CNN algorithm-based off-line map vehicle monitoring and management system

Abstract

The invention discloses an off-line map vehicle monitoring and management system based on a CNN algorithm, and the system achieves the stable positioning in a special scene through embedding an off-line map, and integrates the functions of vehicle positioning, alarm, track playback, personnel authority and the like. In the vehicle real-time positioning function, the CNN algorithm extracts features from positioning data, filters positioning noise caused by environmental factors, and improves accuracy. A CNN algorithm is applied to the track playback function, a large amount of vehicle track data are regarded as two-dimensional plane lines, and the analysis efficiency is improved through classification, anomaly detection and feature extraction. And the vehicle-mounted equipment communicates with the system in real time to obtain the vehicle position, and triggers an alarm to inform personnel when abnormity occurs. The system establishes an intranet and extranet security barrier through a gatekeeper, and ensures efficient real-time and confidentiality of data exchange. In the aspect of personnel login authority management, a CNN algorithm analyzes behavior characteristics such as operation habits of login personnel, whether login behaviors are abnormal or not is judged, and the system safety is enhanced.

Description

Technical Field

[0001] The present invention relates to the technical fields of map development, monitoring, data processing, network security, etc., and specifically to an offline map vehicle monitoring and management system based on a CNN algorithm. Background Art

[0002] With the continuous development of the transportation industry in modern society, the number of vehicles has increased dramatically, and effective monitoring and management of vehicles has become an important demand in many fields. In the logistics industry, efficient vehicle monitoring and management helps optimize transportation routes, improve distribution efficiency, reduce costs and ensure the safety of goods; in the field of public transportation, it can improve the scientific nature of operation scheduling and ensure the safety and convenience of passengers; for corporate fleet management, it can achieve reasonable deployment of vehicles, reduce operational risks and increase vehicle service life.

[0003] Patent No. CN118565499A discloses a road transport vehicle monitoring system and monitoring method. This invention is road transport vehicle monitoring and belongs to the field of road planning. The system includes a main control terminal, a vehicle-mounted terminal and a communication terminal. The main control terminal has a road information module for planning routes, a monitoring module for monitoring vehicle-mounted terminal data and a communication module; the vehicle-mounted terminal includes a driving information module for receiving route information, a vehicle positioning module and a camera module for driving environment monitoring. The system can plan routes in advance, allowing drivers to drive according to the routes and monitor problems during driving in real time.

[0004] Patent number CN112752221B discloses a vehicle monitoring system and method. The invention integrates vehicle data into a graphic display platform (map) to achieve management and monitoring. The vehicle can be located in real time, its data, status, business information can be viewed, and it can be unlocked remotely. The vehicle data can be retrieved and located according to a variety of conditions, and the number of vehicles in each area and their real-time location can be viewed on the web page.

[0005] However, the above technologies have certain defects and shortcomings. Road transport vehicle monitoring lacks the functions of vehicle monitoring systems such as data integration display, remote operation, and multi-condition retrieval. For a vehicle monitoring system and method, the operation of the system in an offline environment is not considered, and there is a lack of feature extraction of large amounts of data. In addition, the system lacks technologies and functions such as trajectory analysis, network security, and personnel authority management. The overall technical depth and functional integrity are insufficient, making it difficult to ensure the efficiency of vehicle monitoring and the security of the system in a complex environment. Summary of the invention

[0006] In view of the defects in the prior art, the present invention provides an offline map vehicle monitoring and management system based on the CNN algorithm, which integrates vehicle data and displays it on a graphic platform, realizes remote operation, and meets the multi-faceted needs of vehicle management. The system embeds an offline map to cope with complex and special environments, uses the CNN algorithm to optimize positioning and trajectory analysis, ensures data security through a network gate, and adds personnel authority management functions, thereby ensuring that the system can operate normally, respond quickly, and have reliable protection capabilities in complex environments.

[0007] The solutions provided in this application are: An offline map vehicle monitoring and management system based on CNN algorithm, the system includes: a vehicle positioning module, a track playback module, a hardware communication module, a data security module, an electronic fence alarm module and a personnel authority management module; The overall architecture includes the interaction between vehicle-mounted devices and the system, and realizes comprehensive monitoring and management of the vehicle through the collaboration of the main control end, the vehicle-mounted end, and the communication end.

[0008] Furthermore, the vehicle positioning module is embedded in an offline map, and the offline map stores preset geographic information data, and the approximate location of the vehicle is determined by the geographic information data in the offline map; The vehicle positioning module uses the CNN algorithm to determine the exact location of the vehicle, identify the positioning noise caused by environmental factors, and filter it out.

[0009] Furthermore, the trajectory playback module is used to convert a large amount of vehicle trajectory data into two-dimensional plane lines, process the two-dimensional plane lines using a CNN algorithm, and classify the two-dimensional plane lines into different types of trajectory patterns.

[0010] Furthermore, the different types of trajectory modes include normal driving trajectory, turning trajectory, acceleration or deceleration trajectory.

[0011] Furthermore, the trajectory replay module analyzes different types of trajectory patterns based on the CNN algorithm to obtain abnormal trajectories; the trajectory replay module extracts features from the trajectory data based on the CNN algorithm, the features including the curvature, length, and direction change information of the trajectory, and further analyzes the vehicle's driving behavior based on the features.

[0012] Furthermore, the personnel authority management module adopts CNN algorithm to analyze the operation behavior characteristics of the logged-in personnel, conducts in-depth analysis of the operation behavior characteristics, and establishes an operation behavior model. When a person logs in, the system will compare the operation behavior of the current logged-in personnel with the operation behavior model. If the operation behavior deviates significantly from normal operation habits, it is judged as abnormal login behavior; the system takes corresponding measures.

[0013] Furthermore, the operation behavior characteristics include login time, operation frequency, and operation sequence information.

[0014] Furthermore, the data security module establishes an internal and external network security barrier through a network gate.

[0015] Furthermore, the electronic fence alarm module is used to detect whether the user is within the area set in the system. When the vehicle travels outside the set area, the electronic fence alarm module triggers the alarm mechanism.

[0016] Furthermore, the hardware communication module is configured on the vehicle, has an intelligent positioning device and a three-axis acceleration sensor, is powered by a vehicle battery, and is used to realize information interaction between the vehicle and the system.

[0017] Compared with the prior art, the present invention has the following beneficial technical effects: By embedding offline maps to achieve stable positioning in special scenarios, operators can obtain vehicle location information in any scenario, thus improving the comprehensiveness of operations.

[0018] In the real-time vehicle positioning function, the CNN algorithm is used to extract features from the positioning data and filter out the positioning noise caused by environmental factors, which effectively improves the positioning accuracy and thus improves the performance of the entire system.

[0019] The trajectory playback function uses the CNN algorithm to treat a large amount of vehicle trajectory data as two-dimensional plane lines. Through operations such as classification, anomaly detection, and feature extraction, it can process trajectory data more efficiently and improve the performance of trajectory playback, such as finding trajectories in a specific time period more quickly or detecting abnormal trajectories more accurately.

[0020] A network gateway is used to establish an internal and external network security barrier to ensure efficient and real-time data exchange. During system operation, the data transmission speed between the vehicle and the system is faster. When the on-board equipment communicates in real time to obtain the vehicle's position, the data can be transmitted quickly. When an abnormality occurs, an alarm can be quickly triggered to notify personnel, thereby improving the response speed of the system.

[0021] In terms of personnel login permission management, the CNN algorithm is used to analyze the operating habits and other behavioral characteristics of the login personnel to determine whether the login behavior is abnormal. This adds a security line for system operation. Operators can manage the system more confidently without worrying about illegal personnel invading the system and performing erroneous operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 The vehicle monitoring and management system architecture diagram of the present invention; Figure 2 It is a schematic diagram of the functional modules of the vehicle monitoring and management system of the present invention; Figure 3 This is a convolutional neural network structure diagram of the vehicle monitoring and management system of the present invention; Figure 4 This is an interface diagram of a vehicle real-time monitoring function module of the vehicle monitoring and management system of the present invention; Figure 5 This is a vehicle execution record interface diagram of the vehicle monitoring and management system of the present invention; Figure 6 This is an interface diagram of the vehicle trajectory playback function module of the vehicle monitoring and management system of the present invention; Figure 7 This is a diagram of the vehicle management function module interface of the vehicle monitoring and management system of the present invention; Figure 8 This is an interface diagram of a vehicle alarm management function module of a vehicle monitoring and management system of the present invention; Fig. 9 This is an interface diagram of the vehicle electronic fence function module of the vehicle monitoring and management system of the present invention. DETAILED DESCRIPTION

[0023] The following will refer to the accompanying drawings given in the embodiments of the present invention to clearly and completely explain the technical solutions in the embodiments of the present invention. It should be clear that the embodiments described here are only a part of all the embodiments of the present invention, and do not cover all the embodiments of the present invention.

[0024] This embodiment provides an offline map vehicle monitoring and management system based on a CNN algorithm, including: Offline map technology is embedded in the system positioning function; The CNN algorithm is used to process vehicle positioning data in the system positioning function; Use CNN algorithm to extract features and mine effective information in the system trajectory playback function; The system uses real-time communication and alarm triggering technology; Use network gates to establish a security barrier in system data security; Incorporating electronic fence technology into the vehicle monitoring system; The CNN algorithm is used to analyze personnel behavior characteristics in system personnel permissions to enhance system security.

[0025] This patent discloses an offline map vehicle monitoring and management system based on CNN algorithm, which covers multiple functional modules, including vehicle positioning, trajectory playback, hardware communication, data security, electronic fence alarm and personnel authority management. The overall architecture includes the interaction between vehicle-mounted equipment and the system, and realizes comprehensive monitoring and management of the vehicle through the collaboration of the main control end, the vehicle-mounted end and the communication end.

[0026] The vehicle positioning module adopts a technical solution of embedding offline maps to solve the problem of unstable vehicle positioning in special scenarios. The offline map pre-stores geographic information data. When the vehicle is in a special scenario (such as remote mountainous areas, underground parking lots and other areas with poor network signals or no network), the system can use the data in the offline map to determine the approximate location of the vehicle. This technical solution does not rely on network connection, ensuring the stability of the vehicle positioning function in various complex environments. In the real-time vehicle positioning function, the CNN (convolutional neural network) algorithm is used. The CNN algorithm has a powerful feature extraction capability, which extracts features from the positioning data. In actual operation, the positioning data may be interfered by environmental factors (such as building occlusion, electromagnetic interference, etc.), resulting in inaccurate positioning. The CNN algorithm can identify these positioning noises caused by environmental factors and filter them out. In this way, the system improves the accuracy of real-time vehicle positioning, making the vehicle's location information more accurate and reliable, and providing an accurate data basis for subsequent monitoring and management.

[0027] The trajectory playback module adopts a technical solution based on the CNN algorithm, which regards a large amount of vehicle trajectory data as two-dimensional plane lines, and uses the advantages of the CNN algorithm in image processing to process these trajectory data. In trajectory classification, the CNN algorithm can identify different types of trajectory patterns for trajectory data. For example, different modes such as normal driving trajectory, turning trajectory, acceleration or deceleration trajectory can be accurately classified. This classification helps to analyze the driving behavior of the vehicle. For example, the frequency of occurrence of different types of trajectories can be counted to evaluate the driver's driving habits or road usage. In anomaly detection, the CNN algorithm can detect abnormal trajectories through in-depth analysis of trajectory data. When the vehicle's driving trajectory deviates greatly from the normal mode, such as suddenly deviating from the scheduled route, staying in a non-parking area for a long time, etc., the system can detect it in time. This is of great significance for ensuring vehicle safety, preventing theft or illegal driving, etc. In feature extraction, the CNN algorithm can also extract useful features from the trajectory data. These features may include information such as the curvature, length, and direction change of the trajectory. By extracting these features, the driving behavior of the vehicle can be further analyzed in depth, such as judging whether the driver is fatigued driving (through features such as irregularity of the trajectory), which improves the efficiency and depth of trajectory analysis.

[0028] The hardware communication module is equipped with an intelligent positioning device and a three-axis acceleration sensor, powered by a vehicle battery, and has a low-energy operation mode when the vehicle is stopped and a low-battery reminder. It is a key component for realizing information exchange between the vehicle and the system. It has the ability to obtain vehicle information in real time, including but not limited to the vehicle's location, speed, engine status, tire pressure and other data related to the vehicle's operating status. The hardware communication module sends the acquired vehicle information to the network gate through a specific communication protocol. Its communication method adopts Beidou and base station positioning technology to ensure that data can be transmitted stably and efficiently.

[0029] The data security module of this patent establishes a security barrier between the internal and external networks through a network gate. The network gate is a security device specially used to isolate the internal and external networks. It ensures the confidentiality of data while ensuring efficient and real-time data exchange. During the operation of the system, vehicle-related data (such as positioning data, trajectory data, vehicle status data, etc.) needs to be exchanged between the internal and external networks, and the network gate strictly controls the flow of these data. It only allows authorized data to pass, prevents illegal intrusion and data leakage from external networks, and protects the security of vehicle-related data in the system.

[0030] The electronic fence alarm module allows users to set a specific area range in the system. When the vehicle runs outside the set area, the module can detect this situation and trigger the alarm mechanism. The working principle of the electronic fence module is based on the vehicle's positioning information, and it determines whether the vehicle has crossed the boundary by comparing it with the preset geographic fence boundary. Once the alarm is triggered, the system will promptly notify relevant personnel (such as drivers, monitoring managers, etc.), and the notification method can be in various forms such as SMS and APP push, so that relevant personnel can take timely measures to track the vehicle or contact the driver to understand the situation.

[0031] The personnel authority management module adopts the CNN algorithm to analyze the operating habits and other behavioral characteristics of the logged-in personnel. The system will record the operating behavior of the logged-in personnel, including information such as login time, operation frequency, and operation sequence. The CNN algorithm conducts in-depth analysis of these behavioral characteristics and establishes a model of normal operating behavior. When a person logs in, the system will compare the operating behavior of the current logged-in personnel with the normal model. If there is a large deviation between the login behavior and the normal operating habits, such as abnormal login time, inconsistent operation sequence, etc., the system will judge it as abnormal login behavior. At this time, the system can take corresponding measures, such as blocking login, secondary verification, system alarm, etc., thereby enhancing the security of the system, preventing illegal personnel from obtaining system permissions, and ensuring the security of vehicle-related data and operations in the system.

[0032] Figure 1This is the architecture diagram of the vehicle monitoring system. The vehicle-mounted positioning terminal is installed in the vehicle, which contains key components such as the intelligent positioning device and the three-axis acceleration sensor. The intelligent positioning device uses Beidou and base station positioning technology to accurately obtain the vehicle's location information. The three-axis acceleration sensor is responsible for detecting the acceleration changes of the vehicle during driving. These data are of great significance for analyzing the vehicle's driving status (such as acceleration, deceleration, turning, etc.). At the same time, the entire vehicle-mounted positioning terminal is powered by the vehicle battery, and in order to adapt to special circumstances during vehicle parking, it also has a low-energy operation mode and a low-battery reminder function.

[0033] The vehicle positioning terminal obtains various operating status-related data of the vehicle in real time, including but not limited to the vehicle's location, speed, engine status, and vehicle status. Inside the vehicle positioning terminal, the collected data is first sorted and formatted, and different types of data are arranged in a predetermined format for subsequent transmission and processing. For example, the location information is standardized in the format of longitude and latitude, and the speed data is converted into a unified unit.

[0034] After preliminary processing, the vehicle data is sent from the vehicle positioning terminal to the external network server through a specific communication protocol to ensure stable and efficient data transmission between the vehicle and the external network server. The communication protocol can encapsulate, verify and correct the data to deal with possible interference and errors during the transmission process. For example, the communication protocol uses data encryption technology to encrypt sensitive information of the vehicle (such as engine status, etc.) to prevent the data from being stolen or tampered with during the transmission process.

[0035] After receiving the data sent from the vehicle positioning terminal, the external network server will further process the data. First, the external data is decrypted, and then the security of the external network server itself is checked to ensure that the process of receiving data is not subject to external malicious attacks. Next, the external network server will classify and store the data, and store different types of data (such as location data, speed data, etc.) in corresponding database tables for subsequent query and processing.

[0036] The data in the external network server needs to pass through the network gate to enter the internal network server. As a security device specially used to isolate the internal and external networks, the network gate plays a vital role in this process. The network gate will strictly check and filter the data transmitted from the external network server. Only authorized data is allowed to pass through. This authorization process is based on pre-set security policies. For example, data that conforms to a specific format, has a legitimate source and has been authenticated can pass through the network gate. When the data passes through the network gate, the network gate will also conduct in-depth security detection on the data, including virus scanning, malicious code detection, etc., to prevent illegal intrusion and data leakage from the external network.

[0037] After the security check and filtering of the network gate, the vehicle data is transmitted to the intranet server. Before receiving the data, the intranet server will interact with the network gate to perform identity authentication and data integrity verification to ensure that the received data is safe, complete and legal. Once the data is successfully transmitted to the intranet server, the intranet server will further integrate and store the data according to its own storage structure and management strategy. For example, different types of data of the same vehicle are stored in association so that they can be easily queried and displayed in the vehicle management system.

[0038] After the vehicle management system receives the data returned from the intranet server, it will perform final processing and display on the data. In order to improve the stability of the system in complex environments, this system is embedded with an offline map. For vehicle location data, the current location of the vehicle will be displayed on the map with a specific icon; for vehicle driving trajectory data, the trajectory will be displayed on the map in the form of lines according to the requirements of the trajectory playback module. Because of the embedded offline map, even in complex environments with poor network signals or no network, the system can still perform operations such as displaying vehicle location and trajectory normally.

[0039] Figure 2 It is the functional module of the vehicle monitoring and management system of the present invention, which includes six functional modules: real-time monitoring, trajectory playback, alarm management, vehicle management, electronic fence, and user authority.

[0040] Figure 3It is a convolutional neural network structure. The CNN convolution algorithm provided in this embodiment is a classic LeNet-5 convolutional neural network CNN structure, which is mainly composed of a convolution layer, a pooling layer and a fully connected layer. The convolution kernel in the convolution layer can automatically learn the local features in the data. Through the stacking of multiple convolution layers (C1, C3, C5 in LeNet-5), the network can gradually extract high-level features (such as edges, textures) from low-level features (such as edges, textures) to high-level features (such as overall characteristics of data, specific patterns, etc.). The pooling layer (S2 and S4) reduces the dimension of the data through downsampling operations, which not only reduces the amount of calculation, but also makes the network have a certain invariance to small changes in the input data (such as translation, rotation, etc.). Due to the dimensionality reduction and feature selection of the pooling layer, the network has a certain robustness to the noise in the input data. In practical applications, LeNet-5 can well ignore the influence of these noises and focus on extracting key features. In the LeNet-5 convolution layer, the convolution kernel performs convolution operations on the overall data, and the same convolution kernel shares parameters at different positions. This greatly reduces the number of parameters in the network. A 5×5 convolution kernel of the LeNet-5 neural network slides in 32×32 data, which only requires 25 parameters, while not using the LeNet-5 neural network requires 32×32 = 1024 parameters. The reduction in parameters means a reduction in the amount of calculation, allowing LeNet-5 to run with relatively low computing resources. This is very important for calculations with high real-time requirements (such as real-time positioning, trajectory playback, and personnel feature recognition), which enables the system to process data faster and obtain results.

[0041] Figure 4 This is the vehicle real-time monitoring function module of the vehicle monitoring and management system of the present invention. On the real-time monitoring map interface, the vehicle location information presents the real-time location of the vehicle in the form of a dynamic icon. Among them, the icon color has a special meaning, which is used to indicate the status of the vehicle: green means that the vehicle is driving normally, red means that the vehicle is in an abnormal state, and gray means that the vehicle is in a parked state. At the same time, dynamic data such as the speed of the vehicle will be displayed in the form of a chart on the monitoring interface, and the administrator can grasp the driving status of the vehicle in real time. In addition, a search box is provided on the left side of the system. The administrator can perform a fuzzy search for the license plate in the search box, so as to accurately locate the corresponding vehicle on the map. Moreover, when the administrator clicks on the vehicle icon in the map, he can view the current status of the vehicle, the execution of the task, and the vehicle's running trajectory and other related data.

[0042] Figure 5 The vehicle execution record of the vehicle monitoring and management system of the present invention records the historical data of all vehicle execution tasks, and the user can view the vehicle historical trajectory through the vehicle execution record or from real-time monitoring.

[0043] Figure 6The vehicle trajectory playback function module of the vehicle monitoring and management system of the present invention treats a large amount of vehicle trajectory data as two-dimensional plane lines, and uses the advantages of the CNN algorithm in image processing to process these trajectory data. In trajectory classification, the CNN algorithm can identify different types of trajectory patterns for trajectory data. For example, different modes such as normal driving trajectory, turning trajectory, acceleration or deceleration trajectory. The vehicle position coordinates at different intervals are recorded through different modes, and the trajectory data is smoothed to remove some abnormal points caused by positioning errors, so that the trajectory is smoother and continuous. After the user selects the vehicle and time period, the processed trajectory data is displayed in the form of lines on the map interface. Important places (such as parking points, alarm points) are marked on the trajectory, and the marked points are marked with different colors (red is the vehicle alarm point, yellow is the parking point). Click these important marked points, and the interface displays detailed information of these places (such as parking time, alarm information, etc.).

[0044] Figure 7 The vehicle management function module of the vehicle monitoring and management system of the present invention is used. The administrator enters the basic information of the vehicle, such as vehicle model, license plate number, frame number, vehicle color, affiliated unit, etc. This information is stored in the basic vehicle information database table of the intranet server. The administrator or authorized user can query the basic information of the vehicle according to the vehicle identification (such as the license plate number), and modify the vehicle information when it changes (such as the change of the affiliated unit of the vehicle). The user can query the historical status data of the vehicle in a certain period of time in the past. For example, check the number of engine starts and average driving speed of a certain vehicle in the past week, so as to analyze the use of the vehicle. Formulate a vehicle scheduling plan according to business needs. For example, arrange a certain vehicle to perform a specific transportation task at a specific time. The scheduling plan includes information such as vehicle arrangement, departure time, destination, etc., which is stored in the vehicle scheduling database table. During the vehicle's execution of the scheduling task, the vehicle's driving condition is monitored through the real-time monitoring function to ensure that the vehicle is driving according to the scheduling plan. If an abnormal situation occurs (such as vehicle failure, traffic jam), the scheduling plan is adjusted in time.

[0045] Figure 8This is the vehicle alarm management function module of the vehicle monitoring and management system of the present invention. In the vehicle management system, the administrator can set the alarm triggering conditions for abnormal vehicle status. For example, when the vehicle speed exceeds the specified maximum speed (such as the speed limit of 120km / h on the highway), the speeding alarm is triggered; when the tire pressure is lower than the safety value, the tire pressure alarm is triggered; when the vehicle deviates from the preset route by a certain distance, the route deviation alarm is triggered. For vehicle-mounted equipment, such as positioning terminals, acceleration sensors, etc., when the equipment has a fault signal (such as positioning signal loss, sensor data abnormality), it is also set as a triggering alarm condition. When the vehicle data meets the alarm triggering condition, an alarm message is generated. The alarm information includes detailed information such as vehicle identification, alarm type, alarm time, and vehicle location. The alarm information can also be sent to relevant personnel through an SMS gateway or an APP push server. For SMS notification, the alarm information is packaged in the SMS format and sent to the pre-set mobile phone number of the administrator. For APP push, the alarm information is pushed to the relevant users of the vehicle management APP, the alarm content is displayed in the notification bar of the APP, and the detailed information can be clicked to view.

[0046] Fig. 9 The electronic fence function module of the vehicle monitoring and management system of the present invention is an authorized user (such as an administrator or a team leader) who can define an electronic fence area through a map interface. The range of the fence can be determined by drawing geometric shapes such as polygons and circles, and multiple fence areas can be set. At the same time, a name (such as "company parking lot", "warehouse area") and related attributes (such as entry permission, exit permission, etc.) are set for each fence area. The parameters of the defined electronic fence area (such as coordinate points, shapes, names, attributes) are stored in the electronic fence database table of the intranet server. The vehicle location information obtained through real-time monitoring is compared with the fence area in the electronic fence database. When a vehicle enters or leaves the fence area, it is determined whether an alarm is triggered according to the attributes of the fence. If the behavior of the vehicle triggers the electronic fence alarm, the alarm information is sent to the relevant personnel according to the alarm transmission method in the alarm management function module. For example, when a vehicle leaves the "company parking lot" without authorization, a text message or APP push notification is sent to the person in charge of the vehicle.

[0047] The user authority module of the vehicle monitoring and management system of the present invention defines different user roles in the vehicle management system, such as administrator, fleet captain, driver, ordinary monitoring personnel, etc. Each role has a different scope of authority. The administrator role has the highest authority and can manage the system comprehensively, including basic vehicle information management, scheduling management, user authority management, etc. The fleet captain can manage the vehicles of the fleet, such as viewing the vehicle status, arranging scheduling tasks, etc. The driver can view the relevant information of the vehicle he drives, such as driving trajectory, real-time location, etc. Ordinary monitoring personnel can only view the basic monitoring information of the vehicle (such as location, status). These authority settings are stored in the user authority database table of the intranet server. When the user logs in to the vehicle management system, first enter the user name and password. The system sends the user name and password to the intranet server for verification. The intranet server checks whether the user name and password match according to the information in the user authority database table, and determines the user's role. The system uses the CNN algorithm to analyze the behavioral characteristics such as the operating habits of the login personnel, and in the process of using the system, the user's operating behavior (such as login time, operation frequency, sequence, etc.) is deeply analyzed. If the operation behavior deviates significantly from the user's normal behavior model, the system's verification steps (verification code input, secondary password verification) will be triggered. If the verification steps are reached three times, the system will automatically exit and issue a system alarm.

Claims

1. An offline map vehicle monitoring and management system based on CNN algorithm, characterized in that: The system includes: a vehicle positioning module, a track playback module, a hardware communication module, a data security module, an electronic fence alarm module and a personnel authority management module; The overall architecture includes the interaction between vehicle-mounted devices and the system, and realizes comprehensive monitoring and management of the vehicle through the collaboration of the main control end, the vehicle-mounted end, and the communication end.

2. According to claim 1, an offline map vehicle monitoring and management system based on CNN algorithm is characterized in that: The vehicle positioning module is embedded in an offline map, and the offline map stores preset geographic information data, and the approximate location of the vehicle is determined by the geographic information data in the offline map; The vehicle positioning module uses the CNN algorithm to determine the exact location of the vehicle, identify the positioning noise caused by environmental factors, and filter it out.

3. According to claim 1, an offline map vehicle monitoring and management system based on CNN algorithm is characterized in that: The trajectory playback module is used to convert a large amount of vehicle trajectory data into two-dimensional plane lines, process the two-dimensional plane lines using a CNN algorithm, and classify the two-dimensional plane lines into different types of trajectory patterns.

4. The offline map vehicle monitoring and management system based on CNN algorithm according to claim 3 is characterized in that: The different types of trajectory modes include normal driving trajectory, turning trajectory, acceleration or deceleration trajectory.

5. The offline map vehicle monitoring and management system based on CNN algorithm according to claim 1 is characterized in that: The trajectory replay module analyzes different types of trajectory patterns based on the CNN algorithm to obtain abnormal trajectories; the trajectory replay module extracts features from trajectory data based on the CNN algorithm, the features including the curvature, length, and direction change information of the trajectory, and further analyzes the vehicle's driving behavior based on the features.

6. The offline map vehicle monitoring and management system based on CNN algorithm according to claim 1 is characterized in that: The personnel authority management module uses the CNN algorithm to analyze the operation behavior characteristics of the login personnel, conducts in-depth analysis of the operation behavior characteristics, and establishes an operation behavior model. When a person logs in, the system will compare the operation behavior of the current login personnel with the operation behavior model. If the operation behavior deviates greatly from the normal operation habits, it is judged as abnormal login behavior; The system takes appropriate measures.

7. The offline map vehicle monitoring and management system based on CNN algorithm according to claim 6 is characterized in that: The operation behavior characteristics include login time, operation frequency, and operation sequence information.

8. The offline map vehicle monitoring and management system based on CNN algorithm according to claim 1 is characterized in that: The data security module establishes an internal and external network security barrier through a network gate.

9. The offline map vehicle monitoring and management system based on CNN algorithm according to claim 1 is characterized in that: The electronic fence alarm module is used to detect whether the user is within the area set in the system. When the vehicle travels outside the set area, the electronic fence alarm module triggers the alarm mechanism.

10. The offline map vehicle monitoring and management system based on CNN algorithm according to claim 1 is characterized in that: The hardware communication module is configured on the vehicle, has an intelligent positioning device and a three-axis acceleration sensor, is powered by a vehicle battery, and is used to realize information interaction between the vehicle and the system.

Images