A method and system for collision detection of moving objects based on real-time coordinates.

Abstract

This invention relates to the field of collision detection technology, specifically to a method and system for moving object collision detection based on real-time coordinate position. The method includes: acquiring various hazardous areas of an overall image and drawing a map; quantifying and analyzing various hazardous areas in the overall image using height information from the overall terrain data, marking static gray-dot areas in the drawn map, and obtaining a hazard assessment index for each road; drawing dynamic gray-dot areas in the map; obtaining an optimal trajectory; obtaining the collision time and collision warning distance based on information between the current vehicle and the vehicle in the nearest dynamic gray-dot area on the optimal trajectory; obtaining the collision factor of the optimal trajectory; and performing early warning detection of moving object collisions using the collision factor of the optimal trajectory and the static gray-dot areas in the drawn map. This invention improves the accuracy of moving object collision detection.

Description

Technical Field

[0001] This invention relates to the field of collision detection technology, and in particular to a method and system for detecting collisions of moving objects based on real-time coordinate positions. Background Technology

[0002] To address the technical challenges faced by vehicles traveling within new energy power stations (such as wind farms), including complex 3D environments, insufficient positioning accuracy, delayed map updates, low dynamic detection efficiency, and limited path planning, this paper proposes a moving object collision detection method based on real-time coordinates. This method achieves high-precision seamless positioning both indoors and outdoors by integrating GPS (Global Positioning System) and UWB (Ultra-Wideband) technologies. It constructs a dynamically updated 3D GIS (Geographic Information System) map and marks static and dynamic hazard areas. Real-time monitoring of the surrounding environment is achieved using multi-sensor data. Dynamic path planning is performed using the DWA (Dynamic Window Approach) algorithm, and accurate risk warnings are provided based on quantitative indicators such as Time to Collision (TTC). This method effectively solves the problems of poor adaptability, delayed warnings, and inflexible path planning in traditional 2D collision detection in complex 3D environments, significantly improving the safety and traffic efficiency of vehicles traveling within power stations. Summary of the Invention

[0003] This invention provides a method and system for detecting collisions of moving objects based on real-time coordinates, which can solve technical problems such as safety, positioning accuracy, path planning and dynamic detection efficiency during vehicle operation in new energy power stations.

[0004] The objective of this invention can be achieved through the following technical solutions:

[0005] The first aspect of this invention is to provide a method for collision detection of moving objects based on real-time coordinates, comprising:

[0006] Acquire overall images and terrain data of the site; based on the overall images, use a deep learning model to obtain various hazardous areas in the overall images; draw a map based on the various hazardous areas in the overall images and the overall terrain data; perform quantitative analysis on various hazardous areas in the overall images using the height information in the overall terrain data, and mark the static gray dot areas in the drawn map; obtain the hazard assessment index for each road based on the hazardous areas in the neighborhood of each road, soil stability, and rainfall.

[0007] Monitor people or vehicles within the site to obtain dynamic gray dot areas in the drawn map;

[0008] A sampling space is constructed using the vehicle's linear velocity and angular velocity at the current moment; several predicted trajectories are obtained by sampling through the sampling space, and the optimal trajectory is selected from these predicted trajectories.

[0009] The collision time is obtained based on the relative speed and distance between the current vehicle and the vehicle in the nearest dynamic gray dot area on the optimal trajectory; the collision warning distance is obtained based on the relative speed between the current vehicle and the vehicle in the nearest dynamic gray dot area on the optimal trajectory, the driver's reaction time, and the vehicle's maximum negative acceleration; the comprehensive risk reduction rate is obtained by quantitatively analyzing the vehicle's risk protection; the collision factor of the optimal trajectory is obtained based on the collision time, the collision warning distance, the comprehensive risk reduction rate, and the hazard assessment index; and the collision factor of the optimal trajectory and the static gray dot area in the map are used for early warning detection of moving object collisions.

[0010] Further, the process includes acquiring an overall image and overall terrain data of the site; obtaining various hazardous areas in the overall image using a deep learning model based on the overall image; and drawing a map based on the various hazardous areas in the overall image and the overall terrain data, including:

[0011] High-altitude aerial photography using drones is used to acquire global terrain data and images of the site; then, vehicle-mounted cameras and LiDAR are used to collect detailed road information to obtain local terrain data and images.

[0012] The field area image and local images are stitched together to obtain the stitched overall image; the global terrain data and local terrain data are stitched together to obtain the stitched overall terrain data.

[0013] Mask R-CNN is trained using a large dataset, and the trained Mask R-CNN is used to identify various danger zones in the overall image.

[0014] By marking various hazardous areas in the overall image onto the overall terrain data, a map can be drawn.

[0015] Furthermore, the method involves quantitatively analyzing various hazardous areas in the overall image using height information from the overall terrain data, and marking static gray dot areas in the map; based on the hazardous areas within the neighborhood of each road, soil stability, and rainfall conditions, a hazard assessment index is obtained for each road, including:

[0016] Quantitative analysis of various hazardous areas in the overall image is performed using height information from the overall terrain data; the specific process is as follows:

[0017] The slope angle for each obstacle's danger zone is calculated, specifically expressed by the formula:

[0018]

[0019] In the formula, This represents the vertical height difference between the highest and lowest points in the danger zone of each obstacle. This represents the difference in horizontal distance between the highest and lowest points within the danger zone of each obstacle. Represents the arctangent function. Indicates the slope angle of each obstacle's danger zone;

[0020] The slope angle corresponding to the danger zone of the vehicle is calculated by the slope angle and recorded as the reference slope angle; when the slope angle of each obstacle danger zone is greater than or equal to the reference slope angle, each obstacle danger zone is recorded as a high-risk zone.

[0021] The depth of each dangerous area of ​​the deep pit is calculated, and the specific formula is as follows:

[0022]

[0023] In the formula, This represents the average height data of the preset length range of each deep pit danger zone. This represents the height data of the lowest point in the danger zone of each deep pit. This indicates the depth of each dangerous zone in the deep pit;

[0024] If the depth of each deep crater danger zone is greater than or equal to half the tire depth of the vehicle, then each deep crater danger zone is recorded as a high-risk zone.

[0025] Mark all high-risk areas as static gray dot areas on the map;

[0026] The risk assessment index is specifically expressed by the formula:

[0027]

[0028] In the formula, This represents the slope angle of the neighborhood of each road. This represents the soil stability coefficient within the neighborhood of each road. This indicates the rainfall within the neighborhood of each road. The weight representing the slope angle. The weights representing the soil stability coefficient, The weight representing the amount of rainfall, This indicates the hazard assessment index for each road. This represents the hyperbolic tangent function, used for data normalization.

[0029] Specifically, if the soil within the neighborhood of each road is loose soil, the soil stability coefficient within the neighborhood of each road is marked as 1; if the soil within the neighborhood of each road is hard rock, the soil stability coefficient within the neighborhood of each road is marked as 0.

[0030] Furthermore, the monitoring of people or vehicles within the area to obtain dynamic gray dot regions in the drawn map includes:

[0031] A monitoring system consisting of several cameras is deployed within the site. Pedestrians and vehicles are tracked and monitored using the monitoring system and target tracking algorithms. The monitored pedestrians and vehicles are combined with GPS and UWB location data and marked as dynamic gray dot areas on the map.

[0032] Furthermore, the construction of the sampling space using the vehicle's linear velocity and angular velocity at the current moment includes:

[0033] Get the linear velocity and angular velocity at the current moment; then get the maximum acceleration corresponding to the linear velocity and angular velocity respectively, and denot them as maximum linear acceleration and maximum angular acceleration;

[0034] The predicted ranges for linear velocity and angular velocity are specifically expressed as follows:

[0035]

[0036] In the formula, Represents the linear velocity at the current moment. This represents the maximum acceleration corresponding to the linear velocity. This represents the angular velocity at the current moment. This represents the maximum acceleration corresponding to the angular velocity. Indicates time interval, The sampled value represents the linear velocity. The sampled value representing the angular velocity;

[0037] The sampling space is formed by predicting the range of linear velocity and angular velocity values.

[0038] Furthermore, the step of obtaining several predicted trajectories through sampling space and selecting the optimal trajectory from the several predicted trajectories includes:

[0039] By sampling the space, several data points are obtained based on linear velocity and angular velocity, and these data points are combined into several combinations.

[0040] The specific formula for obtaining the number of combinations of linear velocity and angular velocity is as follows:

[0041]

[0042] In the formula, This represents the upper limit of linear velocity. This represents the lower limit of linear velocity. This represents the upper limit of angular velocity. This represents the lower limit of angular velocity. The sampling resolution represents the linear velocity. The sampling resolution represents the angular velocity. Indicates to Round up. Indicates to Round up. This indicates the number of combinations of linear velocity and angular velocity;

[0043] Based on the linear velocity and angular velocity in each combination, the kinematic model of the moving vehicle driven by difference is used to predict the attitude at each moment in the future time period. The corresponding predicted trajectory is obtained from the attitudes at all moments in the future time period. The attitude at each moment in the future time period is specifically expressed by the formula:

[0044]

[0045]

[0046]

[0047] In the formula, Indicates the vehicle is in The x-axis at each moment, Indicates the vehicle is in The ordinate at each moment, Indicates the vehicle is in The x-axis at each moment, Indicates the vehicle is in The ordinate at each moment, Indicates the vehicle is in The heading angle at that moment, Indicates the vehicle is in The heading angle at that moment, The sampled value represents the linear velocity. The sampled value represents the angular velocity. Indicates time interval, Represents the cosine function. Represents the sine function;

[0048] By combining each linear velocity and angular velocity, the attitude at several moments in the future is obtained, and a predicted trajectory corresponding to each combination of linear velocity and angular velocity is determined by the attitude at several moments.

[0049] The predicted trajectory with the fewest dynamic gray point areas in the predicted trajectory corresponding to each combination of linear velocity and angular velocity is selected and denoted as the optimal trajectory.

[0050] Further, the step of obtaining the collision time based on the relative speed and distance between the current vehicle and the vehicle in the nearest dynamic gray dot area on the optimal trajectory; and obtaining the collision warning distance based on the relative speed between the current vehicle and the vehicle in the nearest dynamic gray dot area on the optimal trajectory, the driver's reaction time, and the vehicle's maximum negative acceleration, includes:

[0051]

[0052] In the formula, This represents the relative speed between the current vehicle and the vehicle in the nearest dynamic gray point region on the optimal trajectory. This represents the distance between the current vehicle and the vehicle in the nearest dynamic gray dot region on the optimal trajectory. Indicates the time required for the collision;

[0053]

[0054] In the formula, This represents the relative speed between the current vehicle and the vehicle in the nearest dynamic gray dot region on the optimal trajectory. Indicates the driver's reaction time. This indicates the vehicle's maximum negative acceleration. Indicates the collision warning distance.

[0055] Furthermore, the quantitative analysis of vehicle risk protection to obtain the comprehensive risk reduction rate includes:

[0056] The first risk reduction rate is obtained based on the number of accidents occurring when the safety design was used and the number of accidents occurring when the safety design was not used; the first risk reduction rate is specifically expressed by the formula:

[0057]

[0058] In the formula, This indicates the number of incidents that occurred due to the lack of safety design features. This indicates the number of accidents that occurred due to the use of safety design features. Indicates the first risk reduction rate;

[0059] The second risk reduction rate is obtained based on the accident rate after no training or testing and the accident rate after training or testing; the second risk reduction rate is specifically expressed by the formula:

[0060]

[0061] In the formula, This indicates the accident rate after no training or testing was conducted. Indicates the accident rate after training or testing. This indicates the second risk reduction rate;

[0062] The third risk reduction rate is obtained based on the unmanaged accident rate and the managed accident rate; the third risk reduction rate is specifically expressed by the formula:

[0063]

[0064] In the formula, This indicates the rate of unmanaged accidents. Indicates the accident rate after management. Indicates the third risk reduction rate;

[0065] The fourth risk reduction rate is obtained based on the initial failure rate and the post-maintenance failure rate; the fourth risk reduction rate is specifically expressed by the formula:

[0066]

[0067] In the formula, Indicates the initial failure rate. Indicates the failure rate after maintenance. This indicates the fourth risk reduction rate;

[0068] The comprehensive risk reduction rate is obtained based on the first risk reduction rate, the second risk reduction rate, the third risk reduction rate, and the fourth risk reduction rate; the comprehensive risk reduction rate is specifically expressed by the formula:

[0069]

[0070] In the formula, Indicates the first Risk reduction rate Indicates the first The weight of the risk reduction rate Indicates the overall risk reduction rate;

[0071] Each risk reduction rate has a preset weight.

[0072] Further, the step of obtaining the collision factor of the optimal trajectory based on the collision time required, the collision warning distance, the comprehensive risk reduction rate, and the hazard assessment index; and performing early warning detection of moving object collisions using the collision factor of the optimal trajectory and the static gray dot area in the drawn map, includes:

[0073] The collision factor of the optimal trajectory is specifically expressed by the formula:

[0074]

[0075] In the formula, Indicates the time required for the collision. Indicates the collision warning distance. Indicates the overall risk reduction rate. The risk assessment index represents the optimal trajectory; for Functions used to normalize data; The collision factor represents the optimal trajectory; The weights representing the time required for a collision. The weights representing the collision warning distance, The weighting represents the overall risk reduction rate. The weights represent the risk assessment index; where each weight is a preset value.

[0076] When a vehicle encounters a static gray dot area on its optimal trajectory, a warning is triggered, i.e., the system issues an audible and visual alarm. When the collision factor of the optimal trajectory exceeds the preset collision threshold, it indicates that a collision is imminent, and the system needs to automatically send a warning signal to the vehicle and notify the safety management platform for manual intervention.

[0077] A second aspect of the present invention is to provide a moving object collision detection system based on real-time coordinate position, comprising:

[0078] Site Analysis Module: This module acquires overall images and terrain data of the site; uses a deep learning model to identify various hazardous areas within the overall image; draws a map based on these hazardous areas and the overall terrain data; performs quantitative analysis on the various hazardous areas in the overall image using height information from the terrain data; and marks static gray dot areas on the map; and obtains a hazard assessment index for each road based on its neighborhood of hazardous areas, soil stability, and rainfall.

[0079] Dynamic gray dot determination module: used to monitor people or vehicles in the area and obtain dynamic gray dot areas in the drawn map;

[0080] The trajectory analysis module is used to construct a sampling space using the vehicle's linear velocity and angular velocity at the current moment; it then samples the vehicle through this sampling space to obtain several predicted trajectories and selects the optimal trajectory from among them.

[0081] The collision detection and warning module is used to: obtain the collision time required based on the relative speed and distance between the current vehicle and the vehicle in the nearest dynamic gray dot area on the optimal trajectory; obtain the collision warning distance based on the relative speed between the current vehicle and the vehicle in the nearest dynamic gray dot area on the optimal trajectory, the driver's reaction time, and the vehicle's maximum negative acceleration; obtain the comprehensive risk reduction rate by quantitatively analyzing the vehicle's risk protection; obtain the collision factor of the optimal trajectory based on the collision time required, the collision warning distance, the comprehensive risk reduction rate, and the hazard assessment index; and perform early warning detection of moving object collisions using the collision factor of the optimal trajectory and the static gray dot area in the drawn map.

[0082] Compared with existing technologies, the beneficial effects of this invention are: obtaining various hazardous areas of the overall image through a deep learning model; drawing a map based on the various hazardous areas and overall terrain data of the overall image; quantitatively analyzing various hazardous areas in the overall image using height information from the overall terrain data, and marking static gray dot areas in the drawn map; obtaining a hazard assessment index for each road based on the hazardous areas within the neighborhood of each road, soil stability, and rainfall; solving the safety problem of vehicles driving within new energy power stations; monitoring people or vehicles within the area to obtain dynamic gray dot areas in the drawn map; solving the positioning accuracy problem of vehicles driving within new energy power stations; constructing a sampling space using the linear velocity and angular velocity of the vehicle at the current moment; obtaining several predicted trajectories through sampling in the sampling space, and from... The optimal trajectory is selected from several predicted trajectories; the path planning problem of vehicles traveling within new energy vehicle depots is solved; the collision time is obtained based on the relative speed and distance between the current vehicle and the vehicle in the nearest dynamic gray dot area on the optimal trajectory; the collision warning distance is obtained based on the relative speed between the current vehicle and the vehicle in the nearest dynamic gray dot area on the optimal trajectory, the driver's reaction time, and the vehicle's maximum negative acceleration; the comprehensive risk reduction rate is obtained through quantitative analysis of vehicle risk protection; the collision factor of the optimal trajectory is obtained based on the collision time, the collision warning distance, the comprehensive risk reduction rate, and the hazard assessment index; and the collision factor of the optimal trajectory and the static gray dot area in the map are used to perform early warning detection of moving object collisions, thereby improving the accuracy of dynamic detection and early warning during vehicle travel within new energy vehicle depots. Attached Figure Description

[0083] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0084] Figure 1 This invention provides a flowchart illustrating the steps of a method for detecting collisions of moving objects based on real-time coordinates.

[0085] Figure 2 This invention provides a schematic diagram of the module flow of a moving object collision detection system based on real-time coordinate position.

[0086] Figure 3 A schematic diagram illustrating the unified conversion between GPS and UWB coordinates. Detailed Implementation

[0087] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0088] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0089] To address the problems existing in the background technology, a method and system for detecting collisions of moving objects based on real-time coordinates have been developed, which has significant practical implications.

[0090] like Figure 1 As shown, the first aspect of the present invention is to provide a method for collision detection of moving objects based on real-time coordinate position, comprising the following steps:

[0091] Step S001: Collect relevant data within the new energy power station and construct a map.

[0092] It should be noted that in order to perform real-time collision detection of moving objects at their coordinate locations, a map is constructed to identify and detect the risk of collisions with moving objects. Therefore, to obtain the map, it is necessary to collect high-precision geographic information and real-time road condition data of the roads inside the wind farm.

[0093] Specifically, drones are used for high-altitude aerial photography to obtain global terrain data and images of the site; then vehicle-mounted cameras and LiDAR are used to collect detailed road information, thus obtaining local terrain data and images.

[0094] The image of the field area and local images are stitched together to obtain the overall image; the global terrain data and local terrain data are stitched together to obtain the overall terrain data (the terrain data is elevation data). The methods for image stitching and terrain data stitching are well-known techniques and will not be described in detail here.

[0095] The pixels in the overall image correspond to the elevation data in the overall terrain data.

[0096] It should be noted that, in order to determine the impact of road potholes, construction areas, and obstacles, a neural network is first trained using deep learning to identify dangerous areas in the image. By identifying dangerous areas, areas that have a significant impact on driving are determined.

[0097] Specifically, Mask R-CNN (Mask Region-based Convolutional Neural Network) is trained using a large dataset to identify various hazardous regions in the overall image. The loss function used in Mask R-CNN is the cross-entropy loss function, a well-known technique that will not be elaborated upon here. These hazardous regions include potholes, construction areas, and obstacles. Obstacles include hills, pedestrians, vehicles, and buildings.

[0098] Various hazardous areas in the overall image are marked in the overall terrain data to create a map; at this point, the map is complete.

[0099] The elevation information (elevation data) in the overall terrain data is used to quantitatively analyze various hazardous areas in the overall image; the specific process is as follows:

[0100] The slope angle for each obstacle's danger zone is calculated, specifically expressed by the formula:

[0101]

[0102] In the formula, This represents the vertical height difference between the highest and lowest points in the danger zone of each obstacle. This represents the difference in horizontal distance between the highest and lowest points within the danger zone of each obstacle. Represents the arctangent function. This indicates the slope angle of each obstacle's danger zone.

[0103] The slope angle corresponding to the vehicle's danger zone is calculated by the slope angle and recorded as the reference slope angle. When the slope angle of each obstacle danger zone is greater than or equal to the reference slope angle, each obstacle danger zone is recorded as a high-risk zone. When the slope angle of each obstacle danger zone is less than the reference slope angle, each obstacle danger zone is recorded as a low-risk zone.

[0104] The depth of each dangerous area of ​​the deep pit is calculated, and the specific formula is as follows:

[0105]

[0106] In the formula, This represents the average height data of the preset length range of each deep pit danger zone. This represents the height data of the lowest point in the danger zone of each deep pit. This indicates the depth of each deep pit danger zone. In this embodiment, the preset length range for each deep pit danger zone is a preset length range of 5 meters around each deep pit danger zone. The specific preset length range is not limited and can be determined by the implementer according to the specific situation.

[0107] If the depth of each deep crater is greater than or equal to half the tire depth of a vehicle, then each deep crater is designated as a high-risk area; if the depth of each deep crater is less than half the tire depth of a vehicle, then each deep crater is designated as a low-risk area.

[0108] Mark all high-risk areas as static gray dots on the map.

[0109] It should be noted that in order to analyze the danger situation of each road, it is necessary to analyze and determine the danger index of each road through multiple influencing factors.

[0110] Specifically, based on the process of obtaining the slope angle of each obstacle hazard zone, the slope angle of the neighborhood range of each road is obtained; based on the slope angle of the neighborhood range of each road, the soil stability coefficient of the neighborhood range of each road, and the rainfall of the neighborhood range of each road, the hazard assessment index of each road is obtained; the hazard assessment index is specifically expressed by the formula:

[0111]

[0112] In the formula, This represents the slope angle of the neighborhood of each road. This represents the soil stability coefficient within the neighborhood of each road. This indicates the rainfall within the neighborhood of each road. The weight representing the slope angle. The weights representing the soil stability coefficient, The weight representing the amount of rainfall, This indicates the hazard assessment index for each road. This represents the hyperbolic tangent function, used for data normalization.

[0113] It should be noted that when the soil within the neighborhood of each road is loose soil, the soil stability coefficient within the neighborhood of each road is marked as 1; when the soil within the neighborhood of each road is hard rock, the soil stability coefficient within the neighborhood of each road is marked as 0.

[0114] In this embodiment, the neighborhood of each road is a range extending 100 meters along the edge of the road. However, this 100-meter range is not specifically limited in this embodiment and can be determined by the implementer based on specific circumstances. Rainfall data is obtained in real-time through meteorological station monitoring and is data for the current moment.

[0115] Thus, the hazard assessment index for each road is obtained using the methods described above.

[0116] Step S002: High-precision positioning and sensor fusion.

[0117] It should be noted that GPS (Global Positioning System) is generally used for real-time vehicle positioning within new energy vehicle depots. However, GPS signals are susceptible to environmental and geographical influences, exhibiting weaknesses such as weak signal, susceptibility to interference, and signal attenuation. Therefore, UWB (Ultra-Wideband) technology is introduced. UWB offers strong multipath resistance, minimal signal superposition interference, and centimeter-level high-precision positioning. By fusing GPS outdoor positioning data and UWB indoor positioning data, seamless integration of indoor and outdoor positioning is achieved.

[0118] It should be further noted that, in order to ensure the time consistency of GPS and UWB positioning data and seamless fusion of data from different coordinate systems, time synchronization and coordinate system unification are required. The process of time synchronization and coordinate system unification is a well-known technique and will not be detailed here. A schematic diagram of the unified conversion between GPS and UWB coordinates is shown below. Figure 3 As shown.

[0119] It should be further noted that, since there may be other pedestrians and vehicles in the area, in order to avoid collisions caused by the failure to detect other pedestrians and vehicles, it is necessary to monitor other pedestrians or vehicles in the area to prevent such situations from occurring.

[0120] Specifically, a monitoring system consisting of several cameras is deployed within the site. This system, combined with a target tracking algorithm, tracks and monitors pedestrians and vehicles. The monitored pedestrians and vehicles, along with location data obtained from GPS and UWB, are marked as dynamic gray-dot areas on a map. In this embodiment, the target tracking algorithm uses either ByteTrack or CSRT (Channel and Spatial Reliability Tracking). The specific target tracking algorithm is not limited in this embodiment; the implementer can choose according to the specific circumstances. ByteTrack and CSRT are well-known technologies and will not be described in detail here.

[0121] Thus, the dynamic gray dot region in the drawn map is obtained using the above method.

[0122] Step S003: Dynamic path planning.

[0123] It should be noted that due to limitations imposed by vehicle acceleration, environmental obstacles, and the vehicle's own speed, the vehicle's speed within the test area cannot be too high. Therefore, the optimal path is optimized based on vehicle speed.

[0124] Specifically, obtain the linear velocity and angular velocity at the current moment; then obtain the maximum acceleration corresponding to the linear velocity and angular velocity, respectively, and denoted as the maximum linear acceleration and the maximum angular acceleration.

[0125] By using the vehicle's linear velocity and angular velocity, maximum linear acceleration and maximum angular acceleration at the current moment, a sampling space for linear velocity and angular velocity is constructed; the predicted value range for linear velocity and angular velocity is specifically expressed as follows:

[0126]

[0127] In the formula, Represents the linear velocity at the current moment. This represents the maximum acceleration corresponding to the linear velocity. This represents the angular velocity at the current moment. This represents the maximum acceleration corresponding to the angular velocity. Indicates time interval, The sampled value represents the linear velocity. The sampled value represents the angular velocity.

[0128] The sampling space is constructed by predicting the range of linear velocity and angular velocity. Based on the sampling space, several data points are obtained according to the linear velocity and angular velocity, and these data points are combined into several combinations.

[0129] The specific formula for obtaining the number of combinations of linear velocity and angular velocity is as follows:

[0130]

[0131] In the formula, This represents the upper limit of linear velocity. This represents the lower limit of linear velocity. This represents the upper limit of angular velocity. This represents the lower limit of angular velocity. The sampling resolution represents the linear velocity. The sampling resolution represents the angular velocity. Indicates to Round up. Indicates to Round up. This represents the number of combinations of linear velocity and angular velocity. The sampling resolution is the sampling interval.

[0132] Based on the linear velocity and angular velocity in each combination, the kinematic model of the moving vehicle driven by difference is used to predict the attitude at each moment in the future time period. The corresponding predicted trajectory is obtained from the attitudes at all moments in the future time period. The attitude at each moment in the future time period is specifically expressed by the formula:

[0133]

[0134]

[0135]

[0136] In the formula, Indicates the vehicle is in The x-axis at each moment, Indicates the vehicle is in The ordinate at each moment, Indicates the vehicle is in The x-axis at each moment, Indicates the vehicle is in The ordinate at each moment, Indicates the vehicle is in The heading angle at that moment, Indicates the vehicle is in The heading angle at that moment, The sampled value represents the linear velocity. The sampled value represents the angular velocity. Indicates time interval, Represents the cosine function. This represents the sine function.

[0137] Thus, by combining each linear velocity and angular velocity, the attitude at several moments in the future is obtained, and by using the attitude at several moments, a predicted trajectory corresponding to each combination of linear velocity and angular velocity is determined.

[0138] The predicted trajectory with the fewest dynamic gray point areas in the predicted trajectory corresponding to each combination of linear velocity and angular velocity is selected and denoted as the optimal trajectory.

[0139] Thus, the optimal trajectory for the current moment is obtained through the above method.

[0140] Step S004: Collision risk quantification and early warning.

[0141] It should be noted that in full monitoring mode, the system must be equipped with a comprehensive early warning and emergency response mechanism. When the monitoring system detects abnormal situations, such as a vehicle deviating from the planned route, abnormal speed, or other potential hazards, it should promptly issue an early warning signal and immediately activate the corresponding emergency response measures to prevent accidents from occurring to the greatest extent possible.

[0142] Specifically, the relative speed between the current vehicle and the vehicle in the nearest dynamic gray point region on the optimal trajectory is obtained; based on the relative speed and distance between the current vehicle and the vehicle in the nearest dynamic gray point region on the optimal trajectory, the collision time is obtained; wherein, the collision time is specifically expressed by the formula:

[0143]

[0144] In the formula, This represents the relative speed between the current vehicle and the vehicle in the nearest dynamic gray dot region on the optimal trajectory. This represents the distance between the current vehicle and the vehicle in the nearest dynamic gray dot region on the optimal trajectory. Indicates the time required for the collision.

[0145] The collision warning distance is obtained based on the relative speed between the current vehicle and the vehicle in the nearest dynamic gray point area on the optimal trajectory, the driver's reaction time, and the vehicle's maximum negative acceleration; the collision warning distance is specifically expressed by the formula:

[0146]

[0147] In the formula, This represents the relative speed between the current vehicle and the vehicle in the nearest dynamic gray dot region on the optimal trajectory. Indicates the driver's reaction time. This indicates the vehicle's maximum negative acceleration. Indicates the collision warning distance.

[0148] It should be noted that since the condition of the vehicle itself, the driver's condition, and the monitoring of the vehicle can all affect the collision situation, adjustments are made based on the risks involved, and the risk reduction rate is analyzed and evaluated using data before and after the adjustments.

[0149] Specifically, by implementing safety design on the vehicle, the number of accidents occurring with and without the safety design is obtained; based on the number of accidents with and without the safety design, a first risk reduction rate is calculated; the first risk reduction rate is expressed by the following formula:

[0150]

[0151] In the formula, This indicates the number of incidents that occurred due to the lack of safety design features. This indicates the number of accidents that occurred due to the use of safety design features. This indicates the first risk reduction rate.

[0152] Drivers are trained, fatigue driving tests are conducted, and alcohol tests are performed to obtain the accident rate after no training or testing and the accident rate after training or testing. Based on these two rates, a second risk reduction rate is obtained. The second risk reduction rate is specifically expressed by the following formula:

[0153]

[0154] In the formula, This indicates the accident rate after no training or testing was conducted. Indicates the accident rate after training or testing. This indicates the second risk reduction rate.

[0155] A smart transportation system (safety management system) is constructed by deploying cameras and radar on the site to monitor the precise location information of vehicles in real time. The system can provide rapid and accurate early warnings of potential hazards, thereby effectively reducing the probability of accidents and ensuring driving safety. The system obtains the accident rates before and after safety management, i.e., the accident rate before and after management. Based on the accident rates before and after management, a third risk reduction rate is obtained. The third risk reduction rate is specifically expressed by the formula:

[0156]

[0157] In the formula, This indicates the rate of unmanaged accidents. Indicates the accident rate after management. This indicates the third risk reduction rate.

[0158] Regular maintenance is performed on the vehicle to obtain the initial failure rate before and after maintenance, as well as the failure rate after maintenance. Based on the initial failure rate and the failure rate after maintenance, a fourth risk reduction rate is obtained. The fourth risk reduction rate is specifically expressed by the formula:

[0159]

[0160] In the formula, Indicates the initial failure rate. Indicates the failure rate after maintenance. This indicates the fourth risk reduction rate.

[0161] The comprehensive risk reduction rate is obtained based on the first, second, third, and fourth risk reduction rates; the comprehensive risk reduction rate is specifically expressed by the formula:

[0162]

[0163] In the formula, Indicates the first Risk reduction rate Indicates the first The weight of the risk reduction rate This indicates the overall risk reduction rate.

[0164] In this embodiment, the weight of the first risk reduction rate is 0.4, the weight of the second risk reduction rate is 0.3, the weight of the third risk reduction rate is 0.2, and the weight of the fourth risk reduction rate is 0.1. In this embodiment, the weight of each risk reduction rate is not specifically limited, and the implementer can determine it according to the specific circumstances.

[0165] The collision factor for each trajectory is obtained based on the collision time, collision warning distance, overall risk reduction rate, and hazard assessment index; the collision factor for the optimal trajectory is specifically expressed by the formula:

[0166]

[0167] In the formula, Indicates the time required for the collision. Indicates the collision warning distance. Indicates the overall risk reduction rate. The risk assessment index represents the optimal trajectory; for Functions used to normalize data; The collision factor represents the optimal trajectory; The weights representing the time required for a collision. The weights representing the collision warning distance, The weighting represents the overall risk reduction rate. This indicates the weight of the risk assessment index.

[0168] In this embodiment, the weight of the collision time is... Weight of collision warning distance and the weight of the risk assessment index Both are 0.3, representing the weight of the overall risk reduction rate. The value is 0.1; however, in this embodiment, the above weight is not specifically limited, and the implementer can determine it according to the specific circumstances.

[0169] When the vehicle encounters a static gray dot area on its optimal trajectory, a warning is triggered, i.e., the system issues an audible and visual alarm; when the collision factor of the optimal trajectory exceeds a preset collision threshold... At this point, it indicates an impending collision, requiring the system to automatically send a warning signal to the vehicle and notify the safety management platform for manual intervention. In this embodiment, a preset collision threshold is used. In this embodiment, a preset collision threshold is used. No specific restrictions are imposed; implementers can decide based on the specific circumstances.

[0170] This concludes the embodiment.

[0171] like Figure 2 As shown, a second aspect of the present invention is to provide a moving object collision detection system based on real-time coordinate position, comprising the following modules:

[0172] Site Analysis Module: This module acquires overall images and terrain data of the site; uses a deep learning model to identify various hazardous areas within the overall image; draws a map based on these hazardous areas and the overall terrain data; performs quantitative analysis on the various hazardous areas in the overall image using height information from the terrain data; and marks static gray dot areas on the map; and obtains a hazard assessment index for each road based on its neighborhood of hazardous areas, soil stability, and rainfall.

[0173] Dynamic gray dot determination module: used to monitor people or vehicles in the area and obtain dynamic gray dot areas in the drawn map;

[0174] The trajectory analysis module is used to construct a sampling space using the vehicle's linear velocity and angular velocity at the current moment; it then samples the vehicle through this sampling space to obtain several predicted trajectories and selects the optimal trajectory from among them.

[0175] The collision detection and warning module is used to: obtain the collision time required based on the relative speed and distance between the current vehicle and the vehicle in the nearest dynamic gray dot area on the optimal trajectory; obtain the collision warning distance based on the relative speed between the current vehicle and the vehicle in the nearest dynamic gray dot area on the optimal trajectory, the driver's reaction time, and the vehicle's maximum negative acceleration; obtain the comprehensive risk reduction rate by quantitatively analyzing the vehicle's risk protection; obtain the collision factor of the optimal trajectory based on the collision time required, the collision warning distance, the comprehensive risk reduction rate, and the hazard assessment index; and perform early warning detection of moving object collisions using the collision factor of the optimal trajectory and the static gray dot area in the drawn map.

[0176] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, optical storage, etc.) containing computer-usable program code.

[0177] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, systems, and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0178] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0179] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0180] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the present invention.

Claims

1. A method for collision detection of moving objects based on real-time coordinate position, characterized in that, include: Acquire overall images and terrain data of the site; based on the overall image, obtain various hazardous areas in the overall image using a deep learning model; draw a map based on the various hazardous areas and overall terrain data in the overall image; perform quantitative analysis on various hazardous areas in the overall image using height information from the overall terrain data, including calculating the slope angle for each obstacle hazardous area, obtaining the slope angle for each obstacle hazardous area, calculating the slope angle for the vehicle's corresponding hazardous area using the slope angle, and recording it as the reference slope angle; when the slope angle of each obstacle hazardous area is greater than or equal to the reference slope angle, then each obstacle hazardous area is recorded as a high-risk area; The depth of each deep-pit danger zone is calculated. If the depth of any deep-pit danger zone is greater than or equal to half the tire height of a passing vehicle, then that zone is designated as a high-risk zone. All high-risk zones are marked as static gray dots on the map. Based on the danger zones within the vicinity of each road, soil stability, and rainfall conditions, a hazard assessment index is obtained for each road. The specific formula for the hazard assessment index of each road is as follows: In the formula, This represents the slope angle of the neighborhood of each road. This represents the soil stability coefficient within the neighborhood of each road. This indicates the rainfall within the neighborhood of each road. The weight representing the slope angle, The weights representing the soil stability coefficient, The weight representing the amount of rainfall, This indicates the hazard assessment index for each road. denoted as the hyperbolic tangent function, used for data normalization; where, if the soil within the neighborhood of each road is loose soil, the soil stability coefficient within the neighborhood of each road is marked as 1; if the soil within the neighborhood of each road is hard rock, the soil stability coefficient within the neighborhood of each road is marked as 0. Monitor people or vehicles within the site to obtain dynamic gray dot areas in the drawn map; A sampling space is constructed using the vehicle's linear velocity and angular velocity at the current moment; several predicted trajectories are obtained by sampling through the sampling space, and the optimal trajectory is selected from these predicted trajectories. The collision time is determined based on the relative speed and distance between the current vehicle and the vehicle in the nearest dynamic gray-dot area on the optimal trajectory; the collision warning distance is determined based on the relative speed between the current vehicle and the vehicle in the nearest dynamic gray-dot area on the optimal trajectory, the driver's reaction time, and the vehicle's maximum negative acceleration; a comprehensive risk reduction rate is obtained through quantitative analysis of vehicle risk protection, including: obtaining the first risk reduction rate based on the number of accidents occurring when the vehicle's safety design is used and the number of accidents occurring when the safety design is not used; obtaining the accident rate after driver training or fatigue driving detection and the accident rate after training or detection based on driver training or fatigue driving detection; and obtaining the accident rate after driver training or fatigue driving detection based on driver's fatigue driving detection and drunk driving detection. The second risk reduction rate is obtained from the accident rate after no inspection and the accident rate after training or inspection; the third risk reduction rate is obtained from the accident rate before and after vehicle management; the initial failure rate and the failure rate after maintenance are obtained from regular vehicle maintenance; the fourth risk reduction rate is obtained from the initial failure rate and the failure rate after maintenance; the comprehensive risk reduction rate is obtained by weighted summing the first, second, third, and fourth risk reduction rates; the collision factor of the optimal trajectory is obtained from the collision time required, the collision warning distance, the comprehensive risk reduction rate, and the hazard assessment index, and the collision factor of the optimal trajectory is specifically expressed by the formula: In the formula, Indicates the time required for the collision. Indicates the collision warning distance. Indicates the overall risk reduction rate. The risk assessment index represents the optimal trajectory; for Functions used to normalize data; The collision factor represents the optimal trajectory; The weights representing the time required for a collision. The weights representing the collision warning distance, The weighting represents the overall risk reduction rate. The weights represent the risk assessment index; all weights are preset values; the early warning detection of moving object collisions is performed by using the collision factor of the optimal trajectory and the static gray dot area in the map, including triggering an early warning when the vehicle encounters a static gray dot area on the optimal trajectory road during driving, i.e., the system issues an audible and visual alarm; when the collision factor of the optimal trajectory is greater than the preset collision threshold, it indicates that a collision is imminent, i.e., the system needs to automatically send a warning signal to the vehicle and notify the safety management platform for manual intervention.

2. The method for detecting collisions of moving objects based on real-time coordinates according to claim 1, characterized in that, The process includes acquiring an overall image and terrain data of the site; obtaining various hazardous areas from the overall image using a deep learning model; and drawing a map based on the various hazardous areas and terrain data from the overall image, including: High-altitude aerial photography using drones is used to acquire global terrain data and images of the site; then, vehicle-mounted cameras and LiDAR are used to collect detailed road information to obtain local terrain data and images. The field area image and local images are stitched together to obtain the stitched overall image; the global terrain data and local terrain data are stitched together to obtain the stitched overall terrain data. Mask R-CNN is trained using a large dataset, and the trained Mask R-CNN is used to identify various danger zones in the overall image. By marking various hazardous areas in the overall image onto the overall terrain data, a map can be drawn.

3. The method for collision detection of moving objects based on real-time coordinate position according to claim 1, characterized in that, The slope angle of each obstacle danger zone is specifically expressed by the formula: In the formula, This represents the vertical height difference between the highest and lowest points in the danger zone of each obstacle. This represents the difference in horizontal distance between the highest and lowest points within the danger zone of each obstacle. Represents the arctangent function. Indicates the slope angle of each obstacle's danger zone; The specific depth of each dangerous zone in the deep pit is expressed by the following formula: In the formula, This represents the average height data of the preset length range of each deep pit danger zone. This represents the height data of the lowest point in the danger zone of each deep pit. This indicates the depth of each dangerous zone in the deep pit.

4. The method for detecting collisions of moving objects based on real-time coordinates according to claim 1, characterized in that, The monitoring of people or vehicles within the site to obtain dynamic gray dot areas in the drawn map includes: A monitoring system consisting of several cameras is deployed within the site. Pedestrians and vehicles are tracked and monitored using the monitoring system and target tracking algorithms. The monitored pedestrians and vehicles are combined with GPS and UWB location data and marked as dynamic gray dot areas on the map.

5. The method for detecting collisions of moving objects based on real-time coordinates according to claim 1, characterized in that, The construction of the sampling space using the vehicle's linear velocity and angular velocity at the current moment includes: Get the linear velocity and angular velocity at the current moment; then get the maximum acceleration corresponding to the linear velocity and angular velocity respectively, and denot them as maximum linear acceleration and maximum angular acceleration; The predicted ranges for linear velocity and angular velocity are specifically expressed as follows: In the formula, Represents the linear velocity at the current moment. This represents the maximum acceleration corresponding to the linear velocity. This represents the angular velocity at the current moment. This represents the maximum acceleration corresponding to the angular velocity. Indicates time interval, The sampled value represents the linear velocity. The sampled value representing the angular velocity; The sampling space is formed by predicting the range of linear velocity and angular velocity values.

6. The method for collision detection of moving objects based on real-time coordinate position according to claim 1, characterized in that, The step of sampling through the sampling space to obtain several predicted trajectories and selecting the optimal trajectory from these predicted trajectories includes: By sampling the space, several data points are obtained based on linear velocity and angular velocity, and these data points are combined into several combinations. The specific formula for obtaining the number of combinations of linear velocity and angular velocity is as follows: In the formula, This represents the upper limit of linear velocity. This represents the lower limit of linear velocity. This represents the upper limit of angular velocity. This represents the lower limit of angular velocity. The sampling resolution represents the linear velocity. The sampling resolution represents the angular velocity. Indicates to Round up. Indicates to Round up. This indicates the number of combinations of linear velocity and angular velocity; Based on the linear velocity and angular velocity in each combination, the attitude of the moving vehicle is predicted at each moment in the future using a differential driving kinematic model. The predicted trajectory is obtained from the attitudes at all moments in the future. The attitude at each moment in the future is specifically expressed by the following formula: In the formula, Indicates the vehicle is in The x-axis at each moment, Indicates the vehicle is in The ordinate at each moment, Indicates the vehicle is in The x-axis at each moment, Indicates the vehicle is in The ordinate at each moment, Indicates the vehicle is in The heading angle at that moment, Indicates the vehicle is in The heading angle at that moment, The sampled value represents the linear velocity. The sampled value represents the angular velocity. Indicates time interval, Represents the cosine function. Represents the sine function; By combining each linear velocity and angular velocity, the attitude at several moments in the future is obtained, and a predicted trajectory corresponding to each combination of linear velocity and angular velocity is determined by the attitude at several moments. The predicted trajectory with the fewest dynamic gray point areas in the predicted trajectory corresponding to each combination of linear velocity and angular velocity is selected and denoted as the optimal trajectory.

7. The method for detecting collisions of moving objects based on real-time coordinates according to claim 1, characterized in that, The method of obtaining the collision time based on the relative speed and distance between the current vehicle and the vehicle in the nearest dynamic gray dot area on the optimal trajectory; and obtaining the collision warning distance based on the relative speed between the current vehicle and the vehicle in the nearest dynamic gray dot area on the optimal trajectory, the driver's reaction time, and the vehicle's maximum negative acceleration, includes: In the formula, This represents the relative speed between the current vehicle and the vehicle in the nearest dynamic gray point region on the optimal trajectory. This represents the distance between the current vehicle and the vehicle in the nearest dynamic gray dot region on the optimal trajectory. Indicates the time required for the collision; In the formula, This represents the relative speed between the current vehicle and the vehicle in the nearest dynamic gray point region on the optimal trajectory. Indicates the driver's reaction time. This indicates the vehicle's maximum negative acceleration. Indicates the collision warning distance.

8. The method for collision detection of moving objects based on real-time coordinate position according to claim 1, characterized in that, The first risk reduction rate is specifically expressed by the formula: In the formula, This indicates the number of incidents that occurred due to the lack of safety design features. This indicates the number of accidents that occurred due to the use of safety design features. Indicates the first risk reduction rate; The second risk reduction rate is specifically expressed by the formula: In the formula, This indicates the accident rate after no training or testing was conducted. This indicates the accident rate after training or testing. This indicates the second risk reduction rate; The third risk reduction rate is specifically expressed by the formula: In the formula, This indicates the rate of unmanaged accidents. Indicates the accident rate after management. Indicates the third risk reduction rate; The fourth risk reduction rate is specifically expressed by the formula: In the formula, Indicates the initial failure rate. Indicates the failure rate after maintenance. This indicates the fourth risk reduction rate; The comprehensive risk reduction rate is specifically expressed by the following formula: In the formula, Indicates the first Risk reduction rate Indicates the first The weight of the risk reduction rate Indicates the overall risk reduction rate; Each risk reduction rate has a preset weight.

9. A moving object collision detection system based on real-time coordinate position, used to implement the moving object collision detection method based on real-time coordinate position as described in claim 1, characterized in that, include: Site analysis module: used to acquire overall images and terrain data of the site; Based on the overall image, various hazardous areas in the overall image are obtained through a deep learning model; based on the various hazardous areas in the overall image and the overall terrain data, a map is drawn; the various hazardous areas in the overall image are quantitatively analyzed using the height information in the overall terrain data, and the static gray dot areas in the drawn map are marked; based on the hazardous areas in the neighborhood of each road, the soil stability, and the rainfall, the hazard assessment index of each road is obtained. Dynamic gray dot determination module: used to monitor people or vehicles in the area and obtain dynamic gray dot areas in the drawn map; Trajectory Analysis Module: Used to construct a sampling space using the vehicle's linear and angular velocities at the current moment; Several predicted trajectories are obtained by sampling the sampling space, and the optimal trajectory is selected from the several predicted trajectories; The collision detection and warning module is used to: obtain the collision time required based on the relative speed and distance between the current vehicle and the vehicle in the nearest dynamic gray dot area on the optimal trajectory; obtain the collision warning distance based on the relative speed between the current vehicle and the vehicle in the nearest dynamic gray dot area on the optimal trajectory, the driver's reaction time, and the vehicle's maximum negative acceleration; obtain the comprehensive risk reduction rate by quantitatively analyzing the vehicle's risk protection; obtain the collision factor of the optimal trajectory based on the collision time required, the collision warning distance, the comprehensive risk reduction rate, and the hazard assessment index; and perform early warning detection of moving object collisions using the collision factor of the optimal trajectory and the static gray dot area in the drawn map.

Images