Vehicle coal stealing prevention abnormity detection method and system based on artificial intelligence

Abstract

The invention relates to a vehicle coal stealing prevention abnormity detection method and system based on artificial intelligence. The detection method comprises the following steps: acquiring coal transporting vehicle parameter information and transportation task information; determining a to-be-detected area according to the coal transport vehicle parameter information and the transport task information; acquiring data acquired by the first data acquisition module and the second data acquisition module and state information of the first data acquisition module, the second data acquisition module and the first state detection module; identifying whether the coal transport vehicle has a coal quality abnormal risk or not, and dividing risk grades; the detection method is accurate, rapid and strict in logicality, accords with the actual operation condition of the coal transport vehicle, carries out risk level evaluation according to the judgment condition, and has the advantages of intelligence, reasonability and the like.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence and Internet of Things technology, and more specifically to an abnormal detection method and system for preventing theft of coal on a coal transport vehicle. Background Art

[0002] With the rapid development of the energy industry, coal, as an important energy resource, faces severe supervision issues in the process of mining, transportation, storage and use. The traditional supervision method of the fleet responsible for transporting coal often relies on manual inspections and post-event accountability, which is inefficient and difficult to fully cover.

[0003] With the continuous development of artificial intelligence and Internet of Things technology, its application in the field of logistics and transportation supervision is gradually becoming the key to solving these problems. However, the current supervision method only uses GPS positioning points to judge whether the vehicle status is correct. This method has limitations, and the judgment result is not accurate enough to match the actual situation, and the reliability is poor. Summary of the invention

[0004] In view of the deficiencies in the prior art, the present invention proposes a method and system for detecting abnormalities in coal transport vehicles to prevent theft of coal. The detection method has accurate, convenient and fast judgment results.

[0005] To achieve the above object, the present invention adopts the following technical solutions:

[0006] First, the present invention proposes an artificial intelligence-based vehicle anti-theft coal replacement anomaly detection method, comprising the following steps:

[0007] S1. Obtaining coal transport vehicle parameter information and transportation task information;

[0008] S2. Determine the area to be inspected according to the coal transport vehicle parameter information and transport task information;

[0009] S3, obtaining video data or point cloud data in the area to be detected through the first data acquisition module, and obtaining corresponding sampling time, vehicle speed, and vehicle GPS longitude and latitude coordinates through the second data acquisition module; judging the status information of the first data acquisition module and the second data acquisition module through the first state detection module; judging the status information of the second data acquisition module and the first state detection module through the second state detection module;

[0010] S4, identifying whether the coal transport vehicle has a risk of abnormal coal quality and classifying the risk level through the data acquired by the first data acquisition module, the data acquired by the second data acquisition module, the status information of the first data acquisition module, the status information of the second acquisition module and the status information of the first status detection module;

[0011] S5. Provide corresponding alarm level prompts according to the risk level of coal quality abnormality.

[0012] Preferably, in S1, the coal transport vehicle parameter information includes vehicle width, vehicle height, vehicle length, and license plate; the transportation task information includes the starting point name and starting point location coordinates, and the end point name and end point location coordinates.

[0013] Preferably, in S2, the area to be detected includes the restricted driving area of ​​the coal transport vehicle and the restricted video monitoring area behind the coal transport vehicle; the method for setting the restricted driving area of ​​the coal transport vehicle is: inputting the parameter information of the coal transport vehicle and the transportation task information into the navigation software to obtain the restricted driving area of ​​the coal transport vehicle; based on the restricted driving area of ​​the coal transport vehicle, setting the restricted video monitoring area behind the coal transport vehicle.

[0014] Preferably, in S3, the first data acquisition module includes a visual sensor or a laser radar sensor; the second data acquisition module includes a GPS locator or a driving recorder;

[0015] The status information of the first data acquisition module includes: online status, offline status, and abnormal status;

[0016] The status information of the second data acquisition module includes: online status, dormant status, abnormal status;

[0017] The status information of the first status detection module includes: online status, sleep status, and abnormal status.

[0018] The abnormal state of the first data acquisition module includes a black screen, a gray screen, a white screen, or a window of the first data acquisition module is dirty, blocked, or has an abnormal angle;

[0019] The abnormal states of the second data acquisition module and the first state detection module are both states when the timed wake-up function is not executed in the sleep state.

[0020] Preferably, the first state detection module includes a first data acquisition module, the first data acquisition module is used to receive data sent by the first data acquisition module and the second data acquisition module, and the first state detection module determines the state information of the first data acquisition module and the second data acquisition module according to the state of the data received by the first data acquisition module;

[0021] The second status detection module determines the status information of the second data acquisition module and the first data acquisition module according to the heartbeat detection mechanism.

[0022] Preferably, in S4, the levels of coal quality abnormality risk are divided into level one and level two, and the method for identifying whether a coal transport vehicle has coal quality abnormality risk and classifying the risk level specifically includes the following sub-steps:

[0023] S401. Determine whether the first data acquisition module, the second data acquisition module, and the first status detection module meet the first condition, the second condition, or the third condition. If the first condition is met, proceed to S402. If the second condition is met, proceed to S403. If the third condition is met, determine that there is a secondary risk anomaly in the vehicle coal quality.

[0024] The first condition is: the first data acquisition module is in an offline or abnormal state; the second data acquisition module is in an online state; the first status detection module is in an online or offline state.

[0025] The second condition is: the first data acquisition module is in an online state; the second data acquisition module is in an online state; the first status detection module is in an online state.

[0026] The third condition is: within the preset time T4, the second status detection module does not receive the heartbeat messages sent by the second data acquisition module or the first status detection module.

[0027] S402. Determine the current driving state of the coal transport vehicle based on the vehicle speed collected by the second data acquisition module, and determine the parking duration t1 or the driving distance m1 within the first T1 time after the vehicle leaves the starting point through the current driving state. Then compare t1 with t2 or compare m1 with m2, where t2 is the preset parking duration within the first T1 time after the vehicle leaves the starting point; m2 is the preset driving distance within the first T1 time after the vehicle leaves the starting point.

[0028] When t1 > t2, determine the stopping point based on the GPS longitude and latitude coordinates within the parking duration t1. Determine the location information of the stopping point from the stopping point and the map POI information. Match the location information of the stopping point with the data segment of the set common address library. If the location information of the stopping point is not within the data segment of the common address library, determine that there is a secondary risk anomaly in the vehicle coal quality.

[0029] When m1 < m2, determine the vehicle passing center point based on the GPS longitude and latitude coordinates within the first T1 time before the current moment. Determine the location information of the passing center point from the passing center point and the map POI information. Match the location information of the passing center point with the fields in the set common address library. If the location information of the passing center point is not within the fields of the common address library, then determine that there is a secondary risk anomaly in the vehicle coal quality state.

[0030] S403. Determine the current driving state of the coal transport vehicle based on the vehicle speed collected by the second data acquisition module, and determine the parking duration t1 or the driving distance m1 within the first T1 time after the vehicle leaves the starting point through the current driving state. Then compare t1 with t2 or compare m1 with m2.

[0031] When t1 > t2 or m1 > m2, filter out the data within the area to be detected from the data collected by the first data acquisition module; determine the geometric features of each object within the area to be detected based on the data within the area to be detected; and judge the positional relationship between each object and the video monitoring restricted area behind the coal transport vehicle; if the position of the object is directly above the video monitoring restricted area behind the coal transport vehicle, judge the type of the object. If the object is an abnormal object, count the occurrence frequency N or the staying duration T2 of the object; compare N with N1, or compare T2 with T4, where N1 is the preset frequency of the object appearing directly above the video monitoring restricted area behind the coal transport vehicle.

[0032] When N > N1 or T2 > T4, it is determined that there is a first-level risk anomaly in the coal quality of the vehicle.

[0033] When N < N1 or T2 < T4, it is determined that there is a second-level risk anomaly in the coal quality of the vehicle.

[0034] Preferably, the specific method for determining the stopping point or the passing center point is as follows: Apply a clustering algorithm to the GPS longitude and latitude coordinates corresponding to each moment within the already parked duration t1 at the current moment or the GPS longitude and latitude coordinates per unit time within the previous T1 time period at the current moment to obtain the stopping point or the passing center point. The position information of the stopping point or the position information of the passing center point both include GPS longitude and latitude coordinates, and the clustering algorithm selects any one of K-means, K-medoids, and hierarchical clustering.

[0035] Preferably, the specific method for determining the geometric features of each object within the area to be detected based on the data within the area to be detected is as follows: Model the scene based on the point cloud data within the area to be detected, obtain the first feature information of each object through a perception algorithm. The first object feature information includes position, shape, and distance. Based on the first feature information of each object, classify the objects and identify abnormal objects.

[0036] Alternatively, obtain pictures frame by frame based on the video data within the area to be detected, identify each object and the second feature information of the object through an object detection algorithm. The second feature information includes position and distance, and classify the objects based on the second feature information to identify abnormal objects; the abnormal objects specifically include people, the bucket of a loader, a coal loading machine, and a scraper conveyor.

[0037] Secondly, the present invention proposes an artificial intelligence-based vehicle anti-theft coal replacement anomaly detection system, comprising a first data acquisition module, a second data acquisition module, a first state detection module, a second state detection module, a first analysis module, an analysis result marking module, a second data acquisition module, a second analysis module and a result marking warning module; wherein the first data acquisition module is connected to the input of the first state detection module, the second data acquisition module is respectively connected to the inputs of the first state detection module, the second state detection module and the second data acquisition module, the output of the first state detection module is respectively connected to the inputs of the first analysis module, the second state detection module and the second data acquisition module, the output of the first analysis module is respectively connected to the input of the analysis result marking module, the output of the analysis result marking module is connected to the input of the result marking warning module, the output of the second state detection module and the second data acquisition module are respectively connected to the input of the second analysis module, and the output of the second analysis module is connected to the input of the result marking warning module;

[0038] The first analysis module determines whether the coal quality of the coal transport vehicle has an abnormal risk and classifies the risk level according to the received data collected by the first data acquisition module, the data collected by the second data acquisition module, the status information of the first data acquisition module and the status information of the second data acquisition module;

[0039] The analysis result marking module is used to store the coal quality abnormality judgment result of the first analysis module and send the stored information to the second analysis module;

[0040] The second data acquisition module is used to acquire the data collected by the second data acquisition module and the coal quality abnormality judgment result stored by the first state detection module;

[0041] The second analysis module determines whether the coal quality of the coal transport vehicle has an abnormal risk and classifies the risk level according to the status information of the second data acquisition module and the status information of the first status detection module; or determines whether the coal quality of the coal transport vehicle has an abnormal risk and classifies the risk level according to the data collected by the second data acquisition module;

[0042] The result marking warning module is used to store the coal quality abnormality judgment result of the second analysis module or receive the coal quality abnormality judgment result of the second analysis module sent by the analysis result marking module, and issue a corresponding level of alarm;

[0043] The coal quality abnormality judgment result of the first analysis module includes: the license plate, the coal quality abnormality risk level, the video data within the T5 time before and after the current moment saved by the first analysis module, and the GPS longitude and latitude coordinates of the vehicle at the current moment when the coal quality is abnormal;

[0044] The coal quality abnormality judgment result of the second analysis module includes the license plate, the coal quality abnormality risk level, and the vehicle GPS longitude and latitude information most recently received by the second data acquisition module.

[0045] Again, the present invention also proposes a computer program product, including a computer program / instruction, which, when executed by a processor, implements the steps of the method for detecting abnormal coal substitution in a coal transport vehicle as described in any one of claims 1 to 8.

[0046] Compared with the prior art, the present invention has the following beneficial effects:

[0047] (1) The present invention uses a multi-sensor joint detection method to perform abnormal detection to prevent theft of coal, which is convenient and fast, and has a higher accuracy rate than the existing technology of detecting vehicle abnormalities through GPS positioning points. The existing technology only determines whether the GPS positioning point is inside or outside the designated electronic fence for early warning, which is easy to cause false alarms and false alarms. The coal quality abnormality detection algorithm of the present invention has strict logic and is more in line with the actual operation of coal transport vehicles. The method evaluates the risk level according to the judgment conditions, and has the advantages of intelligence and rationality;

[0048] (2) The present invention uses artificial intelligence technology and perception-based target detection technology to obtain the optimal abnormal object detection model through a large amount of preliminary training on abnormal object data. In addition, the present invention clearly defines the key factors involved in the process of stealing coal, namely the human factors and key equipment involved in stealing coal, and detects these key factors in a more targeted manner. It also conducts risk level assessment based on the dwell time or frequency of occurrence of the key factors, which is more reasonable and quicker. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is a flow chart of the detection method of the present invention;

[0050] Figure 2 The monitoring system block diagram of the present invention;

[0051] Figure 3 This is a structural diagram of the target detection model of the present invention; DETAILED DESCRIPTION

[0052] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0053] Example 1

[0054] refer to Figure 1 and Figure 2This example proposes an artificial intelligence-based vehicle anti-theft coal replacement anomaly detection method, the vehicle is equipped with a detection system for implementing the detection method, the detection system includes a first data acquisition module, a second data acquisition module, a first state detection module, a second state detection module, a first analysis module, an analysis result marking module, a second data acquisition module, a second analysis module, and a result marking warning module; the first data acquisition module includes a collection visual sensor or a laser radar sensor; the second data acquisition module includes a GPS locator or a driving recorder, and the first state detection module includes a first data acquisition module; the second data acquisition module and the first state detection module are provided with a backup battery;

[0055] Wherein the first data acquisition module is connected to the input of the first state detection module, the second data acquisition module is respectively connected to the inputs of the first state detection module, the second state detection module and the second data acquisition module, the output of the first state detection module is respectively connected to the inputs of the first analysis module, the second state detection module and the second data acquisition module, the output of the first analysis module is connected to the input of the analysis result marking module, the output of the analysis result marking module is connected to the input of the result marking early warning module, the output of the second state detection module and the second data acquisition module are respectively connected to the input of the second analysis module, and the output of the second analysis module is connected to the input of the result marking early warning module;

[0056] The detection method comprises the following steps:

[0057] S1. Obtaining coal transport vehicle parameter information and transportation task information;

[0058] The parameter information of the coal transport vehicle includes vehicle width, vehicle height, vehicle length, and license plate; the transportation task information includes the starting point name and starting point location coordinates, and the end point name and end point location coordinates.

[0059] S2. Determine the area to be inspected according to the coal transport vehicle parameter information and transport task information;

[0060] The area to be detected includes a limited driving area for coal transport vehicles and a limited video monitoring area behind the coal transport vehicles; the method for setting the limited driving area for coal transport vehicles is as follows: inputting parameter information of coal transport vehicles and transportation task information into navigation software to obtain the limited driving area for coal transport vehicles; based on the limited driving area for coal transport vehicles, setting the limited video monitoring area behind the coal transport vehicles, wherein the navigation software can select any one of Baidu Map Truck Navigation and Amap Truck Edition;

[0061] Based on the navigation map route, an electronic fence is set. By acquiring the data of the electronic fence, the vehicle limited driving area can be determined, and the speed limit of the target coal transport vehicle in the limited driving area is further set. For example, in this embodiment, the limited driving area is a range of 15 meters on both sides of the navigation map route, and does not include the range with the latitude and longitude coordinates of the starting point and the end point as the center and the radius R = 500m;

[0062] The method for defining the restricted area for video monitoring at the rear of a coal transport vehicle is specifically as follows: based on the field of view of a camera installed on the top of the vehicle cab facing the rear of the vehicle, a rectangular frame area in the field of view of the corresponding camera is marked according to the length, width, and height information of the vehicle, and the rectangular frame area and the upper part of the rectangular frame area of ​​the field of view are the restricted area for video monitoring at the rear of the vehicle.

[0063] S3, the first data acquisition module collects video data in the area to be detected through a visual sensor or collects point cloud data in the area to be detected through a laser radar sensor, and the second data acquisition module collects the corresponding sampling time, vehicle speed, and vehicle GPS longitude and latitude coordinates through a GPS locator or a driving recorder; the first data acquisition module transmits the collected data to the first data acquisition module, and the second data acquisition module transmits the collected data to the first data acquisition module and the second data acquisition module, and the first state detection module determines the state information of the first data acquisition module and the second data acquisition module through the state of the data received by the first data acquisition module; the second state detection module determines the state information of the second data acquisition module and the first data acquisition module according to the heartbeat detection mechanism;

[0064] The status information of the first data acquisition module includes: online status, offline status, and abnormal status;

[0065] The status information of the second data acquisition module includes: online status, dormant status, abnormal status;

[0066] The state information of the first state detection module includes: online state, sleep state, abnormal state;

[0067] The abnormal state of the first data acquisition module includes a black screen, a gray screen, a white screen, or a window of the first data acquisition module is dirty, blocked, or has an abnormal angle;

[0068] The abnormal states of the second data acquisition module and the first state detection module are both states when the timed wake-up function is not executed in the sleep state.

[0069] S4. Identify whether the coal transport vehicle has a risk of abnormal coal quality and classify the risk level through the data collected by the first data acquisition module, the data collected by the second data acquisition module, the status information of the first data acquisition module, the status information of the second acquisition module and the status information of the first status detection module; the level of abnormal coal quality risk is divided into level one and level two. The specific method of identifying whether the coal transport vehicle has a risk of abnormal coal quality and classifying the risk level is as follows:

[0070] (1) The first state detection module is online:

[0071] When the first analysis module receives that the first data acquisition module is offline or abnormal, and the second data acquisition module is online, method A is used to determine whether the coal transport vehicle has a risk of abnormal coal quality and classify the risk level. Method A is: determine the current driving state of the coal transport vehicle through the speed collected by the second data acquisition module, and determine the parking time t1 or driving distance m1 within T1 time after the vehicle leaves the starting point through the current driving state; and further compare t1 with t2 or compare m1 with m2, where t2 is the preset parking time within T1 time after the vehicle leaves the starting point; m2 is the preset driving distance within T1 time after the vehicle leaves the starting point; in this embodiment, T1=60s, t2=30s, and m2=250m are used as an example for explanation: the first analysis module makes the following judgment on the vehicle speed V collected by the second device every 60s:

[0072] If V>0, it is defined that the coal transport vehicle is currently in a driving state. Further, the driving distance m1 of the vehicle 60 seconds before the current moment is determined, and m1 is compared with m2.

[0073] If V=0, it is defined that the coal transport vehicle is currently in a parked state. Further, the parking time t1 of the vehicle 60 seconds before the current moment is determined, and t1 is compared with t2.

[0074] When t1>t2, the parking point is determined by the corresponding GPS longitude and latitude coordinates within the parking time t1; the point of interest data is obtained through the GPS longitude and latitude coordinates corresponding to the parking point and the Baidu map POI search interface, and the point of interest data is further filtered to determine the location information of the parking point; the parking point location information is matched with the set common address library data segment; if the parking point location information is not in the common address library data segment, it is judged that the vehicle coal quality has a secondary risk abnormality; the common address library data segment includes key fields such as gas stations, hotels, restaurants, gas stations, highways, national roads, provincial roads, and service areas.

[0075] When m1 < m2, determine the center point passed by the vehicle based on the GPS longitude and latitude coordinates within the previous T1 time of the current moment; determine the position information of the center point passed based on the center point passed and the map POI information; match the position information of the center point passed with the fields in the set common address library; if the position information of the center point passed is not within the fields in the common address library, it is determined that there is a secondary risk anomaly in the coal quality state of the vehicle.

[0076] The specific method for determining the parking point or the center point passed is as follows: Apply the clustering algorithm to the GPS longitude and latitude coordinates corresponding to each moment within the parking duration t1 at the current moment or the GPS longitude and latitude coordinates per unit time within the previous T1 time period at the current moment to obtain the parking point or the center point passed. The position information of the parking point or the center point passed both includes GPS longitude and latitude coordinates. The clustering algorithm selects any one of K-means, K-medoids, and hierarchical clustering.

[0077] When the first analysis module receives that both the first data acquisition module status and the second data acquisition module are in the online state, use method B to determine whether there is a coal quality anomaly risk for the coal transport vehicle and classify the risk level. Method B is as follows: Determine the current driving state of the coal transport vehicle through the vehicle speed collected by the second data acquisition module, and determine the parking duration t1 or the driving distance m1 within the previous T1 time after the vehicle leaves the starting point through the current driving state; and further compare t1 with t2 or compare m1 with m2.

[0078] When t1 > t2 or m1 > m2, screen out the data within the area to be detected from the data collected by the first data acquisition module; determine the geometric features of each object within the area to be detected from the data within the area to be detected; and judge the positional relationship between each object and the video monitoring limited area behind the coal transport vehicle; if the position of the object is directly above the video monitoring limited area behind the coal transport vehicle, judge the object type. If the object is an abnormal object, count the occurrence frequency N or the停留 duration T2 of the object; compare N with N1, or compare T2 with T4, where N1 is the preset frequency of the object appearing directly above the video monitoring limited area behind the coal transport vehicle.

[0079] When N > N1 or T2 > T4, it is determined that there is a primary risk anomaly in the coal quality of the vehicle.

[0080] When N < N1 or T2 < T4, it is determined that there is a secondary risk anomaly in the coal quality of the vehicle.

[0081] The specific method for determining the geometric features of each object within the area to be detected from the data within the area to be detected is as follows: Model the scene based on the point cloud data within the area to be detected, obtain the first feature information of each object through the perception algorithm. The first object feature information includes position, shape, and distance. Based on the first feature information of each object, classify the objects and identify the abnormal objects.

[0082] Alternatively, based on the video data in the area to be detected, pictures are obtained frame by frame, and each object and the second feature information of the object are identified through a target detection algorithm. The second feature information includes position and distance, and the objects are classified based on the second feature information to identify abnormal objects; the abnormal objects specifically include people, loader buckets, coal loaders, and scraper conveyors.

[0083] refer to Figure 3 In this embodiment, the model for identifying and locating abnormal objects such as people, loader buckets, coal loading machines, etc. uses an improved YOLOv11 model; wherein the improved YOLOv11 model mainly selects an improved mechanism for optimizing the network structure of the convolution or backbone network (backbone) or Neck part or the detection head or loss function in the YOLOv11 basic model, or integrates the improved mechanism of the network structure of the feature fusion layer in the backbone and neck structure (Neck). In this embodiment, the improved mechanism of the network structure of the feature fusion layer in the backbone and Neck is preferably integrated.

[0084] The specific improvement strategies are as follows: Figure 3 The backbone network in the improved YOLOv11 model shown in the figure introduces the universal Inverted Bottleneck (UIB, inverted bottleneck search block) structure of MobileNetV4 (lightweight CNN neural network) to the backbone network in the YOLOv11 model, replacing the original attention mechanism module (C2PSA). Although the UIB structure is simple, it successfully unifies several popular architectures, including Inverted Bottleneck (IB, inverted bottleneck), ConvNext (improved convolutional network based on ResNet50 model) and FeedForward Network (FFN, feedforward network), and introduces extra depth convolution (Extra Depthwise) technology, so that the UIB block allows flexible trade-offs between space / channel mixing, receptive field adjustment and computational utilization, so that the network can be adaptively adjusted according to the optimization goal.

[0085] Another improvement is to replace the neck structure of the YOLOv11 model with a BiFPN (bidirectional feature pyramid network) structure. Figure 3The neck structure (Neck) in the improved YOLOv11 model shown in the figure, BiFPN introduces a bidirectional connection, allowing information to propagate bidirectionally between different resolution levels, which helps to better fuse low-level and high-level features, and promotes the contextual propagation of features, improving the accuracy of object detection and segmentation. This improvement method can enable the feature map to be detected to fuse more feature information without increasing too much computational effort;

[0086] (2) The first state detection module is offline

[0087] When the second state detection module detects that the first state detection module is offline, the second analysis module uses the above-mentioned method A to determine whether the coal transport vehicle has a risk of abnormal coal quality and classify the risk level.

[0088] (3) If the second state detection module fails to obtain the heartbeat message sent by the second data acquisition module or the first state detection module within the preset time T4, the second analysis module determines that the coal quality of the vehicle has a secondary risk abnormality; the second data acquisition module and the first state detection module are provided with a backup battery, which can maintain communication with the second state detection module at the interval T4 when the vehicle is powered off.

[0089] S5. If the coal quality is judged to be abnormal through the first analysis module, the first analysis module sends the result of the coal quality judgment to the analysis result marking module, the analysis result marking module stores the result of the coal quality judgment, and sends the stored information to the result marking warning module, the result marking warning module saves the received result of the coal quality judgment and issues an alarm according to the corresponding abnormal risk level; the coal quality judgment result of the first analysis module includes: license plate, coal quality abnormality risk level, video data within T5 time before and after the current moment saved by the first analysis module, and GPS longitude and latitude coordinates of the vehicle at the current moment when the coal quality is abnormal.

[0090] If the coal quality is judged as abnormal through the second analysis module, the second analysis module sends the result of the coal quality judgment to the result marking warning module. The result marking warning module stores the result of the coal quality judgment of the second analysis module and issues an abnormal risk level alarm at the same time. The result of the coal quality judgment of the second analysis module includes the license plate, coal quality abnormality level information, and the vehicle GPS longitude and latitude information most recently received by the second data acquisition module.

[0091] In this embodiment, the result marking warning module is considered to be a coal transport vehicle monitoring and management platform. The transportation platform supervisors can view the information of the alarm vehicle through the result marking warning module. At the same time, the result marking warning module can also sort the vehicle alarm levels and send them to the regulatory agency. When the vehicle enters the destination, it will be inspected.

[0092] Example 2

[0093] This embodiment proposes a computer program product, including a computer program / instruction, which, when executed by a processor, implements the steps of the method for detecting anomalies in coal transport vehicles to prevent theft of coal in Embodiment 1.

[0094] The specific embodiments of the present invention enable those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention.

[0095] It should be understood that the present invention is not limited to what has been described above and that various modifications and changes may be made without departing from its scope. The scope of the present invention is limited only by the appended claims.

Claims

1. A vehicle anti-theft coal replacement anomaly detection method based on artificial intelligence, characterized in that: The following steps are involved: S1. Obtaining coal transport vehicle parameter information and transportation task information; S2. Determine the area to be inspected according to the coal transport vehicle parameter information and transport task information; S3, obtaining video data or point cloud data in the area to be detected through the first data acquisition module, and obtaining corresponding sampling time, vehicle speed, and vehicle GPS longitude and latitude coordinates through the second data acquisition module; judging the status information of the first data acquisition module and the second data acquisition module through the first state detection module; judging the status information of the second data acquisition module and the first state detection module through the second state detection module; S4, identifying whether the coal transport vehicle has a risk of abnormal coal quality and classifying the risk level through the data acquired by the first data acquisition module, the data acquired by the second data acquisition module, the status information of the first data acquisition module, the status information of the second acquisition module and the status information of the first status detection module; S5. Provide corresponding alarm level prompts according to the risk level of coal quality abnormality.

2. According to claim 1, a method for detecting abnormalities in vehicles to prevent theft of coal based on artificial intelligence is characterized in that: In S1, the coal transport vehicle parameter information includes vehicle width, vehicle height, vehicle length, and license plate; the transportation task information includes the starting point name and starting point location coordinates, and the end point name and end point location coordinates.

3. The method for detecting abnormality of vehicle coal substitution based on artificial intelligence according to claim 1 is characterized in that: In S2, the area to be detected includes the limited driving area of ​​the coal transport vehicle and the limited video monitoring area behind the coal transport vehicle; the method for setting the limited driving area of ​​the coal transport vehicle is: inputting the parameter information of the coal transport vehicle and the transportation task information into the navigation software to obtain the limited driving area of ​​the coal transport vehicle; Based on the restricted driving area of ​​coal transport vehicles, a restricted video monitoring area is set up behind the coal transport vehicles.

4. The method for detecting abnormality of coal transportation vehicles against theft of coal according to claim 1, characterized in that: In S3, the first data acquisition module includes a visual sensor or a laser radar sensor; the second data acquisition module includes a GPS locator or a driving recorder; The status information of the first data acquisition module includes: online status, offline status, and abnormal status; The status information of the second data acquisition module includes: online status, dormant status, abnormal status; The status information of the first status detection module includes: online status, sleep status, and abnormal status. The abnormal state of the first data acquisition module includes a black screen, a gray screen, a white screen, or a window of the first data acquisition module is dirty, blocked, or has an abnormal angle; The abnormal states of the second data acquisition module and the first state detection module are both states when the timed wake-up function is not executed in the sleep state.

5. The method for detecting abnormality of coal replacement in coal transport vehicles according to claim 4, characterized in that: The first state detection module includes a first data acquisition module, the first data acquisition module is used to receive data sent by the first data acquisition module and the second data acquisition module, and the first state detection module determines the state information of the first data acquisition module and the second data acquisition module according to the state of the data received by the first data acquisition module; The second status detection module determines the status information of the second data acquisition module and the first data acquisition module according to the heartbeat detection mechanism.

6. The method for detecting abnormality of coal transportation vehicles against theft of coal according to claim 1, characterized in that: In S4, the levels of coal quality abnormality risk are divided into level one and level two. The method for identifying whether a coal transport vehicle has coal quality abnormality risk and classifying the risk level specifically includes the following sub-steps: S401. Determine whether the first data acquisition module, the second data acquisition module, and the first status detection module meet the first condition, the second condition, or the third condition; if the first condition is met, proceed to S402; if the second condition is met, proceed to S403; if the third condition is met, then determine that there is a secondary risk anomaly in the vehicle coal quality; The first condition is: the first data acquisition module is in an offline or abnormal state; the second data acquisition module is in an online state; the first status detection module is in an online or offline state; The second condition is: the first data acquisition module is in an online state; the second data acquisition module is in an online state; the first status detection module is in an online state; The third condition is: within the preset time T4, the second status detection module does not receive the heartbeat packet sent by the second data acquisition module or the first status detection module; S402. Determine the current driving state of the coal transport vehicle based on the vehicle speed collected by the second data acquisition module, and determine the parking duration t1 or the driving distance m1 within the first T1 time after the vehicle leaves the starting point based on the current driving state; And compare t1 with t2 or compare m1 with m2, where t2 is the preset parking duration within the first T1 time after the vehicle leaves the starting point; m2 is the preset driving distance within the first T1 time after the vehicle leaves the starting point; When t1 > t2, determine the parking point based on the GPS longitude and latitude coordinates within the parking duration t1; determine the location information of the parking point from the parking point and the map POI information; Match the location information of the parking point with the data segment of the set common address library; If the location information of the parking point is not within the data segment of the common address library, determine that there is a secondary risk anomaly in the vehicle coal quality; When m1 < m2, determine the vehicle passing center point based on the GPS longitude and latitude coordinates within the first T1 time before the current moment; determine the location information of the passing center point from the passing center point and the map POI information; match the location information of the passing center point with the fields in the set common address library; if the location information of the passing center point is not within the fields of the common address library, then determine that there is a secondary risk anomaly in the vehicle coal quality status; S403. Determine the current driving state of the coal transport vehicle based on the vehicle speed collected by the second data acquisition module, and determine the parking duration t1 or the driving distance m1 within the first T1 time after the vehicle leaves the starting point; And compare t1 with t2 or compare m1 with m2; When t1 > t2 or m1 > m2, screen out the data within the area to be detected from the data collected by the first data acquisition module; determine the geometric features of each object within the area to be detected from the data within the area to be detected; And judge the position relationship between each object and the video monitoring limited area behind the coal transport vehicle; If the object position is directly above the video monitoring limited area behind the coal transport vehicle, judge the object type. If the object is an abnormal object, count the occurrence frequency N or the停留时长 T2 of the object; compare N with N1, or compare T2 with T4, where N1 is the preset frequency of the object appearing directly above the video monitoring limited area behind the coal transport vehicle; When N > N1 or T2 > T4, determine that there is a primary risk anomaly in the vehicle coal quality; When N < N1 or T2 < T4, it is determined that there is a secondary risk anomaly in the coal quality of the vehicle.

7. The method for detecting abnormality of coal transportation vehicles against theft of coal according to claim 6, characterized in that: The specific method for determining the parking point or the passing center point is as follows: Apply the clustering algorithm to the GPS longitude and latitude coordinates corresponding to each moment within the parking duration t1 at the current moment or the GPS longitude and latitude coordinates per unit time in the previous T1 time period at the current moment to obtain the parking point or the passing center point. The position information of the parking point or the position information of the passing center point both include GPS longitude and latitude coordinates. The clustering algorithm can be any one of K-means, K-medoids, and hierarchical clustering.

8. The method for detecting abnormality of coal transportation vehicles against theft of coal according to claim 6, characterized in that: The specific method for determining the geometric features of each object in the area to be detected from the data in the area to be detected is as follows: Based on the point cloud data in the area to be detected, model the scene, and obtain the first feature information of each object through the perception algorithm. The first object feature information includes position, shape, and distance. Based on the first feature information of each object, classify the objects and identify the abnormal objects. Alternatively, obtain pictures frame by frame based on the video data in the area to be detected, identify each object and the second feature information of the object through the object detection algorithm. The second feature information includes position and distance. Based on the second feature information, classify the objects and identify the abnormal objects. The abnormal objects specifically include people, the bucket of a loader, a coal loading machine, and a scraper conveyor.

9. A detection system for implementing the method for detecting abnormality of coal transportation vehicles against theft of coal as claimed in any one of claims 1 to 8, characterized in that: It includes a first data acquisition module, a second data acquisition module, a first status detection module, a second status detection module, a first analysis module, an analysis result marking module, a second data acquisition module, a second analysis module, and a result marking and warning module. Among them, the first data acquisition module is connected to the input of the first status detection module. The second data acquisition module is respectively connected to the inputs of the first status detection module, the second status detection module, and the second data acquisition module. The output of the first status detection module is respectively connected to the inputs of the first analysis module, the second status detection module, and the second data acquisition module. The output of the first analysis module is connected to the input of the analysis result marking module. The output of the analysis result marking module is connected to the input of the result marking and warning module. The outputs of the second status detection module and the second data acquisition module are respectively connected to the input of the second analysis module. The output of the second analysis module is connected to the input of the result marking and warning module. The first analysis module determines whether there is an abnormal risk in the coal quality of the coal transport vehicle and classifies the risk level according to the data collected by the first data acquisition module, the data collected by the second data acquisition module, the status information of the first data acquisition module, and the status information of the second data acquisition module. The analysis result marking module is used to store the coal quality abnormality judgment result of the first analysis module and send the stored information to the second analysis module. The second data acquisition module is used to obtain the data collected by the second data acquisition module and the coal quality abnormality judgment result stored in the first status detection module. The second analysis module determines whether there is an abnormal risk in the coal quality of the coal transport vehicle and classifies the risk level according to the status information of the second data acquisition module and the status information of the first status detection module. or judging whether the coal quality of the coal transport vehicle has an abnormal risk and classifying the risk level according to the data collected by the second data collection module; The result marking warning module is used to store the coal quality abnormality judgment result of the second analysis module or receive the coal quality abnormality judgment result of the second analysis module sent by the analysis result marking module, and issue a corresponding level of alarm; The coal quality abnormality judgment result of the first analysis module includes: the license plate, the coal quality abnormality risk level, the video data within the T5 time before and after the current moment saved by the first analysis module, and the GPS longitude and latitude coordinates of the vehicle at the current moment when the coal quality is abnormal; The coal quality abnormality judgment result of the second analysis module includes the license plate, the coal quality abnormality risk level, and the vehicle GPS longitude and latitude information most recently received by the second data acquisition module.

10. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the steps of the method for detecting abnormal coal substitution in a coal transport vehicle as described in any one of claims 1 to 8 are implemented.

Images