Method and system for cooperatively monitoring illegal behaviors of mud head vehicle based on ai driving

Abstract

The invention provides an ai-driven mud head vehicle illegal behavior cooperative monitoring method and system, and the method comprises the steps: collecting the three-dimensional acceleration and three-dimensional angular velocity data of a mud head vehicle through a vehicle-mounted inertial sensor, and constructing an inertial behavior sequence; establishing an individualized behavior model based on historical normal behavior data, identifying an abnormal behavior period deviating from a normal state, and generating an abnormal score; when the abnormal behavior exists, a vehicle-mounted camera is triggered to collect an environment image, and the legality of the scene where the mud-head vehicle is located is judged through a visual semantic recognition model; and fusing the abnormal score and the scene legality mark, and generating a final violation judgment result of the violation behavior through a joint judgment model. And according to the violation judgment result, triggering a multi-stage linkage response mechanism, including vehicle-mounted real-time feedback, regional monitoring equipment linkage and violation data uploading and evidence storage.

Description

Technical Field

[0001] The present invention belongs to the field of AI drive, and in particular relates to a collaborative monitoring method and system for illegal behaviors of dump trucks based on AI drive. Background Art

[0002] Dump trucks, common large-scale transport vehicles in urban infrastructure construction, are responsible for transporting construction materials such as slag, sand, and gravel. However, due to their large load capacities, frequent operations, and widespread range, and their frequent use in areas with weak oversight, such as the periphery of urban construction and during nighttime construction, dump trucks are prone to becoming a high-incidence source of traffic violations and safety hazards. In practice, dump trucks frequently engage in illegal activities, such as dumping construction waste in non-designated areas, overloading, illegally detouring around controlled routes, driving through restricted areas, and illegally parking. These behaviors not only disrupt the urban environment but also pose a serious threat to road traffic safety.

[0003] Current mainstream methods for monitoring dump trucks typically rely on fixed video surveillance equipment deployed on urban roads or around construction sites, RFID-based time-clocking systems, GPS trajectory analysis systems, and manual inspections. However, these methods often suffer from significant limitations. For one thing, fixed cameras have limited coverage and cannot cover the entire transport route. This is particularly true in remote areas, at night, or in construction blind spots, creating visual blind spots that make it difficult to detect key violations. Furthermore, traditional GPS route analysis lacks accuracy, making it impossible to accurately identify whether a vehicle has actually entered the construction site or unloaded within a prohibited area. Furthermore, some drivers circumvent geofencing or camera recognition by clocking in and leaving immediately or skirting the boundaries of the site. This renders the monitoring system virtually helpless when faced with behaviors that appear to be formally compliant but substantively illegal. Furthermore, existing systems often construct analysis logic from a single-vehicle perspective, lacking the ability to model and identify coordinated violations by multiple vehicles. This makes it difficult to identify highly concealed and regular behaviors such as "cluster detours" and "group dumping."

[0004] In this context, the industry urgently needs a new technology system that can break through the limitations of existing visual / positional information reliance, possess strong environmental adaptability, robust behavior judgment, and support vehicle-to-vehicle / vehicle-road collaborative perception. At the same time, this system must not only be able to identify high-risk behaviors of individual vehicles, but also be able to perceive the semantic context in which the behavior occurs, determining whether it constitutes a true violation of "behavioral violations + illegal scenes," thereby improving the accuracy of judgments and the credibility of law enforcement. In summary, there is still a lack of a technical solution that can effectively integrate multi-source perception, scene semantic understanding, and collaborative analysis of group behavior to support the intelligent monitoring and governance of illegal dump truck behavior. Summary of the Invention

[0005] The purpose of this invention is to propose an AI-driven collaborative monitoring method and system for dump truck violations. This system aims to overcome "perception blind spots" and "formal compliance disguises" by constructing an intelligent recognition framework capable of modeling physical behavior and understanding scene semantics, thereby achieving a joint assessment of the actual behavior of dump trucks and the environment in which they occur.

[0006] In order to achieve the above objectives, a first aspect of the present invention provides an AI-driven collaborative monitoring method for illegal behaviors of dump trucks, the method comprising the following steps:

[0007] S1. 3D acceleration and 3D angular velocity data of the dump truck is collected through on-board inertial sensors to construct an inertial behavior sequence. Based on historical normal behavior data, an individualized behavior model is established to identify abnormal behavior periods that deviate from the norm and generate an anomaly score.

[0008] S2. When abnormal behavior occurs, the vehicle-mounted camera is triggered to collect environmental images and the legality of the scene in which the dump truck is located is determined through the visual semantic recognition model;

[0009] S3. Fusion of the anomaly score and the scenario legitimacy tag to generate a final violation determination result through a joint determination model;

[0010] S4. Based on the violation determination result, a multi-level linkage response mechanism is triggered, including real-time feedback from the vehicle, linkage with regional monitoring equipment, and uploading of violation data for evidence storage.

[0011] Furthermore, the individualized behavior model is obtained by training a Gaussian mixture model, and the anomaly score is determined by calculating the deviation of the current behavior sequence in the historical distribution; the deviation is determined based on the negative log-likelihood value;

[0012] If the deviation is higher than the mean score of normal behavior, it indicates that the behavior deviates greatly from the norm and is thus marked as a possible abnormal behavior;

[0013] Compare the deviation with the mean of historical scores to determine whether it is an abnormal period:

[0014] If the deviation is greater than the sum of the mean of the historical scores and the preset threshold, the current period is marked as an abnormal behavior period.

[0015] Furthermore, the input of the visual semantic recognition model is the environment image, the output scene labels include construction sites, no-dumping zones and roads, and the environment legality labels are dynamically generated based on the scene labels.

[0016] Furthermore, the environmental legitimacy mark is dynamically generated based on the scene label, specifically including:

[0017] Define the scene legality mark. If the construction site is a no-dumping zone, the scene legality mark is illegal, indicating that the dump truck's behavior occurs in the no-dumping zone and violates relevant regulations; if the construction site is a construction site, the scene legality mark is legal, indicating that the area is a legal unloading site.

[0018] Furthermore, the joint judgment model generates a final judgment result by weighting the behavior anomaly score and the scene legitimacy mark, wherein the scene illegal mark triggers an additional penalty weight; the final judgment result is a violation or normal behavior.

[0019] Furthermore, the on-vehicle real-time feedback includes at least one of voice warnings and driving path restrictions, and the regional monitoring equipment linkage includes triggering roadside cameras and drones to collaboratively monitor illegal areas.

[0020] Furthermore, the S4 further includes:

[0021] Calculate a regional risk index that dynamically adjusts monitoring resource allocation based on the number of vehicles violating the regulations and the distribution of violations within the region.

[0022] Furthermore, when the regional risk index exceeds a preset threshold, the deployment density of monitoring equipment in the area is automatically increased, and an enforcement report is generated and pushed to the management platform.

[0023] Furthermore, the inertial sensor includes a three-axis accelerometer and a gyroscope, and the data acquisition frequency is not less than once per second; the on-board camera is a 360-degree panoramic camera, and the image acquisition is strictly synchronized with the timestamp of the abnormal behavior period.

[0024] In a second aspect of the present invention, a collaborative monitoring system for illegal behaviors of dump trucks driven by AI is provided, the system comprising:

[0025] The vehicle-mounted terminal module is used to collect the three-dimensional acceleration and three-dimensional angular velocity data of the dump truck through the vehicle-mounted inertial sensor to construct an inertial behavior sequence. Based on the historical normal behavior data, an individualized behavior model is established to identify abnormal behavior periods that deviate from the normal state and generate an anomaly score.

[0026] The cloud-based analysis platform is used to trigger the on-board camera to collect environmental images when abnormal behavior occurs, and use the visual semantic recognition model to determine the legality of the scene in which the dump truck is located;

[0027] A multi-level response module generates a final violation determination result through a joint determination model based on the integration of the anomaly score and the scenario legitimacy mark;

[0028] The dynamic monitoring and scheduling module triggers a multi-level linkage response mechanism based on the violation determination results, including real-time feedback from the vehicle, linkage with regional monitoring equipment, and uploading of violation data for evidence storage.

[0029] The beneficial technical effects of the present invention are at least as follows:

[0030] In view of the limitations of the existing dump truck supervision system in terms of recognition scope, judgment ability and coordination mechanism, the present invention proposes a collaborative monitoring method and system for dump truck violations driven by artificial intelligence. The system takes breaking through the "perception blind spot" and "formal compliance disguise" as its technical goal, and realizes the joint judgment of the actual behavior of the dump truck and the environment in which it occurs by constructing a set of intelligent recognition frameworks with physical behavior modeling capabilities and scene semantic understanding capabilities. At the same time, the present invention introduces a multi-vehicle collaborative mechanism, which can trigger the linkage perception of surrounding vehicles and environmental nodes after a single-vehicle high-risk event occurs, thereby achieving a larger range and more robust collaborative monitoring efficiency. The core innovation of the present invention lies in the integration of multi-dimensional perception data (such as vehicle inertial behavior, visual semantic features, and spatiotemporal trajectories) to construct a comprehensive judgment model, and by constructing a dynamic risk diffusion mechanism, it realizes the chain tracking capability from "local abnormal behavior" to "system-level high-risk events." BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The present invention is further described with reference to the accompanying drawings. However, the embodiments in the accompanying drawings do not constitute any limitation to the present invention. A person skilled in the art can obtain other drawings based on the following drawings without creative effort.

[0032] Figure 1 This is a flow chart of the collaborative monitoring method for illegal behaviors of dump trucks driven by AI in the present invention. DETAILED DESCRIPTION

[0033] The following describes embodiments of the present invention in detail. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended only to explain the present invention and are not to be construed as limiting the present invention.

[0034] like Figure 1 As shown, an embodiment of the present invention provides an AI-driven collaborative monitoring method for illegal behaviors of dump trucks, the method comprising:

[0035] S1. The three-dimensional acceleration and three-dimensional angular velocity data of the dump truck are collected through the on-board inertial sensor to construct an inertial behavior sequence. Based on the historical normal behavior data, an individualized behavior model is established to identify abnormal behavior periods that deviate from the normal state and generate an abnormality score.

[0036] Specifically, this step uses real-time acceleration and angular velocity data collected by the vehicle's inertial measurement unit (IMU) to detect abnormal behavior during operation, particularly high-risk operations such as illegal dumping. Because illegal dump truck behavior is often accompanied by significant changes in inertia, this step establishes an individualized inertial behavior model to identify periods of significant deviation from normal behavior patterns, which serve as candidate areas for subsequent scenario assessment.

[0037] Furthermore, each dump truck is equipped with a three-axis IMU sensor to collect real-time acceleration and angular velocity data of the vehicle during acceleration, braking, steering, etc. Specifically, the sensor records three-dimensional acceleration (a t ) and angular velocity (ω t ), which constitutes the inertial behavior sequence of the vehicle.

[0038] The data at each time point t includes:

[0039] represent the acceleration along the x, y, and z axes respectively;

[0040] represent the angular velocity along the x, y, and z axes respectively.

[0041] Among them, these acceleration and angular velocity data will be used for subsequent behavior analysis. In order to facilitate processing, a sliding window method is used. Assuming that a window length is τ, the data from time point t-τ to t constitutes a complete inertial behavior segment X t ={a t ,ω t}.

[0042] Furthermore, for each dump truck, the present invention uses its historical normal operation data (excluding violations) to train an individualized behavior model. This model is built using a Gaussian mixture model (GMM), where the normal behavior pattern of each vehicle is represented as a weighted Gaussian distribution.

[0043] For each new behavior period X t , the present invention compares it with the historical normal behavior distribution of the vehicle and calculates its "deviation" in the distribution. The present invention uses the negative log-likelihood value as the deviation score s t , the formula is as follows:

[0044] s t =-logf θ (X t ) (1)

[0045] Among them, f θ (X t) represents the current behavior segment X calculated by the GMM model t The likelihood value of θ is the model parameter, which is adaptively adjusted according to the normal behavior of each vehicle after training.

[0046] If s t The value of is significantly higher than the mean score of normal behavior, indicating that the behavior deviates greatly from the norm and is therefore marked as a possible abnormal behavior.

[0047] Furthermore, in order to identify possible violations, the present invention calculates the current deviation score s t Average of historical ratings Compare and determine whether it is an abnormal period. Set the threshold δ, when , the current period is judged to be a candidate for abnormal behavior.

[0048] Specifically, δ is 1.5 times the standard deviation of historical scores, that is, δ = 1.5 × σ s , where σ s is the standard deviation of historical ratings. If s t If the threshold is exceeded, the behavior period is marked as abnormal, and X t This is recorded as an abnormal period and passed on to the next step.

[0049] The output of this step includes: abnormal behavior period X t , which is a time series containing acceleration and angular velocity data; abnormal behavior score s t , used for subsequent scenario compliance judgment.

[0050] S2. When abnormal behavior occurs, the on-board camera is triggered to collect environmental images, and the legality of the scene in which the dump truck is located is judged through the visual semantic recognition model.

[0051] Specifically, the purpose of this step is to further determine whether these behaviors occur in a legal operating environment by combining the abnormal behavior periods detected in the first step. Dump trucks may exhibit certain abnormal behaviors under normal circumstances, such as sudden braking or acceleration, but whether these behaviors violate regulations depends on the environment in which they occur. Therefore, this step combines inertial sensor data and image data collected by the on-board camera to perform semantic scene understanding to determine whether the behavior occurs in a permitted area (such as a construction site) or an area where unloading is prohibited (such as a no-dumping zone). This not only helps to further confirm whether the behavior is illegal, but also enhances the accuracy and environmental adaptability of the system.

[0052] Furthermore, in step 1, the present invention identifies an abnormal behavior period X tThe acceleration and angular velocity sequence of this period represents the abnormal behavior of the dump truck. Based on the timestamp of the abnormal behavior period, the onboard camera automatically captures the corresponding image I t ,ensures the time synchronization of the images and the abnormal behavior period.,The image acquisition module is usually a 360-degree panoramic camera mounted on a vehicle, which can capture the actual driving status of the,vehicle in different environments.

[0053] Here, X t As the candidate period of abnormal behavior, the time period when the vehicle may commit illegal behavior is identified, and I t The captured environmental image data related to these time periods will be used for scene analysis by the subsequent visual recognition module.

[0054] Semantic scene analysis and environmental legitimacy judgment

[0055] By ImageI t , input it into the trained visual semantic recognition model g φ (I t ), the model can output the scene label c corresponding to the image t , indicating the environment type in the image. Possible environment type labels include:

[0056] c t =Construction site: indicates that the dump truck is unloading normally at the construction site;

[0057] c t = No dumping zone: indicates that the dump truck has entered an area where unloading is prohibited;

[0058] c t =Road: Indicates that the dump truck is driving on an ordinary road.

[0059] Through the image model g φ (I t ), the present invention can accurately identify the environment in which the dump truck is located. After large-scale training, the model can handle images in complex scenes, such as urban construction sites, no-dumping zones, and road areas.

[0060] After determining the environmental label c t After that, the system further performs environmental compliance judgment. The present invention defines the compliance mark l t , if c t = No-falling zone, then l t = illegal, indicating that the dump truck's behavior occurred in a prohibited dumping area, violating relevant regulations; if c t = construction site, then l t =Legal, indicating that the area is a legal unloading site.

[0061] This compliance judgment process not only improves the system's accurate understanding of vehicle behavior, but also closely integrates visual scene and behavioral data, eliminating the misjudgment problem that may be caused by single sensor data.

[0062] Furthermore, in step 1, the present invention obtains a behavioral abnormality score s t , and in this step, the present invention obtains the legality mark l t In order to comprehensively evaluate the compliance of behavior and environment, this paper introduces a joint judgment model h ψ (s t ,l t ), the model will combine s t and l t Generate the final violation judgment result y t The formula is as follows:

[0063]

[0064] In this formula: s t is the abnormal behavior score obtained from step 1, indicating whether the behavior deviates from the normal behavior pattern; t Indicates whether the current environment allows the behavior to occur; w1 and w2 are weight coefficients used to balance the impact between abnormal behavior and the legitimacy of the scene; λ is the regularization factor, especially when l t = illegal, λ increases the impact of the illegal environment on the final violation judgment by adding a penalty term.

[0065] This approach not only combines the abnormality of the behavior itself, but also enhances the role of environmental legitimacy in determining violations. For example, when the behavior score of the dump truck is s t Very high, but if the environment is marked as an "illegal" area (l t = illegal), the judgment of violations will be increased. The addition of this regularization improves the system's sensitivity to the environment.

[0066] Finally, we get the environment scene label c t , indicating the type of environment where the dump truck is located (such as construction site, no-dumping zone, etc.); legality mark l t , indicating whether the environment allows the current behavior of the dump truck to occur; the final violation judgment y t , which will serve as input for subsequent steps (such as regulatory response). The output will be passed to the next step of processing (such as coordinated monitoring and law enforcement response) to determine whether further law enforcement action is needed.

[0067] The design of this step directly addresses a key issue in monitoring dump truck violations: the difficulty of determining environmental compliance based solely on behavioral data. By introducing semantic scene recognition and environmental legitimacy judgment, this invention effectively eliminates the limitations of single sensor data and combines vehicle behavior with actual environmental information for joint judgment. This approach not only improves the accuracy of violation determinations but also enhances the system's adaptability to complex scenarios, allowing dump truck violations to be accurately detected in a more realistic environment.

[0068] S3. Fusion of the anomaly score and the scene legitimacy tag to generate a final violation determination result through a joint determination model.

[0069] Specifically, step three aims to further integrate the data obtained in the previous steps and accurately judge the illegal behavior of the dump truck through the joint modeling of behavior and environment. t and the scene legitimacy label l obtained in step 2 t (and scene label c t ), this step finally outputs the judgment result y t to determine whether a violation has occurred.

[0070] Combining the outputs of these two steps, the present invention performs joint modeling in this step, integrating s t and l t (and c t ) to make the final judgment, specifically:

[0071] In order to comprehensively evaluate the severity of abnormal behavior and the legitimacy of the scenario, the present invention constructs a joint judgment model h ψ (s t ,l t ,c t ), the goal of the model is to output the final violation judgment y t .

[0072] The joint judgment model is based on behavioral abnormality score t , Environmental Legality Mark t and environmental labels c t The combination of these three factors will more accurately determine whether the dump truck has violated the regulations.

[0073] The calculation formula of the joint model is as follows:

[0074]

[0075] Among them, y t is the final violation judgment, y t =1 indicates violation, yt =0 indicates normal behavior. t is the abnormal behavior score obtained from step 1, reflecting the degree of deviation between the current behavior and normal behavior. t The larger the value, the more abnormal the behavior. t is the legality mark obtained from step 2, indicating whether the current environment allows the behavior of the dump truck. t = Illegal, the behavior occurred in a no-dumping zone or other illegal area. t is the scene label obtained from step 2, which indicates the specific type of the current environment. For example, the dump truck is in a construction site or a no-dumping zone. w1, w2, and w3 are weight coefficients used to control the relative importance of behavior score, environment legality, and scene label. λ is the regularization factor, especially in l t = illegal, λ strengthens the response to illegal areas to ensure that violations in illegal areas are given full attention. α is the threshold weighting term, when s t When a certain critical value is exceeded, A value of 1 indicates that the behavior is extremely abnormal and the system gives it additional weighting.

[0076] By introducing the regularization factor λ, the present invention can strengthen the judgment of illegal areas, especially when the dump truck enters the no-dumping area or other illegal areas, even if the behavior score s t The system can still judge it as a violation by judging the legality of the scene, which solves the limitations of traditional methods in complex scenes.

[0077] In order to improve the adaptability and accuracy of the system, this step further considers the impact of different scenarios on behavior judgment. t The introduction of this feature allows the model to fully understand the actual operating scenarios of dump trucks. For example, certain behaviors may be permitted within construction sites and non-prohibited dumping zones, while in prohibited dumping zones, even if the behavior score is low, it should still be considered a violation. This approach solves the problem of misjudgment of behavior that is inherently abnormal but occurs in a legal area.

[0078] Finally, the model h ψ Will generate y t , as the final violation determination. t =1 means the behavior during this period is a violation, y t =0 means the behavior during this period is normal.

[0079] S4. Based on the violation determination result, a multi-level linkage response mechanism is triggered, including real-time feedback from the vehicle, linkage with regional monitoring equipment, and uploading of violation data for evidence storage.

[0080] Specifically, the goal of this step is to ensure timely response and follow-up of illegal behaviors of dump trucks through a multi-level linkage response and collaborative monitoring mechanism. t The system combines a multi-level response system to quickly respond and ensure that violations are dealt with, while also enhancing regional monitoring.

[0081] Once it is determined to be a violation t =1, the system starts the following multi-level response mechanism:

[0082] Local Response Mechanism (In-Vehicle Feedback): On the vehicle terminal, the system alerts the driver of non-compliant behavior through voice warnings and instrument panel displays. Simultaneously, the system can restrict the vehicle's route through the vehicle's intelligent device, such as prohibiting entry into specific areas or limiting speed, to ensure immediate correction of non-compliant behavior.

[0083] Regional linkage response mechanism: When violations occur in sensitive areas (such as no-dumping zones), the system will trigger a regional monitoring response by linking with surrounding intelligent monitoring equipment. For example, roadside cameras, drones, and other sensing devices will simultaneously collect and analyze data to ensure that other vehicles in the area are also alerted and can make corresponding adjustments.

[0084] Data upload and record keeping: All violation-related data (such as acceleration data, environmental image data, violation timestamps, and violation areas) is uploaded to a central platform in real time. This data is used for subsequent audits, penalty decisions, and further supervision. The platform generates detailed violation reports and provides them to traffic management departments for further review.

[0085] The system also has a regional risk index R t By monitoring the violations in a specific area, the risk index of the area is calculated to reflect the degree of violation risk in the area. t Calculated by the following formula:

[0086]

[0087] Where: N is the number of vehicles participating in the monitoring of the area; is the violation judgment of the i-th vehicle at time t, y i,t =1 indicates violation; Is whether the i-th vehicle is in the violation area D.

[0088] When R t When the preset threshold is exceeded, the system will automatically increase the monitoring frequency of the area and deploy more sensors or drones in the area for dynamic monitoring.

[0089] An intelligent feedback mechanism dynamically adjusts monitoring intensity in a specific area based on real-time monitoring data and historical violation records. For areas with frequent violations, the system automatically increases monitoring density there and reduces monitoring frequency in other areas. This dynamic adjustment ensures efficient resource allocation and enables the monitoring system to more flexibly respond to monitoring needs in different areas.

[0090] Automatic enforcement feedback mechanism: When the frequency of violations in an area reaches a certain threshold, the system automatically generates a violation report and triggers an alert in the enforcement system. Enforcement officers can use this report to quickly obtain specific data on the violations and initiate manual intervention if necessary.

[0091] The embodiment of the present invention also provides an AI-driven collaborative monitoring system for illegal behaviors of dump trucks, the system comprising:

[0092] The vehicle-mounted terminal module is used to collect the three-dimensional acceleration and three-dimensional angular velocity data of the dump truck through the vehicle-mounted inertial sensor to construct an inertial behavior sequence. Based on the historical normal behavior data, an individualized behavior model is established to identify abnormal behavior periods that deviate from the normal state and generate an anomaly score.

[0093] The cloud-based analysis platform is used to trigger the on-board camera to collect environmental images when abnormal behavior occurs, and use the visual semantic recognition model to determine the legality of the scene in which the dump truck is located;

[0094] A multi-level response module generates a final violation determination result through a joint determination model based on the integration of the anomaly score and the scenario legitimacy mark;

[0095] The dynamic monitoring and scheduling module triggers a multi-level linkage response mechanism based on the violation determination results, including real-time feedback from the vehicle, linkage with regional monitoring equipment, and uploading of violation data for evidence storage.

[0096] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0097] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the units is merely a division of logical functions. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0098] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0099] Although the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the claims and their equivalents.

Claims

1. The AI-driven collaborative monitoring method for illegal behaviors of dump trucks is characterized by: The method comprises the following steps: S1. 3D acceleration and 3D angular velocity data of the dump truck is collected through on-board inertial sensors to construct an inertial behavior sequence. Based on historical normal behavior data, an individualized behavior model is established to identify abnormal behavior periods that deviate from the norm and generate an anomaly score. S2. When abnormal behavior occurs, the vehicle-mounted camera is triggered to collect environmental images and the legality of the scene in which the dump truck is located is determined through the visual semantic recognition model; S3. Fusion of the anomaly score and the scenario legitimacy tag to generate a final violation determination result through a joint determination model; S4. Based on the violation determination result, a multi-level linkage response mechanism is triggered, including real-time feedback from the vehicle, linkage with regional monitoring equipment, and uploading of violation data for evidence storage.

2. The AI-driven collaborative monitoring method for illegal behaviors of dump trucks according to claim 1 is characterized in that: The individualized behavior model is obtained by training a Gaussian mixture model, and the anomaly score is determined by calculating the deviation of the current behavior sequence in the historical distribution; the deviation is determined based on the negative log-likelihood value; If the deviation is higher than the mean score of normal behavior, it indicates that the behavior deviates greatly from the norm and is thus marked as a possible abnormal behavior; Compare the deviation with the mean of historical scores to determine whether it is an abnormal period: If the deviation is greater than the sum of the mean of the historical scores and the preset threshold, the current period is marked as an abnormal behavior period.

3. The AI-driven collaborative monitoring method for illegal dump truck behavior according to claim 1 is characterized in that: The input of the visual semantic recognition model is the environment image, and the output scene labels include construction sites, no-dumping zones and roads, and the environment legality tags are dynamically generated based on the scene labels.

4. The AI-driven collaborative monitoring method for illegal dump truck behavior according to claim 3 is characterized in that: The environmental legitimacy mark is dynamically generated based on the scene label, specifically including: Define the scene legality mark. If the construction site is a no-dumping zone, the scene legality mark is illegal, indicating that the dump truck's behavior occurs in the no-dumping zone and violates relevant regulations; if the construction site is a construction site, the scene legality mark is legal, indicating that the area is a legal unloading site.

5. The AI-driven collaborative monitoring method for illegal behaviors of dump trucks according to claim 4 is characterized in that: The joint judgment model generates a final judgment result by weighting the behavior anomaly score and the scene legality mark, wherein the scene illegal mark triggers an additional penalty weight; the final judgment result is a violation or normal behavior.

6. The AI-driven collaborative monitoring method for illegal dump truck behavior according to claim 1 is characterized in that: The on-vehicle real-time feedback includes at least one of voice warnings and driving path restrictions, and the regional monitoring equipment linkage includes triggering roadside cameras and drones to coordinate monitoring of illegal areas.

7. The AI-driven collaborative monitoring method for illegal dump truck behavior according to claim 1 is characterized in that: Said S4 further includes: Calculate a regional risk index that dynamically adjusts monitoring resource allocation based on the number of vehicles violating the regulations and the distribution of violations within the region.

8. The AI-driven collaborative monitoring method for illegal dump truck behavior according to claim 6 is characterized in that: When the risk index of the area exceeds the preset threshold, the deployment density of monitoring equipment in the area is automatically increased, and an enforcement report is generated and pushed to the management platform.

9. The AI-driven collaborative monitoring method for illegal dump truck behavior according to claim 1 is characterized in that: The inertial sensor includes a three-axis accelerometer and a gyroscope, and the data acquisition frequency is not less than once per second; the on-board camera is a 360-degree panoramic camera, and the image acquisition is strictly synchronized with the timestamp of the abnormal behavior period.

10. The AI-driven collaborative monitoring system for illegal behaviors of dump trucks is characterized by: The system comprises: The vehicle-mounted terminal module is used to collect the three-dimensional acceleration and three-dimensional angular velocity data of the dump truck through the vehicle-mounted inertial sensor to construct an inertial behavior sequence. Based on the historical normal behavior data, an individualized behavior model is established to identify abnormal behavior periods that deviate from the normal state and generate an anomaly score. The cloud-based analysis platform is used to trigger the on-board camera to collect environmental images when abnormal behavior occurs, and use the visual semantic recognition model to determine the legality of the scene in which the dump truck is located; A multi-level response module is used to generate a final violation violation determination result through a joint determination model based on the fusion of the anomaly score and the scene legitimacy mark; The dynamic monitoring and scheduling module is used to trigger a multi-level linkage response mechanism based on the violation determination results, including real-time feedback from the vehicle, linkage with regional monitoring equipment, and uploading and storing violation data.

Images