Monitoring System and Method for Abnormal Operation Indicators of Customizable Engineering Rental Vehicles

Abstract

The present invention discloses a monitoring system and method for abnormal operation indicators of customizable engineering rental vehicles, which obtains multiple operation indicators of the vehicle in real time based on an in-vehicle intelligent box. The indicator data is sent to the EMQX gateway through MQTT protocol after being load-balanced by Nginx. EMQX verifies the identity information in the MQTT protocol. When the data passes the identity verification, it is sent to the message queue Kafka. The big data real-time processing engine Flink consumes the data in Kafka in real time, integrates various indicator thresholds in Redis and the rule configuration file of Drools, and realizes the alarm of abnormal data. At the same time, the vehicle lessor can dynamically adjust the abnormal thresholds of various indicators according to the differential information of the vehicle, and these values will be stored in Redis and the relational database PostgreSQL respectively. When the vehicle alarms, the vehicle lessor can query the alarm data to achieve real-time control of the vehicle abnormality; focusing on solving the actual application requirements of users, it collects the operation parameters of the vehicle in real time and uses big data technology to monitor the vehicle information in real time.

Description

Technical Field

[0001] The present invention belongs to the field of information technology, and particularly relates to a monitoring system and method for abnormal operation indicators of customizable engineering rental vehicles. Background Art

[0002] With the advent of the Internet and 5G era and the continuous development of big data technology, intelligent technology has brought infinite imagination and possibilities to the automotive industry. This has made the combination of vehicle networking and big data an inevitable choice and is expected to become the next development point and breakthrough point in the automotive industry. The rapid development of vehicle networking will greatly promote the development of future intelligent transportation systems. Vehicle networking is to install electronic sensing vehicles on vehicles and use navigation technology, communication systems, in-vehicle intelligent terminal equipment and server platforms to enable information interconnection and interoperability among vehicles, people, roads and the Internet.

[0003] Currently, the whole society is promoting the integration and innovation of industrial Internet in a wider range, deeper degree and higher level, cultivating and strengthening new drivers of economic development, supporting the realization of high-quality development, and accelerating the construction of new infrastructure. One of its main carriers is engineering vehicles. Engineering vehicles run through the entire life cycle of project construction and play an irreplaceable role in engineering project construction. Their appearance has greatly accelerated the progress of civil engineering projects, greatly reducing the waste of human resources, and usually responsible for the transportation of materials, the excavation of earth and stone materials, road repair, etc. in civil engineering projects.

[0004] However, the purchase of engineering vehicles requires a large amount of funds. With the development of social economy, the asset awareness of enterprises is also changing. Because the significance of vehicle ownership is more reflected at the legal level, while the right of use truly creates value. After many enterprises realize this, they no longer struggle with vehicle ownership, but rather consider more how to obtain the right of use of vehicles at a lower cost. Therefore, more and more engineering vehicle users obtain the right of use of engineering vehicles by leasing rather than direct purchase.

[0005] However, in the current business scenario of engineering vehicle leasing, the following problems exist:

[0006] 1) After an engineering vehicle is leased, the lessor and lessee of the vehicle cannot obtain the vehicle's operation information in real time and understand the vehicle's operation status.

[0007] 2) After an engineering vehicle is leased, when the vehicle has an abnormal operation, the lessor of the vehicle cannot obtain the alarm information and accurately understand the abnormal status of the vehicle.

[0008] 3) After an engineering vehicle is leased, due to the high driving difficulty of large engineering vehicles and the different levels of drivers, there are huge safety hazards during driving.

[0009] Therefore, in such a demand background, a system that can monitor the operation information of engineering vehicles and alarm for abnormal vehicle operation is very necessary. Summary of the Invention

[0010] To solve the problems existing in the existing business requirements, the present invention provides a monitoring system that can customize abnormal operation indicators of engineering rental vehicles, focuses on solving the actual application requirements of users, aims to provide management services for rental vehicles, and at the same time uses an in-vehicle intelligent box to obtain vehicle operation information in real time, and collects operation parameters such as vehicle oil temperature, oil pressure, hydraulic pressure, oil level, working time, etc. in real time, and uses big data technology to monitor vehicle information in real time. By introducing a rule engine to customize personalized warning rules for each vehicle, reasonably warning of vehicle abnormalities, and realizing real-time monitoring and management of rental vehicles by the lessor.

[0011] Based on the in-vehicle intelligent box to obtain vehicle operation information in real time, including operation indicators such as vehicle oil temperature, oil pressure, hydraulic pressure, oil level, working time, etc., the data will be sent to the EMQX gateway cluster after being load-balanced by Nginx in the format of the MQTT protocol. EMQX completes authentication and authorization by comparing the Clinetid / Username and password in the built-in database with the MQTT protocol. When the data authentication passes, the data will be sent to the message queue Kafka. The big data real-time processing engine Flink consumes the data in Kafka in real time, and integrates various index thresholds stored in Redis and the rule configuration file of Drools to realize the screening and alarm of abnormal data. Finally, the vehicle position data will be written into HBase, the index data will be written into Opentsdb, and the abnormal data will be written into PostgreSQL. At the same time, in order to notify the vehicle lessor in time when the vehicle has an abnormality, the abnormal data will also be written back to Kafka, and the system backend will subscribe to the Kafka vehicle abnormality topic and send an alarm text message to the vehicle lessor.

[0012] The engineering vehicle lessor can dynamically adjust various index abnormal thresholds at the front end of the website according to the operation status such as vehicle type and vehicle age. After being processed by the business logic of the website backend, this value will be stored in Redis and the relational database PostgreSQL respectively. When the vehicle alarms, the vehicle lessor role can retrieve the alarm data to realize the real-time grasp of the abnormal vehicle operation.

[0013] To achieve the above object, the technical solution adopted by the present invention is: a monitoring system for customizing abnormal operation indicators of engineering rental vehicles, including a real-time data collection and analysis module and a business logic module; the real-time data collection and analysis module is used to authenticate and analyze the real-time operation information of engineering vehicles collected and transmitted by the in-vehicle intelligent box, persistently store the position data and operation indicator data of the vehicle, and at the same time monitor and give early warnings of abnormal vehicle operation states based on a rule engine. Its structure includes, from top to bottom: a data collection layer, a load balancing layer, a gateway layer, a message queue layer, and a computing engine layer; the business logic module is used to display and meet the business access requirements of system participating roles and the rule setting requirements for vehicle anomaly detection. Its structure includes, from top to bottom: a presentation layer, a reverse proxy layer, a business logic layer, and a data storage layer;

[0014] The data collection layer is used to collect vehicle operation data. An in-vehicle Internet of Things intelligent box is installed on each engineering rental vehicle, and the operation state information of the engineering vehicle is collected in real time based on the in-vehicle Internet of Things intelligent box. The vehicle operation data collected by the in-vehicle Internet of Things intelligent box is transmitted to the Nginx load balancing layer of the system through the MQTT message protocol;

[0015] The load balancing layer is used to perform load balancing on the data transmitted in real time by the data collection layer and undertake the reverse proxy of the front-end and back-end addresses of the website;

[0016] The gateway layer is used to authenticate and authorize the MQTT data processed by the Nginx load balancing and send the MQTT data to the Kafka message queue cluster; the message queue layer uses the message queue Kafka to decouple the identity authentication service and the data calculation service; the engine layer uses the big data real-time computing engine Flink to subscribe to the data in the message queue to filter, clean, and window aggregate the real-time data of engineering rental vehicles; the storage layer stores the vehicle operation indicator data and conventional business data processed by the engine layer based on multiple databases such as the NoSQL database HBase, the time series database Opentsdb, the relational database PostgreSQL, and the in-memory database Redis. The conventional business data includes vehicle details information, box information, vehicle anomaly detection rules, and vehicle anomaly data information. At the same time, the in-memory database Redis and the relational database PostgreSQL are used to store the custom vehicle anomaly detection rule information; the business logic layer provides business logic processing capabilities for the function modules of vehicle management, box management, and rule engine management for business personnel; the presentation layer provides a visual interface for business personnel.

[0017] The data acquisition layer includes an in-vehicle Internet of Things intelligent box, which collects vehicle speed, water temperature, hydraulic oil temperature, fuel level, working hours, fuel consumption, daily working duration, daily idle duration, daily working fuel consumption, and daily idle fuel consumption indicators. When the in-vehicle Internet of Things intelligent box transmits the collected vehicle operation information through the MQTT message protocol, it specifies the client Clientid / Username and password in the MQTT protocol message, completes the binding of the identity of the engineering rental vehicle through the client Clientid / Username and password, and provides a basis for the system to authenticate and authorize the vehicle identity information at the same time.

[0018] The load balancing layer includes an Nginx cluster. The load balancing layer undertakes the real-time vehicle operation data sent by the in-vehicle Internet of Things intelligent box. The Nginx cluster performs load balancing in a round-robin manner, distributes each data to multiple downstream gateway servers one by one in chronological order, and at the same time undertakes the reverse proxy function of the front-end and back-end addresses of the website.

[0019] The gateway layer realizes the authentication / access control of MQTT messages, stores the client Clientid / Username and password in the built-in Mnesia database of EMQX or an external relational database. When the client connects to EMQX, EMQX will obtain the Clientid and Username in the CONNENT message and match them with the password recorded in the database. If the match is successful, the authentication is successful, and the data will be sent to the message queue Kafka, otherwise the authentication fails.

[0020] A vehicle management module, a box management module, and a rule engine management module are set in the business logic layer. The vehicle management module is used for querying, adding, and modifying engineering vehicles; the box management module is used for querying, adding, and modifying in-vehicle Internet of Things intelligent boxes; the rule engine management module is the core module in the business logic layer, mainly responsible for functions such as constructing the real-time vehicle data acquisition and analysis link, formulating alarm rules for vehicle lessors, drawing vehicle electronic fences, and vehicle anomaly warnings. Among them, the rule formulation part is responsible for implementing the setting of the vehicle operation anomaly detection threshold by the vehicle lessor, and the operation indicators include multiple indicators such as the upper and lower limits of water temperature, the upper and lower limits of oil temperature, and the upper and lower limits of speed. At the same time, in order to manage the operation range of engineering rental vehicles, the vehicle lessor draws an electronic fence to manage the operation area of the vehicle.

[0021] The present invention also provides a monitoring method for customizing abnormal operation indicators of engineering rental vehicles, including the following steps:

[0022] Obtain the vehicle operation information data in the message queue in real time;

[0023] Group the vehicle index data based on the vehicle operation information data;

[0024] Aggregate the grouped data based on the sliding window method, and set the window time to 10s;

[0025] Process the aggregated data using a custom aggregation processing class;

[0026] When using the custom aggregation processing class to process the aggregated data, integrate the abnormal index threshold stored in Redis and the Drools rule configuration file to achieve dynamic and personalized monitoring and warning of abnormal data for each engineering rental vehicle; temporarily store the abnormal data in the side output stream, and collect the normal data through the collector. After processing the aggregated data, write the normal location data to Hbase, write the index data such as the water temperature, vehicle speed, and oil temperature of the vehicle to Opentsdb, and write the abnormal alarm data to PostgreSQL. At the same time, in order to notify the vehicle lessor in time when the vehicle has an abnormality, write the abnormal data back to Kafka, and the system backend will subscribe to the Kafka vehicle abnormality topic and send an alarm text message to the vehicle lessor.

[0027] The user realizes the management of vehicles and boxes, the setting of abnormal index thresholds for vehicles, and the drawing of vehicle electronic fences through the visualization interface, realizes the monitoring of vehicle operation information, and directly captures the detailed information of the operation abnormality through the operation abnormality information form when the vehicle operation is abnormal. The front end and back end of the business logic sub-architecture are separated; the front end uses the Vue framework, interacts with the data storage layer server after calling the API interface of the server side and passing through the Nginx cluster reverse proxy.

[0028] Draw a vehicle electronic fence at the presentation layer to realize the user-defined drawing of the operation area of engineering rental vehicles. The specific implementation steps are as follows:

[0029] Step 11, integrate the map API with the Vue framework in terms of the website front-end implementation;

[0030] Step 12, click to add an electronic fence, and the website front-end will display the map to the user;

[0031] Step 13, click any four points on the map to draw a quadrilateral and automatically enter the editing state;

[0032] Step 14, drag any two points on the map to the target position to complete the drawing of the electronic fence range; after clicking save, the range information of the electronic fence is saved to the relational database PostgreSQL and the in-memory database Redis;

[0033] Step 15: Use the big data real-time computing engine Flink to utilize the location data of construction vehicles in real time and perform window aggregation. After receiving the real-time location information of the vehicle, obtain the electronic fence range information for each vehicle from the Redis in-memory database, and determine whether the location of the current vehicle is within the specified area. If the current vehicle crosses the boundary, write the location data of the current vehicle into the relational database PostgreSQL.

[0034] Based on the open-source rule engine Drools, solve the problem that due to the differential information such as the type, model, and vehicle age of construction rental vehicles, the alarm rule management module cannot perform dynamic customization settings for each vehicle. The specific implementation steps are as follows:

[0035] Step 1: Create a Maven project, and then add the relevant dependencies of Drools to its pom.xml file;

[0036] Step 2: Create a Drools native rule file with the suffix.drl for each data metric that needs to be warned in the resources directory;

[0037] Step 3: Specify the rules in the.drl file according to the following structure:

[0038] First, define the package name through the package keyword, which must be placed on the first line of the rule file

[0039] Then, define the rule name through the rule keyword. A rule consists of three parts: the attribute part, which defines some attributes for the execution of the current rule, the condition part, which defines the conditions of the rule through the when keyword, and the result part, which defines the execution operations for the data that meets the conditions through the then keyword.

[0040] Finally, use the end keyword to end a rule;

[0041] Step 4: Obtain the dynamic rule settings for each vehicle from the Redis in-memory database;

[0042] Step 5: Obtain all the currently configured alarm rules, and integrate the dynamic rules obtained in Step 4 with the alarm rules to achieve dynamic customization rule settings for each construction vehicle;

[0043] Step 6: Use the big data real-time computing engine Flink to consume the construction vehicle metric data in Kafka in real time and perform window aggregation, and perform anomaly detection on each aggregated data using the customized rules obtained in Step 5; If an anomaly occurs, write the abnormal data into the relational database.

[0044] Compared with the prior art, the present invention has at least the following beneficial effects:

[0045] Compared with the prior art, a monitoring system for abnormal operation indicators of customizable engineering rental vehicles proposed by the present invention obtains vehicle operation information in real time based on an in-vehicle intelligent box, including operation indicators such as vehicle oil temperature, oil pressure, hydraulic pressure, oil level, working hours, etc. The data will be sent to the EMQX gateway cluster after being load-balanced by Nginx in the format of the MQTT protocol. EMQX completes authentication and authorization by comparing the Clinetid / Username and password in the built-in database with those in the MQTT protocol. When the data authentication passes, the data is sent to the message queue Kafka. The big data real-time processing engine Flink consumes the data in Kafka in real time, and integrates various index thresholds stored in Redis and the rule configuration file of Drools to achieve screening and alarming of abnormal data. Finally, the abnormal data is written into PostgreSQL and the message queue Kafka, the vehicle position data is written into HBase, and the index data is written into Opentsdb.

[0046] The present invention conducts data transmission based on the Message Queuing Telemetry Transport protocol MQTT. Since MQTT can communicate with remote sensors and control vehicles in a low-bandwidth and unreliable network, and has the characteristics of low power consumption and persistent connection, it better meets the characteristics of real-time data transmission of a large number of vehicles;

[0047] The present invention uses Nginx to perform load balancing on the real-time transmitted vehicle operation index data. The Nginx server is an intermediary between the in-vehicle intelligent box and the gateway server. The requests sent by the in-vehicle intelligent box first pass through Nginx, and then Nginx distributes the requests to the corresponding servers according to corresponding rules, solving the problem of the hardware performance bottleneck of a single server and ensuring that all backend servers can fully exert their performance, so as to maintain the overall optimal performance of the server cluster;

[0048] The present invention uses the open-source Internet of Things MQTT message server EMQX developed based on the Erlang / OTP platform. EMQX can achieve high reliability, support MQTT connections carrying a large number of Internet of Things terminals, support 2 million connections for a single server node, support low-latency message routing among a large number of Internet of Things vehicles, support customization of various authentication methods, and efficiently store messages in the backend database, meeting the authentication problem when a large number of vehicle data is connected and information is accessed to the system;

[0049] The present invention uses the big data streaming computing framework Flink to process the real-time data sent by engineering rental vehicles, ensuring high throughput and low latency of the system;

[0050] The present invention uses the rule engine Drools to separate business decisions from the application, reducing the cost and risk of writing "hard - coded" business rules, improving the maintainability and maintenance cost of the system, and realizing the dynamic and personalized specification of vehicle anomaly monitoring rules. Description of the Drawings

[0051] Figure 1 It is an architecture diagram of a monitoring system for customizing abnormal operation indicators of engineering rental vehicles according to the present invention. Detailed Implementation Modes

[0052] The present invention provides a monitoring system for customizing abnormal operation indicators of engineering rental vehicles. Based on the on - vehicle intelligent box, it obtains real - time vehicle operation information, including operation indicators such as vehicle oil temperature, oil pressure, hydraulic pressure, oil level, working hours, etc. The data will be sent to the EMQX gateway cluster after being load - balanced by Nginx in the format of the MQTT protocol. EMQX completes authentication and authorization by comparing with the Clinetid / Username and password in the built - in database and the MQTT protocol. When the data authentication passes, the data is sent to the message queue Kafka. The big - data real - time processing engine Flink consumes the data in Kafka in real - time, and integrates various index thresholds stored in Redis and the rule configuration file of Drools to achieve the screening and alarming of abnormal data. Finally, the vehicle location data is written into HBase, the index data is written into Opentsdb, and the abnormal data is written into PostgreSQL. At the same time, in order to notify the vehicle lessor in time when the vehicle has an anomaly, the abnormal data is also written back to Kafka. The system backend subscribes to the vehicle anomaly topic in Kafka and sends an alarm text message to the vehicle lessor.

[0053] The engineering vehicle lessee can dynamically adjust various index anomaly thresholds according to the operation status such as vehicle type and vehicle age at the front - end of the website. After being processed by the business logic at the back - end of the website, this value is stored in Redis and the relational database PostgreSQL respectively. When the vehicle alarms, the vehicle lessor role can retrieve the alarm data to achieve real - time mastery of the abnormal vehicle operation.

[0054] In order to collect the operation data of engineering rental vehicles, an in-vehicle Internet of Things intelligent box is installed on each engineering rental vehicle, and the in-vehicle Internet of Things intelligent box is used to collect the operation status information of the engineering vehicle in real time. The in-vehicle Internet of Things intelligent box transmits the collected vehicle operation information based on the Message Queuing Telemetry Transport Protocol - MQTT. Specifically, after collecting the vehicle operation data, it is transmitted through the MQTT message protocol, and the client Clientid / Username and password are specified in the MQTT protocol message. The identity of the engineering rental vehicle is bound through the client Clientid / Username and password, and then the data is sent to the Nginx load balancing layer of the system;

[0055] In order to receive the massive real-time vehicle operation data sent by the in-vehicle Internet of Things intelligent box, when the Nginx load balancing layer receives the vehicle data sent by the in-vehicle Internet of Things intelligent box in real time, it performs load balancing in a round-robin manner (default), and distributes each vehicle data to multiple downstream gateway servers one by one in chronological order. At the same time, since the entire website adopts the design concept of separating the front and back ends, the Nginx load balancing layer also undertakes the function of reverse proxy for the front and back end addresses of the website;

[0056] An open-source Internet of Things MQTT message server EMQX is used to build a gateway cluster. In order to implement authentication / access control for MQTT messages, the client Clientid / Username and password can be stored in the built-in Mnesia database of EMQX or an external relational database (such as PostgreSQL). When the client connects to EMQX, the database authentication will obtain the Clientid and Username in the CONNENT message, and then match them with the passwords recorded in the database (Mnesia database or PostgreSQL database). If the match is successful, the authentication is successful, and the real-time operation index data of the vehicle will be sent to the message queue Kafka.

[0057] Otherwise, the authentication fails;

[0058] The message queue Kafka is used to build a message queue cluster to decouple the identity authentication service and the data calculation service. When the data volume increases sharply, it can help the key components withstand the sudden access pressure and will not completely collapse due to sudden overload requests;

[0059] Use the open-source rule engine Drools to build a rule configuration file to solve the problem that due to the differential information such as the type, model, and vehicle age of project rental vehicles, the alarm rule management module cannot perform dynamic customization settings for each vehicle; use the big data real-time computing engine Flink to subscribe to the data in the message queue Kafka to filter, clean, and window aggregate the real-time data of project rental vehicles, obtain the abnormal index thresholds of vehicles stored in the in-memory database Redis, and integrate the abnormal index thresholds with the rule configuration file of the open-source rule engine Drools to achieve dynamic and personalized monitoring and early warning of abnormal data for each project rental vehicle. The abnormal data is temporarily stored in the side output stream, and the normal data is collected through the collector;

[0060] The present invention deploys the NoSQL database HBase, the time series database Opentsdb, the in-memory database Redis, and the relational database PostgreSQL to complete the storage of vehicle location data, vehicle index data, abnormal index threshold data, and vehicle abnormal data.

[0061] Use SpringCloud to build the website backend, and the website backend provides business logic processing capabilities for function modules such as vehicle management, box management, and rule engine management for business personnel. The vehicle management module undertakes the functions of querying, adding, and modifying project vehicles; the box management module undertakes the functions of querying, adding, and modifying in-vehicle Internet of Things intelligent boxes; the rule engine management module is the core module in the system and is responsible for formulating vehicle real-time data rules and data early warning functions. The abnormal rule formulation part is responsible for setting the upper limits of abnormal values of vehicle operation indicators, including operation indicators such as the upper limit of water temperature, the upper limit of oil temperature, the lower limit of rotational speed, and the upper limit of fuel consumption. At the same time, in order to reasonably manage the operation location of project rental vehicles, an electronic fence is drawn to construct a vehicle operation area for monitoring the vehicle's location. The rule information is stored in the relational database PostgreSQL and the in-memory database Redis respectively. Data early warning is used to display the details of alarm records of project vehicles in indicator dimensions such as water temperature, oil temperature, rotational speed, fuel consumption, and location boundary crossing, and provides an API interface for the front end;

[0062] Use Vue to build the website front end to provide a visual interface for business personnel. Through the visual interface, users can manage vehicles and in-vehicle Internet of Things intelligent boxes, set abnormal index thresholds of vehicles, and draw vehicle electronic fences to achieve monitoring of vehicle operation information. When a vehicle runs abnormally, the detailed information of the running abnormality can be directly captured through the running abnormality information form. The business logic sub-architecture separates the front and back ends. The front end uses the Vue framework, interacts with the data storage layer server by calling the API interface of the server side and passing through the Nginx balancing layer for reverse proxy.

Claims

1. A monitoring system for abnormal operation indicators of customizable engineering rental vehicles, characterized in that, It includes a real-time data collection and analysis module and a business logic module; the real-time data collection and analysis module is used to perform identity authentication and data analysis on the real-time operation information of engineering vehicles collected and transmitted by the in-vehicle intelligent box, persistently store the vehicle's location data and the vehicle's operation index data, and at the same time monitor and give early warnings of abnormal vehicle operation states based on the rule engine. Its structure from top to bottom includes: a data collection layer, a load balancing layer, a gateway layer, a message queue layer, and a computing engine layer; the business logic module is used to display and meet the business access requirements of system participants and the rule setting requirements for vehicle anomaly detection. Its structure from top to bottom includes: a presentation layer, a reverse proxy layer, a business logic layer, and a data storage layer; The data collection layer is used to implement the collection of vehicle operation data. An in-vehicle Internet of Things intelligent box is installed on each engineering rental vehicle, and the operation state information of the engineering vehicle is collected in real time based on the in-vehicle Internet of Things intelligent box. The vehicle operation data collected by the in-vehicle Internet of Things intelligent box is transmitted to the Nginx load balancing layer of the system through the MQTT message protocol; The load balancing layer is used to perform load balancing on the data transmitted in real time by the data collection layer and undertake the reverse proxy of the front-end and back-end addresses of the website; The gateway layer is used to authenticate and authorize the MQTT data processed by the Nginx load balancing and send the MQTT data to the Kafka message queue cluster; the message queue layer uses the message queue Kafka to decouple the identity authentication service and the data calculation service; the computing engine layer uses the big data real-time computing engine Flink to subscribe to the data in the message queue to implement filtering, cleaning, and window aggregation of the real-time data of engineering rental vehicles; the data storage layer stores the vehicle operation index data and regular business data processed by the computing engine layer based on multiple databases such as the NoSQL database HBase, the time series database Opentsdb, the relational database PostgreSQL, and the in-memory database Redis. The regular business data includes vehicle details information, box information, vehicle anomaly detection rules, and vehicle anomaly data information. At the same time, the in-memory database Redis and the relational database PostgreSQL are used to store the custom vehicle anomaly detection rule information; the business logic layer provides business logic processing capabilities for the function modules of vehicle management, box management, and rule engine management for business personnel; the presentation layer provides a visual interface for business personnel.

2. The monitoring system for abnormal operation indicators of customizable engineering rental vehicles according to claim 1, wherein, The data acquisition layer includes an in-vehicle Internet of Things intelligent box, which collects vehicle speed, water temperature, hydraulic oil temperature, fuel level, daily working hours, daily idle hours, daily working fuel consumption, and daily idle fuel consumption indicators; when the in-vehicle Internet of Things intelligent box transmits the collected vehicle operation information through the MQTT message protocol, it specifies the client Clientid / Username and password in the MQTT protocol message, and completes the binding of the identity of the engineering rental vehicle through the client Clientid / Username and password, while providing a basis for the system to authenticate and authorize vehicle identity information.

3. The monitoring system for abnormal operation indicators of customizable engineering rental vehicles according to claim 1, characterized in that, The load balancing layer includes an Nginx cluster. The load balancing layer receives the real-time vehicle operation data sent by the in-vehicle Internet of Things intelligent box. The Nginx cluster performs load balancing in a round-robin manner, distributes each data to multiple downstream gateway servers one by one in chronological order, and simultaneously undertakes the reverse proxy function for the front-end and back-end addresses of the website.

4. The monitoring system for abnormal operation indicators of customizable engineering rental vehicles according to claim 1, wherein, The gateway layer realizes the authentication / access control of MQTT messages, stores the client Clientid / Username and password in the built-in Mnesia database of EMQX or an external relational database. When the client connects to EMQX, EMQX will obtain the Clientid and Username in the CONNENT message and match them with the password recorded in the database. If the match is successful, the authentication is successful, and the data will be sent to the message queue Kafka; otherwise, the authentication fails.

5. The monitoring system for abnormal operation indicators of customizable engineering rental vehicles according to claim 1, characterized in that, In the business logic layer, a vehicle management module, a box management module, and a rule engine management module are set up. The vehicle management module is used for querying, adding, and modifying engineering vehicles; the box management module is used for querying, adding, and modifying in-vehicle Internet of Things intelligent boxes; the rule engine management module is the core module in the business logic layer, mainly responsible for building the real-time data collection and analysis link of vehicles, formulating alarm rules for vehicle lessors, drawing vehicle electronic fences, and vehicle anomaly warning functions. Among them, the rule formulation part is responsible for implementing the setting of the vehicle operation anomaly detection threshold by the vehicle lessor, and the operation indicators include multiple indicators such as the upper and lower limits of water temperature, the upper and lower limits of oil temperature, and the upper and lower limits of speed. The vehicle lessor manages the vehicle operation area by drawing an electronic fence.

6. A monitoring method for abnormal operation indicators of customizable engineering rental vehicles, characterized in that, It includes the following steps: Obtain the vehicle operation information data in the message queue in real time; Based on the vehicle operation information data, group the vehicle index data therein; Aggregate the grouped data based on the sliding window method, and set the window time to 10s; Use a custom aggregation processing class to process the aggregated data; When using a custom aggregation processing class to process the aggregated data, the abnormal metric thresholds stored in Redis and the rule configuration files of Drools are integrated to achieve dynamic and personalized monitoring and early warning of abnormal data for each engineering rental vehicle. The abnormal data is temporarily stored in the side output stream, while the normal data is collected through a collector. After processing the aggregated data, the normal location data is written into Hbase, the vehicle water temperature, vehicle speed, and oil temperature metric data are written into Opentsdb, and the abnormal alarm data is written into PostgreSQL. At the same time, to notify the vehicle lessor in a timely manner when an abnormality occurs to the vehicle, the abnormal data is written back to Kafka, and the system backend subscribes to the Kafka vehicle abnormality topic to send alarm messages to the vehicle lessor.

7. The monitoring method according to claim 6, characterized in that, Users can manage vehicles and boxes, set abnormal metric thresholds for vehicles, and draw vehicle electronic fences through a visual interface to monitor vehicle operation information. When a vehicle operates abnormally, detailed abnormal operation information is directly captured through the abnormal operation information form. The front-end and back-end of the business logic sub-architecture are separated. The front-end uses the Vue framework, calls the API interface of the server-side, and interacts with the data storage layer server after reverse proxy through the Nginx cluster.

8. The monitoring method according to claim 7, characterized in that On the presentation layer, draw the vehicle electronic fence to implement the user-defined drawing of the operation area of the engineering rental vehicle. The specific implementation steps are as follows: Step 11, integrate the map API with the Vue framework in terms of the website front-end implementation; Step 12, click to add an electronic fence, and the website front-end presents the map to the user; Step 13, click any four points on the map to draw a quadrilateral and automatically enter the editing state; Step 14, drag any two points on the map to the target position to complete the drawing of the electronic fence range; After clicking to save, the range information of the electronic fence is saved to the relational database PostgreSQL and the in-memory database Redis; Step 15, use the big data real-time computing engine Flink to utilize the position data of the engineering vehicle in real time and perform window aggregation. After receiving the real-time position information of the vehicle, obtain the electronic fence range information for each vehicle from the Redis in-memory database, and determine whether the position of the current vehicle is within the specified area. If the current vehicle crosses the boundary, write the position data of the current vehicle into the relational database PostgreSQL.

9. The monitoring method according to claim 6, wherein Based on the open-source rule engine Drools, solve the problem that due to the different types, models, and vehicle ages of engineering rental vehicles, the alarm rule management module cannot perform dynamic customization settings for each vehicle. The specific implementation steps are as follows: Step 1, create a Maven project, and then add the relevant dependencies of Drools to its pom.xml file; Step 2, create a Drools native rule file with the suffix.drl for each data metric that needs early warning in the resources directory; Step 3, specify the rules in the.drl file according to the following structure: First, define the package name using the "package" keyword, which must be placed on the first line of the rule file Then, define the rule name using the "rule" keyword. A rule consists of three parts: the attribute part, which defines some attributes for the execution of the current rule; the condition part, which defines the conditions of the rule using the "when" keyword; and the result part, which defines the actions to be executed for the data that meets the conditions using the "then" keyword. Finally, use the "end" keyword to end a rule; Step 4: Obtain the dynamic rule settings for each vehicle from the Redis in-memory database; Step 5: Obtain all the currently configured alarm rules, and integrate the dynamic rules obtained in Step 4 with the alarm rules to achieve dynamic customized rule settings for each construction vehicle; Step 6: Use the big data real-time computing engine Flink to consume the construction vehicle metric data in Kafka in real time and perform window aggregation, and perform anomaly detection on each piece of aggregated data using the customized rules obtained in Step 5; if an anomaly occurs, write the abnormal data into the relational database.

Images