Big data-based distributed collaborative inspection method for power machine room

Abstract

The invention provides a big data-based distributed collaborative inspection method for a power machine room, and the method comprises the steps: dividing the power machine room into a plurality of inspection regions according to the related actual information of the power machine room; acquiring physical environment conditions, equipment information and potential safety hazard conditions in each inspection area, and formulating an inspection task of each inspection area; deploying a monitoring device and an inspection robot based on the inspection task of each inspection area; formulating a parallel inspection plan of the inspection robot according to the importance degree and the equipment layout of each inspection area in combination with the monitoring device in each inspection area; and performing routing inspection according to the parallel routing inspection plan to obtain routing inspection data, and determining a final routing inspection result of each routing inspection area in combination with monitoring data acquired by the monitoring device in real time. According to the invention, by dividing the inspection area of the machine room and cooperating with the intelligent monitoring system to execute the inspection task in parallel on the inspection area, the purposes of improving the inspection quality and efficiency and reducing the operation and maintenance cost are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of power maintenance, and in particular to a distributed collaborative inspection method for power machine rooms based on big data. Background Art

[0002] Currently, various industries have strict compliance standards for machine room management, such as the power and medical industries. Inspecting and recording relevant situations is a necessary means to prove compliance, which can avoid regulatory penalties and reputation risks. Therefore, through the power machine room inspection technology, monitoring resource usage and operation conditions can promptly detect and solve existing risks.

[0003] In the prior art, manual inspection is the most commonly used method for machine room inspection. Mainly, inspection personnel enter the machine room to carry out work according to the established inspection routes and cycles. However, manual inspection is easily affected by subjective factors of personnel, resulting in missed inspections and misjudgments, and the efficiency is relatively limited. At the same time, manual inspection has the situation of untimely and inaccurate recording due to relying on manual recording of data and status information, making it difficult to ensure its accuracy when analyzing data later, which is not conducive to making a scientific assessment of the long-term operation status of the machine room.

[0004] To solve the problems of manual inspection, intelligent monitoring systems are widely used to assist in inspecting machine rooms. For example, some research focuses on improving the accuracy and efficiency of inspection through machine vision technology. At the same time, some research is specific to the inspection of specific equipment or systems, or the inspection of a specific scenario. However, although there are many sensors and monitoring means, there may still be some undetected dead corners or complex situations that cannot be accurately monitored, such as extremely subtle hardware changes inside some equipment and data deviations caused by special electromagnetic environment interference. Thus, it can be seen that the intelligent monitoring system-assisted inspection of machine rooms does not consider preventive machine room inspection and operation and maintenance management, lacking the optimal utilization and planning of machine room resources.

[0005] Therefore, there is an urgent need for a new method for inspecting machine rooms. After dividing the inspection areas of the machine room by planning and optimizing the machine room resources, it collaborates with the intelligent monitoring system to parallelly execute inspection tasks for all inspection areas, so as to achieve the purpose of improving the inspection quality and efficiency and reducing the operation and maintenance costs. Summary of the Invention

[0006] The technical problem to be solved by the embodiments of the present invention is to provide a distributed collaborative inspection method for power machine rooms based on big data. After dividing the inspection areas of the machine room by planning and optimizing the machine room resources, it collaborates with the intelligent monitoring system to parallelly execute inspection tasks for all inspection areas, so as to achieve the purpose of improving the inspection quality and efficiency and reducing the operation and maintenance costs.

[0007] To solve the above technical problems, an embodiment of the present invention provides a distributed collaborative inspection method for a power machine room based on big data. The method includes the following steps:

[0008] According to the relevant actual information of the power machine room, divide the power machine room into multiple inspection areas;

[0009] Obtain the physical environment conditions, equipment information, and potential safety hazard situations in each inspection area to formulate physical environment monitoring tasks, equipment status monitoring tasks, and potential safety hazard investigation tasks for each inspection area;

[0010] Based on the physical environment monitoring tasks, equipment status monitoring tasks, and potential safety hazard investigation tasks of each inspection area, a plurality of monitoring devices and an inspection robot are respectively deployed in each inspection area; wherein, the monitoring devices include sensors, cameras, and network monitoring devices installed with monitoring agent software;

[0011] According to the importance level and equipment layout of each inspection area, and in combination with the monitoring devices deployed in each inspection area, formulate a parallel inspection plan for the inspection robots in the power machine room; wherein, the parallel inspection plan includes the inspection task priority, task volume, inspection path, inspection cycle, and inspection time required within each inspection cycle for each inspection robot;

[0012] According to the parallel inspection plan of the inspection robots in the power machine room, drive each inspection robot to conduct inspections to obtain inspection data of each inspection area where the inspection robot is located, and further combine the monitoring data collected in real time by each monitoring device in each inspection area to determine the final inspection results of each inspection area.

[0013] Among them, the specific steps of dividing the power machine room into multiple inspection areas according to the relevant actual information of the power machine room include:

[0014] Obtain the relevant actual information of the power machine room, including overall physical layout information, functional area information, equipment distribution information, and network topology information;

[0015] According to the relevant actual information of the power machine room, form an overall layout diagram of the power machine room, and in combination with the importance level of each functional area in the power machine room, divide the power machine room into several inspection areas;

[0016] Define the boundary range of each inspection area, mark it on the overall layout diagram of the power machine room, and assign each inspection area a unique management number.

[0017] Among them, the specific steps for obtaining the physical environment conditions, equipment information, and potential safety hazards in each inspection area to formulate the physical environment monitoring tasks, equipment status monitoring tasks, and potential safety hazard investigation tasks for each inspection area are as follows:

[0018] Determine the physical environment conditions in each inspection area and the corresponding monitoring work content to be carried out, so as to formulate the environmental parameter monitoring tasks in each inspection area; among them, the environmental parameter monitoring tasks include temperature, humidity, smoke concentration, and dust concentration;

[0019] Determine the equipment in each inspection area and the corresponding status monitoring work content to be carried out, so as to formulate the equipment status monitoring tasks in each inspection area; among them, the equipment status monitoring tasks include voltage, current, switch status, connection status, and operating status;

[0020] Determine the potential safety hazard situations in each inspection area and the corresponding status monitoring work content to be carried out, so as to formulate the potential safety hazard investigation tasks in each inspection area; among them, the potential safety hazard investigation tasks include leakage hazards, cable damage, equipment grounding status, foreign objects, water accumulation, and fire protection facilities.

[0021] Among them, the specific steps for deploying multiple monitoring devices and an inspection robot in each inspection area based on the physical environment monitoring tasks, equipment status monitoring tasks, and potential safety hazard investigation tasks in each inspection area are as follows:

[0022] Based on the physical environment monitoring tasks, equipment status monitoring tasks, and potential safety hazard investigation tasks in each inspection area, first-class sensors or / and second-class sensors are installed at corresponding positions in each inspection area to collect the status of equipment in each inspection area or / and the physical environment outside it; among them, the first-class sensors are installed outside the equipment in each inspection area and include temperature and humidity sensors, smoke sensors, and dust sensors for monitoring the physical environment conditions; the second-class sensors are installed on the equipment in each inspection area and include temperature sensors, current transformers, and vibration sensors for collecting the operating status of the equipment;

[0023] Based on the physical environment monitoring tasks, equipment status monitoring tasks, and potential safety hazard investigation tasks in each inspection area, multiple cameras are installed at corresponding positions in each inspection area to collect images of the appearance of equipment in each inspection area, images of personnel entering and leaving, images of equipment connection conditions, and images of cable connection conditions;

[0024] Based on the physical environment monitoring tasks, equipment status monitoring tasks, and safety hazard investigation tasks in each inspection area, monitoring agent software is installed on the corresponding network monitoring devices in each inspection area to collect the CPU usage rate, memory occupancy, disk I / O situation, network traffic, and process status of each network monitoring device and its interconnected devices; among them, the interconnected devices of the network monitoring devices include various sensors, cameras, and communication devices for data interaction between devices in each inspection area;

[0025] Inspection robots are deployed in each inspection area, and according to the boundary range and equipment distribution information of each inspection area, motion guides for the inspection robots to travel are deployed.

[0026] Among them, the specific steps for formulating the parallel inspection plan of the inspection robots in the power distribution room according to the importance level and equipment layout of each inspection area, and in combination with the monitoring devices deployed in each inspection area include:

[0027] According to the importance level and equipment layout of each inspection area, and in combination with the monitoring devices deployed in each inspection area, determine the inspection task priority, task volume, inspection cycle, and inspection time required for each inspection cycle of each inspection robot;

[0028] According to the task volume and inspection time of each inspection robot, and in combination with the preset data collection frequency of each task and the motion guides in each inspection area, formulate the inspection paths of each inspection robot;

[0029] Based on the inspection task priority, task volume, inspection cycle, and inspection time required for each inspection cycle of each inspection robot, and in combination with the inspection paths of each inspection robot, obtain the parallel inspection plan of the inspection robots in the power distribution room.

[0030] Among them, the specific steps for driving each inspection robot to perform inspections according to the parallel inspection plan of the inspection robots in the power distribution room to obtain the inspection data of each inspection robot for its corresponding inspection area, and further combining the monitoring data collected in real time by each monitoring device in each inspection area to determine the final inspection results of each inspection area include:

[0031] When each inspection robot starts according to the parallel inspection plan, obtain the location, status, task execution progress, and collected data uploaded by each inspection robot, and further obtain the monitoring data collected in real time by each monitoring device in each inspection area;

[0032] Preprocess the location, status, task execution progress uploaded by each inspection robot, as well as the data collected, and preprocess the monitoring data collected in real time by each monitoring device in each inspection area. Further comprehensively analyze the physical environment condition data, equipment operation real-time data, and potential safety hazard situation data in each inspection area to determine the final inspection results of each inspection area.

[0033] Among them, the method further includes:

[0034] Perform chart statistics on the physical environment condition data, equipment operation real-time data, and potential safety hazard situation data in each inspection area obtained through comprehensive analysis to intuitively reflect the final inspection results of each inspection area.

[0035] Among them, the method further includes:

[0036] Use machine learning algorithms to model and train the physical environment condition data, equipment operation real-time data, and potential safety hazard situation data in each inspection area to analyze the anomalies in the final inspection results of each inspection area.

[0037] Among them, the method further includes:

[0038] According to the actual operation requirements, management key points, and resource investment situation of the power distribution room, evaluate the final inspection results of each inspection area to optimize and obtain the next inspection plan for the inspection robots in each inspection area.

[0039] Implementing the embodiments of the present invention has the following beneficial effects:

[0040] Compared with the traditional intelligent monitoring system-assisted inspection of the machine room, after dividing the inspection areas of the machine room by planning and optimizing the machine room resources, the monitoring devices and inspection robots in each inspection area work together simultaneously, and further use big data technology to perform real-time analysis on the multi-modal data collected distributively in each inspection area to quickly locate the final inspection results of each inspection area, so as to achieve the purpose of improving the inspection quality and efficiency and reducing the operation and maintenance costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, obtaining other drawings based on these drawings still belongs to the scope of the present invention.

[0042] Figure 1Flowchart of a distributed collaborative inspection method for power machine rooms based on big data provided by an embodiment of the present invention;

[0043] Figure 2 Distribution map of the distributed areas of the power machine room in the application scenario of a distributed collaborative inspection method for power machine rooms based on big data provided by an embodiment of the present invention;

[0044] Figure 3 For Figure 2 Example diagram of the server numbers in area A01 in Detailed implementation manners

[0045] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings.

[0046] As Figure 1 shown, in an embodiment of the present invention, a distributed collaborative inspection method for power machine rooms based on big data is proposed. The method includes the following steps:

[0047] Step S1: Divide the power machine room into multiple inspection areas according to relevant actual information of the power machine room;

[0048] The specific process is as follows. First, obtain relevant actual information of the power machine room, including but not limited to overall physical layout information, functional area information, equipment distribution information, network topology information, etc.

[0049] Secondly, according to the relevant actual information of the power machine room, form an overall layout diagram of the power machine room, and combine the importance levels of each functional area in the power machine room to divide the power machine room into several inspection areas to ensure that each divided inspection area is relatively independent and convenient for subsequent inspection robots to reach. It should be noted that areas with a large physical scope or a large number of devices can be divided into smaller sub-areas;

[0050] Finally, define the boundary ranges of each inspection area, mark them on the overall layout diagram of the power machine room, and assign each inspection area a unique management number to facilitate subsequent task allocation and corresponding management of data.

[0051] Step S2: Obtain the physical environment conditions, equipment information, and potential safety hazard situations in each inspection area to formulate physical environment monitoring tasks, equipment status monitoring tasks, and potential safety hazard investigation tasks for each inspection area;

[0052] The specific process is as follows. First, determine the physical environment conditions in each inspection area and the corresponding monitoring work contents to be carried out to formulate environmental parameter monitoring tasks in each inspection area; among them, the environmental parameter monitoring tasks include but are not limited to temperature, humidity, smoke concentration, and dust concentration, etc.

[0053] Secondly, determine the equipment within each patrol area and the corresponding status monitoring work content to be carried out, so as to formulate the equipment status monitoring tasks within each patrol area; among them, the equipment status monitoring tasks include but are not limited to voltage, current, switch status, connection status, operation status, etc. For example, the readings of the voltage meter and ammeter, switch status, and connection parts of the power distribution cabinet; another example is the indicator light status and operation status of the server; another example is the temperature setting value, air outlet temperature, and compressor operation status of the refrigeration equipment.

[0054] Finally, determine the safety hazard situations within each patrol area and the corresponding status monitoring work content to be carried out, so as to formulate the safety hazard investigation tasks within each patrol area; among them, the safety hazard investigation tasks include leakage hazards, cable damage, equipment grounding status, foreign objects, water accumulation, and fire protection facilities.

[0055] Step S3: Based on the physical environment monitoring tasks, equipment status monitoring tasks, and safety hazard investigation tasks of each patrol area, a plurality of monitoring devices and a patrol robot are correspondingly deployed in each patrol area; among them, the monitoring devices include sensors, cameras, and network monitoring devices installed with monitoring agent software.

[0056] The specific process is as follows. First, based on the physical environment monitoring tasks, equipment status monitoring tasks, and safety hazard investigation tasks of each patrol area, first-class sensors or / and second-class sensors are installed at corresponding positions in each patrol area to collect the status of the equipment within each patrol area or / and the physical environment outside it; among them, the first-class sensors are installed outside the equipment in each patrol area, including temperature and humidity sensors, smoke sensors, and dust sensors for monitoring the physical environment conditions; the second-class sensors are installed on the equipment in each patrol area, including temperature sensors, current transformers, and vibration sensors for collecting the operation status of the equipment.

[0057] Secondly, based on the physical environment monitoring tasks, equipment status monitoring tasks, and safety hazard investigation tasks of each patrol area, a plurality of cameras (such as high-definition cameras with infrared functions) are installed at corresponding positions in each patrol area to collect images of the appearance of the equipment within each patrol area, images of personnel entering and leaving, images of equipment connections, and images of cable connections, etc.

[0058] Next, based on the physical environment monitoring tasks, equipment status monitoring tasks, and safety hazard investigation tasks in each inspection area, monitoring agent software is installed on the corresponding network monitoring devices in each inspection area to collect the CPU usage rate, memory occupancy, disk I / O conditions, network traffic, and process status of each network monitoring device and its interconnected devices; among them, the interconnected devices of the network monitoring devices include various sensors, cameras, and communication devices for data interaction between devices in each inspection area. It should be noted that the communication devices refer to each node (servers, storage devices, switches, routers, etc.) in the power service system.

[0059] Finally, inspection robots are deployed in each inspection area, and according to the boundary range and equipment distribution information of each inspection area, motion guides for the inspection robots to travel are deployed. It should be noted that the inspection robots in different inspection areas should have good mobility, be able to freely shuttle in the inspection areas, and be able to perform operations across regions according to requirements.

[0060] Step S4: According to the importance level and equipment layout of each inspection area, and in combination with the monitoring devices deployed in each inspection area, formulate a parallel inspection plan for the inspection robots in the power machine room; among them, the parallel inspection plan includes the inspection task priorities, task volumes, inspection paths, inspection cycles, and inspection times required for each inspection cycle of each inspection robot.

[0061] The specific process is as follows. First, according to the importance level and equipment layout of each inspection area, and in combination with the monitoring devices deployed in each inspection area, determine the inspection task priorities, task volumes, inspection cycles, and inspection times required for each inspection cycle of each inspection robot, so as to reasonably arrange the data collection time intervals and sequences of each inspection robot in different inspection areas, ensure that the inspection work in each inspection area can be carried out synchronously, and achieve distributed collaborative inspection. For example, for the high-risk main computer room area (business system equipment) area, key equipment operation parameters and safety protection devices can be focused on for inspection; for the auxiliary area (ordinary office area), mainly check the basic operation conditions of fire protection facilities and electrical equipment, so as to improve the quality of inspection.

[0062] It should be noted that a typical power machine room is generally laid out according to functions, and the machine room is divided into a main machine room area, a support area, and an auxiliary area. The main machine room area is the area where various important devices such as servers and main network devices are placed, mainly used for the installation and operation of information processing, storage, switching, and transmission devices, such as functional areas like the server area, network area, and storage area. This area has high security requirements and strict control over personnel access. The support area is a technical operation place that supports and ensures the completion of the information processing process, such as the transformer substation, UPS room, and fire protection facility room. The auxiliary area is the place for the installation, commissioning, maintenance, operation monitoring, and management of equipment and software, including the incoming line room, monitoring center, spare parts warehouse, staff office, duty room, and maintenance room.

[0063] Secondly, according to the task volume and inspection time of each inspection robot, combined with the preset data collection frequency for each task and the movement guide rails in each inspection area, the inspection paths of each inspection robot are formulated. For example, according to the inspection tasks of the inspection robots in each inspection area, combined with the time requirements and data collection frequency of the inspection tasks, a reasonable inspection path is planned for each inspection robot. When planning the path, safety needs to be considered to avoid damage caused by the inspection robot colliding with equipment, cables, etc. At the same time, the charging requirements of the inspection robot are also taken into account.

[0064] Finally, based on the inspection task priorities, task volumes, inspection cycles of each inspection robot, and the inspection time required within each inspection cycle, and combined with the inspection paths of each inspection robot, a parallel inspection plan for the inspection robots in the power machine room is obtained.

[0065] Step S5: According to the parallel inspection plan of the inspection robots in the power machine room, drive each inspection robot to conduct inspections to obtain the inspection data of each inspection robot for its corresponding inspection area, and further combine the monitoring data collected in real time by each monitoring device in each inspection area to determine the final inspection results of each inspection area.

[0066] The specific process is as follows. First, when each inspection robot starts according to the parallel inspection plan, obtain the position, status, task execution progress, and collected data uploaded by each inspection robot, and further obtain the monitoring data collected in real time by each monitoring device in each inspection area. For example, after each inspection robot starts according to the predetermined parallel inspection plan, it maintains close contact with the data collection system through a real-time communication mechanism, continuously uploading information such as its own position, status, task execution progress, and collected data; the monitoring devices (including sensors, cameras, and monitoring agent software) in each inspection area upload various parameter data collected in real time to the data collection system.

[0067] Finally, perform data preprocessing on the location, status, task execution progress uploaded by each inspection robot, as well as the data collected, and perform preprocessing on the monitoring data collected in real time by each monitoring device in each inspection area. Further comprehensively analyze the physical environment condition data, equipment operation real-time data, and potential safety hazard data in each inspection area to determine the final inspection results for each inspection area. For example, perform preprocessing operations on the massive amount of raw data collected during the inspection process. On this basis, use big data analysis techniques and tools to analyze and mine the data. At this time, the cleaning of the raw data includes removing invalid data, noise data, duplicate data, and obvious abnormal data, etc.; the preprocessing of the raw data includes operations such as data format conversion and standardization processing. The real-time status data analysis is achieved by analyzing and displaying the actually collected data, which can intuitively reflect the physical environment condition data, equipment operation real-time data, and potential safety hazard data in each inspection area, and output the final inspection results.

[0068] In the embodiment of the present invention, the method further includes: performing chart statistics on the physical environment condition data, equipment operation real-time data, and potential safety hazard data in each inspection area obtained through comprehensive analysis to intuitively reflect the final inspection results of each inspection area. For example, adopt statistical analysis methods to calculate the statistical characteristic values of each equipment operation parameter, draw statistical charts of the equipment operation status, intuitively display the operation status and performance change trend of the equipment, and timely discover potential abnormal situations of the equipment.

[0069] In the embodiment of the present invention, the method further includes: using machine learning algorithms to model and train the physical environment condition data, equipment operation real-time data, and potential safety hazard data in each inspection area to analyze the abnormalities in the final inspection results of each inspection area. For example, use machine learning algorithms to model and train equipment failure data to establish equipment failure prediction models and diagnostic models. If it is determined that abnormal data parameters are collected by sensors, cameras, robots, monitoring agent software, etc., and abnormal situations are found when the data processing and analysis system performs data analysis, a real-time warning mechanism will be triggered. The alarm system establishes a collaborative processing mechanism (across regions and devices) for the warning, quickly starts the collaborative processing process to handle the abnormality, and at the same time, the data processing and analysis system will record the abnormal information in detail and upload it to the alarm system, including key information such as the time, location, equipment number, and abnormal parameter values when the abnormality occurs.

[0070] In an embodiment of the present invention, the method further includes: evaluating the final inspection results of each inspection area according to the actual operation requirements, management priorities, and resource investment of the power machine room, so as to optimize and obtain the next inspection plan for the inspection robot in each inspection area. For example, according to the actual operation requirements, management priorities, and resource investment of the power machine room, the inspection work is evaluated periodically, deficiencies and problems are found therein, targeted optimization is carried out, and the results are fed back to the inspection work in the next stage to facilitate the dynamic inspection plan. In addition, according to the preset inspection effect evaluation index system, various factors affecting the evaluation indexes are deeply analyzed to find out the problems and deficiencies in the inspection process.

[0071] Thus, the advantages of big data technology in the inspection of power machine rooms can be fully utilized, the efficient operation of distributed collaborative inspection can be realized, potential problems of the equipment in the machine room can be discovered and solved in a timely manner, and the operation and maintenance management level and reliability of the power machine room can be improved. In practical applications, the above steps and methods can be appropriately adjusted and optimized according to factors such as the specific scale, equipment type, and operating environment of the power machine room to meet the personalized inspection requirements of different machine rooms.

[0072] As Figures 2 to 3 shown, the application scenario of a distributed collaborative inspection method for power machine rooms based on big data provided by an embodiment of the present invention is further described as follows:

[0073] Typical power machine rooms are generally laid out according to functions, and the machine room is divided into a main machine room area, a support area, and an auxiliary area. The main machine room area is the area where various important devices such as servers and main network devices are placed, mainly used for the installation and operation of information processing, storage, switching, and transmission devices, such as functional areas such as the server area, network area, and storage area. This area has high security requirements and strict control over personnel entry and exit. The support area is a technical operation place that supports and guarantees the completion of the information processing process, such as the transformer substation, UPS room, and fire protection facilities room. The auxiliary area is the place for the installation, commissioning, maintenance, operation monitoring, and management of equipment and software, including the incoming line room, monitoring center, spare parts warehouse, staff office, duty room, maintenance room, etc.

[0074] Sensors commonly used for assisting in patrol inspections in power machine rooms can be environmental monitoring types, equipment status monitoring types, video monitoring types, etc. Among the environmental monitoring types, temperature and humidity sensors are used to obtain the temperature and humidity values in the machine room in real time to determine whether the machine room environment is in a state suitable for equipment operation; smoke sensors detect the concentration of smoke to discover potential fire hazards and immediately issue an alarm when the smoke concentration exceeds the threshold; water leakage monitors can detect water leakage in a timely manner and issue an alarm through the monitoring cables and control modules deployed in the areas prone to water leakage in the machine room. Among the equipment status monitoring types, infrared thermal imaging sensors can detect the surface temperature distribution of various equipment and visually display the heating conditions of the equipment through thermal imaging diagrams; partial discharge monitoring sensors, such as transient earth voltage (TEV) sensors, ultrasonic sensors, high-frequency current (HFCT) sensors, etc., can be used to detect the partial discharge conditions of power equipment and can issue early warnings in the early stage of equipment insulation failures; vibration sensors can assist in judging whether there are mechanical failures or unbalanced operation problems in the equipment by detecting the vibration conditions of power equipment; noise sensors can monitor the noise levels in the machine room or equipment in real time to help discover in a timely manner that some equipment failures may cause abnormal noises. The video monitoring type mainly includes various cameras, which can collect the image information of the machine room site in real time. The real-time images of the machine room equipment, environment, and personnel activities can be viewed through the remote monitoring system to help discover problems such as abnormal equipment appearances and personnel's illegal operations in a timely manner. Patrol robots can be regarded as a special type of sensor. Patrol robots are usually equipped with a variety of sensors, such as high-definition cameras, infrared thermal imagers, temperature and humidity sensors, sound collectors, and gas detectors, etc., which can achieve all-round monitoring and data collection of the machine room equipment and environment, replace manual inspections, and improve the inspection efficiency and accuracy.

[0075] A unified numbering rule needs to be formulated for the distributed inspection areas in the power machine room, the equipment and monitoring devices in each area, etc. ①: For the distributed inspection area numbering, first number according to the functional partitions of the machine room in alphabetical order of capital letters. For example, the main machine room area is A, the support area is B, and so on; then divide the specific inspection areas for each functional area and number them with two digits. For example, area A can be divided into A01, A02, …, Aam, and area B can be divided into B01, B02, …, Bbp, where am and bp are the numbers of sub - partitions in area A and area B respectively, and so on. ②: For the equipment numbering in each area, it is composed of the English word abbreviation (in lowercase letters) of the equipment name, a hyphen (-), and two digits. If the equipment name has only one word, take the first two letters; if it has multiple words, take the first letter of each word. For example, the number of the first server in a server area is se - 01; the number of the third data storage in a data storage area is ds - 03, and so on. ③: For the monitoring device numbering, number the various sensors in each area. It is composed of the English word abbreviation of the sensor type (the first letter is capitalized and the rest are lowercase), an underscore (_), and two digits. The abbreviation rule is the same as that of the equipment. For example, the number of the first temperature and humidity sensor in area A01 is Ths_01; the number of the first infrared thermal imaging sensor is Itis_01, and so on. The inspection robot can be numbered as a special monitoring device. For the monitoring device attached to a specific equipment, the equipment number can be added after its own number to indicate this relationship. In practical applications, usually connect the sub - area number with the equipment number or the monitoring device number to accurately determine the specific unit.

[0076] The abnormal handling rules in the power machine room inspection are a series of thresholds, abnormal classifications, and corresponding mechanisms established for data monitoring and abnormal judgment. Usually, thresholds are set for various equipment parameters and environmental indicators in the power machine room, such as server temperature, humidity, voltage, current, power, etc. Based on the normal operating range of the equipment and statistical analysis of historical data, reasonable thresholds are set. When the monitored data exceeds these thresholds, it is determined as an abnormal situation. For example, the normal range of the server CPU temperature is 40°C - 70°C. Once it exceeds 75°C, a high-temperature abnormal alarm is triggered. The abnormal classification standard divides the abnormalities into different levels according to the impact degree and urgency of the abnormal situation on the operation of the machine room equipment. Generally, it can be divided into three levels: minor abnormality, moderate abnormality, and severe abnormality. A minor abnormality may be that the parameters of some non-critical equipment deviate slightly from the normal range, but it does not affect the normal operation of the equipment for the time being. For example, the humidity in a certain area of the machine room is slightly lower than the normal range, but it has not reached the level that may cause electrostatic problems; a moderate abnormality means that some functions of the equipment may be affected to a certain extent and need to be processed within a short time. For example, a small number of bad sectors appear on one hard disk of a certain server, but the data redundancy mechanism can still ensure the normal operation of the business; a severe abnormality means that the equipment failure has seriously affected the overall operation of the machine room and may even cause business interruption. For example, the uninterruptible power supply (UPS) system in the machine room fails and cannot supply power normally. For different levels of abnormalities, corresponding response processes are formulated. A minor abnormality usually only needs to be recorded and a notice is sent to the inspection robot through the monitoring system. The inspection robot can conduct further inspections and processing during the inspection; a moderate abnormality requires the operation and maintenance personnel to immediately conduct a detailed assessment of the abnormal situation, formulate corresponding solutions, and solve the problem within a specified time (such as 1 - 2 hours), and record the processing process and results in the operation and maintenance log; for a severe abnormality, the emergency plan should be immediately activated. On the one hand, organize a professional technical team to repair the faulty equipment with all efforts, and on the other hand, promptly notify the relevant business departments to prepare for business switching or suspension, minimizing the losses caused by equipment failures. During the entire abnormal handling process, maintain communication and coordination with relevant departments and personnel to ensure timely and accurate information transmission.

[0077] To achieve the above objectives, the technical solutions adopted by the present invention are specifically implemented as follows:

[0078] (1) Distributed inspection area division. First, collect detailed information on the physical layout, equipment distribution, network topology, etc. of the power machine room to form an overall layout diagram of the machine room. For example, accurately measure the length, width, and height of the machine room, and record the position coordinates of each server cabinet, communication cabinet, power supply equipment, etc. Considering the characteristics of different functional areas in the machine room, divide the power machine room into several inspection areas, clearly define the boundary ranges of each inspection area, mark them on the overall layout diagram of the machine room, and assign a unique number to each area; then number the equipment in each area and associate it with its position coordinates, and mark it in the corresponding area of the layout diagram. As Figure 2 shown, for a typical power machine room, its main machine room area, support area, and auxiliary area can be determined as areas A, B, and C respectively. Then, the server area, network area, storage area, etc. in area A are divided into A01, A02,..., Aam (am is the number of areas in A); the transformer substation room, UPS room, and fire protection facility room in area B are determined as B01, B02,..., Bbp (bp is the number of areas in B); and area C is divided into C01, C02,..., Ccn (cn is the number of areas in C). Then, number the equipment in the above areas. For example, the servers in A01 are numbered as: se-01, se-02,..., se-sn (sn is the number of servers in area A1), as Figure 3 shown, the dotted arrows in the figure indicate the increasing rule of two digits in the numbering; the network devices in A02 are numbered as: nd-01, nd-02,..., nd-wn (wn is the number of network devices in area A2); the equipment in other areas is numbered according to the predetermined rules, and so on.

[0079] (2) Determine the inspection tasks for each area. According to the division of the distributed inspection areas, there may be differences in the physical environment (area, wiring, voltage, etc.), equipment, facilities, etc. in each area. For the physical environment, specific equipment information, and possible safety hazards in each area, formulate the environmental parameter monitoring tasks, equipment status inspection tasks, and safety hazard investigation tasks to be carried out in each inspection area. Taking area A01 as an example, in the environmental parameter monitoring task, it is necessary to collect temperature, humidity, cleanliness, etc. According to the A-level machine room standard when the machine is turned on: the temperature is 23±1°C, the relative humidity is 40%-55%, and the cleanliness is that the particle size ≥ 0.5 cubic micrometers, and the number ≤ 10,000 particles / dm 3; Consider this area as the server placement area. The device status check task needs to collect hardware status data, system performance data, device connection status, etc. Among them, the hardware status data includes the temperature, fan speed, power status, hard disk health status, etc. of key components such as the server CPU, memory, and hard disk. For example, the CPU temperature generally should not exceed 70℃-80℃; the system performance data includes CPU usage rate, memory usage rate, network bandwidth usage rate, process and service status, network traffic, etc. For example, under normal circumstances, the CPU usage rate should be between 30%-70%; the device connection status mainly checks the connection status of the server with other devices, such as network cables, storage device connections, etc., and the corresponding indicator light status can be collected. For example, when the server network cable interface is normally connected, the power indicator light is green, and the data transmission status indicator light is a flashing orange light; the safety hazard investigation task needs to collect the water leakage situation, smoke concentration, whether the cable outer skin is damaged, whether the grounding is good, etc.

[0080] (3) Deployment of monitoring devices. According to the inspection task requirements of each area, diverse sensors are first installed in each area and on each key device. For example, in Area A01, temperature and humidity sensors and dust sensors for collecting environmental parameters are installed; infrared thermal imaging sensors, current transformers, vibration sensors, etc. are installed at key parts of the server for collecting operation status data of the device; smoke sensors, water leakage sensors, high-definition cameras with infrared function, etc. are installed for collecting safety hazard data. Then, according to the inspection tasks, multiple intelligent inspection robots suitable for the power machine room environment are equipped. They should have good mobility, be able to freely shuttle in the inspection area, and be able to operate across areas according to needs. The robots can carry devices such as high-definition cameras, infrared thermal imagers, temperature and humidity sensors, sound collectors, and gas detectors, and can conduct real-time monitoring of the machine room equipment and environment in all directions without dead angles. Different robots can be configured according to the characteristics of the area and inspection tasks. For example, in Area A01, the configured inspection robot can accurately detect abnormal temperatures of devices through an infrared thermal imager and timely discover potential overheating hazards; the high-definition camera can carefully check the appearance of the device, the status of the indicator lights, etc., and use image recognition technology to automatically judge whether there are appearance damages, indicator light errors, etc. on the device, cable, etc. After deploying various hardware, number each type of hardware and mark its position or the attached device on the layout diagram. For example, the temperature and humidity sensors (Temperature and humidity sensor) in Area A01 are numbered Ths_01, Ths_02, …, Ths_tn (tn is the number of temperature and humidity sensors in Area A01), and the infrared thermal imaging sensors (Infrared Thermal Imaging Sensor) are numbered Itis_01, Itis_02, …, Itis_in (in is the number of infrared thermal imaging sensors in Area A01). If a certain infrared thermal imaging sensor Itis_n1 (n1 ∈ [01, in]) is attached to the server se-n2 (n2 ∈ [01, sn]), its actual number is represented as Itis_n1(se-n2).

[0081] After completing the deployment of the monitoring devices, the next step is to deploy data collection software for the power business system nodes (servers, storage devices, etc.) in each area, that is, deploy lightweight monitoring agent programs in each node to be responsible for collecting various types of information of the node, such as basic operation metrics like CPU usage, memory occupancy, disk I / O situation, network traffic, and process status. For example, lightweight monitoring agent programs (such as Zabbix, Netdata, etc.) need to be deployed on the servers in Area A01 for collecting the operation parameters of each server.

[0082] (4) Patrol task assignment and plan formulation. First, based on the division of the distributed patrol area, the equipment layout within each area, and the expected patrol time, while considering the coordination relationship with other monitoring devices in the area, determine the patrol tasks that the patrol robots in different areas need to complete. For example, in Area A01, the installed temperature and humidity sensors, dust sensors, smoke sensors, water leakage sensors, and high-definition cameras can collect most of the environmental parameters and safety hazard data. The key focus of the patrol robot in this area is the server equipment in the area. Infrared thermal imaging sensors, current transformers, vibration sensors, etc. are installed at key parts of the server to collect relevant data redundantly, improving the accuracy of monitoring the running state of the server. At the same time, carefully check the appearance of the equipment, the status of the indicator lights, the cable connection status, etc. Therefore, the patrol robot in Area A01 needs to traverse each server in the area and also needs to patrol certain monitoring points, such as the air outlet of the air conditioner, the cable gathering point (or communication relay equipment), etc. Next, comprehensively consider the priority, workload, required time, and patrol cycle of the patrol tasks in each area, formulate a detailed parallel patrol plan schedule, and reasonably arrange the data collection time intervals and sequences of each robot in different areas to ensure that the patrol work in each area can be carried out synchronously, realizing distributed collaborative patrol.

[0083] (5) Patrol path planning and parallel collaborative execution. According to the patrol tasks of the robots in each area, combined with the time requirements and data collection frequencies of the tasks, plan a reasonable patrol path for each robot. When planning the path, safety needs to be considered to avoid damage caused by the robot colliding with equipment, cables, etc., and at the same time, the charging requirements of the robot should be taken into account. For example, the patrol robot in Area A01 can determine the initial patrol path by combining the greedy algorithm and the Dijkstra algorithm based on the layout of the area, the number of servers, the detection points, etc. The evaluation function f(n) of this algorithm is used to determine the next node that must be traversed, and f(n) is expressed as:

[0084] f(n) = g(n) + h(n)

[0085] Among them, g(n) represents the actual cost from the starting point to the current node n (such as the distance already traveled), and h(n) is the estimated cost from the current node n to the target node (the virtual target state where all servers have been inspected). By comprehensively considering these two parts of the cost, the optimal path can be searched more efficiently. The algorithm execution process is as follows: ①: Initialization. Set the g value of the starting point to 0, set the h value according to the estimated heuristic distance to other servers, calculate the f value, put the starting point into the priority queue of nodes to be expanded (sorted in ascending order of f value), and at the same time create a set of visited nodes; ②: Select the node to be expanded. Take out the node with the smallest f value from the priority queue as the current node to be expanded, mark it as visited, and examine its adjacent nodes. ③: Update the information of adjacent nodes: For adjacent nodes, calculate the new g value of reaching them through the current node (the g value of the current node plus the edge weight, the higher the priority, the smaller the weight value), and then calculate the f value in combination with the estimated h value. If the new f value is less than the previously recorded f value of this adjacent node, or this adjacent node is discovered for the first time, then update its g, h, and f values, and put it into the priority queue (if it was not there before). ④: Repeat steps ② and ③. Repeat the above operations until all servers have been visited or the priority queue is empty (which means that not all servers can be reached, and in this case, it is necessary to check whether the layout or parameter settings are reasonable). Finally, sort out the inspection path according to the connection order of the nodes.

[0086] After each inspection robot starts according to the predetermined parallel inspection plan and inspection path, it maintains close contact through a real-time communication mechanism and data acquisition system, and continuously uploads information such as its own location, status, task execution progress, and collected data. At the same time, the monitoring devices (various sensors and monitoring agent programs) also upload various types of parameter data collected to the data acquisition system in real time.

[0087] (6) Data preprocessing and analysis. First, the data acquisition system collects various types of raw data from the previous step, cleans and preprocesses the raw data. Cleaning includes removing invalid data, noise data, duplicate data, etc.; data preprocessing includes operations such as data format conversion and standardization. Then, the data processing and analysis system can perform the following operations simultaneously: ①: Real-time status data analysis, analyzing and displaying through the actually collected data to intuitively reflect the real-time situation of the environment and equipment operation in each area at the current moment. For example, through the temperature and humidity sensors, infrared thermal imaging sensors installed in Area A01 and the monitoring agent program running on the server, the environmental temperature, CPU temperature, and temperature distribution of the target server at the current moment can be monitored in real time; ②: Data statistical analysis, using statistical analysis methods to calculate the statistical characteristic values of the operation parameters of each device, draw statistical charts of the device operation status, intuitively display the operation status and performance change trend of the device, and timely discover potential abnormal situations of the device. For example, through the time-series data of the CPU temperature of a certain server collected in Area A01, the data processing and analysis system can draw a corresponding line chart to discover the temperature change of the server CPU over a period of time; ③: Using machine learning algorithms to model and train the device failure data, establishing a device failure prediction model and a diagnostic model. For example, by dividing the data of server failures (performance degradation, system crash, restart, freeze, etc.) caused by overheating of the CPU temperature in Area A01 into a training set, a validation set, and a test set in a ratio of 7:2:1, and using the decision tree algorithm to model and train the server failure data. The training set is used to train the model, the validation set is used to adjust the hyperparameters of the model and select the best model, and the test set is used to evaluate the final performance of the model. After the model training is completed, when the CPU temperature of a certain server overheats, the possible failures can be predicted.

[0088] (7) Real-time warning and collaborative processing. When the sensor, camera, robot, and monitoring agent software collect data parameters, and when the data processing and analysis system performs data analysis, once an abnormal situation is detected, the monitoring device or the data processing and analysis system first records the abnormal information in detail and uploads it to the alarm system, including key information such as the time, location, device number, and abnormal parameter value when the abnormality occurs. For example, Abnormality 1: At 15:29:15.32 on November 29, 2024, the CPU temperature of the server se-01 in Area A01 was 86°C; Abnormality 2: At 16:11:12.23 on November 29, 2024, the power indicator of the network cable interface of the server se-06 in Area A01 was flashing on and off. According to the preset abnormal handling rules and the expert knowledge base, the alarm system automatically generates a preliminary abnormal handling plan. For abnormalities that can be handled by the patrol robot in this area, the patrol robot attempts to handle them. If the robot is unable to handle them or the handling is unsuccessful, the task is assigned to the corresponding operation and maintenance personnel or technical experts. For example, for Abnormality 2, it may be caused by the network cable not being plugged in tightly (poor contact). The robot is commanded to unplug and replug the network cable. If the power indicator stays on constantly after replugging, the fault is repaired; for Abnormality 1, assuming that the CPU overheating is caused by a poor heat dissipation system, then the task is assigned to the corresponding operation and maintenance personnel. After receiving the task, the operation and maintenance personnel quickly go to the site for further verification and handling according to the abnormal information and handling plan provided by the alarm system. For example, through on-site verification, it is found that Abnormality 1 is caused by the abnormal operation of the cooling fan. The operation and maintenance personnel repair it by replacing the cooling fan. During the handling process, they can obtain the latest data and technical support information pushed by the alarm system through the mobile terminal at any time, maintain close communication with other collaborative personnel, and jointly solve the abnormal problem. For warnings related to potential safety hazards, such as the smoke sensor alarm, it is handled according to the fire accident handling plan. After the abnormal handling is completed, the handling process and results are recorded and summarized in detail, and the relevant information is fed back to the alarm system and stored in the equipment maintenance file and the fault case library for reference in future operation and maintenance work.

[0089] (8) Patrol inspection effect evaluation and optimization. According to the predetermined effect evaluation indicators, combined with the actual operation requirements, management priorities, and resource investment of the computer room, the monitoring platform evaluates the patrol inspection work for each cycle, deeply analyzes various factors affecting the evaluation indicators, and finds out the problems and deficiencies in the patrol inspection process; then, targeted optimization is carried out and fed back to the patrol inspection work of the next cycle, so as to dynamically adjust the patrol inspection plan, make full use of big data technology, formulate specific and targeted optimization measures and improvement plans, continuously track and monitor the changes in the evaluation indicators, and verify the effectiveness and actual effects of the optimization measures. For example, based on the analysis of patrol inspection data, it is possible to identify connection points, devices, facilities, etc. that need to be key monitored. When formulating the next round of patrol inspection plan, the patrol inspection tasks and patrol route planning (step 5) can be dynamically adjusted. For example, when planning the patrol route, by adjusting the edge weights of key monitoring points, their priority can be made higher to obtain priority patrol inspection. At the same time, the patrol inspection frequency of this monitoring point can also be increased to detect possible abnormalities earlier.

[0090] Implementing the embodiments of the present invention has the following beneficial effects:

[0091] Compared with the traditional method of using an intelligent monitoring system to assist in patrolling the computer room, the present invention divides the patrol inspection area of the computer room by planning and optimizing the computer room resources, enables the monitoring devices and patrol inspection robots in each patrol inspection area to work together simultaneously, and further uses big data technology to perform real-time analysis on the multi-modal data collected distributively in each patrol inspection area to quickly locate the final patrol inspection results of each patrol inspection area, thereby achieving the purpose of improving the quality and efficiency of patrol inspection and reducing the operation and maintenance costs.

[0092] Those of ordinary skill in the art can understand that all or part of the steps in implementing the above method embodiments can be completed by instructing relevant hardware through a program, and the program can be stored in a computer-readable storage medium, such as ROM / RAM, disk, optical disc, etc.

[0093] The above-disclosed are only the preferred embodiments of the present invention. Of course, the scope of the rights of the present invention cannot be limited thereby. Therefore, equivalent changes made according to the claims of the present invention still fall within the scope covered by the present invention.

Claims

1. A distributed collaborative inspection method for power machine rooms based on big data, characterized in that, The method includes the following steps: According to the relevant actual information of the power machine room, divide the power machine room into multiple inspection areas; Obtain the physical environment conditions, equipment information and potential safety hazard situations in each inspection area to formulate the physical environment monitoring tasks, equipment status monitoring tasks and potential safety hazard investigation tasks for each inspection area; Based on the physical environment monitoring tasks, equipment status monitoring tasks and potential safety hazard investigation tasks for each inspection area, a plurality of monitoring devices and an inspection robot are correspondingly deployed in each inspection area; wherein, the monitoring devices include sensors, cameras and network monitoring devices installed with monitoring agent software; According to the importance degree and equipment layout of each inspection area, and in combination with the monitoring devices deployed in each inspection area, formulate the parallel inspection plan for the inspection robots in the power machine room; wherein, the parallel inspection plan includes the inspection task priorities, task volumes, inspection paths, inspection cycles and the inspection time required within each inspection cycle for each inspection robot; According to the parallel inspection plan for the inspection robots in the power machine room, drive each inspection robot to conduct inspections to obtain the inspection data of each inspection robot for its corresponding inspection area, and further combine the monitoring data collected in real time by each monitoring device in each inspection area to determine the final inspection results of each inspection area.

2. The distributed collaborative inspection method for power machine rooms based on big data according to claim 1, wherein, The specific steps of dividing the power machine room into multiple inspection areas according to the relevant actual information of the power machine room include: Obtain the relevant actual information of the power machine room, including the overall physical layout information, functional area information, equipment distribution information and network topology information; According to the relevant actual information of the power machine room, form an overall layout diagram of the power machine room, and in combination with the importance degree of each functional area in the power machine room, divide the power machine room into several inspection areas; Define the boundary ranges of each inspection area, mark them on the overall layout diagram of the power machine room, and assign each inspection area a unique management number.

3. The distributed collaborative inspection method for power machine rooms based on big data according to claim 2, wherein, The specific steps of obtaining the physical environment conditions, equipment information and potential safety hazard situations in each inspection area to formulate the physical environment monitoring tasks, equipment status monitoring tasks and potential safety hazard investigation tasks for each inspection area include: Determine the physical environment conditions in each inspection area and the corresponding monitoring work contents to be carried out to formulate the environmental parameter monitoring tasks in each inspection area; wherein, the environmental parameter monitoring tasks include temperature, humidity, smoke concentration and dust concentration; Determine the equipment in each inspection area and the corresponding status monitoring work contents to be carried out to formulate the equipment status monitoring tasks in each inspection area; wherein, the equipment status monitoring tasks include voltage, current, switch status, connection status and operation status; Determine the potential safety hazard situations in each inspection area and the corresponding status monitoring work contents to be carried out to formulate the potential safety hazard investigation tasks in each inspection area; wherein, the potential safety hazard investigation tasks include leakage hazards, cable damage, equipment grounding status, foreign objects, water accumulation and fire protection facilities.

4. The distributed collaborative inspection method for power machine rooms based on big data according to claim 3, characterized in that, The specific steps of deploying multiple monitoring devices and one inspection robot in each inspection area based on the physical environment monitoring tasks, equipment status monitoring tasks, and safety hazard detection tasks in each inspection area are as follows: Based on the physical environment monitoring tasks, equipment status monitoring tasks, and safety hazard detection tasks in each inspection area, first-class sensors or / and second-class sensors are installed at corresponding positions in each inspection area to collect the status of equipment in each inspection area or / and the physical environment outside it. Among them, the first-class sensors are installed outside the equipment in each inspection area and include temperature and humidity sensors, smoke sensors, and dust sensors for monitoring the physical environment conditions; the second-class sensors are installed on the equipment in each inspection area and include temperature sensors, current transformers, and vibration sensors for collecting the operating status of the equipment; Based on the physical environment monitoring tasks, equipment status monitoring tasks, and safety hazard detection tasks in each inspection area, multiple cameras are installed at corresponding positions in each inspection area to collect images of the appearance of equipment in each inspection area, images of personnel entering and leaving, images of equipment connections, and images of cable connections; Based on the physical environment monitoring tasks, equipment status monitoring tasks, and safety hazard detection tasks in each inspection area, monitoring agent software is installed on the corresponding network monitoring devices in each inspection area to collect the CPU usage rate, memory occupancy, disk I / O conditions, network traffic, and process status of each network monitoring device and its interconnected devices. Among them, the interconnected devices of the network monitoring devices include various sensors, cameras, and communication devices for data interaction between equipment in each inspection area; An inspection robot is deployed in each inspection area, and according to the boundary range and equipment distribution information of each inspection area, a motion guide rail for the inspection robot to travel is deployed.

5. The distributed collaborative inspection method for power machine rooms based on big data according to claim 4, wherein The specific steps of formulating the parallel inspection plan for the inspection robots in the power distribution room according to the importance level and equipment layout of each inspection area and in combination with the monitoring devices deployed in each inspection area are as follows: According to the importance level and equipment layout of each inspection area and in combination with the monitoring devices deployed in each inspection area, determine the inspection task priority, task volume, inspection cycle, and inspection time required in each inspection cycle for each inspection robot; According to the task volume and inspection time of each inspection robot and in combination with the preset data collection frequency for each task and the motion guide rail in each inspection area, formulate the inspection path for each inspection robot; Based on the inspection task priority, task volume, inspection cycle, and inspection time required in each inspection cycle for each inspection robot and in combination with the inspection path of each inspection robot, obtain the parallel inspection plan for the inspection robots in the power distribution room.

6. The distributed collaborative inspection method for power machine rooms based on big data according to claim 5, wherein, The specific steps of driving each inspection robot to conduct inspections according to the parallel inspection plan for the inspection robots in the power distribution room to obtain the inspection data of each inspection robot for its corresponding inspection area and further combining the monitoring data collected in real time by each monitoring device in each inspection area to determine the final inspection results of each inspection area are as follows: When each inspection robot starts according to the parallel inspection plan, obtain the location, status, task execution progress, and collected data uploaded by each inspection robot, and further obtain the monitoring data collected in real time by each monitoring device in each inspection area; Perform data preprocessing on the location, status, task execution progress, and collected data uploaded by each inspection robot, and perform preprocessing on the monitoring data collected in real time by each monitoring device in each inspection area, and further comprehensively analyze the physical environment condition data, equipment operation real-time data, and potential safety hazard situation data in each inspection area to determine the final inspection result of each inspection area.

7. The distributed collaborative inspection method for power machine rooms based on big data according to claim 6, characterized in that The method further includes: Perform chart statistics on the physical environment condition data, equipment operation real-time data, and potential safety hazard situation data in each inspection area obtained from the comprehensive analysis to visually reflect the final inspection result of each inspection area.

8. The distributed collaborative inspection method for power machine rooms based on big data according to claim 7, characterized in that The method further includes: Use machine learning algorithms to model and train the physical environment condition data, equipment operation real-time data, and potential safety hazard situation data in each inspection area to analyze the anomalies in the final inspection results of each inspection area.

9. The distributed collaborative inspection method for power machine rooms based on big data according to claim 8, wherein The method further includes: Evaluate the final inspection results of each inspection area according to the actual operation requirements, management priorities, and resource investment situation of the power machine room to optimize and obtain the next inspection plan for the inspection robots in each inspection area.

Images