A smart shopping cart charging safety management system and method
Through the series charging and background system detection of the intelligent shopping cart charging safety management system, the problem of overloading of the charging system is solved, real-time monitoring and control of the temperature and charge quantity of the intelligent shopping cart is realized to ensure the safety of the system.
Patent Information
- Application Number
- CN202410929099.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-11
- Publication Date
- 2024-11-12
- Estimated Expiration
- 2044-07-11
AI Technical Summary
The intelligent shopping cart charging system is difficult to effectively control the number of charging vehicles connected at the same time, resulting in overloading of the charging pile's power load, overheating of equipment, line damage and degrading grid stability.
Adopt the intelligent shopping cart charging safety management system, and realize series charging through the front electrode, rear electrode, temperature sensor, front camera, vision sensor and smart terminal. The backend system detects the temperature and number of nodes in the charging chain, and sends a disconnect charging command To avoid overloading.
Real-time monitoring of the temperature of smart shopping carts and precise control of the charging quantity, effectively avoiding the risk of overload and ensuring the safety of the system.
Smart Images

Figure CN118457322B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to a smart shopping cart charging safety management system and method, and belongs to the field of smart shopping carts. Background Art
[0002] The current smart shopping cart charging system generally has the problem of being unable to effectively control the number of vehicles connected to the charging station at the same time. When multiple smart shopping carts are connected to the charging station at the same time, the charging pile may be overloaded, causing safety hazards such as equipment overheating, line damage, and even reduced grid stability. In addition, if the number of sockets is too large, the stability of the vehicle sockets will decrease due to assembly tolerance, resulting in poor contact, excessive contact resistance, heating of the contact points, melting, and other problems. Summary of the invention
[0003] The present invention provides a smart shopping cart charging safety management system and method, which solves the problems disclosed in the background technology.
[0004] In order to solve the above technical problems, the technical solution adopted by the present invention is:
[0005] A smart shopping cart charging safety management system includes a smart shopping cart, a charging pile and a background system; the smart shopping cart is provided with a front electrode, a rear electrode, a temperature sensor, a front camera, a visual sensor and a smart terminal; when charging, a number of smart shopping carts are inserted to form a charging queue, and the front electrode of the first smart shopping cart in the charging queue is connected to the charging pile, and the front electrodes of the smart shopping carts other than the first in the charging queue are used to contact the rear electrodes of the previous smart shopping cart to achieve series charging; the temperature sensor is used to collect the charging temperature of the smart shopping cart; the front camera collects the image of the insertion state of the smart shopping cart; the charging pile and the smart shopping cart are marked with feature numbers, and the visual sensor is used to extract the charging pile or the front electrode. The smart terminal is used to determine whether the smart shopping cart is in the insertion state according to the insertion state image. If the smart shopping cart is in the insertion state, the smart terminal detects whether the smart shopping cart is in the charging state. If the smart shopping cart is in the charging state, the smart terminal obtains the charging temperature and the characteristic number and sends them to the background system; the background system generates a charging chain according to the charging temperature and the characteristic number, and the background system detects each node in the charging chain; the background system sends a disconnect charging instruction to the smart terminal corresponding to the node in the charging chain whose temperature exceeds the threshold or the node at the end of the charging chain exceeds the node number limit; the smart terminal controls the smart shopping cart where it is located to disconnect the charging circuit.
[0006] Furthermore, the smart terminal is provided with an OCR recognition module, and the smart terminal recognizes the feature number through the OCR recognition module.
[0007] Furthermore, the smart terminal is provided with a wireless communication module, and the smart terminal establishes communication with the background system through the wireless communication module.
[0008] Furthermore, the background system is provided with a counting module, and the counting module is used to count the number of smart shopping carts connected to the charging pile.
[0009] Furthermore, the smart terminal is provided with a GPIO interface, and the smart terminal controls the charging switch through the GPIO interface.
[0010] Furthermore, the backend system is a cloud server.
[0011] The present invention also provides a smart shopping cart charging safety management method, comprising the following steps:
[0012] Step 1: Collect the image of the smart shopping cart insertion state;
[0013] Step 2: judging whether the smart shopping cart is in the inserted state according to the inserted state image of the smart shopping cart; if the smart shopping cart is in the inserted state, proceeding to step 3, otherwise returning to step 1;
[0014] Step 3, determine whether the smart shopping cart is in a charging state, if the smart shopping cart is in a charging state, proceed to step 4, otherwise return to step 2 or step 1;
[0015] Step 4: Identify the characteristic number of the charging pile or the preceding smart shopping cart;
[0016] Step 5: Transmit the identified characteristic number and the charging temperature of each smart shopping cart to the backend system;
[0017] Step 6: The background system generates a charging chain according to the charging temperature and characteristic number;
[0018] Step 7: The background system detects whether there are any nodes in the charging chain whose temperature exceeds the threshold or any nodes at the end of the charging chain that exceed the node quantity limit. If so, proceed to step 8; otherwise, return to step 6.
[0019] Step 8, the background system sends a disconnect charging instruction to the smart shopping cart corresponding to the node whose temperature exceeds the threshold in the charging chain or the node at the end of the charging chain exceeds the node quantity limit;
[0020] Step 9: The corresponding smart shopping cart disconnects the charging switch after receiving the disconnect charging instruction.
[0021] Furthermore, in step 2, a visual algorithm is used to process the smart shopping cart insertion state image to determine whether the smart shopping cart is in the insertion state.
[0022] Furthermore, in step 4, an OCR recognition algorithm is used to recognize a characteristic number of the charging pile or the preceding smart shopping cart.
[0023] The beneficial effects achieved by the present invention include: real-time monitoring of the temperature of the smart shopping cart, disconnecting the charging switch when the temperature threshold is exceeded; accurately controlling the number of smart shopping carts connected to the charging pile; and effectively avoiding the risk of overload. BRIEF DESCRIPTION OF THE DRAWINGS
[0024] Figure 1 This is a schematic diagram of the structure of the smart shopping cart in the present invention;
[0025] Figure 2 A schematic diagram of the charging status and identification feature number status of the smart shopping cart in the present invention;
[0026] Figure 3 This is a schematic diagram of the smart shopping cart charging safety management system of the present invention;
[0027] Figure 4 This is a schematic diagram of the process flow of the smart shopping cart charging safety management method of the present invention.
[0028] Reference numerals:
[0029] 1. Front electrode; 2. Rear electrode; 3. Temperature sensor; 4. Front camera; 5. Visual sensor; 6. Smart terminal; A. Charging pile feature number location; B. Smart shopping cart feature number location; C. Visual sensor shooting area; D. Front camera shooting area. DETAILED DESCRIPTION
[0030] The present invention will be further described below in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and cannot be used to limit the protection scope of the present invention.
[0031] like Figure 1As shown, the smart shopping cart used in the smart shopping cart charging safety management system of the present invention is provided with a front electrode 1, a rear electrode 2, a temperature sensor 3, a front camera 4, a visual sensor 5, a smart terminal 6 and a charging switch; when charging, several smart shopping carts are inserted to form a charging queue, and the front electrode 1 of the first smart shopping cart in the charging queue is connected to the charging pile, and the front electrode 1 of the smart shopping carts other than the first one in the charging queue is used to contact the rear electrode 2 of the previous smart shopping cart to achieve series charging; the front electrode 1 and the rear electrode 2 are distributed front and back under the basket of the smart shopping cart; the temperature sensor 3 is arranged at the bottom of the basket of the smart shopping cart, close to the position of the battery compartment of the smart shopping cart, and is used to collect the charging temperature of the smart shopping cart; the front camera 4 is arranged at the front end of the basket of the smart shopping cart, and is used to collect the insertion state image of the smart shopping cart; the smart terminal 6 is arranged at the handle position of the smart shopping cart, and the smart terminal 6 is a tablet computer, and the visual sensor 5 is arranged on the back of the smart terminal 6.
[0032] like Figure 2 As shown, the charging pile feature number position A is located at the top of the charging pile; the smart shopping cart feature number position B is located on the screen of the smart terminal 6; the visual sensor shooting area C covers the charging pile feature number position A or the smart shopping cart feature number position B, which is in the best shooting position; the front camera shooting area D covers the entire rear end of the smart shopping cart, and can clearly determine whether the vehicle is in the plug-in state.
[0033] like Figure 3 As shown, the smart shopping cart charging safety management system of the present invention includes a smart shopping cart, a charging pile and a background system;
[0034] On the charging pile, the characteristic number of the charging pile is marked at a position that can be photographed by the high-precision visual sensor 5. This ensures that the visual sensor 5 can clearly photograph the first smart shopping cart when it is charging.
[0035] When the smart shopping cart is in charging state, the standby page of the screen of the smart terminal 6 (tablet computer) displays the characteristic number of the shopping cart, ensuring that when the rear vehicle is plugged in, the visual sensor 5 of the rear vehicle can clearly capture the characteristic number;
[0036] This feature number is managed and configured in the shopping cart master data in the backend system.
[0037] The front camera 4 of the smart shopping cart captures the insertion status of the shopping cart in real time, and the insertion status is confirmed by image recognition on the smart terminal 6 through the real-time video stream data obtained by the front camera 4. When the insertion and charging status are detected, OCR recognition is turned on.
[0038] When the smart shopping cart is charging, it will identify the charging neighbors (OCR recognizes the feature number on the charging pile and the feature number on the screen of the vehicle in front)
[0039] When the smart shopping cart is charging, the smart terminal 6 regularly reports the charging neighborhood information, which includes its own charging state, temperature, its own characteristic number, and the characteristic number of the preceding vehicle or charging pile.
[0040] The backend system receives the charging neighborhood information, caches it and forms a charging chain;
[0041] The charging chain is as follows:
[0042] Pile N-->Car 1 (number temperature)-->Car 2 (number temperature)-->Car 3 (number temperature)-->Car 4 (number temperature)-->Car 5 (number temperature)--> …………-->Car n (number temperature);
[0043] The backend system continuously receives reports, processes, removes duplications, searches for sources, and forms a cache chain;
[0044] The backend system supports interface viewing and management of the charging chain;
[0045] Based on the system parameters, when the number of vehicles charged in a single chain exceeds the configured threshold, the backend system sends a power-off instruction to all vehicles at the end of the chain that exceed the configured number, or to vehicles whose temperature exceeds the set threshold.
[0046] After receiving the instruction sent by the background system, the intelligent terminal 6 controls the intelligent shopping cart to automatically unlock the socket and stop charging through the GPIO interface.
[0047] After the vehicle switches from charging state to unlocking successfully, the unlocking success information of "not charging" is reported to the background system.
[0048] The smart terminal 6 is embedded with a high-performance wireless communication module (such as 4G / 5G, Wi-Fi, LoRa, etc.) to realize real-time data transmission between the smart shopping cart and the backend system (cloud server). When the data received by the backend system (cloud server) is sorted by an algorithm to form an array model in which the charging piles are formed and the numbered vehicles are connected in sequence, once it is monitored that the number of connected devices of a certain charging device reaches a preset threshold, a control instruction to disconnect the over-limit charging connection is sent to the smart shopping cart with the number through a standardized communication interface.
[0049] Relying on the backend system (cloud server), a vehicle counting algorithm model is constructed. Based on the license plate information recognized by OCR, it is updated to the backend system (cloud server) in real time through the network of the smart shopping cart. The backend system (cloud server) can form the connection information of each vehicle at the previous time, forming an array model in which the numbered vehicles are connected in sequence starting with the charging pile, and the number of vehicles connected to each charging device is ensured to ensure the counting accuracy.
[0050] The smart shopping cart controls the smart switch through the GPIO interface. Under normal circumstances, the smart switch remains in a normally closed state to ensure normal series charging. When the number of smart shopping carts connected to a charging pile or the charging temperature reaches a preset threshold, a control instruction to disconnect the over-limit charging connection is sent to the smart shopping cart with that number through a standardized communication interface. After receiving the instruction to cut off the over-limit charging connection, the internal smart switch is triggered through the GPIO interface to perform a precise disconnection operation. The smart switch matrix accurately locates and instantly cuts off the charging circuit that exceeds the limit based on the instruction content, ensuring that the overall system load is within a safe range and effectively preventing overload risks.
[0051] like Figure 4 As shown, the charging safety management method of the smart shopping cart of the present invention includes the following steps: Step 1, collecting the image of the smart shopping cart insertion state;
[0052] Step 2: Process the image of the smart shopping cart in the inserted state through a visual algorithm to determine whether the smart shopping cart is in the inserted state; if the smart shopping cart is in the inserted state, proceed to step 3, otherwise return to step 1;
[0053] Step 3, determine whether the smart shopping cart is in a charging state, if the smart shopping cart is in a charging state, proceed to step 4, otherwise return to step 2 or step 1;
[0054] Step 4: Enable the OCR recognition algorithm to identify the characteristic number of the charging pile or the smart shopping cart in front;
[0055] Step 5: Transmit the identified feature number and the charging temperature of each smart shopping cart to the backend system (cloud server);
[0056] Step 6: The backend system (cloud server) generates a charging chain based on the charging temperature and feature number;
[0057] Step 7: The backend system (cloud server) detects whether there are any nodes in the charging chain whose temperature exceeds the threshold or any nodes at the end of the charging chain that exceed the node quantity limit. If so, proceed to step 8; otherwise, return to step 6.
[0058] Step 8, the backend system (cloud server) sends a disconnect charging instruction to the smart shopping cart corresponding to the node in the charging chain whose temperature exceeds the threshold or the node at the end of the charging chain that exceeds the node quantity limit;
[0059] Step 9: After receiving the disconnect charging instruction, the corresponding smart shopping cart controls the smart shopping cart to automatically unlock the socket through the GPIO interface, thereby turning the charging circuit into an open circuit.
[0060] In step 2, the visual algorithm uses deep learning models such as ResNet and MobileNet to divide the shopping cart status into two categories: (1) not in the inserted state; (2) in the inserted state.
[0061] Resnet solves the gradient vanishing problem in deep neural network training by introducing residual blocks, making the network depth reach hundreds of layers or even more, thus achieving significant performance improvements in tasks such as image classification and object detection. When applied to the recognition of the charging status of smart shopping carts, Resnet can learn a deeper representation of the shopping cart status features, effectively distinguish the various details of the two states, and maintain high recognition accuracy even in complex scenarios such as light changes and different viewing angles. MobileNet is a lightweight deep learning model designed for mobile devices and embedded systems. It greatly reduces the number of model parameters and computational complexity through technologies such as deep separable convolution, while maintaining good recognition performance. MobileNet is an ideal choice for application scenarios such as smart shopping carts that require instant response and low power consumption. It can quickly and efficiently identify the charging status, ensuring that the smart shopping cart management system can run stably under various hardware configurations while extending the battery life of the device. These two algorithms have their own characteristics and are suitable for different computing resource conditions, but they can effectively improve recognition accuracy and efficiency. In actual deployment, the appropriate model can be selected according to the specific needs and hardware limitations of the smart shopping cart system. If the tablet computer is equipped with a GPU, NPU or other chips dedicated to deep model inference, ResNet is preferred to obtain higher recognition accuracy. If the computing resources and power consumption of the tablet computer are limited, MobileNet is preferred to achieve high-efficiency recognition.
[0062] When implementing, first of all, a large number of images containing shopping carts in two states need to be collected as training and test data sets, and the training and test data sets must be independent of each other and cannot have intersections. In addition, the training and test sets should be able to cover various environmental conditions during charging, such as the external environment of the shopping cart in different supermarkets, various possible lighting conditions such as light and dark, different shopping cart color combinations, and various leftover items such as shopping receipts and empty shopping bags that may be contained in the shopping cart frame. Then, use these data sets to train Resnet or MobileNet, and design data enhancement techniques based on the performance of the model on the test set. Through operations such as rotation, cropping, flipping, changing brightness and contrast, increase sample read positivity, reduce overfitting, and improve the generalization ability of the model in different environments. Furthermore, through model quantization, pruning and other technologies, reduce model size and computing requirements to ensure efficient operation in a limited resource environment. This can not only improve the running speed, but also reduce power consumption and extend the running time of the smart shopping cart. Through fine training set, test set collection and annotation, targeted data enhancement, model optimization and other strategies, the model can accurately and efficiently identify which state the smart shopping cart is in. In addition, in order to reduce the misjudgment of the algorithm, multiple frames of images are used for continuous verification in actual application to improve the robustness of the judgment. Finally, the trained model is integrated into the control system of the smart shopping cart to analyze the images captured by the camera in real time, obtain the charging status of the smart shopping cart, guide the charging process, and improve the automation level of the system and user experience.
[0063] In step 4, the OCR recognition algorithm uses a high-precision visual sensor 5 (camera) on the smart shopping cart to ensure that its viewing angle covers the position of the display screen of the front vehicle and the position of the charging pile. The terminal display screen of the vehicle displays the unique number of each vehicle, and the unique number is marked above the charging pile, using different feature number sequences. Collect and store video stream data in the area in front of the vehicle. Ensure that the image data of the front vehicle or charging pile of the vehicle can be accurately monitored under various lighting environments. Integrate advanced OCR technology to perform real-time deep learning processing on the video stream captured by the camera, accurately identify and extract the number information of the connected charging pile, or the number information of the front vehicle. Different number sequences (numbers are composed of feature sequences-digital sequences) are used to distinguish different types of devices (vehicles or charging piles), and digital interference in the environment can be filtered out. OCR technology should have excellent light adaptability, angle robustness and anti-interference ability in complex environments to ensure fast and accurate recognition of the number information of smart shopping carts or charging piles under various working conditions.
[0064] The above are merely embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention are included in the scope of the claims of the present invention to be approved.
Claims
1. A smart shopping cart charging safety management system, characterized by: The invention comprises a smart shopping cart, a charging pile and a background system; the smart shopping cart is provided with a front electrode (1), a rear electrode (2), a temperature sensor (3), a front camera (4), a visual sensor (5) and a smart terminal (6); when charging, a plurality of smart shopping carts are plugged in to form a charging queue, and the front electrode (1) of the first smart shopping cart in the charging queue is connected to the charging pile, and the front electrodes (1) of the smart shopping carts other than the first one in the charging queue are used to contact the rear electrodes (2) of the preceding smart shopping cart to achieve series charging; the temperature sensor (3) is used to collect the charging temperature of the smart shopping cart; the front camera (4) is used to collect the plug-in state image of the smart shopping cart; the charging pile and the smart shopping cart are both marked with characteristic numbers, and the visual sensor (5) is used to extract the charging pile information. The intelligent terminal (6) is used to determine whether the intelligent shopping cart is in the plug-in state according to the plug-in state image; if the intelligent shopping cart is in the plug-in state, the intelligent terminal (6) detects whether the intelligent shopping cart is in the charging state; if the intelligent shopping cart is in the charging state, the intelligent terminal (6) obtains the charging temperature and the characteristic number and sends them to the background system; the background system generates a charging chain according to the charging temperature and the characteristic number, and the background system detects each node in the charging chain; the background system sends a disconnect charging instruction to the intelligent terminal (6) corresponding to the node in the charging chain whose temperature exceeds the threshold or the node at the end of the charging chain exceeds the node number limit; the intelligent terminal (6) controls the intelligent shopping cart where it is located to disconnect the charging circuit; The smart terminal (6) is provided with an OCR recognition module, and the smart terminal (6) recognizes the feature number through the OCR recognition module; The charging pile feature number position A is located at the top of the charging pile; the smart shopping cart feature number position B is located on the screen of the smart terminal (6); the visual sensor shooting area C covers the charging pile feature number position A or the smart shopping cart feature number position B; the front camera shooting area D covers the entire rear end of the smart shopping cart; The front camera (4) of the smart shopping cart captures the insertion status of the shopping cart in real time, and performs image recognition on the smart terminal 6 through the real-time video stream data obtained by the front camera (4) to confirm the insertion status; when the insertion and charging status are detected, OCR recognition is turned on; The intelligent terminal (6) uses a ResNet or MobileNet deep learning model to classify the shopping cart status into two categories: not in the inserted state or in the inserted state.
2. According to claim 1, the smart shopping cart charging safety management system is characterized by: The smart terminal (6) is provided with a wireless communication module, and the smart terminal (6) establishes communication with the background system via the wireless communication module.
3. The smart shopping cart charging safety management system according to claim 1, characterized in that: The background system is provided with a counting module, and the counting module is used to count the number of smart shopping carts connected to the charging pile.
4. The smart shopping cart charging safety management system according to claim 1, characterized in that: The smart terminal (6) is provided with a GPIO interface, and the smart terminal (6) controls the smart shopping cart to unlock the socket and disconnect the charging circuit via the GPIO interface.
5. According to claim 1, the smart shopping cart charging safety management system is characterized by: The backend system is a cloud server.
6. A smart shopping cart charging safety management method, characterized in that: The steps include: Step 1: Collect the image of the smart shopping cart insertion state; Step 2: judging whether the smart shopping cart is in the inserted state according to the inserted state image of the smart shopping cart; if the smart shopping cart is in the inserted state, proceeding to step 3, otherwise returning to step 1; Step 3, determine whether the smart shopping cart is in a charging state, if the smart shopping cart is in a charging state, proceed to step 4, otherwise return to step 2 or step 1; Step 4: Identify the characteristic number of the charging pile or the smart shopping cart in front; Step 5: Transmit the identified characteristic number and the charging temperature of each smart shopping cart to the backend system; Step 6: The background system generates a charging chain according to the charging temperature and characteristic number; Step 7: The background system detects whether there are any nodes in the charging chain whose temperature exceeds the threshold or any nodes at the end of the charging chain that exceed the node quantity limit. If so, proceed to step 8; otherwise, return to step 6. Step 8, the background system sends a disconnect charging instruction to the smart shopping cart corresponding to the node whose temperature exceeds the threshold in the charging chain or the node at the end of the charging chain exceeds the node quantity limit; Step 9: The corresponding smart shopping cart disconnects the charging switch after receiving the disconnect charging instruction; In step 2, a visual algorithm is used to process the image of the smart shopping cart insertion state to determine whether the smart shopping cart is in the insertion state; In step 2, the visual algorithm uses a ResNet or MobileNet deep learning model to classify the shopping cart status into two categories: not in the inserted state or in the inserted state; First, a large number of images containing shopping carts in two states are collected as training and test data sets, and the training and test data sets are ensured to be independent of each other and have no intersection, and the training and test sets cover various environmental conditions during charging. Then, these data sets are used to train Resnet or MobileNet, and multiple frames of images are used for continuous verification. The trained model is integrated into the control system of the smart shopping cart, and the images captured by the camera are analyzed in real time to obtain the charging status of the smart shopping cart. In step 4, an OCR recognition algorithm is used to recognize the characteristic number of the charging pile or the preceding smart shopping cart.
