbacks the operation results with the updated storage strategy to continuously optimize the storage planning; after completing the storage operation, it automatically analyzes the positive or negative impact of the operation results. For example, if a certain operation significantly improves the storage efficiency, a higher value will be assigned to similar operations in future decisions. By analyzing historical operation data and the current environmental state, the system can automatically generate new storage planning strategies; simulate the effects of different storage strategies in the digital twin virtual warehouse, and compare the simulation results with the actual execution data to improve the overall intelligence level of the storage system.

[0071] This embodiment also provides a computing device applicable to the case of the solder paste intelligent storage method based on image recognition, including:

[0072] A memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the intelligent solder paste storage method based on image recognition as proposed in the above embodiments.

[0073] This embodiment also provides a storage medium, on which a computer program is stored. When the program is executed by a processor, it implements the intelligent solder paste storage method based on image recognition as proposed in the above embodiments.

[0074] The storage medium proposed in this embodiment and the intelligent solder paste storage method based on image recognition proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be referred to in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0075] If the above functions are implemented in the form of software function 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, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0076] Logic and / or steps described in other ways herein, for example, can be considered as a definite sequence list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device), or in combination with these instruction execution systems, apparatus, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.

[0077] It should be understood that various parts of the present invention can be implemented by hardware, software, firmware, or a combination thereof. In the above embodiments, multiple 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 or a combination of the following techniques well known in the art can be used: discrete logic circuits having logic gate circuits for implementing logic functions on data signals, application specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), and the like.

Claims

1. An intelligent solder paste storage method based on image recognition, characterized in that: include: Automatically identify and classify solder paste; After automatic identification and classification, the best storage location is calculated through artificial intelligence optimization algorithm, and the solder paste is transported and stored through the solder paste intelligent storage device; Monitoring based on predictive models and built-in environmental monitoring and control functions of solder paste intelligent storage devices; Integration of manufacturing execution and storage to predict solder paste consumption trends; Adopt self-learning feedback mechanism to improve storage planning, establish digital twin virtual warehouse, and perform anomaly detection.

2. The solder paste intelligent storage method based on image recognition according to claim 1, characterized in that: The automatic recognition uses OCR and image recognition technology to automatically recognize and extract information from the label, barcode, QR code and surface state of the solder paste packaging, and analyzes key information such as the batch number, brand, model and production date of the solder paste through computer vision technology; The classification includes combining infrared imaging and ultraviolet imaging to detect solder paste and perform abnormal classification.

3. The solder paste intelligent storage method based on image recognition according to claim 2, characterized in that: The artificial intelligence optimization algorithm includes a comprehensive calculation based on the current storage state, storage operation selection, reward function, discount factor, state transition rate and future state to obtain the optimal storage location.

4. The method for intelligent storage of solder paste based on image recognition according to claim 3, characterized in that: The prediction model includes a model function for predicting the storage state, taking into account the state transition rate, disturbance coefficient and environmental variables collected in real time under the current storage state; wherein the environmental variables are input into the prediction model for dynamic adjustment, when the storage environment temperature exceeds the safety threshold, the refrigeration device is automatically started to reduce the temperature; when the humidity exceeds the safety range, the regulating device is started to maintain the appropriate humidity; when the concentration of volatile organic compounds in the air exceeds the standard, the ventilation equipment is started to restore the air quality; The environmental monitoring and control function includes real-time collection of temperature and humidity data in the storage environment and detection of air quality in the storage area.

5. The method for intelligent storage of solder paste based on image recognition according to claim 4, characterized in that: The execution and storage integration includes connecting the solder paste intelligent storage device with the standard interface to achieve data intercommunication; The predicted solder paste consumption trend includes predicting future consumption trends based on historical production data and solder paste usage records, including dynamic inventory forecasting and inventory warning and replenishment suggestions; wherein, the dynamic inventory forecast combines the production line work order plan and historical production data to predict future solder paste consumption, and allocates inventory according to the forecast results; when it is identified that the solder paste inventory is lower than the safety inventory, an early warning notification is automatically sent to provide intelligent replenishment suggestions.

6. The method for intelligent storage of solder paste based on image recognition according to claim 5, characterized in that: The self-learning feedback mechanism includes adaptive updating of the value function and intelligent storage planning optimization; wherein, the adaptive updating of the value function includes, after executing the storage operation, collecting feedback data in the storage process in real time, updating the value function, and when the storage operation produces a positive impact, assigning a high value to the storage operation to complete the feedback mechanism; based on the updated value function, reselecting the best storage location; based on the completion of the new storage operation, monitoring the storage results in real time and continuously optimizing.

7. The method for intelligent storage of solder paste based on image recognition according to claim 6, characterized in that: The digital twin virtual warehouse includes a real-time mirror warehouse status, and collects operating status data in the warehouse in real time through sensors, including the location of the solder paste intelligent storage device, equipment operating status, environmental monitoring data and inventory status. The operating status data is input into the digital twin virtual warehouse, and the digital twin technology is used to realize virtual-real linkage, perform anomaly detection, and complete actual detection and optimization.

8. The solder paste intelligent storage device of the solder paste intelligent storage method based on image recognition according to any one of claims 1 to 7, characterized in that: include: The control interface and data intercommunication unit is equipped with a standard interface for data intercommunication with the artificial intelligence optimization algorithm and the digital twin virtual warehouse; Automatic handling and storage actuator, equipped with automatic handling device and built-in navigation and positioning system, to achieve accurate handling and storage of solder paste from the storage location to the optimal storage location; The environmental monitoring and control module has built-in temperature, humidity and air quality sensors to collect and store environmental data in real time, and is linked with the environmental control device to automatically adjust the environmental parameters in the warehouse based on the feedback of the prediction model; Real-time feedback and self-learning support mechanism, equipped with feedback data collection unit, real-time monitoring of handling, storage operations and environmental changes, and transmission of feedback data to the value function for adaptive update; It supports self-learning feedback mechanism and intelligent storage planning optimization, interactively feedbacks operation results and updated storage strategies, and continuously optimizes storage planning.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the solder paste intelligent storage method based on image recognition described in 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 solder paste intelligent storage method based on image recognition described in any one of claims 1 to 7 are implemented.

Images