An autonomous intelligent traffic responsibility determination device
By integrating control modules and WOA algorithms on drones, the problems of traffic congestion and accident handling efficiency in the transportation field are solved, rapid traffic accident handling and responsibility are achieved, transportation efficiency is improved, and new for smart transportation is provided method.
Patent Information
- Application Number
- CN202410631272.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-05-20
- Publication Date
- 2024-11-29
- Estimated Expiration
- 2044-05-20
AI Technical Summary
The existing technology has problems such as traffic congestion, low accident handling efficiency, blind spots in monitoring and determination of responsibilities in the field of transportation. Especially in severe weather conditions, road monitoring of traditional cameras is difficult and high-altitude operations increase construction difficulty. And traditional The accident is determined to be responsible for the traffic police, which is a waste of time.
Design an autonomous intelligent traffic responsibility device, which is set on a drone, including control module, environmental detection unit, camera module, intelligent optimization module, power module, drive module and ground base station, and optimize accident responsibility by using WOA algorithm. Human-machine conducts traffic command, dredging and surveys, quickly collects accident information, and issue warnings through reminder modules to reduce traffic congestion and accident handling time.
It has achieved rapid handling and responsibilities of traffic accidents, reduced traffic accident handling time, improved transportation efficiency, provided a more rapid way to deal with traffic accidents, reduced traffic intersection congestion, and provided new methods for smart transportation construction and New ideas.
Smart Images

Figure CN118506575B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the fields of smart transportation and 5G technology, and in particular to an autonomous intelligent transportation responsibility determination device. Background Art
[0002] In recent years, with the increasing level of domestic scientific and technological development, drones are not only limited to land survey projects. With the continuous in-depth development of transportation construction, the status of traffic drones in the development of the transportation field is increasing. When traffic congestion occurs on a road section that is prone to congestion, the command is issued to the drone through the ground workbench, and the command center can conduct remote high-altitude traffic guidance through the broadcasting system of the traffic drone. In case of emergency, the drone can also restrict the passage of vehicles and reduce traffic congestion through the sirens, signal lights and other devices carried. The existing conventional manual traffic inspection work not only consumes a lot of manpower and material resources, but also has high difficulty and low efficiency under bad weather conditions. The existing technology still uses traditional cameras for road monitoring, which is difficult to set up points, and high-altitude operations increase the difficulty of construction. In addition, the subsequent maintenance and repair requires a lot of money, and there are monitoring blind spots due to technical limitations. In terms of determining responsibility, traditional accident determination requires traffic police to go to the scene to handle, which not only wastes time, but also may cause traffic congestion. When traffic highways are blocked, accidents occur, and some weather and geographical factors endanger transportation safety, it has become a problem for the traffic police department to understand the specific conditions of the road section.
[0003] In response to the above problems, the present invention proposes an autonomous intelligent traffic responsibility determination device, which has a wide field of view, can achieve ultra-long-distance control, and can realize global monitoring of the ground from high altitude. It can also realize flexible inspections, which can make up for the problem of traditional cameras having monitoring blind spots. When a traffic accident occurs, the present invention can quickly arrive at the scene, use broadcasting equipment to direct and clear traffic before the arrival of the traffic police, and conduct surveys and records, and restore road order after completion, and then determine responsibility based on the survey results. In the event of injuries, etc., the present invention can quickly deliver urgently needed materials to the accident site, which can save a lot of time. The present invention provides a faster way to handle traffic accidents, which can effectively reduce the time for handling traffic accidents, reduce traffic congestion at traffic intersections, and improve transportation efficiency, providing a new method and new idea for the construction of smart transportation. Summary of the invention
[0004] Purpose of the invention: In view of the problems existing in the prior art, the present invention provides an autonomous intelligent traffic responsibility determination device and an intelligent traffic management mode, which can effectively reduce the incidence of traffic accidents and provide a new method and new idea for the construction of smart transportation.
[0005] Technical solution: The present invention discloses an autonomous intelligent traffic responsibility determination device, which is arranged on a drone and includes a control module, an environment detection unit, a camera module, an intelligent optimization module, a power module, a drive module, and a ground base station;
[0006] The environmental detection unit is used to collect environmental data, including vehicle speed, accident location, environmental granularity, vehicle density, temperature, and air pressure;
[0007] The camera module transmits the captured images to the intelligent optimization module for intelligent recognition, including the accident object, the distance of the wreckage, the degree of casualties, and the degree of vehicle damage;
[0008] The control module is provided with an intelligent optimization module, which summarizes the information collected by the environment detection unit and optimizes the accuracy of accident responsibility determination using the WOA algorithm based on the images taken by the camera module;
[0009] The database contains the following traffic accident situations: rear-end collision, side collision, head collision, rollover, single-vehicle accident, pedestrian accident, motorcycle accident, bicycle accident, and the corresponding accident coefficients are set;
[0010] The judgment is made based on the accident coefficient results optimized by the intelligent algorithm, and the responsibility determination results are quickly sent to the accident personnel in the form of text messages; if the accident is of high complexity, the picture information and accident data are transmitted to the ground base station for the staff to further determine the responsibility;
[0011] The power module supplies energy to the device; the drive module includes a motor and a propeller, which controls the normal operation of the entire device in the air; the ground base station receives the data information collected by the device, which is processed and analyzed by a computer to help base station personnel handle more complex accidents and make decisions and plans.
[0012] Furthermore, it also includes a reminder module, which exists independently and is connected to the control module, including a loudspeaker, a signal light, a jog control button, and a camera, and the jog control button is connected in parallel with the camera; when the camera captures a traffic accident, the reminder module prompts, issues a loudspeaker warning, and the signal light flashes.
[0013] Furthermore, the following objective function is set in the intelligent optimization module:
[0014]
[0015] Among them, M is the liability coefficient, V is the speed at the time of the accident collision, μ is the degree of deformation of the vehicle, sinθ is the sine value of the collision angle when the accident occurs, γ is the influencing factor of the environmental effect, n is the number of casualties, d is the distance the wreckage flies out, whole is the traffic integrity coefficient, and c is the congestion index.
[0016] Furthermore, the intelligent optimization module uses the WOA algorithm to provide an optimization strategy, and the specific optimization process is as follows;
[0017] 1) Population initialization: At the beginning of the algorithm, a group of populations is generated. During the search process, each population has a corresponding responsibility result;
[0018] 2) Evaluating the objective function: inputting the parameters collected by the environment detection unit and the images captured by the camera device into the objective function to be evaluated;
[0019] 3) Define Mest: Bring the initial population into the objective function for evaluation, sort the population, and take the responsibility coefficient that is closest to the accident coefficient in the database as Mest;
[0020] 4) Calculate the transfer factors A and C. Use the transfer factors to evaluate the moving step size and prediction angle of the population. It is also used to control the balance between the early iterative global search and the later iterative local search. The formula is as follows:
[0021] A=2ar1-a
[0022] C=2r2
[0023] a=2-2t / T max
[0024] Where r1 and r2 are random numbers in the range of (0,1), the value of a decreases from 2 to 0, and t represents the current number of iterations. max Set the maximum number of iterations for the initialization;
[0025] 5) Surrounding prey: The population updates its position by surrounding the prey. The formula is as follows:
[0026] D=|C*Mest-X|
[0027] X(t+1)=Mest-AD
[0028] Among them, X(t+1) represents the position of the population in the current iteration;
[0029] 6) Capture prey: Use spiral motion to move towards the prey. The formula is as follows:
[0030] X(t+1)=Mest-de evit*l cos(2πl)
[0031] Among them, d is the distance the wreckage flew out, e vit Indicates the traffic volume at the intersection. In this formula, it represents the spiral shape. l is a random number in the range of (-1, -1);
[0032] 7) Contraction range: When the population surrounds the prey, it needs to shrink the encirclement circle. It is necessary to select the contraction encirclement mechanism and the spiral model to set the probability P i and 1-P i , the formula is as follows:
[0033]
[0034] 8) Passing probability P i With the change of fluctuation range A, the population will update its position in different ways and move towards the optimal position;
[0035] 9) Update the position of the population so that the target moves to the top ranked accountability coefficient in the search space;
[0036] D=|CX rard -X(t)|
[0037] X(t+1)=X rand -AD
[0038] Among them, X rand For random positions, randomly select an individual and update the positions of other individuals to increase the randomness of exploration;
[0039] 10) Determine whether the set maximum number of iterations has been reached, if not, return to step 3), if yes, go to step 11);
[0040] 11) Output the duty coefficient of the current duty device.
[0041] Furthermore, the search step length is optimized and improved, and the improved formula is as follows:
[0042] R = α × e -(β×g) / Margen
[0043] Among them, α is the step size control factor, β is the exponential control factor, g is the current number of iterations, and Maxge is the maximum number of iterations.
[0044] Furthermore, the algorithm is improved by introducing the Levy flight model to avoid the algorithm from falling into the local optimum during the random exploration or development phase, and to improve the convergence accuracy of the algorithm and the flexibility of the random hiding phase. The model formula is as follows:
[0045]
[0046] σ v=1
[0047] Among them, Γ() is the gamma function, f is the step size, u and v obey the mean value of 0 and the variance is and Gaussian distribution, β represents the relevant parameter;
[0048] Calculate the energy factor R, the calculation formula is as follows:
[0049]
[0050] When R(t)>1, the exploration phase of the algorithm is executed, and when R(t)≤1, the development phase of the algorithm is executed.
[0051] Beneficial effects:
[0052] 1. The new autonomous intelligent traffic responsibility determination device proposed in the present invention can quickly arrive at the scene of a traffic accident, use the broadcasting system to direct and clear the traffic at the accident scene before the traffic police arrive, and quickly collect on-site conditions. The accident can be determined based on the investigation results, and the car owner can be informed of the determination results by text message, which can meet the needs of different scenarios.
[0053] 2. The present invention is provided with an independent module, which ensures that when a traffic accident occurs ahead, the driver can be reminded through the reminder module, and the device will issue a loudspeaker warning and the signal light will flash to prevent the occurrence of a secondary accident.
[0054] 3. The WOA algorithm proposed by the intelligent optimization module of the present invention is highly adaptable to the present invention. In addition, the algorithm has a simple structure, and by adding the Levy flight model, the algorithm's exploration ability is improved, and the algorithm can find the global optimal solution. BRIEF DESCRIPTION OF THE DRAWINGS
[0055] Figure 1 It is a schematic diagram of the overall structure of the present invention;
[0056] Figure 2 It is a structural schematic diagram of the present invention;
[0057] Figure 3 Flow chart of the WOA algorithm of the present invention. DETAILED DESCRIPTION
[0058] The technical solution of the present invention is further described below in conjunction with the accompanying drawings.
[0059] like Figure 1-3 As shown, the present invention discloses an autonomous intelligent traffic responsibility determination device, which is arranged on a drone and includes a control module, an environment detection unit, a camera module, an intelligent optimization module, a power module, a drive module, a reminder module, and a ground base station.
[0060] The environmental detection unit is used to collect environmental data, including vehicle speed, accident location, environmental granularity, vehicle density, temperature, and air pressure. The environmental detection unit includes light sensors, infrared position sensors, temperature sensors, haze sensors, speed sensors, humidity sensors, etc. The environmental detection unit is used to monitor the places where traffic accidents frequently occur in the city.
[0061] The camera module transmits the captured images to the intelligent optimization module for intelligent identification, including the accident object, the distance the wreckage flew, the degree of casualties, and the degree of vehicle damage.
[0062] The control module is provided with an intelligent optimization module, which summarizes the information collected by the environmental detection unit and optimizes the accuracy of accident responsibility determination by using the WOA algorithm based on the images taken by the camera module.
[0063] The database contains the following traffic accident situations: rear-end collision, side collision, head collision, rollover accident, single-vehicle accident, pedestrian accident, motorcycle accident, and bicycle accident, and sets the corresponding accident coefficients.
[0064] The judgment is made based on the accident coefficient results optimized and identified by the intelligent algorithm, and the responsibility determination results are quickly sent to the accident personnel in the form of text messages; if the accident complexity is high, the picture information and accident data are transmitted to the ground base station for the staff to further determine the responsibility.
[0065] The power module supplies energy to the device, using lightweight and environmentally friendly Ni-MH batteries as power. The drive module includes motors and propellers, which control the normal operation of the entire device in the air; the ground base station receives the data collected by the device, which is processed and analyzed by computers to help base station personnel handle more complex accidents and make decisions and plans.
[0066] The intelligent optimization module summarizes the information collected by the sensor module and first uses the WOA algorithm to optimize the accuracy of accident responsibility determination. Then, based on the images taken by the camera module (such as the accident object, the distance of the wreckage flying, the degree of casualties, and the degree of vehicle damage), it is transmitted to the intelligent optimization module for intelligent identification. The database generally contains the following types of traffic accidents: rear-end collisions, side collisions, head collisions, rollover accidents, single-vehicle accidents, pedestrian accidents, motorcycle accidents, and bicycle accidents. Judgments are made based on the results of intelligent algorithm optimization and identification, and the responsibility determination results are quickly sent to the accident personnel in the form of text messages.
[0067] The reminder module exists independently, including a loudspeaker, a signal light, a jog control button, and a camera. The jog control button is connected in parallel with the camera. When the camera captures a traffic accident, the driver can be notified through the reminder module. The device will issue a loudspeaker warning and the signal light will flash to prevent a secondary accident.
[0068] like Figure 3 As shown, the intelligent optimization module uses the WOA algorithm to give an optimization strategy. The specific optimization process is as follows;
[0069] 1) Population initialization: At the beginning of the algorithm, a group of populations is generated. During the search process, each population has a corresponding determination result.
[0070] 2) Evaluate the objective function: Input the parameters collected by the environment detection unit and the images captured by the camera device into the objective function to be evaluated. The objective function formula is as follows:
[0071]
[0072] Among them, M is the liability coefficient, V is the speed at the time of the accident collision, μ is the degree of deformation of the vehicle, sinθ is the sine value of the collision angle when the accident occurs, γ is the influencing factor of the environmental effect, n is the number of casualties, d is the distance the wreckage flies out, whole is the traffic integrity coefficient, and c is the congestion index.
[0073] 3) Define Mest: Bring the initial population into the objective function for evaluation, sort the population, and take the responsibility coefficient that is closest to the accident coefficient in the database as Mest;
[0074] 4) Calculate the transfer factors A and C. The transfer factors can be used to evaluate the moving step size and prediction angle of the population. At the same time, it is used to control the balance between the early iterative global search and the later iterative local search. The formula is as follows:
[0075] A=2ar1-a
[0076] C=2r2
[0077] a=2-2t / T max
[0078] Where r1 and r2 are random numbers in the range of (0,1), the value of a decreases from 2 to 0, and t represents the current number of iterations. max Set the maximum number of iterations for the initialization.
[0079] 5) Surrounding prey: The population updates its position by surrounding the prey. The formula is as follows:
[0080] D=|C*Mest-X|
[0081] X(t+1)=Mest-AD
[0082] Among them, X(t+1) represents the position of the population in the current iteration.
[0083] 6) Capture prey: Use spiral motion to move towards the prey. The formula is as follows:
[0084] X(t+1)=Mest-de evit*l cos(2πl)
[0085] Among them, d is the distance the wreckage flew out, e vit Indicates the traffic volume at the intersection. In this formula, it represents the spiral shape, and l is a random number in the range of (-1, -1).
[0086] 7) Contraction range: When the population surrounds the prey, it needs to shrink the encirclement circle. It is necessary to select the contraction encirclement mechanism and the spiral model to set the probability P i and 1-P i , the formula is as follows:
[0087]
[0088] 8) Passing probability P i With the change of fluctuation range A, the population will update its position in different ways and move towards the optimal position.
[0089] 9) Update the position of the population so that the target moves to the top ranked accountability coefficient in the search space;
[0090] D=|CX rard -X(t)|
[0091] X(t+1)=X rand -AD
[0092] Among them, X rand For random positions, randomly select an individual and update the positions of other individuals to increase the randomness of exploration.
[0093] 10) The search step length is optimized and improved. The improved formula is as follows:
[0094] R = α × e -(β×g) / Margen
[0095] Among them, α is the step size control factor, β is the exponential control factor, g is the current number of iterations, and Maxge is the maximum number of iterations.
[0096] 11) Determine whether the set maximum number of iterations has been reached. If not, return to step 3). If yes, go to step 12).
[0097] 12) Output the duty coefficient of the current duty device.
[0098] The algorithm is improved by introducing the Levy flight model to avoid the algorithm from falling into local optimality during the random exploration or development phase, and to improve the convergence accuracy of the algorithm and the flexibility of the random hiding phase. The model formula is as follows:
[0099]
[0100] σ v =1
[0101] Among them, Γ() is the gamma function, f is the step size, u and v obey the mean value of 0 and the variance is and Gaussian distribution, β represents the correlation parameter, usually set to 1.5.
[0102] 13) Calculate the energy factor R. The calculation formula is as follows:
[0103]
[0104] When R(t)>1, the exploration phase of the algorithm is executed, and when R(t)≤1, the development phase of the algorithm is executed.
[0105] The present invention discloses an autonomous intelligent traffic responsibility determination device. Compared with the prior art, the present invention provides a faster way to handle traffic accidents, which can effectively reduce the time for handling traffic accidents, reduce traffic congestion at traffic intersections, and improve transportation efficiency, thus providing a new method and new idea for the construction of smart transportation.
[0106] The above-mentioned embodiments are only for illustrating the technical concept and features of the present invention, and their purpose is to enable people familiar with this technology to understand the content of the present invention and implement it accordingly, and they cannot be used to limit the protection scope of the present invention. Any equivalent transformation or modification made according to the spirit of the present invention should be included in the protection scope of the present invention. Although the embodiments of the present invention have been shown and described, it can be understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the attached claims and their equivalents.
Claims
1. An autonomous intelligent traffic responsibility determination device, characterized in that: The device is arranged on a drone, and includes a control module, an environment detection unit, a camera module, an intelligent optimization module, a power module, a drive module, and a ground base station; The environmental detection unit is used to collect environmental data, including vehicle speed, accident location, environmental granularity, vehicle density, temperature, and air pressure; The camera module transmits the captured images to the intelligent optimization module for intelligent recognition, including the accident object, the distance of the wreckage, the degree of casualties, and the degree of vehicle damage; The control module is provided with an intelligent optimization module, which summarizes the information collected by the environment detection unit and optimizes the accuracy of accident responsibility determination using the WOA algorithm based on the images taken by the camera module; The database contains the following traffic accident situations: rear-end collision, side collision, head collision, rollover, single-vehicle accident, pedestrian accident, motorcycle accident, bicycle accident, and the corresponding accident coefficients are set; Make judgments based on the accident coefficient results optimized by intelligent algorithms, and quickly send the results of responsibility determination to the accident personnel in the form of text messages; If the accident is of high severity, the image information and accident data will be transmitted to the ground base station for the staff to further determine the responsibility; The following objective function is set in the intelligent optimization module: Among them, M is the liability coefficient, V is the speed at the time of the accident collision, μ is the degree of deformation of the vehicle, sinθ is the sine value of the collision angle at the time of the accident, γ is the influencing factor of the environmental effect, n is the number of casualties, d is the distance the wreckage flew out, whole is the traffic integrity coefficient, and c is the congestion index; The intelligent optimization module uses the WOA algorithm to provide an optimization strategy. The specific optimization process is as follows: 1) Population initialization: At the beginning of the algorithm, a group of populations is generated. During the search process, each population has a corresponding responsibility result; 2) Evaluating the objective function: inputting the parameters collected by the environment detection unit and the images captured by the camera device into the objective function to be evaluated; 3) Define Mest: Bring the initial population into the objective function for evaluation, sort the population, and take the responsibility coefficient that is closest to the accident coefficient in the database as Mest; 4) Calculate the transfer factors A and C. Use the transfer factors to evaluate the moving step size and prediction angle of the population. It is also used to control the balance between the early iterative global search and the later iterative local search. The formula is as follows: A=2ar1-a C=2r2 a=2-2t / T max Where r1 and r2 are random numbers in the range of (0,1), the value of a decreases from 2 to 0, and t represents the current number of iterations. max Set the maximum number of iterations for the initialization; 5) Surrounding prey: The population updates its position by surrounding the prey. The formula is as follows: D=|C*Mest-X| X(t+1)=Mest-AD Among them, X(t+1) represents the position of the population in the current iteration; 6) Capture prey: Use spiral motion to move towards the prey. The formula is as follows: X(t+1)=Mest-of evit*l cos(2πl) Among them, d is the distance the wreckage flew out, e vit Indicates the traffic volume at the intersection. In this formula, it represents the spiral shape. l is a random number in the range of (-1, -1); 7) Contraction range: When the population surrounds the prey, it needs to shrink the encirclement circle. It is necessary to select the contraction encirclement mechanism and the spiral model to set the probability P i and 1-P i , the formula is as follows: 8) Passing probability P i With the change of fluctuation range A, the population will update its position in different ways and move towards the optimal position; 9) Update the position of the population so that the target moves to the top ranked accountability coefficient in the search space; D=∣CX rard -X(t)∣ X(t+1)=X rand -AD Among them, X rand For random positions, randomly select an individual and update the positions of other individuals to increase the randomness of exploration; 10) Determine whether the set maximum number of iterations has been reached, if not, return to step 3), if yes, go to step 11); 11) Output the duty coefficient of the current duty device; The power module supplies energy to the device; the drive module includes a motor and a propeller, which controls the normal operation of the entire device in the air; the ground base station receives the data information collected by the device, which is processed and analyzed by a computer to help base station personnel handle more complex accidents and make decisions and plans.
2. The autonomous intelligent traffic responsibility determination device according to claim 1, characterized in that: It also includes a reminder module, which exists independently and is connected to the control module. It includes a loudspeaker, a signal light, a jog control button, and a camera. The jog control button is connected in parallel with the camera. When the camera captures a traffic accident, the reminder module will prompt, issue a loudspeaker warning, and the signal light will flash.
3. The autonomous intelligent traffic responsibility determination device according to claim 1, characterized in that: The search step length is optimized and improved, and the improved formula is as follows: R=α×e -(β×g) / Margen Among them, α is the step size control factor, β is the exponential control factor, g is the current number of iterations, and Maxge is the maximum number of iterations.
4. The autonomous intelligent traffic responsibility determination device according to claim 1, characterized in that: The algorithm is improved by introducing the Levy flight model to avoid the algorithm from falling into local optimality during the random exploration or development phase, and to improve the convergence accuracy of the algorithm and the flexibility of the random hiding phase. The model formula is as follows: s v =1 Among them, Γ() is the gamma function, f is the step size, u and v obey the mean value of 0 and the variance is and Gaussian distribution, β represents the relevant parameter; Calculate the energy factor R, the calculation formula is as follows: When R(t)>1, the exploration phase of the algorithm is executed, and when R(t)≤1, the development phase of the algorithm is executed.
Citation Information
Patent Citations
Road traffic accident rapid responsibility determination photographing system based on unmanned aerial vehicle
CN107820044A
Method for quickly obtaining evidence for traffic accident
CN107945526A
