Intelligent navigation method and system for transport vehicles in unmanned intelligent cargo stations, and transport vehicles

Abstract

The present invention relates to the field of image processing technology, and provides an intelligent navigation method, system and transport vehicle for transport vehicles in unmanned intelligent cargo stations, including: real-time collection of image information on the travel path of the transport vehicle in the unmanned intelligent cargo station; feature recognition and orientation recognition of objects in the image information to obtain feature information and orientation information of the object; based on the orientation information, determine whether the object is on the travel path; if on the travel path, analyze the feature information based on a preset deep learning model to obtain the obstacle level of the object; based on the obstacle level, obtain a corresponding detour strategy; identify a new route that can be passed based on the image information, and determine whether the new route meets the detour strategy; if it meets the detour strategy, use the new route as a new travel path, and navigate to the new travel path. In the present invention, the intelligent decision-making detour strategy ensures that the transport vehicle can complete the task safely and efficiently even in a complex environment.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and in particular to an intelligent navigation method and system for a transport vehicle of an unmanned intelligent cargo station, and a transport vehicle. Background Art

[0002] In the operating environment of unmanned intelligent cargo stations, efficient and safe cargo handling is the key to improving overall logistics efficiency. Traditional handling methods rely on manual operations, which have problems such as low efficiency, high labor intensity, and prone to safety accidents. With the development of automation and intelligent technology, the use of automatic guided vehicles (AGVs) or other types of unmanned transport vehicles has become an important part of modern logistics and warehousing systems.

[0003] Current automated guided vehicles or other types of unmanned guided vehicles can automatically drive on a preset path. However, this method lacks flexibility. Once an unpredictable obstacle is encountered, the vehicle often cannot make autonomous decisions and cannot bypass the obstacle, requiring human intervention. Summary of the invention

[0004] The main purpose of the present invention is to provide an intelligent navigation method, system and transport vehicle for an unmanned intelligent cargo station, aiming to overcome the defect that the transport vehicle can only travel on a preset path and cannot bypass obstacles.

[0005] To achieve the above object, the present invention provides an intelligent navigation method for a transport vehicle in an unmanned intelligent cargo station, comprising the following steps:

[0006] Real-time collection of image information on the path of the transport vehicle in the unmanned intelligent cargo station; feature recognition and orientation recognition of objects in the image information to obtain feature information and orientation information of the objects;

[0007] Based on the orientation information, determine whether the object is on the travel path; if on the travel path, analyze the feature information based on a preset deep learning model to obtain the obstacle level of the object;

[0008] Based on the obstacle level, a corresponding detour strategy is obtained; based on the image information, a new route that can be passed is identified, and whether the new route conforms to the detour strategy;

[0009] If it complies with the detour strategy, the new route is used as a new travel path and the vehicle is navigated to the new travel path.

[0010] Furthermore, the performing feature recognition and orientation recognition on the object in the image information to obtain feature information and orientation information of the object includes:

[0011] Using an object detection algorithm to locate an object bounding box in the image information;

[0012] Applying a deep convolutional neural network to extract feature information from the object bounding box; wherein the feature information is used to describe the appearance and structural characteristics of the object;

[0013] Based on the position of the object boundary box in the image information and the camera parameters of the transport vehicle, the azimuth angle and the tilt angle of the object relative to the transport vehicle are calculated through triangular geometric relationship as the azimuth information.

[0014] Furthermore, the analyzing the feature information based on a preset deep learning model to obtain the obstacle level of the object includes:

[0015] Inputting the feature information into the preset deep learning model; wherein the preset deep learning model includes a trained size obstacle sub-model, a shape obstacle sub-model and a classification sub-model; the feature information includes size features and shape features;

[0016] Analyzing the size feature based on the size obstacle sub-model to obtain a first obstacle level corresponding to the object size;

[0017] Analyzing the shape feature based on the shape obstacle sub-model to obtain a second obstacle level corresponding to the object shape;

[0018] The first obstacle level and the second obstacle level are input into the classification sub-model to obtain the obstacle level of the object.

[0019] Further, after navigating to a new travel path, the process includes:

[0020] After circumnavigating the object via the new travel path, creating a navigation map in a navigation system of the transport vehicle, and storing the new travel path in the navigation map;

[0021] Sending the navigation map to a management terminal; wherein the management terminal generates a management code for the navigation map and sends it to the transport vehicle;

[0022] When the transport vehicle subsequently requests the navigation map, the management code is sent to the management terminal for authority verification.

[0023] Furthermore, the management terminal generates a management code for the navigation map, including:

[0024] Performing curve simulation on the navigation route in the navigation map to obtain a route curve;

[0025] Acquire identification information of the transport vehicle, and generate a corresponding character array based on the identification information; wherein the character array is composed of a combination of multiple rows and columns of characters;

[0026] The management code is generated based on the route curve and the character array.

[0027] Further, the generating a corresponding character array based on the identification information includes:

[0028] Based on a conversion relationship between preset characters and character combinations, convert each character in the identification information into a character combination;

[0029] Create a blank array, and add the characters in each character combination to the blank array one by one in sequence to form a character array with multiple rows and columns.

[0030] Furthermore, the generating the management code based on the route curve and the character array includes:

[0031] Generate a picture of the character array to obtain an array picture;

[0032] The route curve is superimposed on the array image, and the route curve is proportionally stretched or scaled so that the upper edge and the lower edge of the route curve are tangent to the upper edge and the lower edge of the array image;

[0033] On the array image, characters superimposed on the route curve are obtained, and the characters are combined in sequence according to their order on the route curve to obtain the management code.

[0034] Furthermore, the generating the management code based on the route curve and the character array includes:

[0035] According to the preset selection rules, six key points are selected on the route curve, and they are divided into three groups of key points according to the upper and lower positions of each key point; wherein each group includes two key points;

[0036] For each group of key points, connect the two key points into a straight line; the straight lines obtained from each group are used as target lines;

[0037] Superimposing the target straight line with the character array, and obtaining characters intersecting with each of the target straight lines in the character array as target characters;

[0038] The target characters are combined in sequence according to the order of the target characters in the character array to obtain the management code.

[0039] The present invention also provides an intelligent navigation system for transport vehicles in an unmanned intelligent cargo station, comprising:

[0040] The acquisition unit is used to collect image information on the travel path of the transport vehicle in the unmanned intelligent cargo station in real time; perform feature recognition and orientation recognition on the object in the image information to obtain feature information and orientation information of the object;

[0041] An analysis unit, configured to determine whether the object is on the travel path based on the orientation information; if the object is on the travel path, analyze the feature information based on a preset deep learning model to obtain an obstacle level of the object;

[0042] an identification unit, configured to obtain a corresponding detour strategy based on the obstacle level; identify a new route that can be passed based on the image information, and determine whether the new route complies with the detour strategy;

[0043] The navigation unit is used to take the new route as a new travel path and navigate to the new travel path if it meets the detour strategy.

[0044] The present invention also provides a transport vehicle, comprising a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps of any one of the above methods when executing the computer program.

[0045] The present invention also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of any of the above-mentioned methods are implemented.

[0046] The intelligent navigation method, system and transporter of the transporter in the unmanned intelligent cargo station provided by the present invention include: real-time collection of image information on the travel path of the transporter in the unmanned intelligent cargo station; feature recognition and orientation recognition of the object in the image information to obtain the feature information and orientation information of the object; based on the orientation information, determine whether the object is on the travel path; if it is on the travel path, analyze the feature information based on a preset deep learning model to obtain the obstacle level of the object; based on the obstacle level, obtain the corresponding detour strategy; identify a new route that can be passed based on the image information, and judge whether the new route meets the detour strategy; if it meets the detour strategy, use the new route as a new travel path and navigate to the new travel path. In the present invention, it is possible to collect and analyze image information in real time, accurately identify and locate obstacles, evaluate the degree of their influence on the travel path, and then intelligently decide on the detour strategy to ensure that the transporter can complete the task safely and efficiently in a complex environment. The autonomy and adaptability of the unmanned transporter are significantly improved through the high-precision recognition ability of the deep learning model and the path planning ability of the intelligent algorithm. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1This is a schematic diagram of the steps of an intelligent navigation method for a transport vehicle in an unmanned intelligent cargo station in one embodiment of the present invention;

[0048] Figure 2 This is a structural block diagram of an intelligent navigation system for transport vehicles in an unmanned intelligent cargo station in one embodiment of the present invention;

[0049] Figure 3 It is a schematic block diagram of the structure of a transport vehicle according to an embodiment of the present invention.

[0050] The implementation, functional features and advantages of the present invention will be further described in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0051] 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 intended to limit the present invention.

[0052] Reference Figure 1 In one embodiment of the present invention, a method for intelligent navigation of a transport vehicle in an unmanned intelligent cargo station is provided, comprising the following steps:

[0053] Step S1, collecting image information on the travel path of the transport vehicle in the unmanned intelligent cargo station in real time; performing feature recognition and orientation recognition on the object in the image information to obtain feature information and orientation information of the object;

[0054] Step S2, based on the orientation information, determining whether the object is on the travel path; if on the travel path, analyzing the feature information based on a preset deep learning model to obtain the obstacle level of the object;

[0055] Step S3, obtaining a corresponding detour strategy based on the obstacle level; identifying a new route that can be passed based on the image information, and determining whether the new route complies with the detour strategy;

[0056] Step S4: If the detour strategy is met, the new route is used as a new travel path, and the vehicle is navigated to the new travel path.

[0057] In this embodiment, it specifically includes:

[0058] Real-time collection of image information: The high-definition camera installed on the transporter collects image information on the travel path in real time. The image information contains various details of the surrounding environment, such as shelves, goods, other vehicles or personnel, etc.

[0059] Feature recognition and orientation recognition: Using computer vision technology, the collected images are processed to identify the objects in the images, extract their feature information (such as shape, color, texture, etc.), and determine their position and direction relative to the transporter. This step is the basis for subsequent decision-making and ensures that the transporter has a comprehensive understanding of the environment.

[0060] Determine the positional relationship of objects: Based on the orientation information of the objects, determine which objects are in the path of the transporter and which are not. This helps to distinguish between obstacles of concern and irrelevant environmental elements.

[0061] Obstacle level determination: For objects on the path of travel, the pre-trained deep learning model is used to analyze their feature information, evaluate their potential impact on the movement of the transporter, and determine the obstacle level. The obstacle level can include "minor", "moderate", "serious", etc., to guide the subsequent detour strategy formulation.

[0062] Obtain bypass strategy: According to the obstacle level, select the appropriate bypass strategy from the preset strategy library. The strategy library includes direct bypass, slow down, pause and wait and other strategies to adapt to obstacles of different levels and types.

[0063] New route identification and verification: Based on the current image information, identify the new route that can be passed, and determine whether the new route is safe and feasible in combination with the requirements of the detour strategy. This step ensures the rationality of the detour plan and avoids unnecessary risks. If it does not meet the detour strategy, you can choose to stop moving forward or choose a new route that is closest to the detour strategy.

[0064] Path update and navigation: Once the new route is determined to be in line with the detour strategy, the vehicle’s path is updated and the vehicle is guided along the new route through the onboard navigation system. This step completes the closed loop from perception to decision-making to action, ensuring that the vehicle can quickly and safely adjust its path when encountering obstacles.

[0065] In this embodiment, it is possible to collect and analyze image information in real time, accurately identify and locate obstacles, evaluate their impact on the path of travel, and then make intelligent decisions on detour strategies to ensure that the truck can complete its tasks safely and efficiently in complex environments. The high-precision recognition capabilities of deep learning models and the path planning capabilities of intelligent algorithms can significantly improve the autonomy and adaptability of unmanned trucks. The above technical solution forms an efficient and flexible intelligent navigation system through the four core steps of real-time perception, intelligent analysis, strategy generation, and path updating, enabling trucks to autonomously respond to various complex environments in unmanned smart cargo stations, effectively improving the efficiency and safety of logistics operations.

[0066] In one embodiment, based on the obstacle level, a corresponding detour strategy is obtained; based on the image information, a new route that can be passed is identified, and whether the new route complies with the detour strategy is determined, specifically including:

[0067] In this embodiment, the above technical solution mainly involves the problem of how to dynamically plan a path according to the obstacle level of the current environment in the field of autonomous driving or intelligent robot navigation.

[0068] Obstacle Level: Obstacle level is a classification of the severity of obstacles encountered in the environment, e.g.

[0069] Low-level obstacles: small obstacles, such as pebbles and slight bumps on the road, which do not affect the normal driving of vehicles or robots.

[0070] Medium-level obstacles: Large static obstacles, such as trees and buildings on the roadside, which require changing the driving direction to avoid.

[0071] High-level obstacles: Dynamic obstacles or emergency situations, such as the sudden appearance of pedestrians, animals, traffic accident scenes, etc., require immediate action to avoid collision.

[0072] Detour strategy: Based on the obstacle level, a corresponding detour strategy will be generated. The above strategy may include but is not limited to:

[0073] Low-level obstacles: Maintain the current speed and direction, or slightly adjust the direction to avoid obstacles with a smaller detour distance.

[0074] Medium-level obstacle: slow down and plan a new detour route to avoid the obstacle while minimizing the distance from the original route and increasing the detour distance. The detour distance can be set to be greater than the first preset value.

[0075] High-level obstacles: emergency braking, find the safest avoidance path, stop completely if necessary, and wait for the obstacle to be cleared before continuing. Alternatively, the detour distance can be greatly increased, and the detour distance can be set to be greater than the second preset value. The second preset value is greater than the first preset value.

[0076] Image information recognition and analysis: Use image information collected by on-board cameras and other equipment to recognize and analyze the road conditions ahead in real time. Image processing algorithms can identify obstacles, pedestrians, traffic signs, etc. on the road, and evaluate the type and level of these obstacles.

[0077] Identify new routes: Based on the current location of the vehicle or robot and the specific location of obstacles, calculate alternative routes that can bypass obstacles.

[0078] Evaluate new routes: Analyze the feasibility of each alternative route, considering factors including but not limited to distance, time, road conditions, etc.

[0079] Determine whether the detour strategy is met: Compare each alternative route with the previously formulated detour strategy to ensure that the selected route can effectively avoid obstacles while meeting the requirements of safety and efficiency (mainly comparing the detour distance).

[0080] Execute route change: After selecting the best route, send instructions to the transporter's control system to implement the route change.

[0081] In one embodiment, the performing feature recognition and orientation recognition on the object in the image information to obtain feature information and orientation information of the object includes:

[0082] Using an object detection algorithm to locate an object bounding box in the image information;

[0083] Applying a deep convolutional neural network to extract feature information from the object bounding box; wherein the feature information is used to describe the appearance and structural characteristics of the object;

[0084] Based on the position of the object boundary box in the image information and the camera parameters of the transport vehicle, the azimuth angle and the tilt angle of the object relative to the transport vehicle are calculated through triangular geometric relationship as the azimuth information.

[0085] In this embodiment, it specifically includes:

[0086] Object detection algorithm: First, advanced object detection algorithms (such as YOLO, SSD or Faster R-CNN, etc.) are used to process the image information captured by the truck camera in real time. These algorithms can identify multiple objects in the image and draw bounding boxes for each object, accurately marking the location range of the object in the image.

[0087] Deep Convolutional Neural Networks (DCNNs): DCNNs are further applied to extract feature information of objects from the bounding boxes of objects marked by the target detection algorithm. DCNNs can automatically learn multi-level feature representations from raw image data, including edges, textures, shapes, and colors. These feature information comprehensively describes the appearance and structural characteristics of objects, providing an important basis for subsequent obstacle recognition and classification.

[0088] Calculate the object's orientation information: In order to determine the object's precise orientation relative to the truck, it is necessary to combine the position information of the object's bounding box in the image with the truck's camera's internal parameters (such as focal length, pixel size, etc.). This information together forms the basis for calculating the object's orientation and tilt angle.

[0089] Triangular geometry: Using the principles of triangular geometry, the position of the center point of the object's bounding box in the image coordinate system, combined with the camera parameters, can be used to calculate the object's azimuth (i.e., the relative direction of the object on the horizontal plane) and tilt angle (i.e., the angular offset of the object relative to the vertical direction of the truck). This process involves the inverse solution of perspective projection to restore the object's posture in three-dimensional space.

[0090] Combining the above steps, it is possible to accurately identify the features and positions of objects in the surrounding environment while the transporter is moving. This series of operations provides key data support for subsequent obstacle assessment, detour strategy generation, and path planning, and is an important basis for ensuring the efficient and safe operation of the transporter's intelligent navigation system.

[0091] In summary, through target detection, deep learning feature extraction, and orientation calculation based on triangular geometry, this technical solution effectively enhances the transporter's ability to understand the environment, enabling it to autonomously perceive, analyze and make reasonable decisions in complex logistics environments, significantly improving the operational efficiency and safety of unmanned smart cargo stations.

[0092] In one embodiment, the analyzing the feature information based on a preset deep learning model to obtain the obstacle level of the object includes:

[0093] Inputting the feature information into the preset deep learning model; wherein the preset deep learning model includes a trained size obstacle sub-model, a shape obstacle sub-model and a classification sub-model; the feature information includes size features and shape features;

[0094] Analyzing the size feature based on the size obstacle sub-model to obtain a first obstacle level corresponding to the object size;

[0095] Analyzing the shape feature based on the shape obstacle sub-model to obtain a second obstacle level corresponding to the object shape;

[0096] The first obstacle level and the second obstacle level are input into the classification sub-model to obtain the obstacle level of the object.

[0097] In this embodiment, it specifically includes:

[0098] Size obstacle sub-model: This model focuses on the size characteristics of objects and aims to evaluate the impact of the physical size of an object on its role as an obstacle. It determines whether an object may block the passage of a transporter or restrict the operating space of the transporter by analyzing the length, width, height or other relevant size attributes of the object.

[0099] Shape obstacle sub-model: The shape obstacle sub-model focuses on the shape characteristics of objects, considering the impact of object shapes on the navigation and obstacle avoidance strategies of the transporter. For example, sharp objects may be more difficult to avoid than smooth objects, or may pose a greater risk to the transporter. The model analyzes the outline of the object, shape complexity, etc. to assess its difficulty as an obstacle.

[0100] Classification sub-model: The classification sub-model receives the outputs from the size obstacle sub-model and the shape obstacle sub-model, namely the first obstacle level and the second obstacle level, and combines the comprehensive characteristics of the object to give a final obstacle level assessment. This assessment takes into account the combined impact of the size and shape of the object, and more comprehensively reflects the potential risk of the object as an obstacle.

[0101] Pre-set deep learning model training: In order for these sub-models to accurately assess obstacle levels, they need to be trained with a large amount of training data sets. The training data sets usually contain the size and shape characteristics of various objects, as well as the corresponding obstacle level labels, which are based on expert evaluation or historical operating experience.

[0102] Model architecture: The default deep learning model uses convolutional neural networks (CNN), recurrent neural networks (RNN), or more advanced variants such as Transformer to capture complex relationships between object features.

[0103] Feature extraction: First, the size and shape features of the object are extracted from the image.

[0104] Size and shape obstacle assessment: These features are input into the size obstacle sub-model and the shape obstacle sub-model, respectively, to obtain the first obstacle level and the second obstacle level.

[0105] Comprehensive evaluation of obstacle levels: Finally, the evaluation results of the two levels are input into the classification sub-model to obtain a comprehensive obstacle level, which is used to guide the obstacle avoidance strategy and path planning of the transport vehicle.

[0106] In summary, through the analysis of this series of deep learning models, it is possible to intelligently assess the potential risks of objects as obstacles, help trucks make more accurate and safe decisions, effectively avoid collisions and improve operational efficiency, and play a vital role in the automated logistics scenarios of unmanned smart cargo stations.

[0107] In one embodiment, after navigating to a new travel path, the process includes:

[0108] After circumnavigating the object via the new travel path, creating a navigation map in a navigation system of the transport vehicle, and storing the new travel path in the navigation map;

[0109] Sending the navigation map to a management terminal; wherein the management terminal generates a management code for the navigation map and sends it to the transport vehicle;

[0110] When the transport vehicle subsequently requests the navigation map, the management code is sent to the management terminal for authority verification.

[0111] In this embodiment, when the transporter detects an obstacle during the mission, it will instantly calculate a new path to bypass the obstacle. Once it successfully bypasses and reaches the destination via this new path, the transporter's navigation system will perform the following steps:

[0112] Create a navigation map: Create a new navigation map in the truck's navigation system that incorporates the new path data. This step is critical to documenting valid alternative routes for quick reference and application in similar situations in the future.

[0113] Store new travel path: Store the details of the new path taken to bypass obstacles in the navigation map. This includes the start and end points of the path, as well as key points on the path, ensuring that all relevant information is accurately recorded for later use.

[0114] Secure sharing and authority verification: In order to ensure the security and integrity of information transmission between the transporter and the management system, the solution introduces the concept of management code. The specific process is as follows:

[0115] Navigation map sent to management terminal: The transporter sends the updated navigation map to the central management terminal. The management terminal is responsible for monitoring and scheduling the entire system, ensuring information synchronization and task coordination for all transporters.

[0116] Management code generation and sending: After receiving the navigation map, the management terminal will generate a unique management code for it. This management code is equivalent to the access key of the map and is used for subsequent permission verification. The management terminal then sends the management code back to the transporter to ensure that the transporter can provide correct verification information when requesting maps in the future.

[0117] Permission verification: When the transporter needs to request a previously stored navigation map in a subsequent task, it will send the request to the management terminal together with the management code. The management terminal confirms the legitimacy of the request by verifying the validity of the management code. Only when the management code is correct will the transporter be allowed to access and use the corresponding navigation map.

[0118] In this embodiment, the following beneficial effects are achieved:

[0119] Dynamic adaptability: Through real-time path planning and map updates, the transporter can quickly respond to unforeseen obstacles in the environment, improving operational flexibility and efficiency.

[0120] Enhanced security: The management code mechanism ensures that sensitive path information is only accessed by authorized transporters, preventing unauthorized use or malicious tampering and improving overall security.

[0121] Resource Optimization: Storing and utilizing efficient alternative paths reduces the need for repeated calculations, saving computing resources while also speeding up the decision-making process for moving vehicles.

[0122] In summary, the above technical solution not only improves the autonomous navigation capability of the transporter in complex environments, but also strengthens information security through the management code mechanism. It is a key component for achieving efficient and intelligent logistics operations.

[0123] In one embodiment, the management terminal generates a management code for the navigation map, including:

[0124] Performing curve simulation on the navigation route in the navigation map to obtain a route curve;

[0125] Acquire identification information of the transport vehicle, and generate a corresponding character array based on the identification information; wherein the character array is composed of a combination of multiple rows and columns of characters;

[0126] The management code is generated based on the route curve and the character array.

[0127] In this embodiment, it specifically includes:

[0128] Route curve simulation: First, the management terminal performs curve simulation on the navigation route in the navigation map uploaded by the transporter, aiming to convert discrete path points into smooth continuous curves. This process usually involves mathematical curve fitting algorithms, such as Bezier curves and spline curves, which can effectively capture the overall trend and characteristics of the route. The result of the curve simulation is the route curve, which will become an important part of generating the management code.

[0129] Character array generation: At the same time, the management terminal generates a character array consisting of multiple rows and columns of characters based on the identification information of the transporter. This identification information may include unique identification codes such as the ID, model, and serial number of the transporter. The character array is generated in any way, such as a series of encryption algorithms or hash functions, to ensure that even if the original identification information is the same, different operations or time points can generate different character arrays, increasing the randomness and security of the management code.

[0130] Management code generation: Finally, the management terminal combines the above two elements, the route curve and the character array, and generates the final management code through a preset algorithm or rule. This process can ensure the uniqueness and unpredictability of the management code through complex mathematical operations, logical operations or cryptographic techniques. For example, the generation of the management code is based on the combination of the coordinate value of a specific point on the route curve and the character at the corresponding position in the character array, or an advanced encryption method, such as a public key encryption system (such as RSA), is used to encrypt the route curve information with the character array as the private key to generate a ciphertext management code.

[0131] In this embodiment, technical advantages and security guarantees:

[0132] Dynamic and unpredictable: Since the generation of management codes depends on real-time route information and the identification information of the transport vehicle, this means that even at the same location, different transport vehicles or different time points, the generated management codes will be different, greatly increasing the difficulty of cracking.

[0133] Information protection: By converting sensitive route information into abstract management codes, even if the management codes are intercepted during transmission, it is difficult for attackers to directly obtain the original route data, thereby protecting the core secrets of logistics operations.

[0134] Authentication: The addition of the character array allows the management code to be closely associated with a specific truck, achieving device-based identity authentication and further enhancing the security of the system.

[0135] In summary, the above technical solution generates a highly personalized and secure management code by cleverly integrating route curves and character arrays, effectively ensuring the security of communication between the transport vehicle and the management terminal, and also providing strong technical support for information protection of the intelligent logistics system.

[0136] In one embodiment, the generating a corresponding character array based on the identification information includes:

[0137] Based on a conversion relationship between preset characters and character combinations, convert each character in the identification information into a character combination;

[0138] Create a blank array, and add the characters in each character combination to the blank array one by one in sequence to form a character array with multiple rows and columns.

[0139] In this embodiment, it specifically includes:

[0140] First, a set of preset character conversion rules needs to be stored in advance. These rules define how each character in the identification information is mapped to a character combination. The character combination can be a simple binary code, ASCII code, hexadecimal number or any other form of encoding. The key is that this conversion can ensure that the original characteristics of the identification information are fully preserved, while introducing additional complexity to improve the security of the management code generated subsequently.

[0141] For example, assuming that each character "A" in the identification information is converted into the character combination "10" and "B" is converted into "11", then the identification information "AB" becomes "1011" after conversion.

[0142] Create an empty array and fill it with character combinations:

[0143] Next, a blank array is created, which has a structure of multiple rows and columns, and the specific size depends on the application requirements and security policies. Then, according to the preset order or rules, the characters in the character combination converted in the previous step are added to the blank array one by one to form a complete character array.

[0144] Security and flexibility: This method significantly improves the randomness and complexity of the character array by introducing preset character conversion rules and multi-row and multi-column array structures, thereby improving the information security level of the entire system. At the same time, it is also possible to balance security and efficiency according to the needs of actual application scenarios by flexibly adjusting the conversion rules and array size to meet different levels of security requirements.

[0145] In addition, the character array generation process also provides rich data materials for the subsequent management code generation, ensuring the diversity and unpredictability of the management code, and further strengthening the system's ability to resist attacks. In short, the above technical solution provides an effective encryption method for the identification information of the transport vehicle.

[0146] In one embodiment, generating the management code based on the route curve and the character array includes:

[0147] Generate a picture of the character array to obtain an array picture;

[0148] The route curve is superimposed on the array image, and the route curve is proportionally stretched or scaled so that the upper edge and the lower edge of the route curve are tangent to the upper edge and the lower edge of the array image;

[0149] On the array image, characters superimposed on the route curve are obtained, and the characters are combined in sequence according to their order on the route curve to obtain the management code.

[0150] In this embodiment, it specifically includes:

[0151] Generate images from character arrays: First, convert the character array into a visual image. This usually involves mapping each character or character combination to a specific position in the array, which can be represented by a corresponding symbol or color. In this way, an originally abstract character array is transformed into an intuitive visual image.

[0152] Superposition and adaptation of route curves:

[0153] Next, the predefined route curve is superimposed on the array image generated above. This route curve is generated based on the navigation route and includes a variety of shapes, such as wavy lines, broken lines, spiral lines, etc., but the key point is that it needs to adapt to the boundaries of the array image. Specifically, by stretching or scaling the route curve in equal proportions, ensure that its upper and lower edges are exactly tangent to the upper and lower edges of the array image. This is done to ensure that the coverage of the route curve in the entire array image is maximized while also maintaining the shape characteristics of the curve itself.

[0154] Get the characters superimposed on the route curve:

[0155] After the route curve is successfully superimposed on the array image, the next step is to identify which characters are above the route curve or directly in contact with it. This means that each character in the array image needs to be analyzed to determine whether they intersect with the route curve. Those characters that do intersect with or are on the route curve will be selected as the raw material for generating the management code.

[0156] Generation of management code:

[0157] Finally, the characters superimposed on the route curve are sorted according to their relative positions on the curve, that is, according to the actual direction of the route curve, these characters are combined in sequence to form the final management code. This management code not only contains the information of the character array, but also incorporates the characteristics of the route curve, so it has high uniqueness and security.

[0158] In this embodiment, the beneficial effects of the above technical solution are:

[0159] Enhanced security: By combining character arrays and route curves, the generated management code is highly random and unpredictable, effectively preventing malicious cracking and forgery.

[0160] Flexibility: The shape, direction and position of the route curve can be flexibly changed, so that even if the same character array is used, the management code generated each time may be completely different, which increases flexibility and adaptability.

[0161] Visual verification: Since the management code generation process involves visual array images and route curves, this provides users with an intuitive means of verification, enhancing transparency and trust.

[0162] In summary, the above technical solution cleverly combines graphical elements and logical operations, providing a safe and intuitive management code generation method that is suitable for a variety of scenarios requiring identity authentication and information protection.

[0163] In one embodiment, generating the management code based on the route curve and the character array includes:

[0164] According to the preset selection rules, six key points are selected on the route curve, and they are divided into three groups of key points according to the upper and lower positions of each key point; wherein each group includes two key points;

[0165] For each group of key points, connect the two key points into a straight line; the straight lines obtained from each group are used as target lines;

[0166] Superimposing the target straight line with the character array, and obtaining characters intersecting with each of the target straight lines in the character array as target characters;

[0167] The target characters are combined in sequence according to the order of the target characters in the character array to obtain the management code.

[0168] In this embodiment, it specifically includes:

[0169] Select key points and group them: First, select six key points from the preset route curve. The selection of these key points is based on preset selection rules, such as equidistant distribution, specific geometric characteristics, etc. Then, divide the six key points into three groups according to their top-down positions on the route, each group containing two key points. This grouping method ensures that different parts of the route curve can be covered in subsequent processing, thereby increasing the diversity of management codes. In other embodiments, any two key points can also be grouped together, for a total of three groups.

[0170] Straight line generation and target straight line determination: For each set of key points, a straight line is formed by connecting two points in the same set. In this way, three straight lines are generated from the three sets of key points respectively, and these straight lines are called target straight lines. These target straight lines not only represent specific sections on the route curve, but also have a certain degree of randomness and uniqueness because they are determined by key points.

[0171] Superposition of character array and straight line: Next, these target straight lines are superimposed on the previously defined character array. A character array is a two-dimensional layout that contains a series of characters arranged in a specific order. When the target straight line is superimposed on the character array, the straight line will pass through certain characters, and these characters intersecting with the straight line are marked as target characters. It can be understood that when the above target straight line is superimposed on the character array, due to the characteristics of the straight line, it can be extended so that the straight line intersects with as many characters as possible.

[0172] Generation of management codes: The last step is to combine the target characters in the original order of the character array to form the final management code. This combination ensures that even if the same character array and route curve are used, the generated management code will be different as long as the position or number of key points changes. Therefore, the generation of management codes depends not only on the content of the character array, but also on the route curve and key point selection strategy.

[0173] In this embodiment, the beneficial effects of the above technical solution are:

[0174] High security: Since the generation of the management code depends on multiple random or variable factors (such as the selection of key points and the shape of the route curve), this greatly increases the difficulty of cracking and improves security.

[0175] Flexibility: The selection rules of key points and the design of route curves can be adjusted according to actual needs, providing a high degree of customization.

[0176] Non-repetitive: Even if the same character array is used, due to the random selection of key points, each generated management code will be unique, avoiding the problem of repetitiveness.

[0177] In summary, the above technical solution provides an efficient and secure management code generation mechanism by combining route curves and character arrays in a unique way. It is suitable for applications in scenarios that require high personalization and security, such as password generation, access control, information encryption and other fields.

[0178] Reference Figure 2 In one embodiment of the present invention, there is also provided an intelligent navigation system for transport vehicles of an unmanned intelligent cargo station, comprising:

[0179] The acquisition unit is used to collect image information on the travel path of the transport vehicle in the unmanned intelligent cargo station in real time; perform feature recognition and orientation recognition on the object in the image information to obtain feature information and orientation information of the object;

[0180] An analysis unit, configured to determine whether the object is on the travel path based on the orientation information; if the object is on the travel path, analyze the feature information based on a preset deep learning model to obtain an obstacle level of the object;

[0181] an identification unit, configured to obtain a corresponding detour strategy based on the obstacle level; identify a new route that can be passed based on the image information, and determine whether the new route complies with the detour strategy;

[0182] The navigation unit is used to take the new route as a new travel path and navigate to the new travel path if it meets the detour strategy.

[0183] In this embodiment, for the specific implementation of each unit in the above system embodiment, please refer to the description in the above method embodiment, which will not be repeated here.

[0184] Reference Figure 3 The present invention also provides a transport vehicle in an embodiment, the internal structure of which can be as follows Figure 3 As shown. The transport vehicle includes a processor, a memory, a display screen, an input device, a network interface and a database connected through a system bus. Among them, the computer-designed processor is used to provide computing and control capabilities. The memory of the transport vehicle includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the transport vehicle is used to store the corresponding data in this embodiment. The network interface of the transport vehicle is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, the above method is implemented.

[0185] Those skilled in the art will understand that Figure 3 The structure shown in the figure is merely a block diagram of a part of the structure related to the solution of the present invention, and does not constitute a limitation on the transport vehicle to which the solution of the present invention is applied.

[0186] An embodiment of the present invention further provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the above method is implemented. It can be understood that the computer-readable storage medium in this embodiment can be a volatile readable storage medium or a non-volatile readable storage medium.

[0187] In summary, the intelligent navigation method, system and transport vehicle for the unmanned intelligent cargo station provided in the embodiment of the present invention include: real-time collection of image information on the travel path of the transport vehicle in the unmanned intelligent cargo station; feature recognition and orientation recognition of the objects in the image information to obtain the feature information and orientation information of the object; based on the orientation information, determine whether the object is on the travel path; if it is on the travel path, analyze the feature information based on a preset deep learning model to obtain the obstacle level of the object; based on the obstacle level, obtain the corresponding detour strategy; identify a new route that can be passed based on the image information, and judge whether the new route meets the detour strategy; if it meets the detour strategy, use the new route as a new travel path and navigate to the new travel path. In the present invention, it is possible to collect and analyze image information in real time, accurately identify and locate obstacles, evaluate their impact on the travel path, and then intelligently decide on the detour strategy to ensure that the transport vehicle can complete the task safely and efficiently in a complex environment. The high-precision recognition capability of the deep learning model and the path planning capability of the intelligent algorithm can significantly improve the autonomy and adaptability of the unmanned guided vehicle.

[0188] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media provided by the present invention and used in the embodiments may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double-speed data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM.

[0189] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, device, article or method. In the absence of further restrictions, an element defined by the sentence "includes a ..." does not exclude the presence of other identical elements in the process, device, article or method including the element.

[0190] The above description is only a preferred embodiment of the present invention, and does not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. An intelligent navigation method for a transport vehicle in an unmanned intelligent cargo station, characterized in that: The following steps are involved: Real-time collection of image information on the path of the transporter in the unmanned intelligent cargo station; Performing feature recognition and orientation recognition on the object in the image information to obtain feature information and orientation information of the object; Based on the position information, determining whether the object is on a travel path; If it is on the travel path, the feature information is analyzed based on a preset deep learning model to obtain the obstacle level of the object; Based on the obstacle level, a corresponding detour strategy is obtained; Identifying a new route that can be navigated based on the image information, and determining whether the new route complies with the detour strategy; If the detour strategy is met, the new route is used as a new travel path, and the vehicle is navigated to the new travel path; After navigating to a new path, the method includes: After circumnavigating the object via the new travel path, creating a navigation map in a navigation system of the transport vehicle, and storing the new travel path in the navigation map; Sending the navigation map to a management terminal; wherein the management terminal generates a management code for the navigation map and sends it to the transport vehicle; When the transport vehicle subsequently requests the navigation map, the management code is sent to the management terminal for authority verification; The management terminal generates a management code for the navigation map, including: Performing curve simulation on the navigation route in the navigation map to obtain a route curve; Acquire identification information of the transport vehicle, and generate a corresponding character array based on the identification information; wherein the character array is composed of a combination of multiple rows and columns of characters; generating the management code based on the route curve and the character array; The generating the management code based on the route curve and the character array comprises: Generate a picture of the character array to obtain an array picture; The route curve is superimposed on the array image, and the route curve is proportionally stretched or scaled so that the upper edge and the lower edge of the route curve are tangent to the upper edge and the lower edge of the array image; On the array image, characters superimposed on the route curve are obtained, and the characters are sequentially combined according to their order on the route curve to obtain the management code; Alternatively, generating the management code based on the route curve and the character array includes: According to the preset selection rules, six key points are selected on the route curve, and they are divided into three groups of key points according to the upper and lower positions of each key point; wherein each group includes two key points; For each group of key points, connect the two key points into a straight line; the straight lines obtained from each group are used as target lines; Superimposing the target straight line with the character array, and obtaining characters intersecting with each of the target straight lines in the character array as target characters; The target characters are combined in sequence according to the order of the target characters in the character array to obtain the management code.

2. The intelligent navigation method for transport vehicles in an unmanned intelligent cargo station according to claim 1, characterized in that: The performing feature recognition and orientation recognition on the object in the image information to obtain feature information and orientation information of the object includes: Using an object detection algorithm to locate an object bounding box in the image information; Applying a deep convolutional neural network to extract feature information from the object bounding box; wherein the feature information is used to describe the appearance and structural characteristics of the object; Based on the position of the object boundary box in the image information and the camera parameters of the transport vehicle, the azimuth angle and the tilt angle of the object relative to the transport vehicle are calculated through triangular geometric relationship as the azimuth information.

3. The intelligent navigation method for transport vehicles in an unmanned intelligent cargo station according to claim 1, characterized in that: The analyzing the feature information based on a preset deep learning model to obtain the obstacle level of the object includes: Inputting the feature information into the preset deep learning model; wherein the preset deep learning model includes a trained size obstacle sub-model, a shape obstacle sub-model and a classification sub-model; the feature information includes size features and shape features; Analyzing the size feature based on the size obstacle sub-model to obtain a first obstacle level corresponding to the object size; Analyzing the shape feature based on the shape obstacle sub-model to obtain a second obstacle level corresponding to the object shape; The first obstacle level and the second obstacle level are input into the classification sub-model to obtain the obstacle level of the object.

4. The method for intelligent navigation of a transport vehicle in an unmanned intelligent cargo station according to claim 1, characterized in that: Generating a corresponding character array based on the identification information includes: Based on a conversion relationship between preset characters and character combinations, convert each character in the identification information into a character combination; Create a blank array, and add the characters in each character combination to the blank array one by one in sequence to form a character array with multiple rows and columns.

5. An intelligent navigation system for transport vehicles in an unmanned intelligent cargo station, characterized in that: include: The acquisition unit is used to collect image information on the path of the transport vehicle in the unmanned intelligent cargo station in real time; Performing feature recognition and orientation recognition on the object in the image information to obtain feature information and orientation information of the object; an analyzing unit, configured to determine whether the object is on a travel path based on the position information; If it is on the travel path, the feature information is analyzed based on a preset deep learning model to obtain the obstacle level of the object; an identification unit, configured to obtain a corresponding detour strategy based on the obstacle level; Identifying a new route that can be navigated based on the image information, and determining whether the new route complies with the detour strategy; a navigation unit, configured to take the new route as a new travel path and navigate to the new travel path if the detour strategy is met; After navigating to a new path, the method includes: After circumnavigating the object via the new travel path, creating a navigation map in a navigation system of the transport vehicle, and storing the new travel path in the navigation map; Sending the navigation map to a management terminal; wherein the management terminal generates a management code for the navigation map and sends it to the transport vehicle; When the transport vehicle subsequently requests the navigation map, the management code is sent to the management terminal for authority verification; The management terminal generates a management code for the navigation map, including: Performing curve simulation on the navigation route in the navigation map to obtain a route curve; Acquire identification information of the transport vehicle, and generate a corresponding character array based on the identification information; wherein the character array is composed of a combination of multiple rows and columns of characters; generating the management code based on the route curve and the character array; The generating the management code based on the route curve and the character array comprises: Generate a picture of the character array to obtain an array picture; The route curve is superimposed on the array image, and the route curve is proportionally stretched or scaled so that the upper edge and the lower edge of the route curve are tangent to the upper edge and the lower edge of the array image; On the array image, characters superimposed on the route curve are obtained, and the characters are sequentially combined according to their order on the route curve to obtain the management code; Alternatively, generating the management code based on the route curve and the character array includes: According to the preset selection rules, six key points are selected on the route curve, and they are divided into three groups of key points according to the upper and lower positions of each key point; wherein each group includes two key points; For each group of key points, connect the two key points into a straight line; the straight lines obtained from each group are used as target lines; Superimposing the target straight line with the character array, and obtaining characters intersecting with each of the target straight lines in the character array as target characters; The target characters are combined in sequence according to the order of the target characters in the character array to obtain the management code.

6. A transport vehicle, comprising a memory and a processor, wherein a computer program is stored in the memory, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 4 are implemented.

Images