Intelligent luggage obstacle avoidance method and system based on wireless sensor

Abstract

The invention discloses an intelligent luggage obstacle avoidance method and system based on a wireless sensor, and relates to the technical field of intelligent navigation, and the method comprises the steps: inputting image data into a stacking hourglass network, taking an obstacle region as a guide clue, mapping the detection range of an ultra-wideband sensor to a corresponding ultra-wideband sector, and carrying out the detection of the ultra-wideband sector; binding the comprehensive time difference value with an obstacle area by using area data binding, packaging into focused time difference data, and calculating a three-dimensional coordinate set of the obstacle and the direction of the obstacle; constructing a local grid map, adjusting a pheromone value by using an exponential decay function, setting a weight by using an analytic hierarchy process, performing weighted scoring, and generating an optimal path; a stacked hourglass network is used for marking an obstacle area as a guide clue, interference of irrelevant areas is reduced, a local grid map is constructed, an exponential attenuation function is used for adjusting a pheromone value, pheromones are bound with the obstacle area, an optimized path is generated, and the obstacle avoidance capacity and the autonomous navigation performance of the intelligent luggage are improved.

Description

Technical Field

[0001] The present invention relates to the field of intelligent navigation technology, and in particular to an intelligent luggage obstacle avoidance method and system based on wireless sensors. Background Art

[0002] As a convenient mobile tool, smart bags are increasingly used in daily life and logistics distribution. In the design of smart bags, how to avoid obstacles efficiently and safely, especially to plan paths in complex dynamic environments, has become an important challenge for its intelligence. With the rapid development of smart devices and robotics technology, obstacle avoidance technology based on wireless sensors has been widely used in many fields, especially in the fields of intelligent navigation and autonomous mobile devices. Wireless sensor technology, such as ultra-wideband (UWB) sensors, ultrasonic sensors, and the integration of visual sensors, has become the key to achieving accurate environmental perception and path planning.

[0003] The existing intelligent luggage obstacle avoidance technology has some shortcomings. In scenes with many obstacles and complex spaces, a single sensor data may not be able to provide sufficient accuracy and reliability, which in turn affects the obstacle avoidance effect of the equipment. Existing path planning methods (such as DWA or artificial potential field method) rely on static weights or fixed thresholds in dynamic environments, which are difficult to adapt to changes in obstacle movement speed and distribution. They are prone to falling into local optimality or frequent path adjustments, affecting navigation efficiency. Summary of the invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides an intelligent luggage obstacle avoidance method and system based on wireless sensors to solve the problem that existing path planning methods (such as DWA or artificial potential field method) rely on static weights or fixed thresholds in dynamic environments, are difficult to adapt to changes in obstacle movement speed and distribution, and are prone to falling into local optimality or frequent path adjustments, affecting navigation efficiency.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: In a first aspect, the present invention provides an intelligent luggage obstacle avoidance method based on wireless sensors, which comprises: Collect sensor data, including time difference, echo time, and image data; The image data is input into the stacked hourglass network, the obstacle area is marked, the obstacle area is used as a guiding clue, the detection range of the ultra-wideband sensor is mapped to the corresponding ultra-wideband sector, the comprehensive time difference value is bound to the obstacle area using regional data binding, and encapsulated into focused time difference data, and the three-dimensional coordinate set of the obstacle and the direction of the obstacle are calculated; Construct a local grid map, use an exponential decay function to adjust the pheromone value, combine the pheromone value, use the analytic hierarchy process to set the weight and perform weighted scoring to generate the optimal path; The optimal path is decomposed into small motion instructions and executed, the execution results are monitored and adjusted, and the sensor data generated is stored, collected and analyzed.

[0007] As a preferred solution of the intelligent luggage obstacle avoidance method based on wireless sensors of the present invention, the calculation of the three-dimensional coordinate set of obstacles and the direction of obstacles includes: The wireless sensor includes an ultra-wideband sensor, an ultrasonic sensor and a camera; The distance between the obstacle and the ultrasonic sensor is calculated using the acoustic distance formula, which is defined as the ultrasonic distance; The image data is input into the stacked hourglass network. Through convolution and upsampling, the characteristic regions of obstacles in the image are detected. The characteristic regions are segmented using the semantic segmentation method, and the obstacle regions are marked. Taking the obstacle area as a guiding clue, the detection range of the ultra-wideband sensor is divided into sectors using the edge detection method and mapped to the corresponding ultra-wideband sectors; Use the weighted average method to calculate the comprehensive time difference value of all time difference data for weighted processing, use regional data binding to bind the comprehensive time difference value to the obstacle area, and encapsulate it into focused time difference data; The distance to the obstacle is calculated using the two-way ranging method, which is defined as the time difference distance; Use the maximum entropy threshold method to set the error threshold, use the absolute error method to calculate the error between the ultrasonic distance and the time difference distance, if the error is less than or equal to the error threshold, calculate the ultrasonic distance and the time difference distance as the final distance, if the error is greater than the error threshold, use the ultrasonic distance as the final distance; Taking the smart bag as the origin, the mean of the ultra-wideband sector is calculated as the relative angle of the obstacle, and the two-dimensional coordinates of the obstacle are calculated using the triangulation positioning method; The Canny edge detection method is used to extract the outline of the obstacle, and the outline is converted into the two-dimensional coordinates of the outline. The two-dimensional coordinates of the obstacle are used as a reference, and the two-dimensional coordinates of the outline and the two-dimensional coordinates of the obstacle are converted into three-dimensional coordinates using the parallax method to generate a three-dimensional coordinate set of the obstacle. The bounding box method is used to calculate the outer bounding box of the obstacle, and the principal component analysis method is used to extract the principal axis direction as the direction of the obstacle.

[0008] As a preferred solution of the intelligent luggage obstacle avoidance method based on wireless sensors described in the present invention, the sensor data is first pre-processed.

[0009] As a preferred solution of the intelligent luggage obstacle avoidance method based on wireless sensors of the present invention, the local grid map is constructed, the pheromone value is adjusted by using an exponential decay function, and the weight is set by using the analytic hierarchy process in combination with the pheromone value and weighted scoring is performed to generate the optimal path, including: Use the grid map method to place the smart box at the center of the map as the starting point and build a local grid map. Each grid is assigned an initial pheromone value, and the initial values ​​of all grids are set to "free". The values ​​of the grids corresponding to the three-dimensional coordinate set of the obstacle are changed from "free" to "occupied"; The coordinate averaging method is used to calculate the obstacle center as the reference point, the area expansion method is used to mark the grids in the sector as the affected area, and the exponential decay function is used to adjust the pheromone to obtain the updated local grid map; Use positioning algorithm to calculate current speed; Use kinematic equations to predict the end position of the path; Generate candidate paths using the artificial potential field method, use the three-dimensional coordinate set of obstacles in the local grid map to detect whether the grid corresponding to the candidate path is marked as "occupied", and delete the overlapping "occupied" grids; Select the remaining candidate paths, extract the pheromone corresponding to the end position of the path, use the Euclidean distance formula to calculate the path length, use the nearest point distance method to calculate the minimum distance between the path and the obstacle, use the hierarchical analysis method to set the weight coefficient, use the weighted scoring method to calculate the final score of the path, and sort it from high to low, and select the candidate path corresponding to the highest score as the optimal path.

[0010] As a preferred solution of the intelligent luggage obstacle avoidance method based on wireless sensors of the present invention, the optimal path is decomposed into small motion instructions and executed, including: Use a path decomposition algorithm to decompose the optimal path into small motion instructions; The small motion instructions are transmitted to the smart luggage using a wireless communication protocol. The smart luggage receives the small motion instructions, converts the small motion instructions into control signals using a direct torque control algorithm, and executes them.

[0011] As a preferred solution of the intelligent luggage obstacle avoidance method based on wireless sensors of the present invention, the monitoring and adjustment of the execution results include: Collect feedback position data and preprocess it, and use the YOLO deep learning algorithm to set the target position; Calculate the difference between the target position and the preprocessed feedback position data using the deviation method; Use empirical rules to set the stop threshold, compare the difference with the stop threshold, and when the difference is less than the stop threshold, it is considered normal and monitoring continues; When the difference is greater than or equal to the stop threshold, it is judged as an abnormal state. The PID control algorithm is used to calculate the adjustment proportion, integral and differential terms to obtain the control amount and adjust the optimal path.

[0012] As a preferred solution of the intelligent luggage obstacle avoidance method based on wireless sensors of the present invention, the sensor data generated by storage, collection and analysis includes: The collected video data, abnormal monitoring results generated by analysis, and time indexes are stored in the central database, and security access measures are set. The central database will back up the stored data to the cloud and regularly perform integrity checks on the stored data and backup data. After the test is completed, the integrity test record is generated and stored synchronously in the central database.

[0013] In a second aspect, the present invention provides an intelligent luggage obstacle avoidance system based on wireless sensors, comprising: A collection and calculation module is used to collect sensor data, including time difference, echo time and image data, input the image data into the stacked hourglass network, mark the obstacle area, use the obstacle area as a guide clue, map the detection range of the ultra-wideband sensor to the corresponding ultra-wideband sector, use regional data binding to bind the comprehensive time difference value to the obstacle area, encapsulate it into focused time difference data, and calculate the three-dimensional coordinate set of the obstacle and the direction of the obstacle; The path building module is used to build a local grid map, adjust the pheromone value using an exponential decay function, combine the pheromone value, use the analytic hierarchy process to set the weight and perform weighted scoring to generate the optimal path; The monitoring and storage module is used to decompose the optimal path into small motion instructions and execute them, monitor and adjust the execution results, and store, collect and analyze the generated sensor data.

[0014] In a third aspect, the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, any step of the intelligent luggage obstacle avoidance method based on wireless sensors as described in the first aspect of the present invention is implemented.

[0015] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, any step of the intelligent luggage obstacle avoidance method based on wireless sensors as described in the first aspect of the present invention is implemented.

[0016] The beneficial effects of the present invention are as follows: the present invention uses a stacked hourglass network to mark obstacle areas as guiding clues, reduces interference from irrelevant areas, constructs a local grid map and uses an exponential decay function to adjust the pheromone value, binds the pheromone to the obstacle area, generates an optimized path, and improves the obstacle avoidance capability and autonomous navigation performance of the smart luggage. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0018] Figure 1 This is a flow chart of the intelligent luggage obstacle avoidance method based on wireless sensors in Example 1.

[0019] Figure 2 This is a structural diagram of the intelligent luggage obstacle avoidance system based on wireless sensors in Example 1. DETAILED DESCRIPTION

[0020] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0021] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0022] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0023] Example 1, reference Figure 1 and Figure 2 , which is the first embodiment of the present invention, and provides an intelligent luggage obstacle avoidance method based on wireless sensors, comprising the following steps: S1, collect sensor data, including time difference, echo time and image data; The image data is input into the stacked hourglass network, the obstacle area is marked, the obstacle area is used as a guiding clue, the detection range of the ultra-wideband sensor is mapped to the corresponding ultra-wideband sector, the comprehensive time difference value is bound to the obstacle area using regional data binding, and encapsulated into focused time difference data, and the three-dimensional coordinate set of the obstacle and the direction of the obstacle are calculated; Specifically, the three-dimensional coordinate set of the obstacle and the direction of the obstacle are calculated, including: The wireless sensor includes an ultra-wideband sensor, an ultrasonic sensor and a camera; The distance between the obstacle and the ultrasonic sensor is calculated using the acoustic distance formula, which is defined as the ultrasonic distance; Use linear interpolation to synchronize the collected sensor data, use median filtering to denoise the time difference and echo time data, use Z-score detection to identify abnormal data in the time difference and echo time data, use mean interpolation to fill in the missing data in the time difference and echo time data, use Gaussian blur to denoise the image data, use histogram equalization to enhance the image data, and use the minimum-maximum normalization method to normalize the sensor data; The image data is input into the stacked hourglass network. Through convolution and upsampling, the characteristic regions of obstacles in the image are detected. The characteristic regions are segmented using the semantic segmentation method, and the regions dominated by obstacles are marked. Taking the obstacle area as a guiding clue, the detection range of the ultra-wideband sensor is divided into sectors using the edge detection method and mapped to the corresponding ultra-wideband sectors; Use the weighted average method to calculate the comprehensive time difference value of all time difference data for weighted processing, use regional data binding to bind the comprehensive time difference value to the obstacle area, and encapsulate it into focused time difference data; The distance to the obstacle is calculated using the two-way ranging method, which is defined as the time difference distance. The formula is: , Where d is the distance of the obstacle, c is the signal propagation speed, is the time difference data after focusing; Use the maximum entropy threshold method to set the error threshold, use the absolute error method to calculate the error between the ultrasonic distance and the time difference distance, if the error is less than or equal to the error threshold, calculate the ultrasonic distance and the time difference distance as the final distance, if the error is greater than the error threshold, use the ultrasonic distance as the final distance; Taking the smart bag as the origin, the mean of the ultra-wideband sector is calculated as the relative angle of the obstacle, and the two-dimensional coordinates of the obstacle are calculated using the triangulation positioning method; The Canny edge detection method is used to extract the outline of the obstacle, and the outline is converted into the two-dimensional coordinates of the outline. The two-dimensional coordinates of the obstacle are used as a reference, and the two-dimensional coordinates of the outline and the two-dimensional coordinates of the obstacle are converted into three-dimensional coordinates using the parallax method to generate a three-dimensional coordinate set of the obstacle. The bounding box method is used to calculate the outer bounding box of the obstacle, and the width, height and depth of the bounding box are determined. The principal component analysis method is used to extract the main axis direction as the direction of the obstacle.

[0024] Traditional UWB obstacle avoidance systems (such as separate TDOA processing) usually perform indiscriminate scanning of the entire detection range and are unable to optimize signal processing for specific areas. This step uses the visual area ("30% on the right") as a clue to dynamically adjust the focus of TDOA data processing. Compared with traditional full-domain ranging technology, it significantly reduces interference from irrelevant areas (such as signal noise on the left), improves perception efficiency and accuracy, and focuses on wireless signals under visual guidance, similar to the human perception mechanism of "see first and then touch". This cross-sensory collaboration has no precedent in smart luggage obstacle avoidance. Existing technologies mostly process visual and wireless data in parallel, rather than allowing vision to actively intervene in the priority of wireless signals, breaking through traditional independent operations. The limitations of traditional temporal consistency analysis (such as target tracking in robot navigation) usually rely on a single sensor (such as a camera or radar), while this step combines visual contour changes and wireless distance changes to form a cross-sensory collaborative dynamic inference. Compared with the use of visual tracking (susceptible to occlusion) or wireless ranging (lack of morphological information) alone, this method can more accurately judge the behavior of obstacles. The joint analysis of the time series of visual and wireless data realizes the accurate prediction of the movement trend of obstacles, especially in dynamic scenes (such as areas with dense pedestrians). The traditional obstacle avoidance system has not attempted to deduce dynamic behavior through the synchronous changes of visual and wireless signals. This cross-domain integration breaks through the limitations of a single technology; The multi-scale feature extraction capability of the stacked hourglass network combined with the pixel-level classification of semantic segmentation ensures the accuracy of obstacle area marking. Compared with traditional edge detection or threshold segmentation, it can more accurately distinguish targets (such as pedestrians and luggage) in complex backgrounds. The marked area provides visual clues for ultra-wideband sector division, breaking through the limitation of traditional wireless sensors relying on signal strength. The visually guided sector division enables the ultra-wideband detection range to dynamically focus on the obstacle area, avoiding signal confusion of fixed sectors in multi-target scenarios. Compared with the traditional uniform division method, it significantly improves the positioning targeting. Regional data binding associates the time difference value with a specific obstacle to form "focused data", reducing redundant signal interference and ensuring the accuracy of subsequent distance calculations. It integrates cross-vision and wireless signals and makes full use of the spatial resolution of the image and the time domain of ultra-wideband. Accuracy, achieving deep collaboration among multiple sensors, the maximum entropy threshold method dynamically optimizes the threshold according to the error distribution, and can better adapt to environmental noise (such as sound wave reflection, electromagnetic interference) compared to the fixed threshold method, and improves the reliability of distance fusion. The conversion from two-dimensional to three-dimensional combines contour and depth information to generate a complete spatial representation of obstacles. Compared with the traditional method of only two-dimensional positioning, it can more accurately describe the position and shape of obstacles. The main axis direction extracted by PCA reflects the movement trend of obstacles (such as pedestrian orientation), which provides a prediction basis for dynamic obstacle avoidance and makes up for the shortcomings of static positioning. It not only improves the accuracy and robustness of obstacle detection, but also provides high-quality input for subsequent path planning. Through vision-guided ultra-wideband sector division, regional data binding and maximum entropy threshold optimization, multi-sensor fusion is brought to a new level.

[0025] S2, construct a local grid map, use the exponential decay function to adjust the pheromone value, combine the pheromone value, use the hierarchical analysis method to set the weight and perform weighted scoring to generate the optimal path; Specifically, a local grid map is constructed, the pheromone value is adjusted using an exponential decay function, and the pheromone value is combined with the analytic hierarchy process to set the weight and perform weighted scoring to generate the optimal path, including: Use the grid map method to place the smart box at the center of the map as the starting point and build a local grid map. Each grid is assigned an initial pheromone value. The initial values ​​of all grids are set to "free". Each grid represents a location in the map. There are no obstacles at the time of initialization. The value of the grid corresponding to the obstacle's three-dimensional coordinate set is changed from "free" to "occupied"; The coordinate averaging method is used to calculate the obstacle center as the reference point, the area expansion method is used to mark the grids in the sector as the affected area, and the exponential decay function is used to adjust the pheromone so that the closer the pheromone value is to the obstacle, the faster it decays, and the updated local grid map is obtained; Use the positioning algorithm to calculate the current speed. The formula is: , Where v is the current speed, is the displacement distance of the smart bag, is the time interval; The kinematic equation is used to predict the end position of the path: , , in and is the end point of the path, and are the current position of the path, is the direction of the obstacle; Generate candidate paths using the artificial potential field method, use the three-dimensional coordinate set of obstacles in the local grid map to detect whether the grid corresponding to the candidate path is marked as "occupied", and delete the overlapping "occupied" grids; Select the remaining candidate paths, extract the pheromone corresponding to the end position of the path, use the Euclidean distance formula to calculate the path length, use the closest point distance method to calculate the minimum distance between the path and the obstacle, use the hierarchical analysis method to set the weight coefficient, use the weighted scoring method to calculate the final score of the path, and sort them from high to low, and select the candidate path corresponding to the highest score as the optimal path; In traditional ant colony algorithms, pheromone updates are mostly based on path selection (such as increments) or uniform volatilization (such as time decay), and the dynamic impact of obstacle distance is not directly considered. The exponential decay function (commonly used in the field of signal processing) is introduced into the pheromone mechanism across fields to form a "distance-sensitive pheromone distribution". This allows the pheromone value to not only reflect the path preference, but also directly encode the degree of obstacle threat (the closer the obstacle, the more dangerous it is), providing more fine-grained environmental information for subsequent path selection. Compared with traditional uniform or linear decay, this exponential decay method simulates the characteristic that danger decreases rapidly with distance in reality (such as the threat increases sharply when pedestrians approach luggage), which improves the dynamic adaptability of obstacle avoidance. Adaptability, using pheromone value (reflecting historical experience and obstacle impact) as a scoring factor, combined with path length and minimum distance, to form a multi-dimensional evaluation system. AHP is usually used for strategic decision-making. The technical solution of the present invention introduces it into real-time path selection, adjusts the weight according to the dynamic environment (such as obstacle density, moving speed), and breaks through the limitations of traditional fixed weights. Compared with the static scoring of DWA, the technical solution of the present invention realizes "environmental adaptive scoring" through pheromone and AHP, making path selection more intelligent (such as preferring safe paths in pedestrian-dense areas), improving the obstacle avoidance ability of luggage in complex dynamic environments (such as airports), and avoiding falling into local traps; The local grid map centered on the luggage provides a high-resolution spatial representation. It directly marks "occupancy" in combination with three-dimensional coordinates. Compared with the traditional two-dimensional grid, it can more realistically reflect the distribution of obstacles and improve environmental perception. The initial assignment of "idle" and dynamic update of "occupancy" simplifies the calculation complexity and is suitable for the real-time processing needs of portable devices. The standardized grid structure is easy to integrate with other modules (such as path planning, pheromone update), enhancing the scalability of the solution. The exponential decay function makes the pheromone value decrease rapidly with the distance from the obstacle (such as close to 0 at close range and tending to the initial value of 1 at a distance). Compared with the traditional uniform or linear attenuation, it simulates the threat distribution more realistically (such as the sharp risk when pedestrians approach). Combined with the regional expansion method, pheromone adjustment is not limited to the "occupied" grid. It not only extends to the impact area, but also refines the granularity of threat coding, combines pheromone value (historical experience), path length (efficiency) and minimum distance (safety) to form a comprehensive scoring framework, which significantly improves the intelligence of decision-making compared to traditional single indicators (such as the shortest distance). The hierarchical analysis method adjusts the weights according to environmental dynamics (such as the safety weight increases to 0.6 when the obstacle density is high), breaking through the limitations of fixed weights and realizing adaptive path selection. It shows significant creativity through pheromone adjustment of exponential decay function and path scoring combined with pheromone analytic hierarchy process. The former introduces the concept of signal processing across fields and realizes the spatial coding of dynamic threats. The latter integrates management decision-making methods and bio-inspiration mechanisms to construct an environmentally adaptive scoring system.

[0026] S3, decompose the optimal path into small motion instructions and execute them, monitor and adjust the execution results, store, collect and analyze the generated sensor data; Specifically, the optimal path is decomposed into small motion instructions and executed, including: Use a path decomposition algorithm to decompose the optimal path into small motion instructions; The small motion instructions are transmitted to the smart luggage using a wireless communication protocol. The smart luggage receives the small motion instructions, converts the small motion instructions into control signals using a direct torque control algorithm, and executes them.

[0027] The path decomposition algorithm divides the continuous path into small-step instructions. Compared with directly executing the entire path, it reduces the cumulative error and ensures that the luggage strictly follows the planned trajectory. The decomposed small instructions are easy to interrupt or correct in real time (such as pausing a certain section of instructions when a new obstacle is detected), which enhances the flexibility of obstacle avoidance and is more adaptable than traditional whole-segment path execution. The wireless communication protocol ensures low latency and high stability of instruction transmission. The direct torque control algorithm achieves fast execution of small instructions by adjusting the motor torque in real time. DTC avoids complex vector transformation and directly adjusts the voltage based on torque demand, reducing motor power consumption. The detailed design of the path decomposition algorithm transforms traditional static path planning into a dynamically controllable instruction sequence. Combined with the real-time transmission of the wireless communication protocol, it breaks through the limitations of traditional wired or whole-segment execution. The introduction of the direct torque control algorithm applies advanced technology in the field of motor control to smart luggage across fields, achieving the unity of high precision and high efficiency.

[0028] Furthermore, the execution results are monitored and adjusted, including: Collect feedback position data and preprocess it, and use the YOLO deep learning algorithm to set the target position; Calculate the difference between the target position and the preprocessed feedback position data using the deviation method; Use empirical rules to set the stop threshold, compare the difference with the stop threshold, and when the difference is less than the stop threshold, it is considered normal and monitoring continues; When the difference is greater than or equal to the stop threshold, it is judged as an abnormal state. The PID control algorithm is used to calculate the adjustment proportion, integral and differential terms to obtain the control amount and adjust the optimal path.

[0029] The YOLO algorithm sets the target position through real-time image analysis. Compared with traditional static target setting, it can dynamically adapt to environmental changes (such as pedestrian movement) and improve the targeted monitoring. Through deep learning and fusion of visual feedback, it is not limited to the end of the path, but can also detect potential threats (such as falling objects) to provide forward-looking information for subsequent adjustments. The rule of thumb sets the stop threshold (for example, 0.015m) according to environmental characteristics (such as ground smoothness) to avoid the limitations of fixed thresholds. The PID feedback mechanism can still maintain stable execution under sensor noise or external disturbances, which makes up for the fragility of switch control. The introduction of the YOLO deep learning algorithm upgrades the traditional static target setting to dynamic visual perception. Combined with the deviation method and the rule of thumb, it forms an efficient state monitoring mechanism. The feedback adjustment of the PID control algorithm changes the path adjustment from passive correction to active optimization, integrating target detection and control theory across fields.

[0030] Furthermore, the sensor data generated by the storage, collection and analysis includes: The collected video data, abnormal monitoring results generated by analysis, and time indexes are stored in the central database, and security access measures are set. The central database will back up the stored data to the cloud and regularly perform integrity checks on the stored data and backup data. After the test is completed, the integrity test record is generated and stored synchronously in the central database.

[0031] It simplifies the data storage and retrieval process and also facilitates real-time analysis and monitoring of data. Through the central database, the system can achieve efficient data retrieval, analysis and further data mining, thereby improving the working efficiency of smart devices or monitoring systems. By backing up data in the cloud, it not only improves the availability of data storage, but also enhances the system's disaster recovery capabilities. Integrity detection can ensure that the data in the system is always consistent and accurate, avoiding system failures or decision-making errors caused by data errors. The combination of video data and abnormal monitoring results enables the system to accurately determine whether there are problems or abnormal behaviors, and take timely measures.

[0032] This embodiment also provides an intelligent luggage obstacle avoidance system based on wireless sensors, including: A collection and calculation module is used to collect sensor data, including time difference, echo time and image data, input the image data into the stacked hourglass network, mark the obstacle area, use the obstacle area as a guide clue, map the detection range of the ultra-wideband sensor to the corresponding ultra-wideband sector, use regional data binding to bind the comprehensive time difference value to the obstacle area, encapsulate it into focused time difference data, and calculate the three-dimensional coordinate set of the obstacle and the direction of the obstacle; The path building module is used to build a local grid map, adjust the pheromone value using an exponential decay function, combine the pheromone value, use the analytic hierarchy process to set the weight and perform weighted scoring to generate the optimal path; The monitoring and storage module is used to decompose the optimal path into small motion instructions and execute them, monitor and adjust the execution results, and store, collect and analyze the generated sensor data.

[0033] This embodiment also provides a computer device, which is suitable for the case of an intelligent luggage obstacle avoidance method based on wireless sensors, including: a memory and a processor; the memory is used to store computer executable instructions, and the processor is used to execute computer executable instructions to implement the intelligent luggage obstacle avoidance method based on wireless sensors as proposed in the above embodiment.

[0034] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covered on the display screen, or a key, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

[0035] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by the processor, the intelligent luggage obstacle avoidance method based on wireless sensors proposed in the above embodiment is implemented; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, referred to as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, referred to as EEPROM), erasable programmable read-only memory (Erasable Programmable Read Only Memory, referred to as EPROM), programmable read-only memory (Programmable Red-Only Memory, referred to as PROM), read-only memory (Read-Only Memory, referred to as ROM), magnetic storage, flash memory, magnetic disk or optical disk.

[0036] In summary, the present invention uses a stacked hourglass network to mark obstacle areas as guiding clues to reduce interference from irrelevant areas. By constructing a local grid map and using an exponential decay function to adjust the pheromone value, the pheromone is bound to the obstacle area to generate an optimized path, thereby improving the obstacle avoidance capability and autonomous navigation performance of the smart luggage.

[0037] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. An intelligent luggage obstacle avoidance method based on wireless sensors, characterized in that: include, Collect sensor data, including time difference, echo time, and image data; The image data is input into the stacked hourglass network, the obstacle area is marked, the obstacle area is used as a guiding clue, the detection range of the ultra-wideband sensor is mapped to the corresponding ultra-wideband sector, the comprehensive time difference value is bound to the obstacle area using regional data binding, and encapsulated into focused time difference data, and the three-dimensional coordinate set of the obstacle and the direction of the obstacle are calculated; Construct a local grid map, use an exponential decay function to adjust the pheromone value, combine the pheromone value, use the analytic hierarchy process to set the weight and perform weighted scoring to generate the optimal path; The optimal path is decomposed into small motion instructions and executed, the execution results are monitored and adjusted, and the sensor data generated is stored, collected and analyzed.

2. The intelligent luggage obstacle avoidance method based on wireless sensors as claimed in claim 1, characterized in that: The calculating the three-dimensional coordinate set of the obstacle and the direction of the obstacle includes: The wireless sensor includes an ultra-wideband sensor, an ultrasonic sensor and a camera; The distance between the obstacle and the ultrasonic sensor is calculated using the acoustic distance formula, which is defined as the ultrasonic distance; The image data is input into the stacked hourglass network. Through convolution and upsampling, the characteristic regions of obstacles in the image are detected. The characteristic regions are segmented using the semantic segmentation method, and the obstacle regions are marked. Taking the obstacle area as a guiding clue, the detection range of the ultra-wideband sensor is divided into sectors using the edge detection method and mapped to the corresponding ultra-wideband sectors; Use the weighted average method to calculate the comprehensive time difference value of all time difference data for weighted processing, use regional data binding to bind the comprehensive time difference value to the obstacle area, and encapsulate it into focused time difference data; The distance to the obstacle is calculated using the two-way ranging method, which is defined as the time difference distance; Use the maximum entropy threshold method to set the error threshold, use the absolute error method to calculate the error between the ultrasonic distance and the time difference distance, if the error is less than or equal to the error threshold, calculate the ultrasonic distance and the time difference distance as the final distance, if the error is greater than the error threshold, use the ultrasonic distance as the final distance; Taking the smart bag as the origin, the mean of the ultra-wideband sector is calculated as the relative angle of the obstacle, and the two-dimensional coordinates of the obstacle are calculated using the triangulation positioning method; The Canny edge detection method is used to extract the outline of the obstacle, and the outline is converted into the two-dimensional coordinates of the outline. The two-dimensional coordinates of the obstacle are used as a reference, and the two-dimensional coordinates of the outline and the two-dimensional coordinates of the obstacle are converted into three-dimensional coordinates using the parallax method to generate a three-dimensional coordinate set of the obstacle; The bounding box method is used to calculate the outer bounding box of the obstacle, and the principal component analysis method is used to extract the principal axis direction as the direction of the obstacle.

3. The intelligent luggage obstacle avoidance method based on wireless sensors as claimed in claim 2, characterized in that: The method of constructing a local grid map, adjusting the pheromone value using an exponential decay function, combining the pheromone value, using a hierarchical analysis method to set weights and perform weighted scoring, and generating an optimal path includes: Use the grid map method to place the smart box at the center of the map as the starting point and build a local grid map. Each grid is assigned an initial pheromone value, and the initial value of all grids is set to "free". The value of the grid corresponding to the obstacle's three-dimensional coordinate set is changed from "free" to "occupied"; The coordinate averaging method is used to calculate the obstacle center as the reference point, the area expansion method is used to mark the grids in the sector as the affected area, and the exponential decay function is used to adjust the pheromone to obtain the updated local grid map; Use positioning algorithm to calculate current speed; Use kinematic equations to predict the end position of the path; Generate candidate paths using the artificial potential field method, use the three-dimensional coordinate set of obstacles in the local grid map to detect whether the grid corresponding to the candidate path is marked as "occupied", and delete the overlapping "occupied" grids; Select the remaining candidate paths, extract the pheromone corresponding to the end position of the path, use the Euclidean distance formula to calculate the path length, use the nearest point distance method to calculate the minimum distance between the path and the obstacle, use the hierarchical analysis method to set the weight coefficient, use the weighted scoring method to calculate the final score of the path, and sort it from high to low, and select the candidate path corresponding to the highest score as the optimal path.

4. The intelligent luggage obstacle avoidance method based on wireless sensors as claimed in claim 3, characterized in that: Decomposing the optimal path into small motion instructions and executing them includes: Use a path decomposition algorithm to decompose the optimal path into small motion instructions; The small motion instructions are transmitted to the smart luggage using a wireless communication protocol. The smart luggage receives the small motion instructions, converts the small motion instructions into control signals using a direct torque control algorithm, and executes them.

5. The intelligent luggage obstacle avoidance method based on wireless sensors as claimed in claim 4, characterized in that: The monitoring and adjustment of the execution results include: Collect feedback position data and preprocess it, and use the YOLO deep learning algorithm to set the target position; Calculate the difference between the target position and the preprocessed feedback position data using the deviation method; Use empirical rules to set the stop threshold, compare the difference with the stop threshold, and when the difference is less than the stop threshold, it is considered normal and monitoring continues; When the difference is greater than or equal to the stop threshold, it is judged as an abnormal state. The PID control algorithm is used to calculate the adjustment proportion, integral and differential terms to obtain the control amount and adjust the optimal path.

6. The intelligent luggage obstacle avoidance method based on wireless sensors as claimed in claim 1, characterized in that: After the sensor data is received, a pre-processing operation is performed.

7. The intelligent luggage obstacle avoidance method based on wireless sensors as claimed in claim 5, characterized in that: The storage, collection and analysis of generated sensor data includes: The collected video data, abnormal monitoring results generated by analysis, and time indexes are stored in the central database, and security access measures are set. The central database will back up the stored data to the cloud and regularly perform integrity checks on the stored data and backup data. After the test is completed, the integrity test record is generated and stored synchronously in the central database.

8. An intelligent luggage obstacle avoidance system based on wireless sensors, based on the intelligent luggage obstacle avoidance method based on wireless sensors according to any one of claims 1 to 7, characterized in that: include, A collection and calculation module is used to collect sensor data, including time difference, echo time and image data, input the image data into the stacked hourglass network, mark the obstacle area, use the obstacle area as a guide clue, map the detection range of the ultra-wideband sensor to the corresponding ultra-wideband sector, use regional data binding to bind the comprehensive time difference value to the obstacle area, encapsulate it into focused time difference data, and calculate the three-dimensional coordinate set of the obstacle and the direction of the obstacle; The path building module is used to build a local grid map, adjust the pheromone value using an exponential decay function, combine the pheromone value, use the analytic hierarchy process to set the weight and perform weighted scoring to generate the optimal path; The monitoring and storage module is used to decompose the optimal path into small motion instructions and execute them, monitor and adjust the execution results, and store, collect and analyze the generated sensor data.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the intelligent luggage obstacle avoidance method based on wireless sensors described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the intelligent luggage obstacle avoidance method based on wireless sensors described in any one of claims 1 to 7 are implemented.

Images