Equipment fault intelligent detection method and system based on rule analysis and sliding window algorithm

Abstract

The invention belongs to the technical field of equipment fault detection, and particularly relates to an equipment fault intelligent detection method and system based on rule analysis and a sliding window algorithm. The method comprises the following steps: collecting operation data of multi-source equipment; establishing a network data transmission channel between the traffic cloud and the government affair cloud; storing equipment operation plaintext data of the government affair cloud into a memory database and a relational database; based on the stored data, the fault state of the equipment is judged through a rule engine and a sliding window algorithm; the operation state of the equipment is displayed in a visual mode; and pushing the fault equipment. The system comprises a data acquisition subsystem, a data sending and receiving subsystem, a data storage subsystem, a data analysis subsystem and a data display subsystem. According to the equipment operation monitoring work, manual inspection is converted into automatic inspection, the equipment inspection efficiency is improved, and the problem that a traditional equipment fault detection method is low in efficiency is solved.

Description

Technical Field

[0001] This invention belongs to the field of equipment fault detection technology, specifically relating to an intelligent equipment fault detection method and system based on rule analysis and sliding window algorithm. Background Technology

[0002] With the rapid development of highway construction and the increasing demands for operation and management, the types and quantities of electromechanical equipment are constantly increasing, placing higher requirements on equipment operation monitoring and management. Traditional monitoring methods often rely on manual inspections, resulting in low efficiency and slow response times. Traditional network connectivity monitoring mainly uses Ping to monitor the network connectivity of on-site electromechanical equipment, which cannot detect issues such as missing or inaccurate monitoring data or failure to respond to control commands when equipment is online. Adapting to the needs of operation and management requires modifying business logic code, but this results in long response times and insufficient flexibility. Furthermore, the temporal sequence of equipment operating status means that scheduling maintenance tasks for occasional faults wastes operation and maintenance resources and costs. Summary of the Invention

[0003] This invention addresses the problems of low efficiency and high maintenance costs in traditional equipment fault detection methods. It proposes an intelligent equipment fault detection method and system based on rule analysis and sliding window algorithm to monitor the operating status of highway electromechanical equipment, thereby reducing costs and increasing efficiency in the operation and management of highway electromechanical equipment.

[0004] The technical solution claimed by this invention is as follows:

[0005] A method for intelligent detection of equipment faults based on rule analysis and sliding window algorithm includes the following steps:

[0006] S1: Collect multi-source device operation data through acquisition and reception, and perform data cleaning, data conversion and data storage on the collected data;

[0007] S2: A physical connection between the transportation cloud and the government cloud is established across cloud environments by using a dedicated network and configuring address and port mapping for the firewalls on the transportation cloud side and the government cloud side, thus establishing a network data transmission channel between the transportation cloud and the government cloud. The client deployed in the transportation cloud environment encrypts the device operation data obtained in S1 and sends it to the receiving server deployed in the government cloud environment via the aforementioned network data transmission channel. After receiving the data, the server decrypts it to obtain the plaintext data of device operation.

[0008] S3: Store the plaintext data of the device operation obtained in S2 into an in-memory database and a relational database;

[0009] S4: Process the data stored in the in-memory database and relational database. Based on the data obtained from the data processing, classify and analyze the devices through the rule engine to obtain the device operating status. Use the sliding window algorithm to process the results of the device operating status, obtain the most recent status data of a certain device from the current time, and if the proportion of the statistical fault status exceeds the threshold, the device is judged to be in a fault state.

[0010] S5: Convert the device location data into coordinate data, and then display the device information on the map through the map display function. Combined with the device operating status data obtained in S4, display the device operating status in a visual way, and provide prompts for devices that are judged to be faulty.

[0011] S6: Push the faulty device.

[0012] Preferably, the multi-source device operation data in S1 includes: camera online status, vehicle detector online status and vehicle detection data, meteorological detector online status and meteorological index data, variable message sign online status and control status data, traffic survey station online status and traffic survey data, and basic device information.

[0013] Preferably, the acquisition and reception methods described in S1 include: detecting the online status of the device, acquiring data from a read-only table, acquiring data according to an interface protocol, and periodically acquiring data proactively.

[0014] Preferably, the data processing in S4 includes: data extraction, data transformation, and data loading.

[0015] Preferably, the analysis rules predetermined by the rule engine in S4 are as follows: the device operation status model is used to analyze the plaintext data of device operation stored in the memory database and relational database in S3. The analysis is performed according to the different types of devices. All types of devices are treated as network devices. The network connectivity detection method is adopted. The network connectivity between two devices is determined by sending ICMP echo request messages to the target device and receiving echo response messages. On this basis, the weather detector and variable message sign send self-test commands through the interface protocol to obtain the operation status. The vehicle detector and traffic survey station analyze the device operation status by the time interval and data quality of the received data.

[0016] Preferably, the sliding window algorithm described in S4 processes the results of the device's operating status, obtains the 10 most recent status data of a certain device starting from the current time, and if the proportion of the statistical fault status exceeds 20%, the device is judged to be in a fault state.

[0017] Preferably, the steps in S4 where the rule engine judges the device to determine its operating status include:

[0018] Step 1: First, obtain the device's network online status data based on the device number. If the device is offline, set its status to "Device Offline"; if the device is online, determine its status to "Device Online" and proceed with subsequent checks.

[0019] Step 2: Based on the device type information, determine whether it is a controllable device. If it is not a controllable device, query its operating data. If it is a controllable device, send a device status query command to determine if the device communication is abnormal. If the device returns abnormal status data, set its status to "communication failure". If the device is not a controllable device, query its operating data to determine if its data interruption has not exceeded 1 hour. If it has exceeded 1 hour, set its status to "communication failure". If the device communication is normal, analyze the device's operating data.

[0020] Step 3: Based on the latest equipment operation data, determine whether all data indicators are empty. If they are all empty, set their status to "equipment failure"; if they are not all empty, determine whether the equipment operation data exceeds the normal range.

[0021] Step 4: Based on the set normal range of indicator data, determine whether the equipment operation data exceeds the range. If it does, set its status to "equipment failure"; if it does not exceed the range, determine whether the data are contradictory.

[0022] Step 5: Based on the correlation between the set indicators, determine whether the indicator values ​​of the equipment operation data are contradictory. If they are contradictory, set the status to "equipment failure"; if they are not contradictory, set the status to "equipment normal".

[0023] Step 6: Analyze the most recent 10 device status data using the sliding window algorithm. If the failure rate exceeds 20%, the device status is determined to be "device faulty"; otherwise, it is "device normal".

[0024] Preferably, the prompts in S5 include: sound and display reminders.

[0025] Preferably, step S5 further includes: manually confirming and correcting the equipment determined to be faulty.

[0026] This invention also provides an intelligent equipment fault detection system based on rule analysis and sliding window algorithm, comprising: a data acquisition subsystem for acquiring multi-source equipment operation data through acquisition and reception, and performing data cleaning, data conversion, and data storage on the acquired data; a data sending and receiving subsystem for establishing a network data transmission channel between a transportation cloud and a government cloud via a dedicated network and transmitting data acquired by the multi-source equipment operation data acquisition module from the transportation cloud to the government cloud through the network data transmission channel; a data storage subsystem for storing equipment operation data obtained from the government cloud into an in-memory database and a relational database; a data analysis subsystem for classifying and analyzing and judging equipment through a rule engine, obtaining equipment operation status, and processing the results of equipment operation status using a sliding window algorithm to obtain the most recent status data of a certain equipment from the current time and judging the fault status of the equipment based on the status data; and a data display subsystem for visually displaying the equipment operation status and providing prompts for equipment judged to be faulty.

[0027] Beneficial effects

[0028] This invention provides an intelligent equipment fault detection method based on rule analysis and a sliding window algorithm. It collects multi-source equipment operation data through acquisition and reception, and performs data cleaning, transformation, and storage on the collected data, unifying and integrating the multi-source equipment data. A physical connection between the transportation cloud and the government cloud is established through a dedicated network and address and port mapping of the firewalls on both the transportation cloud and government cloud sides. This establishes a network data transmission channel between the two clouds, enabling data transmission across cloud computing environments and meeting the needs of system security and performance expansion. It also allows the information system on the government cloud to access software deployed on the transportation cloud and to collect, transmit, store, and analyze field equipment operation data, achieving cross-cloud transmission and unified management of equipment operation data. Furthermore, a rule engine is used to classify and analyze the equipment. The system determines the equipment's operating status. Using a sliding window algorithm, it processes the results, retrieving the most recent status data for a given device from the current time. If the percentage of faulty states exceeds a threshold, the device is considered faulty. The business rule engine is based on separating business rules from application source code, enabling flexible configuration of business logic without modifying the application source code or interrupting application operation, thus meeting changing management requirements. This application introduces a rule engine to make the device operating status judgment logic more flexible and reduce dependence on system code. The sliding window algorithm comprehensively analyzes the device status over a recent period, improving fault diagnosis accuracy, reducing maintenance costs, and addressing the high maintenance costs of traditional equipment fault detection methods. This application transforms equipment operation monitoring from manual to automatic inspection, improving inspection efficiency and solving the problem of low efficiency in traditional equipment fault detection methods.

[0029] Furthermore, highway operation and management units have different management requirements for different monitoring equipment, requiring verification methods that support automatic detection and flexible configuration to determine the equipment's operating status. This application supports different types of equipment and different status acquisition methods through various detection and judgment methods such as network connectivity detection, equipment status request and reception, and sensor data analysis; it uses a rule engine to flexibly configure the equipment operating status judgment logic to meet the management requirements in different scenarios; and it introduces a sliding window algorithm to perform delayed analysis on fault equipment information after rule analysis, reducing the false alarm rate and meeting the collaborative management needs of equipment operation monitoring data and operation and maintenance management system.

[0030] This invention, without requiring repeated investment in upgrading equipment monitoring software, uses an effective method to connect with existing equipment monitoring software and aggregate basic equipment data and operational data. It then performs intelligent analysis of the equipment operational data, providing a basis for scheduling equipment maintenance tasks. Attached Figure Description

[0031] Figure 1 This is a flowchart of the intelligent device fault detection method based on rule analysis and sliding window algorithm according to an embodiment of the present invention.

[0032] Figure 2 This is a network architecture diagram of an intelligent device fault detection system based on rule analysis and sliding window algorithm according to an embodiment of the present invention.

[0033] Figure 3 This is a diagram illustrating the overall logical architecture of the intelligent equipment fault detection system based on rule analysis and sliding window algorithm, as described in an embodiment of the present invention.

[0034] Figure 4 This is a functional architecture diagram of an intelligent device fault detection system based on rule analysis and sliding window algorithm according to an embodiment of the present invention.

[0035] Figure 5 This is a data flow architecture diagram of an intelligent equipment fault detection system based on rule analysis and sliding window algorithm, according to an embodiment of the present invention.

[0036] Figure 6 is a data storage architecture diagram of the intelligent equipment fault detection system based on rule analysis and sliding window algorithm according to an embodiment of the present invention.

[0037] Figure 7 This is a schematic diagram of the device operation data analysis rules in an embodiment of the present invention.

[0038] Figure 8 This is a flowchart of the device operation status judgment model in an embodiment of the present invention.

[0039] Figure 9 This is a schematic diagram illustrating the fault state analysis of the sliding window algorithm in an embodiment of the present invention.

[0040] Figure 10 This is a schematic diagram of the sliding window algorithm in an embodiment of the present invention.

[0041] Figure 11 This is a schematic diagram of the business rule analysis engine in an embodiment of the present invention. Specific implementation methods

[0042] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions will be further described clearly and completely below with reference to the accompanying drawings.

[0043] First set of embodiments: Intelligent device fault detection method based on rule analysis and sliding window algorithm

[0044] This set of embodiments provides a method for intelligent detection of device faults based on rule analysis and sliding window algorithm, such as Figure 1 As shown, it includes the following steps:

[0045] S1: Multi-source device operation data acquisition steps: Multi-source device operation data is acquired through acquisition and reception, and the acquired data is cleaned, transformed, and saved. Data acquisition and reception are completed using read-only data tables, interface protocol-based acquisition, and periodic active acquisition. The acquired and received data is then further cleaned, transformed, and saved. The main data to be collected includes: camera online status, vehicle detector online status and vehicle detection data, meteorological detector online status and meteorological index data, variable message sign online status and control status data, and traffic survey station online status and traffic survey data. In a specific embodiment of this invention, device online status is determined by network connectivity. This network connectivity determination uses the ICMP protocol, sending request messages in batches via multi-threading. To avoid misjudging network failure due to poor network conditions, the response message timeout is set to 5 seconds, and the number of echo requests is 10. Based on this, the device response data packet loss rate is analyzed to determine the device online status. If the loss rate is 100%, the device is determined to be offline. Some external data, such as meteorological data and traffic survey data, are read by calling database views through external system data reading services to retrieve the equipment operation data gathered from external systems and save it to the local database.

[0046] Equipment control status data, such as variable message sign control status data, is sent to online remote controllable devices via the equipment status command control service. The devices then return their current status, such as communication failure or normal operation.

[0047] S2: Equipment operation data sending and receiving steps: as follows Figure 2 As shown, a physical connection between the transportation cloud and the government cloud is achieved through a dedicated network and by configuring address and port mapping for the firewalls on the transportation cloud side and the government cloud side, thus establishing a network data transmission channel between the transportation cloud and the government cloud. The client deployed in the transportation cloud environment encrypts the device operation data obtained in S1 and sends it to the receiving server deployed in the government cloud environment through the aforementioned network data transmission channel. After receiving the data, the server decrypts it to obtain the plaintext data of device operation.

[0048] S3: Device Operation Data Storage Steps: The plaintext device operation data obtained in S2 is stored in both an in-memory database and a relational database. In a specific embodiment of this invention, the plaintext device operation data received and decrypted by the government cloud server is stored in both an in-memory database and a relational database. The in-memory database stores the latest device status and operation data, while the relational database stores the latest and historical data. When performing data analysis and display, the device operation data is queried first from the in-memory database to improve data analysis efficiency using its high performance. If the target data cannot be found, the historical device operation data is queried from the relational database. Figure 6 As shown.

[0049] S4: Equipment Operation Data Analysis Steps: Data processing is performed on the data stored in the in-memory database and relational database. Based on the processed data, a rule engine is used to classify and analyze the equipment to determine its operating status. A sliding window algorithm is used to process the results of the equipment operating status, obtaining the most recent status data for a given device from the current time. If the percentage of faulty states exceeds a threshold, the device is considered faulty. The data processing includes: data extraction, data transformation, and data loading. Data extraction is responsible for extracting the online status, operational data, and controllable device status from the in-memory database and relational database. Data transformation is responsible for matching the extracted data with device IDs from different source systems and their corresponding road segments, and unifying device status codes. Data loading provides the data in a unified format to the next data processing step.

[0050] First, the data stored in the in-memory database and relational database undergoes the aforementioned data processing to obtain data with a unified format. Then, based on the data obtained from the data processing, a rule engine is used to classify, analyze, and judge the devices to obtain their operating status, such as... Figure 8 As shown; in a specific embodiment of the present invention, the rule engine's predetermined analysis rules are as follows: The stored device operation data is analyzed using a device operation status model. Different analysis methods are applied based on different types of devices. All types of devices are treated as network devices. Common network connectivity detection methods are used, and the network connectivity between two devices is determined by sending ICMP echo request messages to the target device and receiving echo response messages. Based on this, weather detectors and variable message signs obtain their operating status by sending self-test commands through interface protocols. Vehicle detectors and traffic survey stations analyze the device operating status by the time interval and data quality of the received data. Figure 7As shown, the rule engine employs targeted data analysis rules for different types of equipment. It comprehensively judges the operating status of equipment through methods such as single-value threshold range judgment, multi-value cumulative verification, and data value correlation logic analysis. Specifically, vehicle detectors and traffic survey stations use the data valid value range analysis method. By analyzing the lane cumulative flow, occupancy rate, average vehicle speed, and traffic flow data of various types output by the equipment, if the data exceeds the valid value range, the equipment is determined to be in a fault state. At the same time, the cumulative vehicle flow is compared with the sum of traffic flow of various types of vehicles. If they are inconsistent, the equipment is determined to be in a fault state. The meteorological detector uses a data valid value range analysis method. It analyzes data such as atmospheric temperature, relative humidity, wind direction, wind speed, precipitation, visibility, road ice and snow, road wetness, and road dryness output by the device. If the data exceeds the valid value range, the device is judged to be in a faulty state. At the same time, it uses the logical relationship between data such as precipitation greater than 0 and road dryness, precipitation greater than 0 and road wetness, road dryness and road wetness, and road dryness and road ice and snow to determine the operating status of the device. If they contradict each other, the device is judged to be in a faulty state.

[0051] The equipment operation status model is described as follows: Based on the online status of cameras, vehicle detectors and vehicle detection data, weather detectors and weather index data, variable message signs and control status data, and traffic survey stations and traffic survey data, logical judgments are made according to business management rules to finally obtain the equipment operation status. The judgment steps are as follows: Figure 8 As shown:

[0052] 1. First, obtain the device's network online status data based on the device number. If the device is offline, set its status to "device offline"; if the device is online, determine its status to "device online" and proceed with subsequent judgments.

[0053] 2. Based on the device type information, determine if it is a controllable device. If not, query the device's operating data. If it is a controllable device, send a device status query command to determine if the device communication is abnormal. If the device returns abnormal status data, set its status to "communication failure". If the device is not a controllable device, query its operating data to determine if its data interruption has not exceeded 1 hour (configurable according to management needs). If it has exceeded 1 hour, set its status to "communication failure". If the device's communication is normal, analyze the device's operating data.

[0054] 3. Based on the latest equipment operation data, determine whether all data indicators are empty. If they are all empty, set their status to "equipment failure". If they are not all empty, determine whether the equipment operation data exceeds the normal range.

[0055] 4. Based on the normal range of the set indicator data, determine whether the equipment operation data exceeds the range. If it does, set its status to "equipment failure"; if it does not exceed the range, determine whether the data are contradictory.

[0056] 5. Based on the correlation between the set indicators, determine whether the indicator values ​​of the equipment operation data are contradictory. If they are contradictory, set the status to "equipment failure"; if they are not contradictory, set the status to "equipment normal".

[0057] 6. Analyze the most recent 10 device status data using a sliding window algorithm. If the failure rate exceeds 20%, the device status is determined to be "device faulty"; otherwise, it is determined to be "device normal".

[0058] For example: Based on the device number of a weather detector, the network online status is obtained as "device online," indicating the weather detector is a controllable device. A device status query command is actively sent to the device, and the device returns normal status data. Further analysis of the device's operational data, combined with the device status judgment rules, shows that all indicators of the weather detector are within the normal range. However, the analysis reveals inconsistencies: the rainfall is displayed as 2 mm / min, but the ground is dry ("yes"). Therefore, the weather detector's operational status is judged as "device malfunction." Using a sliding window algorithm, the proportion of the last 10 statuses indicating "device malfunction" for this weather detector is 30%. Based on this, the overall judgment is that the weather detector's operational status is "device malfunction."

[0059] The business rule analysis engine (rule engine) is primarily based on the idea of ​​separating business rules from application source code. This allows for flexible configuration of business logic without modifying the application source code or interrupting application operation, thus meeting the needs of changing management requirements. A diagram illustrating the principle of the business rule analysis engine is shown below. Figure 11 As shown.

[0060] (1) Data income is used to receive and preprocess the input data and check the compatibility of the data format with the rule engine.

[0061] (2) The working memory receives input data and stores data objects.

[0062] (3) The rule base stores the rule records. Each rule consists of conditions and actions.

[0063] (4) The pattern matcher compares and matches the data objects in the working memory with the rules in the rule base, and puts the rules that meet the conditions into the activation queue.

[0064] (5) The conflict resolution module is used to handle conflicts between multiple rules that are activated at the same time.

[0065] (6) The activation queue is used to store rules that have been activated and are yet to be executed.

[0066] (7) Execution rules are used to execute the rules in the activation queue.

[0067] (8) The execution result is the result produced after the rule is executed.

[0068] The sliding window algorithm is used to process the device operating status results determined by the rule engine. For a given device, the 10 most recent status data points from the current time are retrieved. If the percentage of faulty states exceeds 20%, the device is considered faulty. Figure 9 As shown.

[0069] The sliding window algorithm solves specific types of problems by defining a window, sliding it across an array, and processing data step by step. Each time the window slides, it updates its content by adding new elements and deleting old ones, and then calculates the result of the window's content. This invention uses a fixed-length sliding window. The principle of the sliding window algorithm is as follows: Figure 10 As shown, the main steps include:

[0070] Step 1: Initialize the window: Determine the initial position and size of the window, as well as the initial positions of the left and right pointers.

[0071] Step 2: Enter the window: According to the question requirements, move the right pointer to the right to bring the new element into the window.

[0072] Step 3: Conditional Judgment: Check if the elements in the window meet the problem requirements. If not, perform an exit window operation; if they meet, update the results or perform other operations according to the problem requirements.

[0073] Step 4: Leave the window: Move the left pointer to the right to remove the old element from the window. This step is usually performed when the elements in the window no longer meet the requirements of the problem.

[0074] Step 5: Looping process: Repeat steps 2 to 4 until the right pointer reaches the end of the array.

[0075] S5: Equipment Operation Data Display Steps: The equipment location data is converted into coordinate data, and then the equipment information is displayed on a map using the map display function. Combined with the equipment operation status data obtained in S4, the equipment operation status is displayed in a visual manner, and prompts are given for equipment determined to be faulty. In a specific embodiment of the invention, the location data in the equipment's basic information is converted into coordinate data, and then the equipment information is displayed on a map using the map display function. Combined with the equipment operation status data obtained from the equipment operation data analysis steps, the equipment operation status is displayed in a visual manner. Simultaneously, the system provides audible and visual alerts for equipment determined to be faulty, allowing staff to confirm the fault and eliminate false alarms. The equipment location data conversion uses a map conversion service to convert the equipment's location station number to latitude and longitude coordinates; the equipment information display uses a data display service to render and display road and equipment location markers on the map.

[0076] S6: Fault Data Push Step: Push the faulty equipment; In a specific embodiment of the present invention, the confirmed fault is pushed to the electromechanical operation and maintenance management system by a manual secondary confirmation method, so that the operation and maintenance personnel can further process it, thereby realizing cross-system data interaction and business process collaboration; The push is made by calling the operation and maintenance system interface through the equipment fault information push service and sending the equipment fault information to the operation and maintenance system for processing.

[0077] The second set of embodiments: Intelligent equipment fault detection system based on rule analysis and sliding window algorithm

[0078] This set of embodiments provides an intelligent device fault detection system based on rule analysis and sliding window algorithm, such as Figure 3-5 As shown, the system includes, in sequence, a data acquisition subsystem for collecting multi-source device operation data through acquisition and reception, and for cleaning, transforming, and storing the collected data; a data sending and receiving subsystem for establishing a network data transmission channel between the transportation cloud and the government cloud via a dedicated network and transmitting the data collected by the multi-source device operation data acquisition module from the transportation cloud to the government cloud via the network data transmission channel; a data storage subsystem for storing the device operation data obtained from the government cloud into an in-memory database and a relational database; a data analysis subsystem for classifying, analyzing, and judging devices through a rule engine, obtaining device operation status, processing the results of device operation status using a sliding window algorithm to obtain the most recent status data of a device from the current time, and judging the fault status of the device based on the status data; and a data display subsystem for visually displaying the device operation status and providing prompts for devices judged to be faulty.

[0079] In specific embodiments of the present invention, such as Figure 4As shown, the data acquisition subsystem provides network connectivity judgment service, external system data reading service, and device status command control service. The network connectivity judgment service uses the ICMP protocol and sends request messages in batches via multiple threads. To avoid misjudging network connectivity due to poor network conditions, the response message timeout is set to 5 seconds, and the number of echo requests is 10. Based on this, the loss rate of device response data packets is analyzed to determine the device's online status. If the loss rate is 100%, the device is considered offline, thus improving inspection efficiency while also considering the aging state of network hardware. The external system data reading service reads device operating data aggregated from external systems by calling database views and stores it in the local database, thus maintaining the independence between systems and the reliability of the original data. The device status command control service sends status control commands to online remotely controllable devices, which then return their current status, such as communication failure or device normal operation. The data sending and receiving subsystem provides data encryption service, data sending service, data receiving service, and data decryption service. The data storage subsystem includes an in-memory database and a relational database. The data analysis subsystem includes data processing service, rule engine, and device... The system includes an operational status model and an equipment operational status query service. The data processing service comprises data extraction, data transformation, and data loading. Data extraction is responsible for extracting online equipment status, operational data, and controllable equipment status from the data storage subsystem. Data transformation involves matching the above data with equipment IDs from different source systems and their respective road segments, and unifying equipment status codes. Data loading provides data in a unified format to the next processing step. The data display subsystem includes a map conversion service, a data display service, and an equipment fault information push service. The map conversion service converts the equipment's location station number to latitude and longitude coordinates. The data display service renders and displays road and equipment location markers on the map. The equipment fault information push service calls the maintenance system interface and sends equipment fault information to the maintenance system for processing.

[0080] In specific embodiments of the present invention, such as Figure 3 As shown, the system integrates and analyzes multi-source data, including the online status of cameras, the online status and vehicle detection data of vehicle detectors, the online status and meteorological index data of meteorological detectors, the online status and control status data of variable message signs, and the online status and traffic survey data of traffic survey stations. Based on rule logic, it determines the equipment operating status and performs comprehensive analysis through a sliding window algorithm, providing reference and judgment basis for equipment operation monitoring and equipment maintenance.

[0081] like Figure 5As shown, the data acquisition subsystem includes network connectivity assessment services, external system data reading services, and equipment status command control services. The data transmission and reception subsystem includes data encryption services, data transmission services, data reception services, and data decryption services. The data storage subsystem consists of a high-performance in-memory database and a relational database. The data analysis subsystem includes data preprocessing services, a rule engine, a sliding window algorithm, an equipment operating status model, a sliding window algorithm, and an equipment operating status query service. The data display subsystem includes map conversion services, data display services, and equipment fault information push services. These five subsystems are connected through business logic and data relationships to form an intelligent monitoring system for highway electromechanical equipment.

[0082] like Figure 5 As shown, equipment operation data is acquired and aggregated by the data acquisition subsystem, then enters the data storage subsystem through a secure channel composed of data sending and receiving subsystems, and is stored by both an in-memory database and a relational database. The data analysis subsystem analyzes the equipment operation data, determines the equipment's operating status, and marks the status accordingly. The data display subsystem marks points on the electronic map interface based on the equipment's coordinate data, uses different colors to identify the equipment and reflect its operating status, and sends faulty equipment confirmed by manual verification to the electromechanical maintenance system for further processing via the equipment fault information push service, achieving cross-system business collaboration.

[0083] The above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail through the above preferred embodiments, those skilled in the art should understand that various changes can be made to it in form and detail without departing from the scope defined by the claims of the present invention.

Claims

1. A method for intelligent detection of equipment faults based on rule analysis and sliding window algorithm, characterized in that, Includes the following steps: S1: Collect multi-source device operation data through acquisition and reception, and perform data cleaning, data conversion and data storage on the collected data; S2: A physical connection between the transportation cloud and the government cloud is established across cloud environments by using a dedicated network and configuring address and port mapping for the firewalls on the transportation cloud side and the government cloud side, thus establishing a network data transmission channel between the transportation cloud and the government cloud. The client deployed in the transportation cloud environment encrypts the device operation data obtained in S1 and sends it to the receiving server deployed in the government cloud environment via the aforementioned network data transmission channel. After receiving the data, the server decrypts it to obtain the plaintext data of device operation. S3: Store the plaintext data of the device operation obtained in S2 into an in-memory database and a relational database; S4: Process the data stored in the in-memory database and relational database. Based on the data obtained from the data processing, classify and analyze the devices through the rule engine to obtain the device operating status. Use the sliding window algorithm to process the results of the device operating status, obtain the most recent status data of a certain device from the current time, and if the proportion of the statistical fault status exceeds the threshold, the device is judged to be in a fault state. S5: Convert the device location data into coordinate data, and then display the device information on the map through the map display function. Combined with the device operating status data obtained in S4, display the device operating status in a visual way, and provide prompts for devices that are judged to be faulty. S6: Push the faulty device.

2. The intelligent equipment fault detection method based on rule analysis and sliding window algorithm according to claim 1, characterized in that, The multi-source device operation data mentioned in S1 includes: camera online status, vehicle detector online status and vehicle detection data, meteorological detector online status and meteorological index data, variable message sign online status and control status data, traffic survey station online status and traffic survey data, and basic device information.

3. The intelligent equipment fault detection method based on rule analysis and sliding window algorithm according to claim 1, characterized in that, The acquisition and reception methods described in S1 include: acquisition from read-only data tables, acquisition according to interface protocols, and periodic active acquisition.

4. The intelligent equipment fault detection method based on rule analysis and sliding window algorithm according to claim 2 or 3, characterized in that, The rule engine in S4 predefines the following analysis rules: it uses the device operation status model to analyze the plaintext data of device operation stored in the memory database and relational database in S3. It performs corresponding analysis according to different types of devices. All types of devices are treated as network devices. The network connectivity detection method is used to determine whether the network between two devices is normally connected by sending ICMP echo request messages to the target device and receiving echo response messages. On this basis, the weather detector and variable message sign send self-test commands through the interface protocol to obtain the operation status. The vehicle detector and traffic survey station analyze the device operation status by the time interval and data quality of the received data.

5. The intelligent equipment fault detection method based on rule analysis and sliding window algorithm according to claim 4, characterized in that, The data processing described in S4 includes: data extraction, data transformation, and data loading.

6. The intelligent equipment fault detection method based on rule analysis and sliding window algorithm according to claim 5, characterized in that, The sliding window algorithm described in S4 processes the results of the device's operating status, obtains the 10 most recent status data of a certain device starting from the current time, and judges the device to be in a fault state if the proportion of the statistical fault state exceeds 20%.

7. The intelligent equipment fault detection method based on rule analysis and sliding window algorithm according to claim 6, characterized in that, In S4, the rule engine determines the device's operating status through the following steps: Step 1: First, obtain the device's network online status data based on the device number. If the device is offline, set its status to "Device Offline"; if the device is online, determine its status to "Device Online" and proceed with subsequent checks. Step 2: Based on the device type information, determine whether it is a controllable device. If it is not a controllable device, query the device's operating data. If it is a controllable device, send a device status query command to determine if the device communication is abnormal. If the device returns abnormal status data, set its status to "communication failure". If the device is not a controllable device, query the device's operating data to determine if its data interruption has not exceeded 1 hour. If it has exceeded 1 hour, set its status to "communication failure". If the device communication is normal, analyze the device's operating data. Step 3: Based on the latest equipment operation data, determine whether all data indicators are empty. If they are all empty, set their status to "equipment failure"; if they are not all empty, determine whether the equipment operation data exceeds the normal range. Step 4: Based on the set normal range of indicator data, determine whether the equipment operation data exceeds the range. If it does, set its status to "equipment failure"; if it does not exceed the range, determine whether the data are contradictory. Step 5: Based on the correlation between the set indicators, determine whether the indicator values ​​of the equipment operation data contradict each other. If they contradict each other, set its status to "equipment failure"; if they do not contradict each other, set its status to "equipment normal". Step 6: Analyze the most recent 10 device status data using a sliding window algorithm. If the failure rate exceeds 20%, the device status is determined to be "device faulty"; otherwise, it is "device normal".

8. The intelligent equipment fault detection method based on rule analysis and sliding window algorithm according to claim 2, characterized in that, The prompts mentioned in S5 include: sound and display reminders.

9. The intelligent equipment fault detection method based on rule analysis and sliding window algorithm according to claim 8, characterized in that, Step S5 also includes: manually verifying and correcting equipment identified as faulty.

10. A device fault intelligent detection system based on rule analysis and sliding window algorithm, characterized in that, The system includes, in sequence, a data acquisition subsystem for collecting multi-source device operation data through acquisition and reception, and for cleaning, transforming, and storing the collected data; a data transmission and reception subsystem for establishing a network data transmission channel between the transportation cloud and the government cloud via a dedicated network and transmitting the data collected by the multi-source device operation data acquisition module from the transportation cloud to the government cloud via the network data transmission channel; a data storage subsystem for storing the device operation data obtained from the government cloud into an in-memory database and a relational database; a data analysis subsystem for classifying, analyzing, and judging devices through a rule engine, obtaining device operation status, processing the results of device operation status using a sliding window algorithm to obtain the most recent status data of a device from the current time, and judging the fault status of the device based on the status data; and a data display subsystem for visually displaying the device operation status and providing prompts for devices judged to be faulty.

Images