Vehicle overload detection method and system

Abstract

The invention relates to the technical field of vehicle detection, and discloses a vehicle overload detection method and system, and the method comprises the steps: obtaining a vehicle point cloud, determining the static information and vibration information of a vehicle based on the vehicle point cloud, and determining the final predicted weight of the vehicle based on the static information and vibration information, and vehicle overload judgment is carried out based on the final predicted weight of the vehicle and the rated weight of the vehicle. According to the scheme, the adaptive capacity of vehicle overload detection to different environments and the accuracy of vehicle weight prediction are improved, the accuracy of vehicle overload detection is improved, and the misjudgment rate is reduced.

Description

Technical Field

[0001] This invention relates to the field of vehicle inspection technology, and in particular to a method and system for detecting vehicle overload. Background Technology

[0002] Currently, vehicle overload detection mainly relies on methods such as weighbridge inspection, embedded sensor inspection, and visual inspection. Weighbridge inspection requires the installation of fixed weighing facilities on the road, which presents problems such as high construction and maintenance costs and disruption to normal traffic flow. Embedded sensor inspection estimates vehicle weight by disrupting the existing road structure and laying pressure or bending sensors under the road surface, which suffers from difficulties in later maintenance and high sensor failure rates. Visual inspection uses cameras to measure vehicle subsidence and cargo loading to estimate weight, but due to the lack of vehicle characteristic parameters, it is prone to large estimation errors. Furthermore, the estimation results of visual inspection are greatly affected by lighting, weather, and weather obstruction, resulting in a high false positive rate and poor accuracy. Summary of the Invention

[0003] In view of this, the present invention proposes a vehicle overload detection method and system, which solves the technical problems of high misjudgment rate and poor accuracy when estimating vehicle weight through existing vehicle overload detection methods.

[0004] On one hand, embodiments of the present invention provide a method for detecting vehicle overload, including: Acquire vehicle point cloud data, and determine the static and vibration information of the vehicle based on the vehicle point cloud data; Based on the static information and the vibration information, the final predicted weight of the vehicle is determined; The vehicle overload determination is based on the vehicle's final predicted weight and the vehicle's rated weight.

[0005] In some implementations, the static information includes static sinkage and total tire contact area; the vibration information includes a vibration sequence of vehicle height over time; and determining the final predicted weight of the vehicle based on the static information and the vibration information includes: Based on the static subsidence, the first predicted weight of the vehicle is determined; Based on the total tire contact area, the second predicted weight of the vehicle is determined; Based on the vibration sequence, the third predicted weight of the vehicle is determined; The final predicted weight of the vehicle is determined based on the first predicted weight, the second predicted weight, and the third predicted weight.

[0006] In some implementations, determining the final predicted weight of the vehicle based on the first predicted weight, the second predicted weight, and the third predicted weight includes: Based on the vehicle's speed, the type of road surface the vehicle is traveling on, and weather conditions, the weights assigned to the first predicted weight, the second predicted weight, and the third predicted weight are determined respectively. The final predicted weight of the vehicle is determined based on the first predicted weight, the second predicted weight, and the third predicted weight, and their respective weights.

[0007] In some implementations, before determining the static and vibration information of the vehicle based on the vehicle point cloud, the method further includes: acquiring a side image of the vehicle and determining the vehicle model based on the side image.

[0008] Determining the first predicted weight of the vehicle based on the static sinking includes: obtaining the suspension stiffness of the vehicle based on the vehicle model, and determining the first predicted weight of the vehicle based on the static sinking and the suspension stiffness.

[0009] In some implementations, determining the second predicted weight of the vehicle based on the total tire contact area includes: The standard tire pressure of the vehicle is obtained based on the vehicle model, and the second predicted weight of the vehicle is determined based on the total tire contact area and the standard tire pressure.

[0010] In some implementations, after acquiring the vehicle side image, the method further includes: determining the tire sidewall deformation information of the vehicle based on the vehicle side image; Determining the second predicted weight of the vehicle based on the total tire contact area and the standard tire pressure includes: The standard tire aspect ratio of the vehicle is obtained based on the vehicle model. Based on the tire sidewall deformation information, the actual aspect ratio of the tire is determined; The tire pressure correction factor is determined based on the actual aspect ratio of the tire and the standard tire aspect ratio. The second predicted weight of the vehicle is determined based on the tire pressure correction factor, the standard tire pressure, and the total tire contact area.

[0011] In some implementations, determining the third predicted weight of the vehicle based on the vibration sequence includes: Based on the vibration sequence, the damping ratio when the vehicle vibrates is determined; Based on the damping ratio, the natural frequency of the vehicle when it vibrates is determined; The third predicted weight of the vehicle is determined based on the natural frequency and the vehicle's suspension stiffness.

[0012] In some implementations, acquiring vehicle point clouds and determining the vehicle's static and vibration information based on the vehicle point clouds includes: Acquire vehicle point cloud data while the vehicle is traveling in a flat detection area, and determine the ground reference plane and vehicle contour point cloud based on the vehicle point cloud data. Based on the vehicle contour point cloud and the ground reference plane, determine the static subsidence and total tire contact area in the static information; The vehicle point cloud sequence after the vehicle passes through the vibration detection area is obtained, and based on the vehicle point cloud sequence, the vibration sequence of the vehicle height changing over time is determined. The vibration detection area is equipped with a road surface excitation device.

[0013] In some implementations, determining the static subsidence and total tire contact area in the static information based on the vehicle contour point cloud and the ground reference plane includes: Based on the height of the vehicle contour point cloud relative to the ground reference plane, determine the static subsidence amount in the static information; Extract the vehicle tire point cloud clusters that are in contact with the ground reference plane from the vehicle outline point cloud; Based on the extracted vehicle tire point cloud clusters and the ground reference plane, the total tire ground contact area in the static information is determined.

[0014] In some embodiments, the vehicle overload detection method provided by the present invention further includes: determining a system deviation based on the vehicle's historical predicted weight, and correcting the basic parameter information and weights corresponding to the vehicle in response to the existence of a system deviation.

[0015] On the other hand, embodiments of the present invention also provide a vehicle overload detection system, which includes a vehicle information acquisition device and a terminal electronic device connected to the vehicle information acquisition device. The vehicle information acquisition device is used to acquire vehicle point clouds, and the terminal electronic device is configured to execute the vehicle overload detection method as described in any of the above embodiments.

[0016] The present invention has at least the following beneficial effects: This invention provides a vehicle overload detection method and system. By acquiring vehicle point clouds, determining the vehicle's static and vibration information based on the vehicle point clouds, determining the vehicle's final predicted weight based on the static and vibration information, and making a vehicle overload judgment based on the vehicle's final predicted weight and the vehicle's rated weight, this technical solution improves the adaptability of vehicle overload detection to different environments, as well as the accuracy of vehicle weight prediction, thereby improving the accuracy of vehicle overload detection and reducing the false judgment rate. Attached Figure Description

[0017] 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 embodiments can be obtained based on these drawings without creative effort.

[0018] Figure 1 A flowchart of a vehicle overload detection method provided in an embodiment of the present invention; Figure 2 This is a flowchart illustrating a method for determining static and vibration information of a vehicle in an embodiment of the present invention for vehicle overload detection. Figure 3 A flowchart illustrating a method for determining the final predicted weight of a vehicle in a vehicle overload detection method provided in an embodiment of the present invention; Figure 4 This is a flowchart of a method for correcting the second predicted weight of a vehicle in a vehicle overload detection method provided in an embodiment of the present invention; Figure 5 This is a schematic diagram of a vehicle overload detection system provided in an embodiment of the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be further described in detail below with reference to specific examples and the accompanying drawings.

[0020] It should be noted that all uses of "first" and "second" in the embodiments of the present invention are for the purpose of distinguishing two entities or parameters with the same name but different names. It is clear that "first" and "second" are only for the convenience of expression and should not be construed as limiting the embodiments of the present invention. Subsequent embodiments will not explain this in detail.

[0021] The present invention will now be described in detail with reference to the embodiments and accompanying drawings.

[0022] The first aspect of this invention provides a method for detecting vehicle overload, such as... Figure 1 As shown, the method includes steps S10 to S30.

[0023] S10. Obtain vehicle point cloud and determine vehicle static and vibration information based on vehicle point cloud.

[0024] Specifically, in this embodiment of the invention, a detection area can be pre-demarcated on the traffic road, and one or more radars can be installed on the side or directly above the designated detection area. The radar can be lidar, millimeter-wave radar, etc. For ease of understanding, this embodiment of the invention will use lidar as an example. In this embodiment, for vehicles passing through the detection area, the installed lidar can collect three-dimensional point cloud data (i.e., vehicle point cloud) of the vehicle as it passes through the detection area.

[0025] In this embodiment of the invention, if there is a radar on the side or directly above the road in the defined detection area, it is not necessary to follow the radar; the vehicle point cloud when the vehicle passes through the detection area can be obtained directly from the existing radar.

[0026] The vehicle's static information may include information unaffected by road conditions, such as vehicle height, static sinking, and total tire contact area. The vehicle's vibration information may include a sequence of changes in vehicle height over time when the vehicle vibrates. In some embodiments, to ensure that the vehicle vibrates, a road excitation device (e.g., a speed bump) may be placed in the detection area. This causes the vehicle to vibrate after passing over the speed bump, allowing for the visualization of a vibration sequence showing the change in vehicle point cloud height over time when the vehicle vibrates.

[0027] In this embodiment of the invention, after obtaining the vehicle point cloud, calculations can be performed on the vehicle point cloud to obtain the vehicle's static and vibration information. This embodiment of the invention improves the adaptability of vehicle overload detection to different environments and reduces the false positive rate of vehicle overload detection by using vehicle point cloud data to calculate the vehicle's static and vibration information.

[0028] S20. Based on static and vibration information, determine the final predicted weight of the vehicle.

[0029] Specifically, a vehicle weight prediction model that represents the relationship between vehicle static information and vehicle weight, and a vehicle weight prediction model that represents the relationship between vehicle vibration information and vehicle weight can be constructed. The vehicle static information and vibration information obtained from the vehicle point cloud are then input into the corresponding vehicle weight prediction models to predict the vehicle weight. The predicted weights are then fused to obtain the fused weight of the vehicle (i.e., the final predicted weight of the vehicle). This multi-dimensional vehicle weight prediction improves the accuracy of vehicle weight prediction.

[0030] S30. Determine if a vehicle is overloaded based on its final predicted weight and its rated weight.

[0031] Specifically, the vehicle model can be obtained, and based on the vehicle model, the vehicle's rated weight can be obtained. Based on the final predicted weight of the vehicle and the vehicle's rated weight, an overload determination can be made.

[0032] When obtaining a vehicle's model, you can first acquire a vehicle image and then perform image recognition to determine the model. The method for acquiring vehicle images can be similar to acquiring vehicle point clouds, the difference being that the image acquisition device is a camera or other video recording equipment.

[0033] When determining overload, for example, the final predicted weight of the vehicle can be compared with its rated weight. If the final predicted weight exceeds the rated weight, the vehicle is considered overloaded; otherwise, it is considered not overloaded. Alternatively, an overload threshold can be set based on the vehicle's rated weight, and the final predicted weight can be compared with this threshold. If the final predicted weight exceeds the threshold, the vehicle is considered overloaded; otherwise, it is considered not overloaded.

[0034] The present invention provides a technical solution that improves the adaptability of vehicle overload detection to different environments and the accuracy of vehicle weight prediction by acquiring vehicle point clouds, determining the final predicted weight of the vehicle based on the vehicle point clouds, and determining vehicle overload based on the final predicted weight and the vehicle's rated weight. This improves the accuracy of vehicle overload detection and reduces the false judgment rate.

[0035] In some embodiments of the present invention, the vehicle license plate number can also be determined by acquiring vehicle images and recognizing the acquired vehicle images, thereby realizing the vehicle marking and the association between the collected multi-frame data.

[0036] Specifically, in this embodiment, the vehicle image can be an image acquired along the length of the vehicle to facilitate the identification of the vehicle license plate number. The identification method can employ an OCR recognition algorithm to obtain an accurate vehicle license plate number.

[0037] In this embodiment of the invention, at least two industrial cameras can be installed on the side or directly above the road in the detection area. One camera is used to capture the full view of the side of the vehicle to identify the vehicle model and tire shape; the other camera is used to capture the front or rear of the vehicle to identify the license plate information.

[0038] In this embodiment of the invention, data can be collected synchronously using radar and cameras. The radar continuously acquires a complete point cloud sequence of a vehicle from entering to leaving the detection area, while the camera simultaneously captures an image sequence containing clear license plates and vehicle bodies. The cloud device acquires the point cloud sequence and images collected synchronously by the radar and cameras, creates a unique identifier (such as "timestamp + license plate number") for each passing vehicle, and binds and stores the corresponding point cloud sequence and image sequence to ensure data consistency.

[0039] In some embodiments of the present invention, such as Figure 2 As shown, the static information and vibration information of the vehicle can be determined according to steps S210~S230.

[0040] S210. Obtain the vehicle point cloud while the vehicle is traveling in a flat detection area, and determine the ground reference plane and vehicle contour point cloud based on the vehicle point cloud.

[0041] The detection area includes a flat detection area and a vibration detection area. When the vehicle travels in the flat detection area, no vibration occurs. The ground reference plane and vehicle contour point cloud can be determined based on the vehicle point cloud obtained at this time. The vibration detection area is equipped with road excitation devices (such as speed bumps). Therefore, when the vehicle travels in the vibration detection area, it will experience bumps, which will cause the vehicle to vibrate. This allows the acquisition of the vibration sequence of the vehicle point cloud height over time.

[0042] In this embodiment of the invention, the ground reference plane and vehicle contour point cloud can be segmented from the acquired vehicle point cloud using algorithms such as RANSAC (Random Sample Consensus), clustering algorithms, and semantic segmentation algorithms.

[0043] In this embodiment of the invention, noise can also be eliminated from the vehicle contour point cloud through denoising. This improves the quality of the vehicle contour point cloud, enhances the accuracy of subsequent calculations, improves the accuracy of vehicle weight prediction, and improves the accuracy of vehicle overload determination. Statistical filtering algorithms, radius filtering algorithms, voxel grid filtering algorithms, etc., can be used for vehicle contour point cloud denoising, but are not limited to these; other algorithms for point cloud denoising can also be employed.

[0044] In this embodiment of the invention, for vehicle wheel point clouds, inter-frame registration can be used to eliminate noise between consecutive frame point cloud sequences, thereby eliminating small displacement and rotation errors during vehicle movement, improving the accuracy of subsequent calculations, the accuracy of vehicle weight prediction, and the accuracy of vehicle overload determination. When performing inter-frame registration of consecutive frame point cloud sequences, the ICP (Iterative Closest Point) algorithm can be used, but it is not limited to this; other algorithms suitable for inter-frame registration of point cloud sequences can also be employed.

[0045] S220. Based on the vehicle contour point cloud and the ground reference plane, determine the static subsidence and total tire ground contact area in the static information.

[0046] Static information may include static subsidence and total tire contact area.

[0047] In some embodiments of the present invention, the static sinking of a vehicle can be determined based on the height of the vehicle contour point cloud relative to a ground reference plane.

[0048] Specifically, when determining the static sinking of a vehicle, the overall profile height can be calculated by analyzing the average height of the point clouds representing key components of the vehicle's profile height relative to a reference plane. This calculated profile height is then compared to the standard height of the corresponding vehicle model to obtain the static sinking. Key components can include the vehicle roof, chassis, front, and rear. The vehicle model can be determined based on vehicle type recognition of the acquired vehicle images. For example, by calculating the average height of the point cloud representing the vehicle roof relative to the ground reference plane, the average height of the point cloud representing the vehicle chassis relative to the ground reference plane, and the average height of the point cloud representing the front of the vehicle relative to the ground reference plane, the overall profile height of the vehicle is obtained. This overall profile height is then compared to the standard height of the corresponding vehicle model to obtain the static sinking. .

[0049] In this embodiment of the invention, a database can also be constructed to store basic parameter information of various vehicle models. The basic parameter information may include basic parameter information such as vehicle profile height, vehicle suspension stiffness, vehicle rated weight, vehicle standard tire aspect ratio, and standard tire pressure, but is not limited to these, and may also include other basic parameter information of the vehicle.

[0050] In this embodiment of the invention, to improve calculation speed while ensuring accuracy when determining the static sinking of a vehicle, multiple key point points can be selected from the point cloud of key parts of the vehicle for static sinking calculation. For example, the overall height profile of the vehicle can be calculated by calculating the average height of the top of the vehicle point cloud and key points of the chassis (such as the centerline of the carriage and the position of the main beam) relative to the ground reference plane. The static sinking is then obtained by comparing it with the standard height of the corresponding vehicle model in the database. h.

[0051] In some embodiments of the present invention, a cluster of vehicle tire point clouds that is in contact with a ground reference plane is extracted from the vehicle contour point cloud; based on the extracted cluster of vehicle tire point clouds and the ground reference plane, the total tire ground contact area in the static information is determined.

[0052] Specifically, when determining the total tire contact area, the contact area of ​​each tire of the vehicle can be calculated, and then the total tire contact area can be determined based on the sum of the contact areas of all tires of the vehicle.

[0053] In this embodiment of the invention, when determining the contact area of ​​each tire, point cloud clusters of the area where the bottom of the vehicle tire contacts the ground can be identified and extracted. By projecting these clusters onto the ground plane and calculating their convex hull area, the contact area of ​​each tire can be estimated, and finally, the total contact area of ​​the tires can be obtained by summing them up. .

[0054] S230. Obtain the vehicle point cloud sequence after the vehicle passes through the vibration detection area, and determine the vibration sequence of the vehicle height changing over time based on the vehicle point cloud sequence.

[0055] Vibration information includes a vibration sequence showing the vehicle's height changing over time. Specifically, a low (e.g., 1-2 cm) gentle speed bump can be set up on the road in the vibration detection area, and the sequence of point cloud height changes over time after the vehicle passes over the speed bump can be extracted. This refers to the vibration sequence, which can be used as the input signal for subsequent vehicle dynamic vibration analysis.

[0056] This invention provides a technical solution that involves acquiring vehicle point clouds while the vehicle is traveling in a flat detection area, determining a ground reference plane and a vehicle contour point cloud based on these points, determining the static sinking and total tire contact area in the static information based on the vehicle contour point cloud and the ground reference plane, and acquiring a vehicle point cloud sequence after the vehicle has passed through a vibration detection area, and determining the vibration sequence of the vehicle height changing over time based on this sequence. This approach can accurately determine the vehicle's static sinking, total tire contact area, and vibration sequence of the vehicle height changing over time, facilitating accurate prediction of vehicle weight and accurate determination of vehicle overload.

[0057] In some embodiments of the present invention, such as Figure 3 As shown, the final predicted weight of the vehicle can be determined according to steps S310~S340.

[0058] S310. Based on the static subsidence, determine the first predicted weight of the vehicle.

[0059] In this embodiment of the invention, the suspension stiffness of the vehicle can be obtained according to the type of the vehicle, and the first predicted weight of the vehicle can be determined based on the static sag and the suspension stiffness.

[0060] Specifically, a side view image of the vehicle can be acquired, and vehicle model recognition can be performed on the image to obtain the vehicle model. Then, based on the identified vehicle model, the suspension stiffness of the vehicle can be obtained. By calculating the product of the static sag and the suspension stiffness, the first predicted weight of the vehicle can be obtained, thereby improving the accuracy of vehicle weight prediction and, consequently, improving the accuracy of subsequent overload determination based on the predicted vehicle weight.

[0061] S320. Based on the total tire contact area, determine the second predicted weight of the vehicle.

[0062] In some embodiments of the present invention, the standard tire pressure of a vehicle can be obtained according to the type of vehicle, and a second predicted weight of the vehicle can be determined based on the total tire contact area and the standard tire pressure.

[0063] Specifically, a side view image of the vehicle can be acquired, and vehicle model recognition can be performed on the image to obtain the vehicle model. Then, based on the identified vehicle model, the standard tire pressure is obtained, and the second predicted weight of the vehicle is calculated by multiplying the total tire contact area by the standard tire pressure. In this embodiment, the vehicle weight is predicted using the total tire contact area and standard tire pressure, thereby improving the accuracy of vehicle prediction and, consequently, the accuracy of subsequent overload determination based on the predicted vehicle weight.

[0064] The standard tire pressure is the tire pressure measured when the vehicle is unloaded. Therefore, when using the standard tire pressure to predict the weight of a non-unloaded vehicle, the predicted result will deviate from the actual vehicle weight. To eliminate this deviation, in some embodiments of the present invention, the standard tire pressure can be corrected based on the vehicle's tire load information, and the second predicted weight of the vehicle is obtained by calculating the product of the corrected tire pressure and the total tire contact area. In this embodiment, the vehicle weight is predicted using the total tire contact area and the corrected tire pressure, thereby further improving the accuracy of vehicle prediction and, consequently, the accuracy of subsequent overload determination based on the predicted vehicle weight.

[0065] In some embodiments of the present invention, such as Figure 4 As shown, the second predicted weight of the vehicle can be determined through steps S321 to S324.

[0066] S321. Obtain the vehicle's standard tire aspect ratio and tire sidewall deformation information.

[0067] Specifically, the standard tire aspect ratio of a vehicle can be obtained as follows: acquire a side image of the vehicle, perform vehicle model recognition on the side image to obtain the vehicle model number, and then obtain the standard tire aspect ratio based on the identified vehicle model number.

[0068] The specific method for obtaining tire sidewall deformation information of a vehicle can be as follows: acquire a side image of the vehicle and perform target recognition on the side image to identify the image area of ​​each tire. In the tire image area, measure the vertical height of the tire sidewall (i.e., the vertical distance from the edge of the wheel hub to the ground) and the maximum horizontal width of the tire. The vertical height of the tire sidewall and the maximum horizontal width of the tire are the tire sidewall deformation information to be obtained.

[0069] S322. Determine the actual aspect ratio of the tire based on the tire sidewall deformation information.

[0070] Specifically, tire sidewall deformation information includes the vertical height of the tire sidewall and the maximum horizontal width of the tire. Therefore, the actual aspect ratio of the tire can be determined based on the vertical height of the tire sidewall and the maximum horizontal width of the tire. More specifically, the actual aspect ratio of the tire can be determined based on the ratio of the vertical height of the tire sidewall to the maximum horizontal width of the tire.

[0071] S323. Determine the tire pressure correction factor based on the actual aspect ratio of the tire and the standard tire aspect ratio.

[0072] Specifically, the tire pressure correction factor is determined based on the ratio of the standard tire aspect ratio to the actual tire aspect ratio.

[0073] S324. Determine the second predicted weight of the vehicle based on the tire pressure correction factor, standard tire pressure, and total tire contact area.

[0074] Specifically, the second predicted weight of the vehicle can be obtained by multiplying the tire pressure correction factor, the standard tire pressure, and the total tire contact area.

[0075] In this embodiment of the invention, by acquiring the standard tire aspect ratio and tire sidewall deformation information of the vehicle; determining the actual tire aspect ratio based on the tire sidewall deformation information; determining the tire pressure correction factor based on the actual tire aspect ratio and the standard tire aspect ratio; and determining the second predicted weight of the vehicle based on the tire pressure correction factor, the standard tire pressure, and the total tire contact area, the vehicle weight prediction is realized, thereby improving the accuracy of vehicle prediction and thus improving the accuracy of subsequent overload determination based on the predicted vehicle weight.

[0076] S330. Based on the vibration sequence, determine the third predicted weight of the vehicle.

[0077] Specifically, when a vehicle travels over a bumpy road, it will generate vertical vibrations. After traveling over a bumpy road, it will enter a free vibration state. By obtaining the point cloud-time vibration sequence of the vehicle's height after traveling over a bumpy road, the vehicle's third predicted weight can be accurately predicted.

[0078] In some embodiments of the present invention, the damping ratio when the vehicle vibrates can be determined based on the vibration sequence; the natural frequency when the vehicle vibrates can be determined based on the damping ratio; and the third predicted weight of the vehicle can be determined based on the natural frequency and the suspension stiffness of the vehicle.

[0079] Specifically, bandpass filtering can be applied to the vibration sequence to eliminate high-frequency noise and low-frequency drift, resulting in a clear attenuated vibration waveform. Subsequently, the amplitudes of two consecutive vibration peaks are automatically acquired on the waveform to determine the attenuation rate of the vehicle under vibration. These two consecutive vibration peaks may or may not contain other peaks. In some embodiments, to improve the accuracy and robustness of the calculation, the amplitudes of two peaks spaced n periods apart can be acquired on the vibration waveform, denoted as... and This is used to calculate the vehicle's attenuation rate under vibration. The logarithmic attenuation rate is an indicator of the rate at which vibration energy decays; therefore, it can be calculated based on the amplitudes of two wave peaks spaced n periods apart. The damping ratio was calculated based on the logarithmic decay rate. , For vehicle suspension systems, low damping is typically used. (<0.3), therefore an approximate formula can also be used: This represents the vehicle's damping ratio. Then, based on the sequence... By performing spectral analysis using FFT (Fast Fourier Transform), the frequency point with the highest energy can be found, which is the damped natural frequency of the vehicle. and the damped natural frequency Convert to angular frequency , Then, the frequency is corrected using the damping ratio, resulting in the undamped natural frequency ( ) and damped natural frequency ( There are conversion relationships between them. Since the undamped natural frequency can be obtained .

[0080] The vertical vibration (bounce mode) of a vehicle traversing a bumpy road is precisely modeled as a single-degree-of-freedom "mass-spring-damped" system. Its physical motion follows the following second-order ordinary differential equation:

[0081] in, This refers to the equivalent mass, i.e., the sprung mass of the vehicle. This represents the equivalent stiffness, i.e., the stiffness of the vehicle suspension. , This represents the damping coefficient, the physical damping coefficient of the vibration damper, measured in N·s / m, and signifies its energy dissipation capacity. External stimuli, namely the instantaneous impact force generated by road bumps (such as pre-set speed bumps) on the vehicle, are the triggers for this vibration.

[0082] Once the vehicle overcomes the bumps and enters the free vibration phase, the external excitation F(t) disappears, and the equation simplifies to a homogeneous equation:

[0083] Undamped Situation:

[0084] Let the solution be Substituting, we get:

[0085] Introducing Undamped Natural Frequency Then the vehicle mass .

[0086] Undamped natural frequency Substituting the formula for calculating weight into the formula for calculating weight, we can obtain: .

[0087] in, The third predicted weight of the vehicle is represented by g, which represents the acceleration due to weight.

[0088] This embodiment obtains the natural frequency of the vehicle when it vibrates by collecting and processing the vehicle vibration signal. Based on the natural frequency and the vehicle's suspension stiffness, it achieves accurate prediction of the vehicle's third predicted weight, thereby improving the accuracy of subsequent overload determination based on the predicted vehicle weight.

[0089] S340. Based on the first predicted weight, the second predicted weight, and the third predicted weight, determine the final predicted weight of the vehicle.

[0090] In this embodiment of the invention, weights can be assigned to the first predicted weight, the second predicted weight, and the third predicted weight respectively, and the final predicted weight of the vehicle can be determined based on the first predicted weight, the second predicted weight, and the third predicted weight and their respective weights.

[0091] Specifically, assume that the weight assigned to the first predicted weight is the first weight, the weight assigned to the second predicted weight is the second weight, and the weight assigned to the third predicted weight is the third weight, and the sum of the first weight, the second weight, and the third weight equals 1. Then, the final predicted weight of the vehicle can be obtained by summing the results of the calculation: first predicted weight × first weight + second predicted weight × second weight + third predicted weight × thirteenth weight.

[0092] This invention provides a technical solution for predicting vehicle weight by analyzing a point cloud-time vibration sequence showing the changes in vehicle height over time based on three dimensions: static sinking, tire contact area, and vehicle height after vibration. By assigning corresponding weights to the predicted weights from these three dimensions, a fused vehicle weight is obtained. This achieves accurate vehicle weight prediction, improves the adaptability of vehicle overload detection to different environments, enhances the accuracy of vehicle weight prediction, and reduces the false positive rate.

[0093] In some embodiments of the present invention, the first weight, the second weight, and the third weight may be the same, and the sum of the first weight, the second weight, and the third weight is equal to 1.

[0094] In some embodiments of the present invention, the first weight, second weight, and third weight may be different, and the sum of the different first weight, second weight, and third weight is equal to 1. In this case, the weights assigned to the first predicted weight, second predicted weight, and third predicted weight can be determined based on the vehicle's speed, the type of road surface area the vehicle is traveling on, and weather conditions. For example, when the vehicle speed is low (e.g., below 30 km / h), the static sinking is more accurate, so the first weight is increased. When the vehicle speed is high, the dynamic vibration characteristics are more pronounced, so the second weight is increased. When the vehicle passes through a road section with standard speed bumps, the second weight is increased. When the vehicle is traveling on an extremely flat road surface, the second weight is decreased. When encountering rain or snow, the quality of ground point clouds may decrease, affecting the accuracy of the calculation of the total tire contact area, so the third weight is appropriately decreased.

[0095] In this embodiment of the invention, when it is necessary to adjust the first weight, the second weight, and the third weight, the adjustment can be made within the range of 1 / 3 to 1 / 2 to ensure the accuracy of vehicle weight prediction.

[0096] In this embodiment of the invention, the first weight, the second weight, and the third weight are adaptively adjusted according to the actual driving conditions of the vehicle, which further improves the adaptability of vehicle overload detection to different environments and the accuracy of vehicle weight prediction, thereby improving the accuracy of vehicle overload detection and reducing the false judgment rate.

[0097] The following specific embodiment illustrates the process of predicting the final weight of a vehicle. It should be understood that the following embodiment is for illustrative purposes only and is not intended to limit the invention.

[0098] First, a static suspension model, a tire ground pressure model, and a dynamic vibration model are constructed.

[0099] The static suspension model can be represented as:

[0100] This indicates the static subsidence of the vehicle. This indicates the suspension stiffness.

[0101] The tire contact pressure model can be expressed as:

[0102] in, This indicates the second predicted weight of the vehicle. This indicates the total contact area of ​​the vehicle's tires with the ground. Indicates standard tire pressure. This indicates the standard tire aspect ratio. This indicates the actual aspect ratio of the vehicle's tires, where... , in, Indicates the maximum horizontal width of the tire. This indicates the vertical height of the tire sidewall.

[0103] The dynamic vibration model can be represented as:

[0104] in, This indicates the vehicle's third predicted weight. Indicates suspension stiffness. Indicates the damping ratio. Represents gravitational acceleration. This indicates the damped natural frequency of the vehicle, expressed as angular frequency.

[0105] in, , This indicates a damped natural frequency, based on the sequence. By performing spectral analysis using FFT (Fast Fourier Transform), the frequency point with the highest energy is obtained. .

[0106] in, , Represents the logarithmic decay rate. ,in, express The amplitude of the i-th peak on the waveform, express The amplitude of the (i+n)th peak on the waveform.

[0107] Subsequently, the vehicle weight was calculated in parallel using the three physical models mentioned above to improve the robustness of the results.

[0108] Specifically, based on the static deflection and suspension stiffness, a static suspension model is derived to obtain the vehicle's first predicted weight. The actual aspect ratio of the tires is calculated based on the vertical height of the tire sidewalls and the maximum horizontal width of the tires. Using the actual aspect ratio, standard tire aspect ratio, standard tire pressure, and total tire contact area, a tire contact model is derived to obtain the vehicle's second predicted weight. Based on the vibration sequence... The damping ratio and the vehicle's damped natural frequency in terms of angular frequency are obtained. Based on the suspension stiffness, damping ratio, gravitational acceleration, and the vehicle's damped natural frequency in terms of angular frequency, a dynamic vibration model is obtained, and the vehicle's third predicted weight is obtained.

[0109] Subsequently, the final predicted weight of the vehicle is determined based on the first, second, and third predicted weights. Among them, the final predicted weight It can be represented as:

[0110] in, These represent the first weight, the second weight, and the third weight, respectively. ,in, It can be dynamically adjusted according to real-time operating conditions.

[0111] This invention provides a technical solution for vehicle weight fusion by constructing a static suspension model, a tire ground pressure model, and a dynamic vibration model, and predicting vehicle weight for each model. The predicted weights from the three models are then assigned corresponding weights. This achieves accurate vehicle weight prediction, improves the adaptability of vehicle overload detection to different environments, enhances the accuracy of vehicle weight prediction, and reduces the false positive rate.

[0112] In some embodiments of the present invention, when determining vehicle overload based on the vehicle's final predicted weight and its rated weight, the rated weight of the vehicle can be obtained based on the identified license plate number, and then determined based on the final predicted weight. and rated weight Calculate overload rate The formula for calculating the overload rate is: .

[0113] Overload thresholds set according to national or local regulations (If it exceeds 10% of the rated total mass) and the calculated overload rate are used for overload judgment. > If the vehicle is found to be overloaded, its information (license plate, image, estimated weight, overload rate) will be recorded and reported to the management platform.

[0114] This invention, in its embodiments, obtains the vehicle's rated weight based on the identified license plate number and then calculates the final predicted weight. and rated weight Calculate overload rate and overload threshold The system uses the calculated overload rate (e.g., exceeding 10% of the rated total mass) to determine overload, thus improving the accuracy of vehicle overload determination.

[0115] In this embodiment of the invention, the standard parameters used to predict vehicle weight can also be updated through a closed-loop feedback update mechanism to ensure the accuracy of weight prediction and overload determination.

[0116] In some embodiments of the present invention, the basic parameter information and weights of the vehicle can be updated based on the vehicle's final predicted weight and the vehicle's actual weight.

[0117] Specifically, vehicles that have undergone inspection can be periodically guided to a nearby weighbridge for accurate weighing to obtain their true weight. The final predicted weight will be... Compared to actual weight Compare the data and use it to assign weights to the vehicles. and basic parameter information (such as) The values ​​of the weights are optimized, and the final predicted weight is calculated based on the optimized weights and basic parameter information, thereby improving the accuracy of vehicle weight prediction.

[0118] In this embodiment of the invention, the weight is the weight used to calculate the final predicted weight of the vehicle, and the basic parameter information is the basic parameter information of the vehicle model. Therefore, the weight can be correlated with the vehicle model, that is, the vehicle model, the basic parameter information, and the weight can correspond. In a specific embodiment, the final predicted weight of vehicles under different vehicle models can be compared. Compared to actual weight Based on the comparison results (e.g., the deviation between the two is greater than a threshold), the basic parameter information and weights corresponding to each vehicle model are optimized and iteratively updated in batches, thereby improving the accuracy of vehicle weight prediction for each vehicle model.

[0119] This invention implements a weighting system by linking the vehicle's final predicted weight with the vehicle's actual weight measured by the weighbridge. The updating of basic parameter information enables the vehicle prediction process to self-iterate and optimize, maintaining a high level of detection over a long period, resulting in more accurate vehicle weight prediction results.

[0120] In some embodiments of the present invention, a system deviation can be determined based on the vehicle's historical predicted weight, and in response to the existence of a system deviation, the basic parameter information and weights corresponding to the vehicle can be corrected.

[0121] Specifically, systematic deviation refers to the average deviation between the vehicle's historical predicted weight and its actual weight. When determining systematic deviation based on the vehicle's historical predicted weight and actual weight, this can be done by comparing the average deviation with a set deviation threshold (this threshold can be a value close to 0, such as 0.1, 0.2, 0.3, 0.4, 0.5, or 0.6). If the average deviation is greater than the set deviation threshold, a systematic deviation is determined to exist; if the average deviation is not greater than the set deviation threshold, no systematic deviation is determined to exist. When a systematic deviation is determined to exist, the basic parameter information and weights corresponding to the vehicle can be corrected. In some specific embodiments, as mentioned above, the basic parameter information and weights corresponding to the vehicle refer to the basic parameter information and weights corresponding to the vehicle model to which the vehicle belongs.

[0122] In some specific embodiments, the historical predicted weight of vehicles under each vehicle model can be continuously analyzed. If a systematic deviation in the predicted weight of a certain vehicle model is found over a long period, the weight and basic parameter information of that vehicle model (such as...) are triggered. , Suggestions for correcting the value are updated after manual review.

[0123] In some specific embodiments, the records of the last 100 times a vehicle of a certain model passes through the weighbridge can be collected as the actual weight of the vehicle. The error between the final predicted weight of the vehicle and the actual weight of the vehicle can be calculated. If the error is close to 0, it means that there is no systematic error. If the error is not close to 0 and is a stable positive value (e.g., 500 kg larger on average) and has a small standard deviation, it means that it is not a random fluctuation but caused by a systematic error, i.e., a systematic error exists.

[0124] In some specific embodiments, when a system deviation is determined to exist and the basic parameter information and weights corresponding to the vehicle are corrected, the basic parameter information and weights can be corrected according to the actual situation. The correction process is described below through specific embodiments.

[0125] Example 1: Weight correction.

[0126] The final predicted weight calculation formula is as follows:

[0127] In the above formula, the weights assigned to the static suspension model, the tire contact model, and the dynamic vibration model are... , , The initial weights may be set based on experience (e.g., 0.4, 0.3, 0.3), but historical data may reveal that the vehicle weight predicted by the tire contact model is closer to the actual vehicle weight than that predicted by the static suspension model. Therefore, it is advisable to appropriately increase the weights. The value of .

[0128] Example 2: Suspension Stiffness Corrections.

[0129] In the above formulas, the expressions for the static suspension model and the dynamic vibration model are as follows:

[0130]

[0131] As can be seen, in the embodiments of the present invention, suspension stiffness is used when predicting vehicle weight using both the static suspension model and the dynamic vibration model. If the vehicle weight predicted by the static suspension model and / or dynamic vibration model is consistently lower than the actual weight over a long period, and the deviation is stable, it should be recommended to increase the weight of that vehicle model. .

[0132] Example 3: Standard Tire Pressure Corrections.

[0133] In real-world applications, a vehicle's tire pressure may be higher than its standard tire pressure when unloaded. This results in a smaller tire contact patch, leading to an underestimation of the vehicle weight calculated based on this patch. Therefore, if the vehicle weight predicted by the tire contact patch model is consistently lower than the actual weight over a long period, and the deviation is stable, it is recommended to increase the standard tire pressure for that vehicle model. .

[0134] This invention analyzes historical vehicle detection data to update weights and basic parameter information, enabling the vehicle prediction process to self-iterate and optimize, maintain a high detection level over a long period, and make the vehicle weight prediction results more accurate.

[0135] Based on the same inventive concept, according to another aspect of the present invention, embodiments of the present invention also provide a vehicle overload detection system, such as... Figure 5As shown, the vehicle overload detection system 10 includes a vehicle information acquisition device 11 and a terminal electronic device 12 connected to the vehicle information acquisition device 11. The vehicle information acquisition device 11 is used to acquire vehicle point clouds, and the terminal electronic device 12 is configured to execute the vehicle overload detection method as described in any of the above embodiments.

[0136] In this embodiment of the invention, vehicle point clouds can be collected by vehicle information acquisition device 11 and sent to terminal electronic device 12 for processing. Terminal electronic device 12 determines the vehicle's static and vibration information based on the vehicle point clouds, and determines the vehicle's final predicted weight based on the static and vibration information. Based on the final predicted weight and the vehicle's rated weight, it makes a vehicle overload judgment, thereby improving the adaptability of vehicle overload detection to different environments and the accuracy of vehicle weight prediction, thus improving the accuracy of vehicle overload detection and reducing the false judgment rate.

[0137] The vehicle overload detection system provided in this embodiment of the invention can accurately estimate the total weight of a vehicle and identify overload without contacting the vehicle or damaging the road structure.

[0138] In some embodiments of the present invention, the vehicle information acquisition device 11 may include a LiDAR and an industrial camera. Specifically, a LiDAR (using a 32-line or higher LiDAR to ensure point cloud density) can be installed on a gantry frame on the side or directly above the road to accurately capture 3D point cloud data as the vehicle passes. Two industrial cameras are installed at the same location: one to capture the full side view of the vehicle and identify the vehicle type and tire shape; the other to capture the front or rear of the vehicle, specifically for identifying license plate information. The LiDAR and industrial camera are connected to an edge computing terminal via a network interface to achieve synchronous triggering and local processing of data acquisition.

[0139] The LiDAR acts as a trigger; when it detects an object point cloud within a designated area, the system determines that a vehicle has entered. The system immediately activates the LiDAR and all cameras to simultaneously acquire data. The LiDAR continuously outputs a complete point cloud sequence of the vehicle from its entry into the detection area to its departure, while the cameras simultaneously capture image sequences containing clear license plates and vehicle bodies. The system creates a unique identifier for each passing vehicle and binds and stores the corresponding point cloud sequence and image sequence to ensure data consistency.

[0140] This invention, through the fusion of multi-source data from lidar and cameras, predicts vehicle weight, effectively overcoming the problem of single sensors being greatly affected by the environment and improving the system's adaptability to different environments.

[0141] Finally, it should be noted that those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium for the program can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc. The above computer program embodiments can achieve the same or similar effects as any of the corresponding foregoing method embodiments.

[0142] Those skilled in the art will also understand that the various exemplary logic blocks, modules, circuits, and algorithm steps described in conjunction with the disclosure herein can be implemented as electronic hardware, computer software, or a combination of both. To clearly illustrate this interchangeability between hardware and software, the functionality of various illustrative components, blocks, modules, circuits, and steps has been generally described. Whether this functionality is implemented as software or as hardware depends on the specific application and the design constraints imposed on the system as a whole. Those skilled in the art can implement the functionality in various ways for each specific application, but such implementation decisions should not be construed as departing from the scope of the embodiments disclosed herein.

[0143] The above are exemplary embodiments disclosed in this invention. However, it should be noted that various changes and modifications can be made without departing from the scope of the embodiments of this invention as defined by the claims. The functions, steps, and / or actions of the methods according to the disclosed embodiments described herein do not need to be performed in any particular order. The sequence numbers of the disclosed embodiments of this invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments. Furthermore, although the elements disclosed in the embodiments of this invention may be described or claimed individually, they may be understood as multiple unless explicitly limited to a singular number.

[0144] It should be understood that, as used herein, the singular form “a” is intended to include the plural form as well, unless the context clearly supports an exception. It should also be understood that, as used herein, “and / or” refers to any and all possible combinations of one or more of the associated listed items.

[0145] Those skilled in the art should understand that the discussion of any of the above embodiments is merely exemplary and is not intended to imply that the scope of the invention (including the claims) is limited to these examples. Within the framework of the invention, technical features of the above embodiments or different embodiments can be combined, and many other variations of different aspects of the invention exist, which are not provided in the details for the sake of brevity. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the invention should be included within the protection scope of the invention.

Claims

1. A method for detecting vehicle overload, characterized in that, include: Acquire vehicle point cloud data, and determine the static and vibration information of the vehicle based on the vehicle point cloud data; Based on the static information and the vibration information, the final predicted weight of the vehicle is determined; The vehicle overload determination is based on the vehicle's final predicted weight and the vehicle's rated weight.

2. The method according to claim 1, characterized in that, The static information includes static subsidence and total tire contact area; the vibration information includes a vibration sequence of vehicle height over time. Based on the static information and the vibration information, determining the final predicted weight of the vehicle includes: Based on the static subsidence, the first predicted weight of the vehicle is determined; Based on the total tire contact area, the second predicted weight of the vehicle is determined; Based on the vibration sequence, the third predicted weight of the vehicle is determined; The final predicted weight of the vehicle is determined based on the first predicted weight, the second predicted weight, and the third predicted weight.

3. The method according to claim 2, characterized in that, Based on the first predicted weight, the second predicted weight, and the third predicted weight, the final predicted weight of the vehicle is determined by: Based on the vehicle's speed, the type of road surface the vehicle is traveling on, and weather conditions, the weights assigned to the first predicted weight, the second predicted weight, and the third predicted weight are determined respectively. The final predicted weight of the vehicle is determined based on the first predicted weight, the second predicted weight, and the third predicted weight, and their respective weights.

4. The method according to claim 2, characterized in that, Before determining the static and vibration information of the vehicle based on the vehicle point cloud, the method further includes: acquiring a side image of the vehicle and determining the vehicle model based on the side image of the vehicle. Determining the first predicted weight of the vehicle based on the static subsidence includes: The suspension stiffness of the vehicle is obtained based on the vehicle model, and the first predicted weight of the vehicle is determined based on the static sag and the suspension stiffness.

5. The method according to claim 4, characterized in that, Determining the second predicted weight of the vehicle based on the total tire contact area includes: The standard tire pressure of the vehicle is obtained based on the vehicle model, and the second predicted weight of the vehicle is determined based on the total tire contact area and the standard tire pressure.

6. The method according to claim 5, characterized in that, After acquiring the vehicle side image, the method further includes: determining the tire sidewall deformation information of the vehicle based on the vehicle side image; Determining the second predicted weight of the vehicle based on the total tire contact area and the standard tire pressure includes: The standard tire aspect ratio of the vehicle is obtained based on the vehicle model. Based on the tire sidewall deformation information, the actual aspect ratio of the tire is determined; The tire pressure correction factor is determined based on the actual aspect ratio of the tire and the standard tire aspect ratio. The second predicted weight of the vehicle is determined based on the tire pressure correction factor, the standard tire pressure, and the total tire contact area.

7. The method according to claim 4, characterized in that, Based on the vibration sequence, determining the third predicted weight of the vehicle includes: Based on the vibration sequence, the damping ratio when the vehicle vibrates is determined; Based on the damping ratio, the natural frequency of the vehicle when it vibrates is determined; The third predicted weight of the vehicle is determined based on the natural frequency and the vehicle's suspension stiffness.

8. The method according to claim 2, characterized in that, Acquiring vehicle point clouds and determining the vehicle's static and vibration information based on these point clouds includes: Acquire vehicle point cloud data while the vehicle is traveling in a flat detection area, and determine the ground reference plane and vehicle contour point cloud based on the vehicle point cloud data. Based on the vehicle contour point cloud and the ground reference plane, determine the static subsidence and total tire contact area in the static information; The vehicle point cloud sequence after the vehicle passes through the vibration detection area is obtained, and based on the vehicle point cloud sequence, the vibration sequence of the vehicle height changing over time is determined. The vibration detection area is equipped with a road surface excitation device.

9. The method according to claim 8, characterized in that, Based on the vehicle contour point cloud and the ground reference plane, determining the static subsidence and total tire contact area in the static information includes: Based on the height of the vehicle contour point cloud relative to the ground reference plane, determine the static subsidence amount in the static information; Extract the vehicle tire point cloud clusters that are in contact with the ground reference plane from the vehicle outline point cloud; Based on the extracted vehicle tire point cloud clusters and the ground reference plane, the total tire ground contact area in the static information is determined.

10. The method according to claim 1, characterized in that, Also includes: Based on the vehicle's historical predicted weight, a system deviation is determined, and in response to the existence of a system deviation, the basic parameter information and weights corresponding to the vehicle are corrected.

11. A vehicle overload detection system, characterized in that, The device includes a vehicle information collection device and a terminal electronic device connected to the vehicle information collection device. The vehicle information collection device is used to collect vehicle point clouds, and the terminal electronic device is configured to perform the vehicle overload detection method as described in any one of claims 1 to 10.

Images