Intelligent logistics park transport vehicle management method and system

Abstract

The invention is applicable to the technical field of logistics transport vehicle management, and provides an intelligent logistics park transport vehicle management method and system, and the method comprises the steps: judging whether an obstacle is related to an artificial dynamic behavior or not when a target transport vehicle encounters the obstacle on a preset transport route, and if the obstacle is determined to be related to the artificial dynamic behavior, determining that the obstacle is not related to the artificial dynamic behavior; if yes, setting the obstacle as a target obstacle; identifying a human control behavior feature category of the target obstacle and a designated area where the target obstacle is located, and obtaining a historical dynamic visual data set of the designated area, a dynamic obstacle event change log and a current initial path deviation control quantity of the target transport vehicle; according to the method, the path deviation control quantity of the transport vehicle is optimized by combining the dual characteristics of the spatial dynamic behavior and the removal response time of the target obstacle. Through comprehensive adjustment of the path deviation correction amplitude and the time correction deviation, multi-dimensional optimization of obstacle avoidance path planning is realized, and the accuracy and adaptability of dynamic obstacle avoidance can be effectively improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of logistics transport vehicle management, and particularly relates to a method and system for managing transport vehicles in an intelligent logistics park. Background Art

[0002] In modern logistics parks, unmanned transport vehicles, as an important link in realizing intelligent logistics, have achieved automated goods transportation functions to a certain extent. Existing technologies enable unmanned transport vehicles to bypass static or simple dynamic obstacles on a preset transport route through path planning and obstacle avoidance algorithms. However, these technologies mainly rely on real-time sensor data (such as lidar, cameras, etc.) and combine fixed path planning models to make obstacle avoidance decisions, lacking behavior prediction and response optimization for complex dynamic obstacles. This makes the accuracy and adaptability of obstacle avoidance of existing technologies still insufficient when facing artificial dynamic obstacles in complex logistics environments.

[0003] Specifically, when an unmanned transport vehicle encounters an artificial dynamic obstacle (such as forklift operation or temporarily stacked items), existing technologies usually only rely on real-time detection of its position and morphological changes and adjust the obstacle avoidance path based on static rules. This method fails to fully utilize historical data to predict and optimize the expansion trend and removal response time of obstacles, resulting in difficulties in balancing the safety and efficiency of path adjustment. In addition, when facing obstacles with a long removal time or a fast dynamic expansion speed, existing technologies cannot comprehensively evaluate the influence of spatial and temporal factors, easily causing phenomena such as excessive or insufficient path deviation, thereby increasing the risk of energy consumption and task delay. Summary of the Invention

[0004] The purpose of the present invention is to provide a method and system for managing transport vehicles in an intelligent logistics park, aiming to solve the problems raised in the background art.

[0005] The present invention is implemented as follows. An intelligent logistics park transport vehicle management method, the method comprising:

[0006] When a target transport vehicle encounters an obstacle on a preset transport route, determine whether the obstacle is related to an artificially induced dynamic behavior. If it is determined to be related, set it as a target obstacle;

[0007] Identify the human-controlled behavior feature category of the target obstacle and the designated area where it is located, obtain the historical dynamic visual data set, dynamic obstacle event change log of the designated area, and the current initial path deviation control amount of the target transport vehicle;

[0008] Analyze the historical dynamic visual data set, extract the target behavior visual data set with the same feature category as the human control behavior of the target obstacle, intelligently analyze the dynamic change trend of the space occupancy of the target obstacle based on the target behavior visual data set, and quantify the correction amplitude of its path deviation;

[0009] Parse the dynamic obstacle event change log, calculate the average removal time of historical obstacles of the same type after receiving the removal instruction, and compare it with the preset standard removal time to determine the time correction deviation;

[0010] Based on the path deviation correction amplitude and time correction deviation, the initial path deviation control amount is optimized and adjusted to generate the path deviation optimization control amount, which is then applied to the actual obstacle avoidance execution.

[0011] As a further limitation of the technical solution of the embodiment of the present invention, the steps of parsing the historical dynamic visual data set, extracting the target behavior visual data set with the same category as the human control behavior feature of the target obstacle, intelligently analyzing the dynamic change trend of the spatial occupancy of the target obstacle based on the target behavior visual data set, and quantifying the path deviation correction amplitude include:

[0012] Analyze the historical dynamic visual data set of the designated area where the target obstacle is located, use the target detection and classification algorithm to filter out the local video clips that meet the human control behavior feature category of the target obstacle, and generate the target behavior visual data set after filtering out irrelevant content;

[0013] In the target behavior visual dataset, the spatial occupied area of ​​historical obstacles is analyzed frame by frame. The dynamic expansion boundary and movement trajectory of historical obstacles are extracted by using visual segmentation and trajectory tracking technology. The expansion degree of historical obstacles in each local video clip is calculated, and a time series polyline is generated. The abnormal data points are removed by combining denoising processing to obtain a smooth trend of expansion changes.

[0014] Based on the expansion change curve, the average slope of the expansion trend is calculated to quantify the dynamic expansion behavior of historical obstacles and use it as the path deviation correction amplitude.

[0015] As a further limitation of the technical solution of the embodiment of the present invention, the steps of parsing the dynamic obstacle event change log, calculating the average removal time of the same type of historical obstacles after receiving the removal instruction, and comparing it with the preset standard removal time, and determining the time correction deviation include:

[0016] Extract the historical records with the same category as the human control behavior feature of the target obstacle in the dynamic obstacle event change log, including the timestamp of the historical obstacle receiving the removal instruction and the timestamp of the actual removal completion, calculate the removal duration of each historical record, and form a removal duration dataset;

[0017] Statistically analyze the removal duration dataset, eliminate outliers or extreme data points through denoising processing, calculate the average value of the removal duration, and obtain the average removal duration;

[0018] Compare the average removal duration with the preset standard removal duration, and calculate the time correction deviation.

[0019] As a further limitation of the technical solution of the embodiment of the present invention, based on the path deviation correction amplitude and the time correction deviation, the steps of optimizing and adjusting the initial path deviation control amount, generating a path deviation optimized control amount, and applying it to the actual obstacle avoidance execution include:

[0020] Retrieve the optimized calculation formula for the path deviation control amount, and optimize and adjust the initial path deviation control amount in combination with the path deviation correction amplitude and the time correction deviation;

[0021] Apply the calculated path deviation optimized control amount to the actual obstacle avoidance execution.

[0022] As a further limitation of the technical solution of the embodiment of the present invention, the optimized calculation formula for the path deviation control amount is: , where is the path deviation optimized control amount, is the initial path deviation control amount, is the path deviation correction amplitude, is the weight value of the path deviation correction amplitude, is the time correction deviation, is the weight value of the time correction deviation;

[0023] In the optimized calculation formula for the path deviation control amount, , where n is the total number of local video segments, is the change in the expansion area of the historical obstacle in the i-th period of time, is the time used for expansion;

[0024] , where is the average removal duration of the historical obstacle, is the preset standard removal duration.

[0025] An intelligent logistics park transportation vehicle management system, the system includes: a target obstacle determination module, a data acquisition module, a path deviation correction amplitude determination module, a time correction deviation determination module, and a path deviation control amount adjustment module, where:

[0026] The target obstacle determination module is used to determine whether the obstacle is related to the artificially induced dynamic behavior when the target transport vehicle encounters an obstacle on the preset transport route. If it is determined to be related, it is set as the target obstacle;

[0027] A data acquisition module, configured to identify the human-controlled behavior feature category of the target obstacle and the specified area where it is located, and acquire the historical dynamic vision data set of the specified area, the dynamic obstacle event change log, and the current initial path deviation control amount of the target transport vehicle;

[0028] A path deviation correction amplitude determination module, configured to analyze the historical dynamic vision data set, extract the target behavior vision data set with the same human-controlled behavior feature category as the target obstacle, intelligently analyze the dynamic change trend of the spatial occupancy of the target obstacle based on the target behavior vision data set, and quantify its path deviation correction amplitude;

[0029] A time correction deviation determination module, configured to analyze the dynamic obstacle event change log, calculate the average removal duration of the same type of historical obstacles after receiving the removal instruction, and compare it with the preset standard removal duration to determine the time correction deviation;

[0030] A path deviation control amount adjustment module, configured to optimize and adjust the initial path deviation control amount based on the path deviation correction amplitude and the time correction deviation, generate a path deviation optimized control amount, and apply it to the actual obstacle avoidance execution.

[0031] As a further limitation of the technical solution of the embodiment of the present invention, the path deviation correction amplitude determination module specifically includes:

[0032] A video screening unit, configured to analyze the historical dynamic vision data set of the specified area where the target obstacle is located, and use a target detection and classification algorithm to screen out local video segments that conform to the human-controlled behavior feature category of the target obstacle, and generate a target behavior vision data set after filtering out irrelevant content;

[0033] A polyline generation unit, configured to parse the spatial occupancy area of the historical obstacle frame by frame on the target behavior vision data set, use vision segmentation and trajectory tracking technologies to extract the dynamic expansion boundary and movement trajectory of the historical obstacle, calculate the expansion degree of the historical obstacle in each local video segment, and generate a time series polyline, and combine denoising processing to remove abnormal data points to obtain a smooth trend of expansion change;

[0034] An average slope calculation unit, configured to calculate the average slope of the expansion trend based on the expansion change polyline, so as to quantify the dynamic expansion behavior of the historical obstacle, and use it as the path deviation correction amplitude.

[0035] As a further limitation of the technical solution of the embodiment of the present invention, the time correction deviation determination module specifically includes:

[0036] A historical record screening unit, configured to extract historical records in the dynamic obstacle event change log that are of the same category as the target obstacle manual control behavior characteristics, including the time stamp when the historical obstacle receives a removal instruction and the time stamp when the actual removal is completed, calculate the removal duration of each historical record, and form a removal duration data set;

[0037] An average removal duration determination unit, configured to perform statistical analysis on the removal duration data set, eliminate outliers or extreme data points through denoising processing, calculate the average value of the removal duration, and obtain the average removal duration;

[0038] A time correction deviation calculation unit, configured to compare the average removal duration with a preset standard removal duration and calculate the time correction deviation.

[0039] As a further limitation of the technical solution of the embodiment of the present invention, the path deviation control amount adjustment module specifically includes:

[0040] A control amount adjustment unit, configured to retrieve an optimized calculation formula for the path deviation control amount, and optimize and adjust the initial path deviation control amount in combination with the path deviation correction amplitude and the time correction deviation;

[0041] An optimized control amount application unit, configured to apply the calculated path deviation optimized control amount to the actual obstacle avoidance execution.

[0042] As a further limitation of the technical solution of the embodiment of the present invention, the optimized calculation formula for the path deviation control amount is: , where is the path deviation optimized control amount, is the initial path deviation control amount, is the path deviation correction amplitude, is the weight value of the path deviation correction amplitude, is the time correction deviation, is the weight value of the time correction deviation;

[0043] In the optimized calculation formula for the path deviation control amount, , where n is the total number of local video segments, is the change in the expansion area of the historical obstacle in the i-th period of time, is the time used for expansion;

[0044] , where is the average removal duration of the historical obstacle, is the preset standard removal duration.

[0045] Compared with the prior art, the present invention has the following beneficial effects:

[0046] The present invention optimizes the path deviation control amount of a transport vehicle by combining the dual characteristics of the spatial dynamic behavior of a target obstacle and the removal response time. Through the comprehensive adjustment of the path deviation correction amplitude and the time correction deviation, multi-dimensional optimization of obstacle avoidance path planning is achieved, which can effectively improve the accuracy and adaptability of dynamic obstacle avoidance. Compared with the single-dimensional obstacle avoidance adjustment scheme, the present invention ensures both the safety and efficiency of the transport task in the complex logistics park scenario, and achieves a better balance between energy consumption and path adjustment amplitude, with high practical value and reliability. Description of the Drawings

[0047] Figure 1 is a flowchart of the method provided by an embodiment of the present invention;

[0048] Figure 2 is a flowchart of analyzing the dynamic change trend of the spatial occupancy of a target obstacle and quantifying its path deviation correction amplitude in the method provided by an embodiment of the present invention;

[0049] Figure 3 is a flowchart of determining the time correction deviation based on the dynamic obstacle event change log in the method provided by an embodiment of the present invention;

[0050] Figure 4 is a flowchart of optimizing and adjusting the initial path deviation control amount in the method provided by an embodiment of the present invention;

[0051] Figure 5 is an application architecture diagram of the system provided by an embodiment of the present invention;

[0052] Figure 6 is a structural block diagram of the path deviation correction amplitude determination module in the system provided by an embodiment of the present invention;

[0053] Figure 7 is a structural block diagram of the time correction deviation determination module in the system provided by an embodiment of the present invention;

[0054] Figure 8 is a structural block diagram of the path deviation control amount adjustment module in the system provided by an embodiment of the present invention. Detailed Embodiments

[0055] In order to make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0056] Figure 1 shows a flowchart of the method provided by an embodiment of the present invention.

[0057] Specifically, a method for managing transport vehicles in an intelligent logistics park, the method specifically includes the following steps:

[0058] Step S100, when a target transport vehicle encounters an obstacle on a preset transport route, determine whether the obstacle is related to an artificially induced dynamic behavior. If it is determined to be related, set it as the target obstacle.

[0059] Step S200, identify the human-controlled behavior feature category of the target obstacle and the designated area where it is located, and obtain the historical dynamic visual data set of the designated area, the dynamic obstacle event change log, and the current initial path deviation control amount of the target transport vehicle.

[0060] In the embodiments of the present invention, the target transport vehicle in the present invention refers to a vehicle applied to specific transport tasks, mainly including autonomous driving vehicles, semi-autonomous driving vehicles, or other transport tools with path planning and environmental perception functions. These vehicles are widely used in scenarios such as intelligent logistics parks, logistics distribution, port transportation, and warehouse management. Especially in an intelligent logistics park, the target transport vehicle undertakes tasks such as in-park cargo transfer, container handling, and material allocation, and relies on a preset route to execute tasks, with the ability of autonomous obstacle avoidance and path adjustment. The preset transport route is a fixed path planned by the system according to transport requirements and task scenarios, usually designed based on real-time map data, traffic rules, environmental conditions, and the flow direction of goods in the park, and is used to guide the vehicle to efficiently complete the transport task according to the optimal plan.

[0061] When a target transport vehicle encounters an obstacle on a preset transport route, the behavior characteristics of the obstacle are captured through multi-modal sensors (such as lidar, cameras, ultrasonic radars, etc.), including the movement trajectory, morphological changes, and behavior patterns. Combining artificial intelligence algorithms, the dynamic behavior of the obstacle is analyzed to determine whether it belongs to an artificially induced behavior (such as manually operating an object into the route, deliberately blocking the vehicle's forward path, etc.). If the algorithm detects that the behavior characteristics of the obstacle are related to an artificial dynamic behavior, set it as the target obstacle for subsequent processing.

[0062] In order to identify the human-controlled behavior feature category of the target obstacle and the designated area where it is located, the system uses multi-modal fusion technology to process the dynamic behavior data of the obstacle and performs matching analysis in combination with the historical data of the designated area. The historical dynamic visual data set of the designated area is usually collected when the vehicle is deployed and contains records of the obstacle behavior patterns in the area. The dynamic obstacle event change log records the obstacle position, behavior changes, and related time information, and extracts the specific behavior characteristics of the target obstacle by comparing with the current event.

[0063] The initial path deviation control amount of the target transport vehicle is calculated based on the deviation between the vehicle's real-time position data and the preset route. This belongs to the path tracking and deviation control method widely used in the prior art, and its core technology relies on high-precision positioning and environmental perception systems. The current position of the vehicle is obtained in real time through high-precision positioning technologies (such as RTK GPS, IMU inertial navigation system), and compared with the geometric model of the preset route to calculate the path deviation amount of the vehicle. In addition, by fusing various sensor data (such as wheel encoders, steering wheel angle sensors, lidar, cameras, etc.), the positioning accuracy and path correction ability of the vehicle are further improved.

[0064] In the specific implementation process, RTK GPS provides centimeter-level positioning accuracy, and the IMU inertial navigation system provides attitude change information. The two are combined to achieve highly robust real-time position measurement; the wheel encoder monitors the moving distance and direction of the vehicle, and the steering wheel angle sensor records the steering situation of the vehicle. Based on the vehicle's dynamic model, these data are filtered and fused to generate high-precision deviation measurement values. At the same time, using vehicle control algorithms (such as PID control or MPC model predictive control), the system adjusts the steering angle or speed according to the deviation amount to correct the path deviation in real time and ensure that the vehicle always travels along the preset route.

[0065] Furthermore, the intelligent logistics park transport vehicle management method further includes the following steps:

[0066] Step S300, parse the historical dynamic vision dataset, extract the target behavior vision dataset with the same human control behavior feature category as the target obstacle, intelligently analyze the dynamic change trend of the space occupancy of the target obstacle based on the target behavior vision dataset, and quantify its path deviation correction amplitude.

[0067] Specifically, Figure 2 The flowchart shows the analysis of the dynamic change trend of the space occupancy of the target obstacle and the quantification of its path deviation correction amplitude.

[0068] Among them, parsing the historical dynamic vision dataset, extracting the target behavior vision dataset with the same human control behavior feature category as the target obstacle, intelligently analyzing the dynamic change trend of the space occupancy of the target obstacle based on the target behavior vision dataset, and quantifying its path deviation correction amplitude specifically include the following steps:

[0069] Step S301, parse the historical dynamic vision dataset of the specified area where the target obstacle is located, use the target detection and classification algorithm to screen out the local video segments that meet the human control behavior feature category of the target obstacle, and generate the target behavior vision dataset after filtering out irrelevant content;

[0070] Step S302: On the target behavior visual dataset, parse the spatial occupancy area of historical obstacles frame by frame. Utilize visual segmentation and trajectory tracking technologies to extract the dynamic expansion boundary and movement trajectory of historical obstacles, calculate the expansion degree of historical obstacles in each local video segment, generate a time-series broken line, and combine denoising processing to eliminate abnormal data points to obtain a smooth trend of expansion changes.

[0071] Step S303: Based on the expansion change broken line, calculate the average slope of the expansion trend to quantify the dynamic expansion behavior of historical obstacles and use it as the path deviation correction amplitude.

[0072] In the embodiments of the present invention, to implement the parsing of the historical dynamic visual dataset of the specified area where the target obstacle is located, it is necessary to rely on deep learning and computer vision technologies. First, the historical dynamic visual dataset is collected by the monitoring equipment in the target area or the camera system of the vehicle itself. Through object detection and classification algorithms (such as YOLO, FasterR-CNN and other models based on deep learning), the segments with specific human-controlled behavior characteristics are screened out from the video data. These characteristics are defined through pre-trained models or specifically labeled datasets to ensure that the selected local video segments conform to the behavior categories of the target obstacles. During the screening process, the system will filter out irrelevant static backgrounds or other irrelevant dynamic objects to improve the accuracy and pertinence of the target behavior visual dataset.

[0073] In the parsing of the target behavior visual dataset, frame-by-frame analysis is a key step. Through visual segmentation technologies (such as Mask R-CNN, DeepLab, etc.), the obstacle areas in each frame of the picture are segmented to clarify their spatial occupancy ranges. At the same time, combined with trajectory tracking algorithms (such as Kalman filtering, SORT or DeepSORT), the movement trajectories and dynamic expansion boundaries of obstacles in time series are extracted. The characteristics of the dynamic expansion behavior are obtained by calculating the change in the expansion area of the obstacle in each frame and the time taken for the expansion. These data are organized in the form of a time series to generate an expansion change broken line, showing the change trend of the expansion degree of the obstacle over time. To eliminate noise interference, the system will apply denoising processing (such as moving average filtering or time series smoothing technology) to eliminate abnormal data points to obtain a smoother expansion trend.

[0074] Based on the expansion change broken line, calculate the average slope of the expansion trend to quantify the dynamic expansion behavior of historical obstacles. This slope value can reflect the overall trend of the obstacle expansion speed and is directly related to the vehicle path correction requirements. The ratio of the increase in the expansion area to the time taken for the expansion, that is, the expansion speed, is the core parameter for calculating the slope.

[0075] The reason for analyzing the dynamic trend of the spatial occupancy of the target obstacle and quantifying the path deviation correction amplitude is that it can intuitively reflect the degree of continuous interference of the obstacle on the path. By quantifying the expansion behavior, the system can predict in advance the impact that the obstacle may have on the path, and adjust the path deviation correction strategy of the target transport vehicle accordingly. The advantage of this method is its real-time and dynamic nature, avoiding the limitation of traditional static obstacle handling solutions that may ignore dynamic expansion behavior. In addition, this method makes full use of the contextual information provided by historical data, and can make more accurate predictions on the changing trends of future scenarios, thereby improving the accuracy of path correction and transportation efficiency.

[0076] Furthermore, the intelligent logistics park transport vehicle management method further comprises the following steps:

[0077] Step S400 , analyzing the dynamic obstacle event change log, calculating the average removal time of the same type of historical obstacles after receiving the removal instruction, and comparing it with the preset standard removal time to determine the time correction deviation.

[0078] Specifically, Figure 3 A flow chart for determining a time correction deviation based on a dynamic obstacle event change log is shown.

[0079] The following steps are used to analyze the dynamic obstacle event change log, calculate the average removal time of the same type of historical obstacles after receiving the removal instruction, and compare it with the preset standard removal time to determine the time correction deviation:

[0080] Step S401, extracting historical records of the same category as the human control behavior feature category of the target obstacle in the dynamic obstacle event change log, including the timestamp of the historical obstacle receiving the removal instruction and the timestamp of the actual removal completion, calculating the removal duration of each historical record, and forming a removal duration data set;

[0081] Step S402, statistically analyzing the removal duration data set, removing outliers or extreme data points through denoising, calculating the average removal duration, and obtaining the average removal duration;

[0082] Step S403, comparing the average removal time with the preset standard removal time, and calculating the time correction deviation.

[0083] In the embodiments of the present invention, to extract relevant historical records from the change log of dynamic obstacle events, database retrieval and event log parsing technologies are required. The system filters the change log of dynamic obstacle events to locate records that match the human-controlled behavior feature categories of the target obstacle. The matching criteria are based on predefined behavior category tags or descriptive information recorded in the log. For qualified records, the timestamp of receiving the removal instruction and the actual removal completion timestamp of each historical obstacle are extracted. The single removal duration is calculated through the time difference between the two, forming a removal duration data set.

[0084] After the removal duration data set is generated, statistical analysis is performed to ensure the reliability and representativeness of the data. First, through data cleaning and denoising processes (such as the box plot method based on statistics or the outlier detection method based on cluster analysis), outliers or extreme data points are removed. These outliers may result from recording errors, system failures, or other uncontrollable factors. Subsequently, the average value of the removal duration data set is calculated to obtain the average removal duration of the target category.

[0085] By comparing the average removal duration with the preset standard removal duration, the time correction deviation can be calculated through a simple time difference formula. The preset standard removal duration usually comes from prior experience values, laboratory simulation test results, or industry standards. For example, for the removal operation of typical obstacles in a specific scenario (such as an intelligent logistics park or an industrial transportation environment), the system determines a benchmark duration through a large number of tests and sets it as the standard value.

[0086] Adopting the method of determining the time correction deviation based on the change log of dynamic obstacle events can make full use of the statistical value of historical data and accurately reflect the actual time requirements of different obstacle removal operations. The advantage of this method lies in its data-driven characteristics. It can not only capture the time consumption rules in actual operation but also dynamically adjust the deviation correction scheme, improving the adaptability and decision-making accuracy of the system. In addition, by comparing with the standard value, it can help identify abnormal obstacle behaviors or improve existing removal strategies, thereby enhancing the overall operation efficiency and safety of the target transport vehicle.

[0087] Furthermore, the method for managing transport vehicles in the intelligent logistics park further includes the following steps:

[0088] Step S500, based on the path deviation correction amplitude and the time correction deviation, optimize and adjust the initial path deviation control amount to generate a path deviation optimization control amount, and apply it to the actual obstacle avoidance execution.

[0089] Specifically, Figure 4 shows a flowchart for optimizing and adjusting the initial path deviation control amount.

[0090] Among them, based on the path deviation correction amplitude and the time correction deviation, the initial path deviation control amount is optimized and adjusted to generate an optimized path deviation control amount, and its application to the actual obstacle avoidance execution specifically includes the following steps:

[0091] Step S501, retrieve the optimized calculation formula for the path deviation control amount, and optimize and adjust the initial path deviation control amount in combination with the path deviation correction amplitude and the time correction deviation;

[0092] Step S502, apply the calculated optimized path deviation control amount to the actual obstacle avoidance execution.

[0093] The optimized calculation formula for the path deviation control amount is: , where is the optimized path deviation control amount, is the initial path deviation control amount, is the path deviation correction amplitude, is the weight value of the path deviation correction amplitude, is the time correction deviation, is the weight value of the time correction deviation;

[0094] In the optimized calculation formula for the path deviation control amount, , where n is the total number of local video segments, is the change in the expansion area of the historical obstacle in the i-th period of time, is the time used for expansion;

[0095] , where is the average removal duration of the historical obstacle, is the preset standard removal duration.

[0096] In the embodiments of the present invention, by combining the path deviation correction amplitude and the time correction deviation, the process of optimizing the path deviation control amount reflects the advantages of collaborative optimization of multi-dimensional factors. The path deviation correction amplitude evaluates the dynamic expansion behavior of the obstacle and its direct impact on the path from the spatial dimension, while the time correction deviation analyzes the impact of the delay in the obstacle removal response on the obstacle avoidance requirement from the time dimension. When the two correction factors act simultaneously, it is not only possible to dynamically adjust the obstacle avoidance path spatially, but also to anticipate the continuous impact of the obstacle temporally, providing more accurate guidance for path planning.

[0097] This comprehensive correction avoids the limitations of a single dimension. For example, relying solely on the path deviation correction amplitude may lead to excessive obstacle avoidance when obstacles are removed quickly, while simply relying on the time correction deviation may ignore the spatial risks of dynamic expansion. Through the synergistic effect of both, the system can optimize the safety and efficiency of the obstacle avoidance path simultaneously, ensuring the smooth execution of tasks in complex scenarios, effectively balancing energy consumption and resource use, and achieving overall performance improvement.

[0098] Furthermore, Figure 5 The application architecture diagram of the system provided by the embodiment of the present invention is shown.

[0099] Among them, in another preferred embodiment provided by the present invention, an intelligent logistics park transportation vehicle management system includes:

[0100] A target obstacle determination module 100, configured to, when a target transportation vehicle encounters an obstacle on a preset transportation route, determine whether the obstacle is related to an artificially induced dynamic behavior. If it is determined to be related, set it as a target obstacle;

[0101] A data acquisition module 200, configured to identify the human control behavior feature category of the target obstacle and the designated area where it is located, obtain the historical dynamic visual data set, dynamic obstacle event change log of the designated area, and the current initial path deviation control amount of the target transportation vehicle.

[0102] In the embodiment of the present invention, the target transportation vehicle in the present invention refers to a vehicle applied to specific transportation tasks, mainly including autonomous driving vehicles, semi-autonomous driving vehicles, or other transportation tools with path planning and environmental perception functions. These vehicles are widely used in scenarios such as intelligent logistics parks, logistics distribution, port transportation, and warehouse management. Especially in intelligent logistics parks, the target transportation vehicle undertakes tasks such as in-park cargo transfer, container handling, and material allocation, relies on the preset route to execute tasks, and has the ability of autonomous obstacle avoidance and path adjustment. The preset transportation route is a fixed path planned by the system according to transportation needs and task scenarios, usually designed based on real-time map data, traffic rules, environmental conditions, and cargo flow directions in the park, and is used to guide the vehicle to efficiently complete transportation tasks according to the optimal plan.

[0103] When a target transportation vehicle encounters an obstacle on a preset transportation route, the behavior characteristics of the obstacle, including the movement trajectory, morphological changes, and behavior patterns, are captured through multi-modal sensors (such as lidar, cameras, ultrasonic radars, etc.). Combining artificial intelligence algorithms, the dynamic behavior of the obstacle is analyzed to determine whether it belongs to an artificially induced behavior (such as manually operating an object into the route, deliberately blocking the vehicle's forward path, etc.). If the algorithm detects that the behavior characteristics of the obstacle are related to human dynamic behavior, it is set as a target obstacle for subsequent processing.

[0104] To identify the human-controlled behavior feature categories of the target obstacle and the designated area where it is located, the system uses multi-modal fusion technology to process the dynamic behavior data of the obstacle and performs matching analysis in combination with the historical data of the designated area. The historical dynamic vision dataset of the designated area is usually collected during vehicle deployment and contains records of the behavior patterns of obstacles in this area. The dynamic obstacle event change log records the obstacle position, behavior changes, and related time information. By comparing with the current event, specific behavior features of the target obstacle are extracted.

[0105] The initial path deviation control amount of the target transport vehicle is calculated based on the deviation between the vehicle's real-time position data and the preset route. This belongs to the path tracking and deviation control method widely used in the prior art, and its core technology relies on high-precision positioning and environmental perception systems. The current position of the vehicle is obtained in real time through high-precision positioning technologies (such as RTK GPS, IMU inertial navigation system) and compared with the geometric model of the preset route to calculate the path deviation amount of the vehicle. In addition, by fusing various sensor data (such as wheel encoders, steering wheel angle sensors, lidar, cameras, etc.), the positioning accuracy and path correction ability of the vehicle are further improved.

[0106] In the specific implementation process, RTK GPS provides centimeter-level positioning accuracy, and the IMU inertial navigation system provides attitude change information. The two are combined to achieve highly robust real-time position measurement; the wheel encoder monitors the moving distance and direction of the vehicle, and the steering wheel angle sensor records the steering condition of the vehicle. Based on the vehicle's dynamic model, these data are filtered and fused to generate high-precision deviation measurement values. At the same time, using vehicle control algorithms (such as PID control or MPC model predictive control), the system adjusts the steering angle or speed according to the deviation amount to correct the path deviation in real time and ensure that the vehicle always travels along the preset route.

[0107] Furthermore, the intelligent logistics park transport vehicle management system further includes:

[0108] A path deviation correction amplitude determination module 300, which is used to analyze the historical dynamic vision dataset, extract the target behavior vision dataset with the same human-controlled behavior feature category as the target obstacle, intelligently analyze the dynamic change trend of the spatial occupancy of the target obstacle based on the target behavior vision dataset, and quantify its path deviation correction amplitude.

[0109] Specifically, Figure 6 Fig. shows the structural block diagram of the path deviation correction amplitude determination module 300 in the system provided by the embodiment of the present invention.

[0110] Among them, in the preferred implementation manner provided by the present invention, the path deviation correction amplitude determination module 300 specifically includes:

[0111] The video screening unit 301 is configured to parse the historical dynamic visual data set of the specified area where the target obstacle is located, and use target detection and classification algorithms to screen out local video segments that conform to the human-controlled behavior feature categories of the target obstacle, and generate a target behavior visual data set after filtering out irrelevant content;

[0112] The polyline generation unit 302 is configured to parse the spatial occupancy area of the historical obstacle frame by frame on the target behavior visual data set, use visual segmentation and trajectory tracking technologies to extract the dynamic expansion boundary and movement trajectory of the historical obstacle, calculate the expansion degree of the historical obstacle in each local video segment, and generate a time series polyline, and combine denoising processing to eliminate abnormal data points to obtain a smooth trend of expansion change;

[0113] The average slope calculation unit 303 is configured to calculate the average slope of the expansion trend based on the expansion change polyline, so as to quantify the dynamic expansion behavior of the historical obstacle, and use it as the path deviation correction amplitude.

[0114] In the embodiments of the present invention, to implement parsing the historical dynamic visual data set of the specified area where the target obstacle is located, deep learning and computer vision technologies are relied on. First, the historical dynamic visual data set is collected by the monitoring equipment in the target area or the camera system of the vehicle itself. Through target detection and classification algorithms (such as YOLO, FasterR-CNN and other models based on deep learning), segments with specific human-controlled behavior characteristics are screened out from the video data. These characteristics are defined through pre-trained models or specifically labeled data sets to ensure that the selected local video segments conform to the behavior categories of the target obstacle. During the screening process, the system will filter out irrelevant static backgrounds or other irrelevant dynamic objects to improve the accuracy and pertinence of the target behavior visual data set.

[0115] In the parsing of the target behavior visual data set, frame-by-frame analysis is a key step. Through visual segmentation technologies (such as Mask R-CNN, DeepLab, etc.), the obstacle area in each frame of the picture is segmented to clarify its spatial occupancy range. At the same time, combined with trajectory tracking algorithms (such as Kalman filtering, SORT or DeepSORT), the movement trajectory and dynamic expansion boundary of the obstacle in time series are extracted. The characteristics of the dynamic expansion behavior are obtained by calculating the change in the expansion area of the obstacle in each frame and the time used for expansion. These data are sorted in the form of a time series to generate an expansion change polyline, showing the change trend of the obstacle expansion degree over time. To eliminate noise interference, the system will apply denoising processing (such as moving average filtering or time series smoothing technology) to eliminate abnormal data points and obtain a smoother expansion trend.

[0116] Based on the expansion change curve, the average slope of the expansion trend is calculated to quantify the dynamic expansion behavior of historical obstacles. This slope value can reflect the overall trend of the obstacle expansion speed, which is directly related to the need for vehicle path correction. The ratio of the increase in the expansion area to the time taken for expansion, that is, the expansion speed, is the core parameter for calculating the slope.

[0117] The reason for analyzing the dynamic trend of the spatial occupancy of the target obstacle and quantifying the path deviation correction amplitude is that it can intuitively reflect the degree of continuous interference of the obstacle on the path. By quantifying the expansion behavior, the system can predict in advance the impact that the obstacle may have on the path, and adjust the path deviation correction strategy of the target transport vehicle accordingly. The advantage of this method is its real-time and dynamic nature, avoiding the limitation of traditional static obstacle handling solutions that may ignore dynamic expansion behavior. In addition, this method makes full use of the contextual information provided by historical data, and can make more accurate predictions on the changing trends of future scenarios, thereby improving the accuracy of path correction and transportation efficiency.

[0118] Furthermore, the intelligent logistics park transport vehicle management system also includes:

[0119] The time correction deviation determination module 400 is used to parse the dynamic obstacle event change log, calculate the average removal time of the same type of historical obstacles after receiving the removal instruction, and compare it with the preset standard removal time to determine the time correction deviation.

[0120] Specifically, Figure 7 It shows a structural block diagram of the time correction deviation determination module 400 in the system provided by an embodiment of the present invention.

[0121] Among them, in the preferred implementation manner provided by the present invention, the time correction deviation determination module 400 specifically includes:

[0122] The history record screening unit 401 is used to extract the history records of the same category as the human control behavior feature category of the target obstacle in the dynamic obstacle event change log, including the timestamp of the historical obstacle receiving the removal instruction and the timestamp of the actual removal completion, calculate the removal time of each history record, and form a removal time data set;

[0123] The average removal time determination unit 402 is used to perform statistical analysis on the removal time data set, remove outliers or extreme data points through denoising, calculate the average removal time, and obtain the average removal time;

[0124] The time correction deviation calculation unit 403 is used to compare the average removal time with the preset standard removal time to calculate the time correction deviation.

[0125] In the embodiments of the present invention, to extract relevant historical records from the change log of dynamic obstacle events, database retrieval and event log parsing technologies are required. The system screens the change log of dynamic obstacle events to locate records that match the target obstacle manual control behavior feature categories. The matching criteria are based on predefined behavior category tags or description information recorded in the log. For eligible records, the receipt removal instruction timestamp and the actual removal completion timestamp of each historical obstacle are extracted. The single removal duration is calculated through the time difference between the two, forming a removal duration data set.

[0126] After the removal duration data set is generated, statistical analysis is performed to ensure the reliability and representativeness of the data. First, through data cleaning and denoising processes (such as the box plot method based on statistics or the outlier detection method based on cluster analysis), outliers or extreme data points are removed. These outliers may result from recording errors, system failures, or other uncontrollable factors. Subsequently, the average value of the removal duration data set is calculated to obtain the average removal duration of the target category.

[0127] By comparing the average removal duration with the preset standard removal duration, the time correction deviation can be calculated through a simple time difference formula. The preset standard removal duration usually comes from prior experience values, laboratory simulation test results, or industry standards. For example, for the removal operation of typical obstacles in a specific scenario (such as an intelligent logistics park or an industrial transportation environment), the system determines a benchmark duration through a large number of tests and sets it as the standard value.

[0128] Adopting the method of determining the time correction deviation based on the change log of dynamic obstacle events can fully utilize the statistical value of historical data and accurately reflect the actual time requirements of different obstacle removal operations. The advantage of this method lies in its data-driven characteristics. It can not only capture the time consumption rules in actual operation but also dynamically adjust the deviation correction scheme to improve the adaptability and decision-making accuracy of the system. In addition, by comparing with the standard value, it can help identify abnormal obstacle behaviors or improve existing removal strategies, thereby enhancing the overall operation efficiency and safety of the target transport vehicle.

[0129] Furthermore, the intelligent logistics park transport vehicle management system further includes:

[0130] A path deviation control amount adjustment module 500, which is used to optimize and adjust the initial path deviation control amount based on the path deviation correction amplitude and the time correction deviation, generate a path deviation optimized control amount, and apply it to the actual obstacle avoidance execution.

[0131] Specifically, Figure 8 The structural block diagram of the path deviation control amount adjustment module 500 in the system provided by the embodiments of the present invention is shown.

[0132] Among them, in the preferred embodiment provided by the present invention, the path deviation control amount adjustment module 500 specifically includes:

[0133] A control amount adjustment unit 501, configured to retrieve an optimized calculation formula for the path deviation control amount, and optimize and adjust the initial path deviation control amount in combination with the path deviation correction amplitude and the time correction deviation;

[0134] An optimized control amount application unit 502, configured to apply the calculated optimized path deviation control amount to the actual obstacle avoidance execution.

[0135] The optimized calculation formula for the path deviation control amount is: , where is the optimized path deviation control amount, is the initial path deviation control amount, is the path deviation correction amplitude, is the weight value of the path deviation correction amplitude, is the time correction deviation, is the weight value of the time correction deviation;

[0136] In the optimized calculation formula for the path deviation control amount, , where n is the total number of local video segments, is the change in the expansion area of the historical obstacle within the i-th time period, is the time used for expansion;

[0137] , where is the average removal duration of the historical obstacle, is the preset standard removal duration.

[0138] In the embodiments of the present invention, the process of optimizing the path deviation control amount by combining the path deviation correction amplitude and the time correction deviation reflects the advantage of collaborative optimization of multi-dimensional factors. The path deviation correction amplitude evaluates the dynamic expansion behavior of the obstacle and its direct impact on the path from the spatial dimension, while the time correction deviation analyzes the impact of the delay in the obstacle removal response on the obstacle avoidance requirement from the time dimension. When the two correction factors act simultaneously, it can not only dynamically adjust the obstacle avoidance path spatially, but also anticipate the continuous impact of the obstacle temporally, providing more accurate guidance for path planning.

[0139] This comprehensive correction avoids the limitations of a single dimension. For example, relying solely on the path deviation correction amplitude may lead to excessive obstacle avoidance when obstacles are removed quickly, while simply depending on the time correction deviation may ignore the spatial risks of dynamic expansion. Through the synergistic effect of the two, the system can optimize the safety and efficiency of the obstacle avoidance path simultaneously, ensuring the smooth execution of tasks in complex scenarios, effectively balancing energy consumption and resource utilization, and achieving an overall performance improvement.

[0140] It should be understood that although the steps in the flowcharts of the embodiments of the present invention are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless there is a clear indication in this article, the execution of these steps has no strict order limit, and these steps can be executed in other orders. Moreover, at least a part of the steps in each embodiment may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed alternately or in turn with at least a part of other steps or sub-steps or stages of other steps.

[0141] Those of ordinary skill in the art can understand that all or part of the processes of implementing the methods in the above embodiments can be completed by instructing relevant hardware through a computer program. The program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, storage, database, or other medium used in the embodiments provided in the present application can include non-volatile and / or volatile memories. Non-volatile memories can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memories can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and Rambus dynamic RAM (RDRAM), etc.

[0142] The technical features of the above-described embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as falling within the scope described in this specification.

[0143] The above-described embodiments merely represent several implementation manners of the present invention. The description thereof is relatively specific and detailed, but it should not be construed as a limitation to the scope of the patent for the present invention. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can still be made, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent for the present invention shall be subject to the appended claims.

[0144] The above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent replacements, and improvements made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for managing transport vehicles in an intelligent logistics park, characterized in that: The method comprises: When the target transport vehicle encounters an obstacle on the preset transport route, it is determined whether the obstacle is related to the artificially induced dynamic behavior. If it is determined to be related, it is set as the target obstacle; Identify the human-controlled behavior feature category of the target obstacle and the designated area where it is located, obtain the historical dynamic visual data set of the designated area, the dynamic obstacle event change log, and the current initial path deviation control amount of the target transport vehicle; Analyze the historical dynamic visual data set, extract the target behavior visual data set with the same feature category as the human control behavior of the target obstacle, intelligently analyze the dynamic change trend of the space occupancy of the target obstacle based on the target behavior visual data set, and quantify the correction amplitude of its path deviation; Parse the dynamic obstacle event change log, calculate the average removal time of historical obstacles of the same type after receiving the removal instruction, and compare it with the preset standard removal time to determine the time correction deviation; Based on the path deviation correction amplitude and time correction deviation, the initial path deviation control amount is optimized and adjusted to generate the path deviation optimization control amount, which is then applied to the actual obstacle avoidance execution.

2. The intelligent logistics park transportation vehicle management method according to claim 1 is characterized in that: The steps of parsing the historical dynamic visual data set, extracting the target behavior visual data set with the same characteristic category as the human control behavior of the target obstacle, intelligently analyzing the dynamic change trend of the space occupancy of the target obstacle based on the target behavior visual data set, and quantifying the path deviation correction amplitude include: Analyze the historical dynamic visual data set of the designated area where the target obstacle is located, use the target detection and classification algorithm to filter out the local video clips that meet the human control behavior feature category of the target obstacle, and generate the target behavior visual data set after filtering out irrelevant content; In the target behavior visual dataset, the spatial occupied area of ​​historical obstacles is analyzed frame by frame. The dynamic expansion boundary and movement trajectory of historical obstacles are extracted by using visual segmentation and trajectory tracking technology. The expansion degree of historical obstacles in each local video clip is calculated, and a time series polyline is generated. The abnormal data points are removed by combining denoising processing to obtain a smooth trend of expansion changes. Based on the expansion change curve, the average slope of the expansion trend is calculated to quantify the dynamic expansion behavior of historical obstacles and use it as the path deviation correction amplitude.

3. The intelligent logistics park transportation vehicle management method according to claim 2 is characterized in that: The steps of parsing the dynamic obstacle event change log, calculating the average removal time of the same type of historical obstacles after receiving the removal instruction, and comparing it with the preset standard removal time to determine the time correction deviation include: Extract the historical records with the same category as the human control behavior feature of the target obstacle in the dynamic obstacle event change log, including the timestamp of the historical obstacle receiving the removal instruction and the timestamp of the actual removal completion, calculate the removal duration of each historical record, and form a removal duration dataset; Perform statistical analysis on the removal time dataset, remove outliers or extreme data points through denoising, calculate the average removal time, and obtain the average removal time; The average removal time is compared with the preset standard removal time, and the time correction deviation is calculated.

4. The intelligent logistics park transportation vehicle management method according to claim 3 is characterized in that: The steps of optimizing and adjusting the initial path deviation control amount based on the path deviation correction amplitude and the time correction deviation, generating the path deviation optimization control amount, and applying it to the actual obstacle avoidance execution include: Retrieve the optimization calculation formula of the path deviation control amount, and optimize and adjust the initial path deviation control amount by combining the path deviation correction amplitude and the time correction deviation; The calculated path deviation optimization control amount is applied to the actual obstacle avoidance execution.

5. The method for managing transportation vehicles in an intelligent logistics park according to claim 4, characterized in that: The optimization calculation formula of the path deviation control amount is: ,in is the path deviation optimization control quantity, is the initial path deviation control amount, is the path deviation correction amplitude, is the weight value of the path deviation correction amplitude, Correct the deviation for time, The weight value for time correction deviation; In the optimization calculation formula of path deviation control amount, , where n is the total number of local video clips, is the expansion area change of the historical obstacle in the i-th period, The time taken for expansion; ,in is the average time of removing historical obstacles, Remove duration for preset criteria.

6. An intelligent logistics park transport vehicle management system, characterized in that: The system comprises: a target obstacle determination module, a data acquisition module, a path deviation correction amplitude determination module, a time correction deviation determination module and a path deviation control amount adjustment module, wherein: A target obstacle determination module is used to determine whether the obstacle is related to the artificially induced dynamic behavior when the target transport vehicle encounters an obstacle on the preset transport route. If it is determined to be related, it is set as a target obstacle; The data acquisition module is used to identify the human control behavior feature category of the target obstacle and the designated area where it is located, obtain the historical dynamic visual data set of the designated area, the dynamic obstacle event change log, and the current initial path deviation control amount of the target transport vehicle; The path deviation correction amplitude determination module is used to parse the historical dynamic visual data set, extract the target behavior visual data set with the same feature category as the human control behavior of the target obstacle, intelligently analyze the dynamic change trend of the spatial occupancy of the target obstacle based on the target behavior visual data set, and quantify its path deviation correction amplitude; The time correction deviation determination module is used to parse the dynamic obstacle event change log, calculate the average removal time of the same type of historical obstacles after receiving the removal instruction, and compare it with the preset standard removal time to determine the time correction deviation; The path deviation control amount adjustment module is used to optimize and adjust the initial path deviation control amount based on the path deviation correction amplitude and time correction deviation, generate the path deviation optimization control amount, and apply it to the actual obstacle avoidance execution.

7. The intelligent logistics park transportation vehicle management system according to claim 6 is characterized in that: The path deviation correction amplitude determination module specifically includes: The video screening unit is used to analyze the historical dynamic visual data set of the designated area where the target obstacle is located, and use the target detection and classification algorithm to screen out the local video clips that meet the human control behavior feature category of the target obstacle, and generate the target behavior visual data set after filtering out irrelevant content; The polyline generation unit is used to analyze the spatial occupied area of ​​historical obstacles frame by frame on the target behavior visual data set, extract the dynamic expansion boundary and movement trajectory of historical obstacles by using visual segmentation and trajectory tracking technology, calculate the expansion degree of historical obstacles in each local video clip, and generate a time series polyline. Combined with denoising, it removes abnormal data points to obtain a smooth trend of expansion changes; The average slope calculation unit is used to calculate the average slope of the expansion trend based on the expansion change line, so as to quantify the dynamic expansion behavior of the historical obstacles and use it as the path deviation correction amplitude.

8. The intelligent logistics park transportation vehicle management system according to claim 7 is characterized in that: The time correction deviation determination module specifically includes: A history record screening unit is used to extract the history records of the same category as the human control behavior feature category of the target obstacle in the dynamic obstacle event change log, including the timestamp of the historical obstacle receiving the removal instruction and the timestamp of the actual removal completion, calculate the removal time of each history record, and form a removal time data set; The average removal time determination unit is used to perform statistical analysis on the removal time data set, remove outliers or extreme data points through denoising, calculate the average removal time, and obtain the average removal time; The time correction deviation calculation unit is used to compare the average removal time with the preset standard removal time to calculate the time correction deviation.

9. The intelligent logistics park transportation vehicle management system according to claim 8, characterized in that: The path deviation control amount adjustment module specifically includes: A control amount adjustment unit is used to retrieve a path deviation control amount optimization calculation formula, and optimize and adjust the initial path deviation control amount by combining the path deviation correction amplitude and the time correction deviation; The optimized control amount application unit is used to apply the calculated path deviation optimized control amount to the actual obstacle avoidance execution.

10. The intelligent logistics park transportation vehicle management system according to claim 9, characterized in that: The optimization calculation formula of the path deviation control amount is: ,in is the path deviation optimization control quantity, is the initial path deviation control amount, is the path deviation correction amplitude, is the weight value of the path deviation correction amplitude, Correct the deviation for time, The weight value for time correction deviation; In the optimization calculation formula of path deviation control amount, , where n is the total number of local video clips, is the expansion area change of the historical obstacle in the i-th period, The time taken for expansion; ,in is the average time of removing historical obstacles, Remove duration for preset criteria.

Images