Intelligent operation and maintenance system and method based on big data

Abstract

The invention discloses an intelligent operation and maintenance system and method based on big data, and relates to the technical field of intelligent operation and maintenance. The method comprises the main steps of calculating the importance of each piece of maintenance data to obtain a maintenance scheme; constructing a maintenance scheme template according to the main steps; the effectiveness of each maintenance scheme is calculated, the maximum effectiveness is extracted as the optimal maintenance scheme, and the optimal maintenance scheme of different fault reasons under each fault phenomenon is obtained; collecting data values of all fault phenomena when the equipment works in real time, and judging whether the equipment breaks down or not; a maintenance scheme is selected from an automatic maintenance scheme database to be output, different maintenance schemes are flexibly formulated according to different fault reasons, different types of faults of equipment can be subjected to targeted maintenance scheme adjustment, tedious maintenance steps can be simplified, the fault maintenance efficiency is greatly improved, and the maintenance cost is reduced. And the conformity of the maintenance scheme and the fault is greatly improved.

Description

Technical Field

[0001] The present invention relates to the field of intelligent operation and maintenance technology, and specifically to an intelligent operation and maintenance system and method based on big data. Background Art

[0002] With the introduction and development of the concepts of industry and intelligent manufacturing, industries such as manufacturing have put forward higher requirements for the intelligent management and efficient operation and maintenance of production equipment. Enterprises need to use advanced technologies to achieve real-time monitoring, fault prediction and rapid repair of equipment in order to improve production efficiency, reduce costs and enhance market competitiveness, which has prompted the emergence and development of intelligent operation and maintenance technology. Internet of Things technology enables equipment to be interconnected, and a large number of sensors can be deployed on various devices to collect equipment operation data in real time, such as temperature, pressure, vibration, etc. These data provide a rich information basis for intelligent operation and maintenance, making it possible to fully perceive the status of equipment. In recent years, with the deep integration of technologies such as artificial intelligence, big data, and the Internet of Things, intelligent operation and maintenance has become a development trend in the field of operation and maintenance. The intelligent operation and maintenance system can not only realize real-time monitoring and fault alarm of equipment, but also predict and evaluate the operating status of equipment through data analysis and machine learning algorithms, discover potential fault risks in advance, and automatically generate corresponding maintenance plans and suggestions. However, with the development of science and technology, various equipment has become more and more sophisticated and complex. Maintenance after equipment failure often requires specific technicians to perform. Ordinary staff cannot maintain the equipment, which greatly increases the difficulty and cost of equipment operation and maintenance. Moreover, since there are often fewer technicians, the equipment cannot be maintained in time when multiple failures occur. Summary of the invention

[0003] The purpose of the present invention is to provide an intelligent operation and maintenance system and method based on big data to solve the problems raised in the prior art.

[0004] To achieve the above object, the present invention provides the following technical solutions: An intelligent operation and maintenance method based on big data, the method comprising the following steps: S100, collecting records of equipment failures during historical operation and maintenance by technicians, collecting failure data in the records, analyzing the failure data to obtain failure phenomena in each record, and classifying equipment failures according to the failure phenomena; Furthermore, the specific steps for classifying equipment failures according to the failure phenomena are as follows: S101. Collect the records of equipment failures during the operation and maintenance of the equipment by the technicians in the past, collect and record each type of failure data of the equipment during the operation process, and assume that each type of failure data that changes with the operation of the equipment is {S1, S2, S3, ..., S n}, S1, S2, S3, ..., S n Indicates collecting the 1st, 2nd, 3rd, ..., nth data values ​​of each type of fault data that changes as the equipment works, where n is a positive integer; using the b data values ​​of each type of fault data collected to draw a curve graph, the abscissa is the working time of the equipment, and the ordinate is the data value of each type of fault data; S102, analyzing the drawn curve graph, extracting the peak value and the low estimate value in the curve graph, judging the data value of each point in the curve graph, and x-1 x >S x+1 When S x is the peak value in the curve. x-1 >S x x+1 When S x is the underestimation in the curve graph; let the peak value in the curve graph be {f1, f2, f3, ..., f k}, f1, f2, f3, ..., f k represents the 1st, 2nd, 3rd, ..., kth peaks in the curve graph obtained by judgment, where k is a positive integer; suppose the low estimate in the curve graph obtained by judgment is {d1, d2, d3, ..., d p}, d1, d2, d3, ..., d p It represents the 1st, 2nd, 3rd, ..., pth underestimated values ​​in the curve graph obtained by judgment, where p is a positive integer; and the average data value of the fault data in the curve graph is calculated as Sp, and the absolute difference between each peak value, underestimated value and average data value is calculated, and the formula is: ; In the formula, Sc i Represents the absolute difference between the i-th peak and the average data value in the curve graph, f i Indicates the i-th peak in the curve graph; Sc j Represents the absolute difference between the jth underestimation and the average value in the curve graph, d j Indicates the jth underestimation in the curve graph; after calculation, the absolute difference between k peak values ​​and the average value data value, and the absolute difference between p underestimation values ​​and the average value data are obtained; the maximum value is selected from the k+p absolute differences as the change feature of the corresponding fault data; the change features of all fault data in the record are calculated in turn to obtain n types of change features of fault data, and the maximum value is selected as the fault phenomenon of the record of the historical operation and maintenance equipment failure; ​​S103, all the records of equipment failures collected in the historical operation and maintenance are calculated to obtain the fault phenomenon of each equipment failure record, and all the records with the same fault phenomenon are classified into the same fault type, and finally the equipment failure is classified into {L1, L2, L3, ..., L m}, L1, L2, L3, ..., L m Indicates the 1st, 2nd, 3rd, ...mth fault category of the equipment, where m is a positive integer.

[0005] After extracting fault phenomena and classifying faults using fault data from historical operation and maintenance, when a device fails, the system can quickly narrow the scope of investigation based on the fault category, greatly shortening the fault location time; based on the fault classification, a more targeted maintenance plan can be formulated, and different maintenance methods can be adopted for different fault phenomena, which can greatly increase the efficiency of equipment maintenance; S200, after classifying the equipment failures, collect the equipment operation data when each failure occurs in the historical operation and maintenance, calculate the correlation between each failure phenomenon and all equipment operation data, and find all failure causes of each failure phenomenon; Furthermore, the specific steps for finding all the causes of each fault phenomenon are as follows: S201, after classifying the equipment failures, collect the equipment operation data when each failure phenomenon occurs in the historical operation and maintenance as {Y1, Y2, Y3, ..., Y h}, Y1, Y2, Y3, ..., Y h Indicates the 1st, 2nd, 3rd, ..., hth equipment operation data collected in the historical operation and maintenance when each fault phenomenon occurs, where h is a positive integer; extracts the fault phenomenon data value when each fault occurs, and draws a scatter plot with the fault phenomenon data value as the horizontal axis and the collected equipment operation data as the vertical axis; draws a scatter plot of h kinds of equipment operation data and fault phenomenon data values ​​in each fault; uses the linear regression algorithm to fit the h kinds of scatter plots, and obtains the linear function in the scatter plot as , in which Sx represents the fault phenomenon of each fault; a represents the slope of the linear function, and b represents the intercept of the linear function; S202, for each of the h scatter plots, the corresponding linear function is calculated, and the slope of each linear function is extracted as the correlation between the equipment operation data and the fault phenomenon in the scatter plot. The correlation between all the equipment operation data and the fault phenomenon is assumed to be {a1, a2, a3, ..., a h}, a1, a2, a3, ..., a h Indicates the correlation between the 1st, 2nd, 3rd, ..., hth equipment operation data and fault phenomena, calculates the average value of the h kinds of correlation as ap, and makes a judgment on each correlation. , determine that the equipment operation data corresponding to correlation a is the cause of the fault. When the equipment operation data corresponding to the correlation a is judged not to be the cause of the fault; after judgment, the cause of each fault phenomenon is {ay1, ay2, ay3, ..., ay t}, ay1, ay2, ay3,..., ay t It indicates that the 1st, 2nd, 3rd, ..., tth fault causes of the fault phenomenon are judged, and t is a positive integer; the fault causes of the fault phenomenon of each fault are judged to obtain all the fault causes.

[0006] Find the cause of each fault phenomenon of the equipment. After clarifying the different causes corresponding to each fault, maintenance personnel can take accurate maintenance measures. The analysis results of the fault causes can provide valuable feedback to the equipment design department to optimize the equipment according to the fault causes. S300, collecting maintenance plans for equipment failures by technicians in historical operation and maintenance, extracting all maintenance data in the maintenance plan, calculating the importance of each maintenance data, and obtaining the main steps of the maintenance plan; and constructing a maintenance plan template according to the main steps; Furthermore, the specific steps for constructing a maintenance plan template according to the main steps are: S301, collecting maintenance plans for equipment failures by technicians in historical operation and maintenance, extracting all maintenance data in the maintenance plans, comparing the collected maintenance data in the maintenance plans for equipment failures by technicians in historical operation and maintenance, and extracting the maintenance data existing in all maintenance plans as {W1, W2, W3, ..., W g}, W1, W2, W3, ..., W g Indicates the 1st, 2nd, 3rd, ..., gth types of maintenance data that exist in all maintenance plans, where g is a positive integer; extract the equipment performance before and after the maintenance data that exist in all maintenance plans are executed in historical operation and maintenance, and calculate the importance of each maintenance data that exists. The formula is: ; In the formula, MS u represents the importance of the maintenance data that exists in the u-th type, P u represents the equipment performance after the u-th type of maintenance data is executed, P u ’ Equipment performance before the execution of maintenance data of the uth type; X u Indicates the amount of information of the maintenance data that exists in the u-th type; S302. Obtain the importance of each type of maintenance data through calculation, and calculate the average value of all existing maintenance data, select maintenance data with importance greater than the average importance as the main steps of the maintenance plan, and construct a maintenance plan template using all the main steps obtained through judgment.

[0007] By extracting the main steps of the maintenance plan in the historical operation and maintenance and generating the maintenance plan template, different maintenance plans can be flexibly formulated for different fault causes. Targeted maintenance plans can be adjusted for different types of equipment faults, and cumbersome maintenance steps can be simplified, which greatly increases the efficiency of fault maintenance and greatly increases the fit between the maintenance plan and the fault. S400, for all the fault causes of each fault phenomenon, the maintenance plans of technicians in historical operation and maintenance are collected, the effectiveness of each maintenance plan is calculated, the best maintenance plan with the greatest effectiveness is extracted, the main steps in the best maintenance plan are extracted and input into the plan template, the best maintenance plan for different fault causes under each fault phenomenon is obtained, and an automatic maintenance plan database is generated; Furthermore, the specific steps of generating the automatic maintenance solution database are: S401. For all the fault causes of each fault phenomenon, the maintenance plans of the technicians in the historical operation and maintenance are collected. The execution time of all the maintenance plans of the technicians in the historical operation and maintenance and the number of faults caused by the same fault cause after execution are collected. The effectiveness of different maintenance plans for the same fault cause is calculated. The formula is: In the formula, Ef represents the effectiveness of different maintenance plans for the same fault cause, T represents the execution time of different maintenance plans by technicians in the collected historical operation and maintenance, and C represents the number of faults caused by the same fault cause after different maintenance plans are executed by technicians in the collected historical operation and maintenance; the maintenance direction with the minimum effectiveness is selected as the best maintenance plan for the corresponding fault cause under the corresponding type of fault phenomenon; S402. After determining the best maintenance plan for each cause of failure, extract the maintenance data contained in the main steps of the best maintenance plan, input the extracted maintenance data of the main steps into the maintenance plan template, generate a maintenance plan for each cause of failure, and unify the maintenance plans for all causes of failure to generate an automatic maintenance plan database.

[0008] Select the best maintenance plan to generate an automatic maintenance plan database. When it is determined that the equipment fails, the corresponding maintenance plan can be automatically generated, so that the staff can maintain the equipment failure according to the displayed maintenance plan. The equipment can be maintained without the need for technical personnel, which reduces the maintenance cost caused by the specialization of operation and maintenance, greatly reduces the maintenance time of equipment failure, and avoids irreversible damage to the equipment; S500, calculate the fault threshold value according to the corresponding data value of each fault phenomenon during historical operation and maintenance, collect the data values ​​of all fault phenomena when the equipment is working in real time, and judge whether the equipment has a fault; collect all real-time fault cause data to judge the real-time fault cause, and select the maintenance plan in the automatic maintenance plan database for output; Furthermore, the specific steps of selecting a maintenance plan for output in the automatic maintenance plan database are: S501. Collect data values ​​when each fault phenomenon occurs in historical operation and maintenance, and calculate fault thresholds of different fault phenomena. When the maximum absolute difference when judging the fault phenomenon in S100 is the peak value, the formula is: ; When the maximum absolute difference in judging the fault phenomenon is an underestimate, the formula is: ; In the formula, Gy represents the calculated fault threshold of different fault phenomena, Sx_p represents the average data value of different fault phenomena, and Sx_st represents the standard deviation of the data value of different fault phenomena; S502, when the device is working, collect the real-time data value of each fault phenomenon as Sxs, use the corresponding fault threshold to judge the real-time fault phenomenon, when the maximum absolute difference value when judging the fault phenomenon is the peak value, , determine that the equipment has a corresponding fault phenomenon; When the maximum absolute difference in judging the fault phenomenon is an underestimate, , determine that the equipment has a corresponding fault phenomenon; When it is determined that the equipment has failed, the data values ​​of all the fault causes under the fault phenomenon are collected. When the data value of the fault cause changes, the real-time fault cause of the equipment fault phenomenon that occurs in real time is determined; the maintenance plan is selected and output in the automatic maintenance plan database according to the real-time fault cause.

[0009] S600, collect all the operation and maintenance knowledge of the equipment, build a visual automatic operation and maintenance platform, combine the operation and maintenance knowledge to carry out fault warning in the automatic operation and maintenance platform, and display the fault phenomenon, fault cause and maintenance plan of the fault warning. The staff will perform fault maintenance on the equipment according to the automatic operation and maintenance platform and learn operation and maintenance knowledge.

[0010] Furthermore, the specific steps to build a visual automatic operation and maintenance platform are: S601. Collect all operation and maintenance knowledge of the equipment, use Django technology to build a visual automatic operation and maintenance platform, input and store all operation and maintenance knowledge for users to query and learn; when it is determined that the equipment fails, the failure phenomenon, cause of the failure and recommended maintenance plan will be displayed in the automatic operation and maintenance platform, and the failure will be warned to remind users to perform maintenance.

[0011] Through the automatic operation and maintenance platform, all staff can monitor and maintain the equipment operation, and they can also learn professional knowledge on the platform, which increases their professional capabilities in daily fault maintenance and reduces the cost investment in equipment maintenance.

[0012] An intelligent operation and maintenance system based on big data, the intelligent operation and maintenance system includes a data collection module, a fault classification module, a fault cause finding module, a maintenance plan generation module, a fault judgment module and a visualization module; The data collection module is used to collect fault data, equipment operation data and maintenance plans when equipment fails in historical operation and maintenance; The fault classification module is used to extract fault phenomena based on the fault data of equipment failures in historical operation and maintenance, and classify equipment failures according to the fault phenomena; The fault cause finding module is used to analyze the equipment operation data during historical operation and maintenance, calculate the correlation between each equipment operation data and the fault phenomenon, and determine all the fault causes under each fault phenomenon; The maintenance plan generation module is used to analyze the maintenance plans in historical operation and maintenance, extract the main steps to build a maintenance plan template, and find the best maintenance plan for each fault cause, extract the main steps and input them into the maintenance plan template to generate a maintenance plan; The fault judgment module is used to calculate the fault thresholds of different fault phenomena according to the data values ​​of the fault phenomena, judge whether the equipment has a fault based on the real-time fault phenomenon data values, and find the real-time fault cause; The visualization module is used to build a visualized automatic operation and maintenance platform, perform fault warning in the automatic operation and maintenance platform in combination with operation and maintenance knowledge, and display the fault phenomenon, fault cause and maintenance plan of the fault warning.

[0013] The maintenance plan generation module includes a maintenance plan template generation unit, an optimal maintenance plan search unit and a maintenance plan database unit; The maintenance plan template generation unit is used to analyze the maintenance plans in historical operation and maintenance, extract the main steps and construct the maintenance plan template; The optimal maintenance solution search unit is used to calculate the effectiveness of different maintenance solutions for the same fault cause and determine the optimal maintenance solution; The maintenance solution database unit is used to generate a maintenance solution by inputting the main steps of the optimal maintenance solution into a maintenance solution template, and to construct an automatic maintenance solution database.

[0014] The fault judgment module includes a fault threshold calculation unit, a fault judgment unit and a fault cause finding unit; The fault threshold calculation unit is used to calculate the fault threshold corresponding to each fault phenomenon using the data value of each fault phenomenon; The fault judgment unit is used to judge the collected real-time fault phenomenon data using the fault threshold of each fault phenomenon to judge whether the equipment fails; The fault cause search unit is used to determine all fault cause data values ​​after determining that a device has failed, and to use the changed fault cause data as the real-time fault cause.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1. After the present invention uses fault data from historical operation and maintenance to extract fault phenomena and classify faults, when a device fails, the system can quickly narrow the scope of investigation according to the fault category, greatly shortening the fault location time; according to the fault classification, a more targeted maintenance plan can be formulated, and different maintenance methods can be adopted for different fault phenomena, which can greatly increase the efficiency of equipment maintenance.

[0016] 2. The present invention extracts the main steps of the maintenance plan in the historical operation and maintenance, and after generating the maintenance plan template, different maintenance plans can be flexibly formulated according to different causes of faults. Targeted maintenance plans can be adjusted for different types of equipment faults, and cumbersome maintenance steps can be simplified, which greatly increases the efficiency of fault maintenance and greatly increases the fit between the maintenance plan and the fault.

[0017] 3. The present invention selects the best maintenance plan to generate an automatic maintenance plan database. When it is determined that a device fails, a corresponding maintenance plan can be automatically generated, allowing staff to maintain the device failure according to the displayed maintenance plan. The equipment can be maintained without the need for technical personnel, reducing maintenance costs caused by the specialization of operation and maintenance. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 This is a module distribution diagram of an intelligent operation and maintenance system based on big data of the present invention; Figure 2 The figure is a schematic diagram of the steps of an intelligent operation and maintenance method based on big data of the present invention. DETAILED DESCRIPTION

[0019] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0020] Example: Figure 1-Figure 2As shown, the present invention provides a technical solution. An intelligent operation and maintenance method based on big data, the method comprising the following steps: S100, collecting records of equipment failures during historical operation and maintenance by technicians, collecting failure data in the records, analyzing the failure data to obtain failure phenomena in each record, and classifying equipment failures according to the failure phenomena; The specific steps to classify equipment failures according to the failure phenomenon are: S101. Collect the records of equipment failures during the operation and maintenance of the equipment by the technicians in the past, collect and record each type of failure data of the equipment during the operation process, and assume that each type of failure data that changes with the operation of the equipment is {S1, S2, S3, ..., S n}, S1, S2, S3, ..., S n Indicates collecting the 1st, 2nd, 3rd, ..., nth data values ​​of each type of fault data that changes as the equipment works, where n is a positive integer; using the b data values ​​of each type of fault data collected to draw a curve graph, the abscissa is the working time of the equipment, and the ordinate is the data value of each type of fault data; S102, analyzing the drawn curve graph, extracting the peak value and the low estimate value in the curve graph, judging the data value of each point in the curve graph, and x-1 x >S x+1 When S x is the peak value in the curve. x-1 >S x x+1 When S x is the underestimation in the curve graph; let the peak value in the curve graph be {f1, f2, f3, ..., f k}, f1, f2, f3, ..., f k represents the 1st, 2nd, 3rd, ..., kth peaks in the curve graph obtained by judgment, where k is a positive integer; suppose the low estimate in the curve graph obtained by judgment is {d1, d2, d3, ..., d p}, d1, d2, d3, ..., d p It represents the 1st, 2nd, 3rd, ..., pth underestimated values ​​in the curve graph obtained by judgment, where p is a positive integer; and the average data value of the fault data in the curve graph is calculated as Sp, and the absolute difference between each peak value, underestimated value and average data value is calculated, and the formula is: ; In the formula, Sc i Represents the absolute difference between the i-th peak and the average data value in the curve graph, f i ​​Indicates the i-th peak in the curve graph; Sc j Represents the absolute difference between the jth underestimation and the average value in the curve graph, d j Indicates the jth underestimation in the curve graph; after calculation, the absolute difference between k peak values ​​and the average value data value, and the absolute difference between p underestimation values ​​and the average value data are obtained; the maximum value is selected from the k+p absolute differences as the change feature of the corresponding fault data; the change features of all fault data in the record are calculated in turn to obtain n types of change features of fault data, and the maximum value is selected as the fault phenomenon of the record of the historical operation and maintenance equipment failure; S103, all the records of equipment failures collected in the historical operation and maintenance are calculated to obtain the fault phenomenon of each equipment failure record, and all the records with the same fault phenomenon are classified into the same fault type, and finally the equipment failure is classified into {L1, L2, L3, ..., L m}, L1, L2, L3, ..., L m Indicates the 1st, 2nd, 3rd, ...mth fault category of the equipment, where m is a positive integer. After extracting the fault phenomenon and classifying the fault using the fault data in the historical operation and maintenance, when the equipment fails, the system can quickly narrow the scope of investigation according to the fault category, greatly shortening the fault location time; according to the fault classification, a more targeted maintenance plan can be formulated, and different maintenance methods can be adopted for different fault phenomena, which can greatly increase the efficiency of equipment maintenance; S200, after classifying the equipment failures, collect the equipment operation data when each failure occurs in the historical operation and maintenance, calculate the correlation between each failure phenomenon and all equipment operation data, and find all failure causes of each failure phenomenon; The specific steps to find all the causes of each fault phenomenon are as follows: S201, after classifying the equipment failures, collect the equipment operation data when each failure phenomenon occurs in the historical operation and maintenance as {Y1, Y2, Y3, ..., Y h}, Y1, Y2, Y3, ..., Y h Indicates the 1st, 2nd, 3rd, ..., hth equipment operation data collected in the historical operation and maintenance when each fault phenomenon occurs, where h is a positive integer; extracts the fault phenomenon data value when each fault occurs, and draws a scatter plot with the fault phenomenon data value as the horizontal axis and the collected equipment operation data as the vertical axis; draws a scatter plot of h kinds of equipment operation data and fault phenomenon data values ​​in each fault; uses the linear regression algorithm to fit the h kinds of scatter plots, and obtains the linear function in the scatter plot as , in which Sx represents the fault phenomenon of each fault; a represents the slope of the linear function, and b represents the intercept of the linear function; S202, for each of the h scatter plots, the corresponding linear function is calculated, and the slope of each linear function is extracted as the correlation between the equipment operation data and the fault phenomenon in the scatter plot. The correlation between all the equipment operation data and the fault phenomenon is assumed to be {a1, a2, a3, ..., a h}, a1, a2, a3, ..., a h Indicates the correlation between the 1st, 2nd, 3rd, ..., hth equipment operation data and fault phenomena, calculates the average value of the h kinds of correlation as ap, and makes a judgment on each correlation. , determine that the equipment operation data corresponding to correlation a is the cause of the fault. When the equipment operation data corresponding to the correlation a is judged not to be the cause of the fault; after judgment, the cause of each fault phenomenon is {ay1, ay2, ay3, ..., ay t}, ay1, ay2, ay3,..., ay t It indicates that the 1st, 2nd, 3rd, ..., tth fault causes of the fault phenomenon are judged, and t is a positive integer; the fault causes of the fault phenomenon of each fault are judged to obtain all the fault causes.

[0021] Find the cause of each fault phenomenon of the equipment. After clarifying the different causes corresponding to each fault, maintenance personnel can take accurate maintenance measures. The analysis results of the fault causes can provide valuable feedback to the equipment design department to optimize the equipment according to the fault causes. S300, collecting maintenance plans for equipment failures by technicians in historical operation and maintenance, extracting all maintenance data in the maintenance plan, calculating the importance of each maintenance data, and obtaining the main steps of the maintenance plan; and constructing a maintenance plan template according to the main steps; The specific steps to build a maintenance plan template according to the main steps are: S301, collecting maintenance plans for equipment failures by technicians in historical operation and maintenance, extracting all maintenance data in the maintenance plans, comparing the collected maintenance data in the maintenance plans for equipment failures by technicians in historical operation and maintenance, and extracting the maintenance data existing in all maintenance plans as {W1, W2, W3, ..., W g}, W1, W2, W3, ..., W g Indicates the 1st, 2nd, 3rd, ..., gth types of maintenance data that exist in all maintenance plans, where g is a positive integer; extract the equipment performance before and after the maintenance data that exist in all maintenance plans are executed in historical operation and maintenance, and calculate the importance of each maintenance data that exists. The formula is: ; In the formula, MS u represents the importance of the maintenance data that exists in the u-th type, Pu represents the equipment performance after the u-th type of maintenance data is executed, P u ’ Equipment performance before the execution of maintenance data of the uth type; X u Indicates the amount of information of the maintenance data that exists in the u-th type; S302. Obtain the importance of each type of maintenance data through calculation, and calculate the average value of all existing maintenance data, select maintenance data with importance greater than the average importance as the main steps of the maintenance plan, and construct a maintenance plan template using all the main steps obtained through judgment.

[0022] By extracting the main steps of the maintenance plan in the historical operation and maintenance and generating the maintenance plan template, different maintenance plans can be flexibly formulated for different fault causes. Targeted maintenance plans can be adjusted for different types of equipment faults, and cumbersome maintenance steps can be simplified, which greatly increases the efficiency of fault maintenance and greatly increases the fit between the maintenance plan and the fault. S400, for all the fault causes of each fault phenomenon, the maintenance plans of technicians in historical operation and maintenance are collected, the effectiveness of each maintenance plan is calculated, the best maintenance plan with the greatest effectiveness is extracted, the main steps in the best maintenance plan are extracted and input into the plan template, the best maintenance plan for different fault causes under each fault phenomenon is obtained, and an automatic maintenance plan database is generated; The specific steps to generate the automatic maintenance program database are: S401. For all the fault causes of each fault phenomenon, the maintenance plans of the technicians in the historical operation and maintenance are collected. The execution time of all the maintenance plans of the technicians in the historical operation and maintenance and the number of faults caused by the same fault cause after execution are collected. The effectiveness of different maintenance plans for the same fault cause is calculated. The formula is: In the formula, Ef represents the effectiveness of different maintenance plans for the same fault cause, T represents the execution time of different maintenance plans by technicians in the collected historical operation and maintenance, and C represents the number of faults caused by the same fault cause after different maintenance plans are executed by technicians in the collected historical operation and maintenance; the maintenance direction with the minimum effectiveness is selected as the best maintenance plan for the corresponding fault cause under the corresponding type of fault phenomenon; S402. After determining the best maintenance plan for each cause of failure, extract the maintenance data contained in the main steps of the best maintenance plan, input the extracted maintenance data of the main steps into the maintenance plan template, generate a maintenance plan for each cause of failure, and unify the maintenance plans for all causes of failure to generate an automatic maintenance plan database.

[0023] Select the best maintenance plan to generate an automatic maintenance plan database. When it is determined that the equipment fails, the corresponding maintenance plan can be automatically generated, so that the staff can maintain the equipment failure according to the displayed maintenance plan. The equipment can be maintained without the need for technical personnel, which reduces the maintenance cost caused by the specialization of operation and maintenance, greatly reduces the maintenance time of equipment failure, and avoids irreversible damage to the equipment; S500, calculate the fault threshold value according to the corresponding data value of each fault phenomenon during historical operation and maintenance, collect the data values ​​of all fault phenomena when the equipment is working in real time, and judge whether the equipment has a fault; collect all real-time fault cause data to judge the real-time fault cause, and select the maintenance plan in the automatic maintenance plan database for output; The specific steps for selecting a maintenance plan for output in the automatic maintenance plan database are as follows: S501. Collect data values ​​when each fault phenomenon occurs in historical operation and maintenance, and calculate fault thresholds of different fault phenomena. When the maximum absolute difference when judging the fault phenomenon in S100 is the peak value, the formula is: ; When the maximum absolute difference in judging the fault phenomenon is an underestimate, the formula is: ; In the formula, Gy represents the calculated fault threshold of different fault phenomena, Sx_p represents the average data value of different fault phenomena, and Sx_st represents the standard deviation of the data value of different fault phenomena; S502, when the device is working, collect the real-time data value of each fault phenomenon as Sxs, use the corresponding fault threshold to judge the real-time fault phenomenon, when the maximum absolute difference value when judging the fault phenomenon is the peak value, , determine that the equipment has a corresponding fault phenomenon; When the maximum absolute difference in judging the fault phenomenon is an underestimate, , determine that the equipment has a corresponding fault phenomenon; When it is determined that the equipment has failed, the data values ​​of all the fault causes under the fault phenomenon are collected. When the data value of the fault cause changes, the real-time fault cause of the equipment fault phenomenon that occurs in real time is determined; the maintenance plan is selected and output in the automatic maintenance plan database according to the real-time fault cause.

[0024] S600, collect all the operation and maintenance knowledge of the equipment, build a visual automatic operation and maintenance platform, combine the operation and maintenance knowledge to carry out fault warning in the automatic operation and maintenance platform, and display the fault phenomenon, fault cause and maintenance plan of the fault warning. The staff will perform fault maintenance on the equipment according to the automatic operation and maintenance platform and learn operation and maintenance knowledge.

[0025] The specific steps to build a visual automatic operation and maintenance platform are: S601. Collect all operation and maintenance knowledge of the equipment, use Django technology to build a visual automatic operation and maintenance platform, input and store all operation and maintenance knowledge for users to query and learn; when it is determined that the equipment fails, the failure phenomenon, cause of the failure and recommended maintenance plan will be displayed in the automatic operation and maintenance platform, and the failure will be warned to remind users to perform maintenance.

[0026] Through the automatic operation and maintenance platform, all staff can monitor and maintain the equipment operation, and they can also learn professional knowledge on the platform, which increases their professional capabilities in daily fault maintenance and reduces the cost investment in equipment maintenance.

[0027] An intelligent operation and maintenance system based on big data, the intelligent operation and maintenance system includes a data collection module, a fault classification module, a fault cause finding module, a maintenance plan generation module, a fault judgment module and a visualization module; The data collection module is used to collect fault data, equipment operation data and maintenance plans when equipment fails in historical operation and maintenance; The fault classification module is used to extract fault phenomena based on the fault data of equipment failures in historical operation and maintenance, and classify equipment failures according to the fault phenomena; The fault cause finding module is used to analyze the equipment operation data during historical operation and maintenance, calculate the correlation between each equipment operation data and the fault phenomenon, and determine all the fault causes under each fault phenomenon; The maintenance plan generation module is used to analyze the maintenance plans in historical operation and maintenance, extract the main steps to build a maintenance plan template, and find the best maintenance plan for each fault cause, extract the main steps and input them into the maintenance plan template to generate a maintenance plan; The fault judgment module is used to calculate the fault thresholds of different fault phenomena according to the data values ​​of the fault phenomena, judge whether the equipment has a fault based on the real-time fault phenomenon data values, and find the real-time fault cause; The visualization module is used to build a visualized automatic operation and maintenance platform, perform fault warning in the automatic operation and maintenance platform in combination with operation and maintenance knowledge, and display the fault phenomenon, fault cause and maintenance plan of the fault warning.

[0028] The maintenance plan generation module includes a maintenance plan template generation unit, an optimal maintenance plan search unit and a maintenance plan database unit; The maintenance plan template generation unit is used to analyze the maintenance plans in historical operation and maintenance, extract the main steps and construct the maintenance plan template; The optimal maintenance solution search unit is used to calculate the effectiveness of different maintenance solutions for the same fault cause and determine the optimal maintenance solution; The maintenance solution database unit is used to generate a maintenance solution by inputting the main steps of the optimal maintenance solution into a maintenance solution template, and to construct an automatic maintenance solution database.

[0029] The fault judgment module includes a fault threshold calculation unit, a fault judgment unit and a fault cause finding unit; The fault threshold calculation unit is used to calculate the fault threshold corresponding to each fault phenomenon using the data value of each fault phenomenon; The fault judgment unit is used to judge the collected real-time fault phenomenon data using the fault threshold of each fault phenomenon to judge whether the equipment fails; The fault cause search unit is used to determine all fault cause data values ​​after determining that a device has failed, and to use the changed fault cause data as the real-time fault cause.

[0030] Example: Now monitor the equipment failure in a certain manufacturing plant, and extract the failure data including equipment weight, equipment color, and equipment vibration from a historical operation and maintenance record; Draw a curve chart of the above four fault data, find the peaks and valleys of the three fault data, and calculate the absolute differences from the average values, which are 10, 12, and 30 respectively; select equipment vibration as the fault phenomenon recorded in this fault, and the absolute difference of equipment vibration is the absolute difference between the peak value and the average value; Through historical operation and maintenance, we found that there are two different types of equipment failures, the failure phenomena are severe vibration of the equipment and intermittent working frequency of the equipment; For example, the collected equipment operation data includes unstable placement, voltage, hardware damage, temperature, etc.; scatter plots of each equipment operation data and equipment severe vibration are generated respectively, and the linear function is generated using the phenomenon regression curve: , , , The final calculated slope average is 1.1, and it is determined that the unstable placement and hardware damage are the causes of the equipment's severe vibration. Similarly, the cause of the equipment's intermittent operating frequency is searched; The reasons for the equipment's severe vibration are unstable placement and hardware damage. The reasons for the equipment's intermittent operating frequency are excessive temperature and unstable voltage. According to historical operation and maintenance, the best maintenance plans for the two failure causes of severe vibration of the equipment were calculated and determined to be Plan 1 and Plan 2 respectively; According to the data value of the fault phenomenon in the historical operation and maintenance records, the fault threshold of severe vibration of the equipment is calculated to be 55; the vibration of the real-time acquisition equipment is 60. Since the severe vibration fault phenomenon is the absolute difference between the peak value and the average value when it is selected, 60>55, and it is judged that the equipment has a severe vibration fault phenomenon; Two fault causes were extracted and found to be stable. The changing data was hardware quality. Maintenance Plan 2 was selected for display. Ordinary staff maintained the equipment according to the displayed Plan 2.

[0031] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. An intelligent operation and maintenance method based on big data, characterized by: The method comprises the following steps: S100, collecting records of equipment failures during historical operation and maintenance by technicians, collecting failure data in the records, analyzing the failure data to obtain failure phenomena in each record, and classifying equipment failures according to the failure phenomena; S200, after classifying the equipment failures, collect the equipment operation data when each failure occurs in the historical operation and maintenance, calculate the correlation between each failure phenomenon and all equipment operation data, and find all failure causes of each failure phenomenon; S300, collecting maintenance plans for equipment failures by technicians in historical operation and maintenance, extracting all maintenance data in the maintenance plan, calculating the importance of each maintenance data, and obtaining the main steps of the maintenance plan; and constructing a maintenance plan template according to the main steps; S400, for all the fault causes of each fault phenomenon, the maintenance plans of technicians in historical operation and maintenance are collected, the effectiveness of each maintenance plan is calculated, the best maintenance plan with the greatest effectiveness is extracted, the main steps in the best maintenance plan are extracted and input into the plan template, the best maintenance plan for different fault causes under each fault phenomenon is obtained, and an automatic maintenance plan database is generated; S500, calculate the fault threshold value according to the corresponding data value of each fault phenomenon during historical operation and maintenance, collect the data values ​​of all fault phenomena when the equipment is working in real time, and judge whether the equipment has a fault; collect all real-time fault cause data to judge the real-time fault cause, and select the maintenance plan in the automatic maintenance plan database for output; S600, collect all the operation and maintenance knowledge of the equipment, build a visual automatic operation and maintenance platform, combine the operation and maintenance knowledge to carry out fault warning in the automatic operation and maintenance platform, and display the fault phenomenon, fault cause and maintenance plan of the fault warning. The staff will perform fault maintenance on the equipment according to the automatic operation and maintenance platform and learn operation and maintenance knowledge.

2. According to claim 1, the intelligent operation and maintenance method based on big data is characterized in that: The specific steps of classifying the equipment faults according to the fault phenomena in S100 are: S101. Collect the records of equipment failures during the operation and maintenance of the equipment by the technicians in the past, collect and record each type of failure data of the equipment during the operation process, and assume that each type of failure data that changes with the operation of the equipment is {S1, S2, S3, ..., S n }, S1, S2, S3, ..., S n Indicates collecting the 1st, 2nd, 3rd, ..., nth data values ​​of each type of fault data that changes as the equipment works, where n is a positive integer; using the b data values ​​of each type of fault data collected to draw a curve graph, the abscissa is the working time of the equipment, and the ordinate is the data value of each type of fault data; S102, analyzing the drawn curve graph, extracting the peak value and the low estimate value in the curve graph, judging the data value of each point in the curve graph, and x-1 x >S x+1 When S x is the peak value in the curve. x-1 >S x x+1 When S x is the underestimation in the curve graph; let the peak value in the curve graph be {f1, f2, f3, ..., f k }, f1, f2, f3, ..., f k represents the 1st, 2nd, 3rd, ..., kth peaks in the curve graph obtained by judgment, where k is a positive integer; suppose the low estimate in the curve graph obtained by judgment is {d1, d2, d3, ..., d p }, d1, d2, d3, ..., d p It represents the 1st, 2nd, 3rd, ..., pth underestimated values ​​in the curve graph obtained by judgment, where p is a positive integer; and the average data value of the fault data in the curve graph is calculated as Sp, and the absolute difference between each peak value, underestimated value and average data value is calculated, and the formula is:​​ ； In the formula, Sc i Represents the absolute difference between the i-th peak and the average data value in the curve graph, f i Indicates the i-th peak in the curve graph; Sc j Represents the absolute difference between the jth underestimation and the average value in the curve graph, d j Indicates the jth underestimation in the curve graph; after calculation, the absolute difference between k peak values ​​and the average value data value, and the absolute difference between p underestimation values ​​and the average value data are obtained; the maximum value is selected from the k+p absolute differences as the change feature of the corresponding fault data; the change features of all fault data in the record are calculated in turn to obtain n types of change features of fault data, and the maximum value is selected as the fault phenomenon of the record of the historical operation and maintenance equipment failure; S103, all the records of equipment failures collected in the historical operation and maintenance are calculated to obtain the fault phenomenon of each equipment failure record, and all the records with the same fault phenomenon are classified into the same fault type, and finally the equipment failure is classified into {L1, L2, L3, ..., L m }, L1, L2, L3, ..., L m Indicates the 1st, 2nd, 3rd, ...mth fault category of the equipment, where m is a positive integer.

3. The intelligent operation and maintenance method based on big data according to claim 2 is characterized by: The specific steps of searching for all fault causes of each fault phenomenon in S200 are: S201, after classifying the equipment failures, collect the equipment operation data when each failure phenomenon occurs in the historical operation and maintenance as {Y1, Y2, Y3, ..., Y h }, Y1, Y2, Y3, ..., Y h Indicates the 1st, 2nd, 3rd, ..., hth equipment operation data collected in the historical operation and maintenance when each fault phenomenon occurs, where h is a positive integer; extracts the fault phenomenon data value when each fault occurs, and draws a scatter plot with the fault phenomenon data value as the horizontal axis and the collected equipment operation data as the vertical axis; draws a scatter plot of h kinds of equipment operation data and fault phenomenon data values ​​in each fault; uses the linear regression algorithm to fit the h kinds of scatter plots, and obtains the linear function in the scatter plot as , in which Sx represents the fault phenomenon of each fault; a represents the slope of the linear function, and b represents the intercept of the linear function; S202, for each of the h scatter plots, the corresponding linear function is calculated, and the slope of each linear function is extracted as the correlation between the equipment operation data and the fault phenomenon in the scatter plot. The correlation between all the equipment operation data and the fault phenomenon is assumed to be {a1, a2, a3, ..., a h }, a1, a2, a3, ..., a h Indicates the correlation between the 1st, 2nd, 3rd, ..., hth equipment operation data and fault phenomena, calculates the average value of the h kinds of correlation as ap, and makes a judgment on each correlation. , determine that the equipment operation data corresponding to correlation a is the cause of the fault. When the equipment operation data corresponding to the correlation a is judged not to be the cause of the fault; after judgment, the cause of each fault phenomenon is {ay1, ay2, ay3, ..., ay t }, ay1, ay2, ay3,..., ay t It indicates that the 1st, 2nd, 3rd, ..., tth fault causes of the fault phenomenon are judged, and t is a positive integer; the fault causes of the fault phenomenon of each fault are judged to obtain all the fault causes.

4. The intelligent operation and maintenance method based on big data according to claim 1, characterized in that: The specific steps of constructing the maintenance plan template according to the main steps in S300 are: S301, collecting maintenance plans for equipment failures by technicians in historical operation and maintenance, extracting all maintenance data in the maintenance plans, comparing the collected maintenance data in the maintenance plans for equipment failures by technicians in historical operation and maintenance, and extracting the maintenance data existing in all maintenance plans as {W1, W2, W3, ..., W g }, W1, W2, W3, ..., W g Indicates the 1st, 2nd, 3rd, ..., gth types of maintenance data that exist in all maintenance plans, where g is a positive integer; extract the equipment performance before and after the maintenance data that exist in all maintenance plans are executed in historical operation and maintenance, and calculate the importance of each maintenance data that exists. The formula is: ； In the formula, MS u represents the importance of the maintenance data that exists in the u-th type, P u represents the equipment performance after the u-th type of maintenance data is executed, P u ’ Equipment performance before the execution of maintenance data of the uth type; X u Indicates the amount of information of the maintenance data that exists in the u-th type; S302. Obtain the importance of each type of maintenance data through calculation, and calculate the average value of all existing maintenance data, select maintenance data with importance greater than the average importance as the main steps of the maintenance plan, and construct a maintenance plan template using all the main steps obtained through judgment.

5. The intelligent operation and maintenance method based on big data according to claim 4 is characterized in that: The specific steps of generating the automatic maintenance solution database in S400 are: S401. For all the fault causes of each fault phenomenon, the maintenance plans of the technicians in the historical operation and maintenance are collected. The execution time of all the maintenance plans of the technicians in the historical operation and maintenance and the number of faults caused by the same fault cause after execution are collected. The effectiveness of different maintenance plans for the same fault cause is calculated. The formula is: In the formula, Ef represents the effectiveness of different maintenance plans for the same fault cause, T represents the execution time of different maintenance plans by technicians in the collected historical operation and maintenance, and C represents the number of faults caused by the same fault cause after different maintenance plans are executed by technicians in the collected historical operation and maintenance; the maintenance direction with the minimum effectiveness is selected as the best maintenance plan for the corresponding fault cause under the corresponding type of fault phenomenon; S402. After determining the best maintenance plan for each cause of failure, extract the maintenance data contained in the main steps of the best maintenance plan, input the extracted maintenance data of the main steps into the maintenance plan template, generate a maintenance plan for each cause of failure, and unify the maintenance plans for all causes of failure to generate an automatic maintenance plan database.

6. The intelligent operation and maintenance method based on big data according to claim 5 is characterized by: The specific steps of selecting a maintenance plan from the automatic maintenance plan database for output in S500 are: S501. Collect data values ​​when each fault phenomenon occurs in historical operation and maintenance, and calculate fault thresholds of different fault phenomena. When the maximum absolute difference when judging the fault phenomenon in S100 is the peak value, the formula is: ; When the maximum absolute difference in judging the fault phenomenon is an underestimate, the formula is: ; In the formula, Gy represents the calculated fault threshold of different fault phenomena, Sx_p represents the average data value of different fault phenomena, and Sx_st represents the standard deviation of the data value of different fault phenomena; S502, when the device is working, collect the real-time data value of each fault phenomenon as Sxs, use the corresponding fault threshold to judge the real-time fault phenomenon, when the maximum absolute difference value when judging the fault phenomenon is the peak value, , determine that the equipment has a corresponding fault phenomenon; When the maximum absolute difference in judging the fault phenomenon is an underestimate, , determine that the equipment has a corresponding fault phenomenon; When it is determined that the equipment has failed, the data values ​​of all the fault causes under the fault phenomenon are collected. When the data value of the fault cause changes, the real-time fault cause of the equipment fault phenomenon that occurs in real time is determined; the maintenance plan is selected and output in the automatic maintenance plan database according to the real-time fault cause.

7. The intelligent operation and maintenance method based on big data according to claim 6 is characterized by: The specific steps of building a visualized automatic operation and maintenance platform in S600 are: S601. Collect all operation and maintenance knowledge of the equipment, use Django technology to build a visual automatic operation and maintenance platform, input and store all operation and maintenance knowledge for users to query and learn; When it is determined that a device failure has occurred, the failure phenomenon, cause of the failure and recommended maintenance plan will be displayed on the automatic operation and maintenance platform, and an early warning will be issued for the failure to remind the user to perform maintenance.

8. An intelligent operation and maintenance system based on big data, characterized by: The intelligent operation and maintenance system includes a data collection module, a fault classification module, a fault cause finding module, a maintenance plan generation module, a fault judgment module and a visualization module; The data collection module is used to collect fault data, equipment operation data and maintenance plans when equipment fails in historical operation and maintenance; The fault classification module is used to extract fault phenomena based on the fault data of equipment failures in historical operation and maintenance, and classify equipment failures according to the fault phenomena; The fault cause finding module is used to analyze the equipment operation data during historical operation and maintenance, calculate the correlation between each equipment operation data and the fault phenomenon, and determine all the fault causes under each fault phenomenon; The maintenance plan generation module is used to analyze the maintenance plans in historical operation and maintenance, extract the main steps to build a maintenance plan template, and find the best maintenance plan for each fault cause, extract the main steps and input them into the maintenance plan template to generate a maintenance plan; The fault judgment module is used to calculate the fault thresholds of different fault phenomena according to the data values ​​of the fault phenomena, judge whether the equipment has a fault based on the real-time fault phenomenon data values, and find the real-time fault cause; The visualization module is used to build a visualized automatic operation and maintenance platform, perform fault warning in the automatic operation and maintenance platform in combination with operation and maintenance knowledge, and display the fault phenomenon, fault cause and maintenance plan of the fault warning.

9. The intelligent operation and maintenance system based on big data according to claim 8, characterized in that: The maintenance plan generation module includes a maintenance plan template generation unit, an optimal maintenance plan search unit and a maintenance plan database unit; The maintenance plan template generation unit is used to analyze the maintenance plans in historical operation and maintenance, extract the main steps and construct the maintenance plan template; The optimal maintenance solution search unit is used to calculate the effectiveness of different maintenance solutions for the same fault cause and determine the optimal maintenance solution; The maintenance solution database unit is used to generate a maintenance solution by inputting the main steps of the optimal maintenance solution into a maintenance solution template, and to construct an automatic maintenance solution database.

10. The intelligent operation and maintenance system based on big data according to claim 8, characterized in that: The fault judgment module includes a fault threshold calculation unit, a fault judgment unit and a fault cause search unit; The fault threshold calculation unit is used to calculate the fault threshold corresponding to each fault phenomenon using the data value of each fault phenomenon; The fault judgment unit is used to judge the collected real-time fault phenomenon data using the fault threshold of each fault phenomenon to judge whether the equipment fails; The fault cause search unit is used to determine all fault cause data values ​​after determining that a device has failed, and to use the changed fault cause data as the real-time fault cause.

Images