Operation and maintenance method and system for organic light-emitting device

Abstract

The invention relates to the technical field of organic light-emitting devices, and discloses an operation and maintenance method and system for an organic light-emitting device, and the method comprises the steps: presetting a plurality of monitoring point positions of the organic light-emitting device, extracting historical fault features, and determining the feature monitoring data and monitoring time nodes of the corresponding monitoring point positions; acquiring and analyzing real-time feature monitoring data according to the monitoring time node, and generating a risk coefficient according to an analysis result; judging whether operation and maintenance are needed or not according to the risk coefficient, if so, screening risk factors corresponding to the monitoring point positions, performing fault feature prediction, and configuring a first operation and maintenance strategy according to a prediction result; simulation operation and maintenance are carried out on the corresponding monitoring point positions according to the first operation and maintenance strategy, performance testing is carried out on the monitoring point positions after simulation operation and maintenance, a simulation application coefficient is generated according to a testing result, whether the first operation and maintenance strategy is corrected or not is judged, and the state judgment precision and the operation and maintenance efficiency of the organic light-emitting device are improved.

Description

Technical Field

[0001] The present application relates to the technical field of organic electroluminescent devices, and in particular, to an operation and maintenance method and system for organic electroluminescent devices. Background Art

[0002] The operation and maintenance methods of organic electroluminescent devices mostly rely on data monitoring and manual maintenance. However, there are a large number of operation data of organic electroluminescent devices and the accuracy rate is low, resulting in deviations in the data monitoring results, unable to detect abnormal states in time or issue incorrect operation and maintenance instructions, reducing the operation and maintenance efficiency. Therefore, there is an urgent need for an operation and maintenance method and system for organic electroluminescent devices to improve the accuracy of state judgment and the application effect of operation and maintenance strategies. Summary of the Invention

[0003] To solve the above technical problems, the present application provides an operation and maintenance method and system for organic electroluminescent devices. By setting multiple monitoring points, determining the characteristic monitoring data and monitoring time nodes of each monitoring point, obtaining real-time characteristic monitoring data according to the monitoring time nodes and generating the risk coefficient and fault characteristic prediction result of the corresponding monitoring point, configuring the corresponding first operation and maintenance strategy, and performing simulation operation and maintenance on the first operation and maintenance strategy to obtain the simulation application coefficient, the accuracy of state judgment and the operation and maintenance efficiency of organic electroluminescent devices are improved.

[0004] In some embodiments of the present application, an operation and maintenance method for organic electroluminescent devices is provided, including: Predetermine multiple monitoring points of the organic electroluminescent device, extract the historical fault characteristics of each monitoring point, and determine the characteristic monitoring data and monitoring time nodes of the corresponding monitoring point according to the historical fault characteristics; Obtain the real-time characteristic monitoring data of each monitoring point according to the monitoring time node, analyze the real-time characteristic monitoring data, and generate the risk coefficient of each monitoring point according to the analysis result; Judge whether operation and maintenance are required according to the risk coefficient. If so, screen out the risk factors of the corresponding monitoring point and perform fault characteristic prediction, and configure the first operation and maintenance strategy according to the prediction result; Perform simulation operation and maintenance on the corresponding monitoring point according to the first operation and maintenance strategy, and perform performance testing on the monitoring point after simulation operation and maintenance. Generate the simulation application coefficient according to the test result and judge whether to correct the first operation and maintenance strategy. If so, obtain the second operation and maintenance strategy and issue the operation and maintenance instruction.

[0005] In some embodiments of the present application, determining the characteristic monitoring data and monitoring time nodes of the corresponding monitoring point according to the historical fault characteristics includes: Obtain a number of historical operation and maintenance logs for each monitoring point, and extract historical fault features from each historical operation and maintenance log. The historical fault features include historical fault types, historical fault data, historical fault time periods, historical fault coefficients, and historical fault impact coefficients; Compare the historical fault data involved in the same historical fault type for each monitoring point to obtain the historical fault frequency of each historical fault data; Calculate the historical data difference between the same historical fault data and the corresponding standard data range, and generate the historical impact degree of the corresponding historical fault data on the corresponding historical fault coefficient based on multiple historical data differences and the corresponding historical fault coefficients; Set the historical fault data with a historical fault frequency greater than the preset fault frequency threshold and a historical impact degree greater than the preset impact degree threshold for each historical fault type as the characteristic monitoring data for the corresponding monitoring point, and generate the first attention evaluation value for the corresponding characteristic monitoring data based on the historical fault frequency and the historical impact degree; Set the weight coefficient for the corresponding characteristic monitoring data according to the historical impact degree of the characteristic monitoring data on the historical fault coefficient, and generate the second attention evaluation value for the corresponding characteristic monitoring data in combination with the historical fault impact coefficient of the corresponding historical fault type of the characteristic monitoring data; Obtain multiple historical fluctuation degrees of the same characteristic monitoring data in multiple historical fault time periods of the corresponding historical fault type, and generate the average historical fluctuation degree of the corresponding characteristic monitoring data based on the multiple historical fluctuation degrees; Generate the third attention evaluation value for the corresponding characteristic monitoring data based on the average historical fluctuation degree; Generate the comprehensive attention evaluation value for the corresponding characteristic monitoring data based on the first attention evaluation value, the second attention evaluation value, and the third attention evaluation value, and set the time interval between adjacent monitoring time nodes for the corresponding characteristic monitoring data according to the comprehensive attention evaluation value.

[0006] In some embodiments of the present application, setting the time interval between adjacent monitoring time nodes for the corresponding characteristic monitoring data according to the comprehensive attention evaluation value includes: Preset the first preset comprehensive attention evaluation value interval, the second preset comprehensive attention evaluation value interval, the third preset comprehensive attention evaluation value interval, and the fourth preset comprehensive attention evaluation value interval; When the comprehensive attention evaluation value of the characteristic monitoring data is within the first preset comprehensive attention evaluation value interval, select the fourth preset time interval as the time interval between adjacent monitoring time nodes for the corresponding characteristic monitoring data; When the comprehensive attention evaluation value of the characteristic monitoring data is within the second preset comprehensive attention evaluation value interval, select the third preset time interval as the time interval between adjacent monitoring time nodes for the corresponding characteristic monitoring data; When the comprehensive attention evaluation value of the feature monitoring data is within the third preset comprehensive attention evaluation value range, the second preset time interval is selected as the time interval between adjacent monitoring time nodes of the corresponding feature monitoring data; When the comprehensive attention evaluation value of the feature monitoring data is within the fourth preset comprehensive attention evaluation value range, the first preset time interval is selected as the time interval between adjacent monitoring time nodes of the corresponding feature monitoring data.

[0007] In some embodiments of the present application, generating a risk coefficient for each monitoring point according to the analysis result includes: Setting multiple monitoring time nodes of a plurality of feature monitoring data for each monitoring point at a preset time interval, and collecting real-time feature monitoring data for each monitoring point according to the monitoring time nodes; Comparing the real-time feature monitoring data with the corresponding standard data range to obtain a real-time feature monitoring data difference; Dividing the real-time feature monitoring data into abnormal feature monitoring data, suspected abnormal feature monitoring data, and normal feature monitoring data according to the real-time feature monitoring data difference, and dividing according to the corresponding fault types to obtain a set of real-time feature monitoring data for each fault type; Setting a data sorting value for the real-time feature monitoring data according to the data status and data weight, and sorting the real-time feature monitoring data in the set of real-time feature monitoring data according to the data sorting value; Generating a failure probability for the corresponding failure type according to the number of abnormal data, the number of suspected abnormal data, the number of normal data, the corresponding real-time feature monitoring data difference, and the weight coefficient of the corresponding real-time feature monitoring data in the set of real-time feature monitoring data for each failure type; The calculation formula of the failure probability is: ; where P is the failure probability, m1, m2, and m3 are respectively the number of abnormal data, the number of suspected abnormal data, and the number of normal data in the set of real-time feature monitoring data for the corresponding failure type, is the difference of the v1-th abnormal real-time feature monitoring data, is the weight coefficient of the difference of the v1-th abnormal real-time feature monitoring data, is the difference of the v2-th suspected abnormal real-time feature monitoring data, is the weight coefficient of the difference of the v2-th suspected abnormal real-time feature monitoring data, is the difference of the v3-th normal real-time feature monitoring data, is the weight coefficient of the difference of the v3-th normal real-time feature monitoring data, x1 is the first failure probability conversion coefficient, x2 is the second failure probability conversion coefficient, and x3 is the third failure probability conversion coefficient; Preset the fault probability threshold for each type of fault, and calculate the fault probability difference for each type of fault; Generate the risk coefficient for the corresponding monitoring point according to the fault probability differences of multiple types of faults at the same monitoring point and the corresponding weight coefficients.

[0008] In some embodiments of the present application, screen out the risk factors for the corresponding monitoring point and perform fault feature prediction, including: Preset the risk coefficient threshold; If the risk coefficient is less than the risk coefficient threshold, no operation and maintenance are required; If the risk coefficient is greater than the risk coefficient threshold, screen out the types of faults whose fault probabilities are greater than the corresponding fault probability thresholds; Set the number of data extractions for the real-time feature monitoring data set of the corresponding type of fault according to the fault probability differences of the screened types of faults and the corresponding weight coefficients; Extract the feature monitoring data in the real-time feature monitoring data set of the corresponding type of fault according to the number of data extractions and the sorting result, and set the risk factors for the corresponding monitoring point according to the data extraction results of the screened types of faults at the same monitoring point. The risk factors are several real-time feature monitoring data extracted; Compare the risk factors of the same type of fault at the current monitoring point with the feature monitoring data of the corresponding type of fault, and generate the credibility of the corresponding type of fault according to the comparison result; If the credibility is less than the preset credibility threshold, eliminate the corresponding type of fault and the risk factors involved in the corresponding type of fault; If the credibility is greater than the preset credibility threshold, retain the risk factors involved and set the corresponding type of fault as the predicted type of fault, and determine the predicted fault feature of the corresponding predicted type of fault according to the data status and data change characteristics of the risk factors involved in the current predicted type of fault; Among them, when the data status of the risk factors involved in the predicted type of fault is all abnormal, determine the predicted fault feature according to the sorting result of the risk factors involved and the real-time feature monitoring data difference; When the data status of the risk factors involved in the predicted type of fault has a suspected abnormal status or a normal status, construct a real-time data change curve for the corresponding risk factor, and obtain the data change characteristics. The data change characteristics include the change trend, change rate, and change magnitude; Curve extrapolation is performed based on the data change characteristics and the real-time data change curve to obtain the predicted data change curve of the corresponding risk factor in the preset time period. The predicted characteristic monitoring data difference, the real-time characteristic monitoring data difference of the risk factor in the abnormal state, and the sorting result are used to determine the predicted fault characteristics, where the predicted fault characteristics include predicted fault factors, predicted fault time periods, predicted fault coefficients, and predicted fault impact coefficients; Generate the prediction result of the current monitoring point, where the prediction result includes a predicted fault sequence, the predicted fault sequence includes several predicted fault types and the corresponding predicted fault characteristics, and several predicted fault types are sorted according to the credibility of the predicted fault types.

[0009] In some embodiments of the present application, configure the first operation and maintenance strategy according to the prediction result, including: Construct an operation and maintenance strategy reference library for the corresponding monitoring point. The operation and maintenance strategy reference library includes several preset fault types, each fault type includes several preset fault characteristics, and each preset fault characteristic is mapped to a corresponding preset operation and maintenance strategy; Perform similarity analysis on several preset fault types and the corresponding preset fault characteristics in the operation and maintenance strategy reference library and several predicted fault types and the corresponding predicted fault characteristics in the predicted fault sequence of the corresponding monitoring point to obtain the similarity between the preset fault characteristics and the predicted fault characteristics of the same fault type; Set the preset operation and maintenance strategy mapped by the preset fault characteristic with the largest similarity as the operation and maintenance sub-strategy corresponding to the predicted fault characteristic; Generate the operation and maintenance sub-strategies of the predicted fault characteristics of each predicted fault type in sequence; Judge whether several operation and maintenance sub-strategies need to be optimized. If so, construct the first operation and maintenance strategy for the corresponding monitoring point according to the optimized operation and maintenance sub-strategy.

[0010] In some embodiments of the present application, before performing performance testing on the monitoring point after simulation operation and maintenance, include: Preset several performance evaluation indicators, and construct a performance evaluation index tree based on the importance of each performance evaluation indicator, the correlation relationship and the degree of correlation between different performance evaluation indicators; Among them, the performance evaluation index tree includes a preset main trunk. The connection relationship and the connection distance between the corresponding performance evaluation indicator and the preset main trunk are set according to the importance of the performance evaluation indicator, and the connection relationship and the connection distance between different performance evaluation indicators are set according to the correlation relationship and the degree of correlation between different performance evaluation indicators; Set the first weight coefficient of each performance evaluation indicator according to the connection information and the position information of each performance evaluation indicator in the performance evaluation index tree; Determine the predicted failure type corresponding to the monitoring point and the predicted negative impact degree of the corresponding predicted failure characteristics on several performance evaluation indicators based on the historical operation and maintenance logs of the monitoring points, and set the second weight coefficient of each performance evaluation indicator according to the predicted negative impact degree; Generate the comprehensive weight coefficient of the corresponding performance evaluation indicator according to the first weight coefficient and the second weight coefficient, set the performance evaluation indicators with the comprehensive weight coefficient greater than the preset weight coefficient threshold as the concerned performance evaluation indicators, and sort the concerned performance evaluation indicators according to the comprehensive weight coefficient; Generate the performance test characteristics of each concerned performance evaluation indicator in turn according to the sorting result, set the standard test response characteristics corresponding to the performance test characteristics, and generate the performance test instruction according to the sorting result and the performance test characteristics of each concerned performance evaluation indicator; Among them, the performance test characteristics include test behavior, test frequency and test intensity, and the standard test response characteristics include standard test response data, standard test response frequency and standard test response intensity.

[0011] In some embodiments of the present application, generating a simulation application coefficient according to the test result includes: Obtain the location information, structure information and environmental information of the monitoring points that need operation and maintenance, and build a simulation operation model of the corresponding monitoring points in combination with several predicted failure types and corresponding predicted failure characteristics of the corresponding monitoring points; Obtain the operation and maintenance information to be set according to the first operation and maintenance strategy, obtain the operation relationship between the operation and maintenance information to be set and each characteristic monitoring data of the corresponding monitoring points, and import it into the simulation operation model of the corresponding monitoring points to obtain the simulated monitoring points after simulation operation and maintenance; Obtain the simulated characteristic monitoring data of the simulated monitoring points, calculate the difference of the simulated characteristic monitoring data of each simulated characteristic monitoring data, and generate the simulated operation evaluation value of the simulated monitoring points according to several differences of the simulated characteristic monitoring data; Generate the simulated change characteristics of the predicted failure type and the corresponding predicted failure characteristics of the corresponding monitoring points according to the simulated characteristic monitoring data, and generate the compensation coefficient of the simulated operation evaluation value according to the simulated change characteristics; Perform performance testing on the simulated monitoring points according to the performance test instructions to obtain the simulated test response characteristics corresponding to the performance test characteristics of each concerned performance evaluation indicator, and the simulated test characteristics include simulated test response data, simulated test response frequency and simulated test response intensity; Compare the simulated test response characteristics with the standard test response characteristics of the corresponding performance test characteristics to obtain the difference of the simulated test response characteristics; Generate the simulation performance evaluation value of the simulation monitoring point according to the differences in the simulation test response characteristics of several performance evaluation indicators of concern and the corresponding comprehensive weight coefficients; Generate the simulation application coefficient of the first operation and maintenance strategy for the corresponding monitoring point according to the simulation operation evaluation value and the simulation performance evaluation value; The calculation formula of the simulation application coefficient is: ; Where Y is the simulation application coefficient, a1 is the weight coefficient of the simulation operation evaluation value, y1 is the first simulation application conversion coefficient, r is the compensation coefficient, n1 is the total number of characteristic monitoring data of the current monitoring point, is the difference in the simulation characteristic monitoring data of the i-th simulation characteristic monitoring data, q1i is the weight coefficient of the i-th simulation characteristic monitoring data, a2 is the simulation performance evaluation value, y2 is the second simulation application conversion coefficient, n2 is the total number of performance evaluation indicators of concern, is the difference in the simulation test response characteristics of the s-th performance evaluation indicator of concern, and q2s is the comprehensive weight coefficient of the s-th performance evaluation indicator of concern.

[0012] In some embodiments of the present application, determining whether to correct the first operation and maintenance strategy, and if so, obtaining the second operation and maintenance strategy and issuing an operation and maintenance instruction, includes: Preset a simulation application coefficient threshold; If the simulation application coefficient is greater than the simulation application coefficient threshold, do not correct the first operation and maintenance strategy and issue an operation and maintenance instruction; If the simulation application coefficient is less than the simulation application coefficient threshold, screen out the to-be-optimized characteristic monitoring data according to the simulation operation evaluation value, screen out the to-be-optimized performance evaluation indicators according to the simulation performance evaluation indicators, and correct the corresponding operation and maintenance information in the first operation strategy according to the to-be-optimized characteristic monitoring data and the to-be-optimized performance evaluation indicators, obtain the second operation and maintenance strategy, and issue an operation and maintenance instruction.

[0013] In some embodiments of the present application, there is also an operation and maintenance system for an organic electroluminescent device: A determination module, configured to preset a plurality of monitoring points of the organic electroluminescent device, extract the historical fault characteristics of each monitoring point, and determine the characteristic monitoring data and the monitoring time node of the corresponding monitoring point according to the historical fault characteristics; An analysis module, configured to obtain the real-time characteristic monitoring data of each monitoring point according to the monitoring time node, analyze the real-time characteristic monitoring data, and generate a risk coefficient for each monitoring point according to the analysis result; A judgment module, configured to judge whether operation and maintenance is required according to the risk coefficient, and if so, screen out the risk factors of the corresponding monitoring point and perform fault characteristic prediction, and configure the first operation and maintenance strategy according to the prediction result; An operation and maintenance module is used to perform simulation operation and maintenance on corresponding monitoring points according to the first operation and maintenance strategy, perform performance testing on the monitoring points after simulation operation and maintenance, generate a simulation application coefficient based on the test results, and determine whether to modify the first operation and maintenance strategy. If so, obtain the second operation and maintenance strategy and issue an operation and maintenance instruction.

[0014] Compared with the prior art, a method and system for operation and maintenance of an organic electroluminescent device according to an embodiment of the present application have the following beneficial effects: By setting multiple monitoring points, determining the characteristic monitoring data and monitoring time nodes of each monitoring point, obtaining real-time characteristic monitoring data according to the monitoring time nodes, generating a risk coefficient and a fault characteristic prediction result for the corresponding monitoring points, configuring the corresponding first operation and maintenance strategy, and performing simulation operation and maintenance on the first operation and maintenance strategy to obtain a simulation application coefficient, the state judgment accuracy and operation and maintenance efficiency of the organic electroluminescent device are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 is a schematic flowchart of a method for operation and maintenance of an organic electroluminescent device according to an embodiment of the present application; Figure 2 is a schematic diagram of a system for operation and maintenance of an organic electroluminescent device according to an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0016] The following describes in further detail the specific embodiments of the present application with reference to the drawings and embodiments. The following embodiments are used to illustrate the present application, but are not used to limit the scope of the present application.

[0017] In the description of the present application, it should be understood that the terms "center", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as limiting the present application.

[0018] The terms "first" and "second" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of such features. In the description of the present application, unless otherwise specified, "a plurality of" means two or more.

[0019] In the description of the present application, it should be noted that unless otherwise clearly specified and limited, the terms "installation", "connection", and "linkage" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, and it can be the communication inside two components. For those of ordinary skill in the art, the specific meanings of the above terms in the present application can be understood according to specific circumstances.

[0020] As Figure 1 shown, a method for operation and maintenance of an organic electroluminescent device according to an embodiment of the present application includes: Step S101: Predetermine a plurality of monitoring points of the organic electroluminescent device, extract the historical fault characteristics of each monitoring point, and determine the characteristic monitoring data and monitoring time nodes corresponding to the monitoring points according to the historical fault characteristics; Step S102: Obtain the real-time characteristic monitoring data of each monitoring point according to the monitoring time nodes, analyze the real-time characteristic monitoring data, and generate a risk coefficient for each monitoring point according to the analysis result; Step S103: Judge whether operation and maintenance are required according to the risk coefficient. If so, screen out the risk factors of the corresponding monitoring points and perform fault characteristic prediction, and configure a first operation and maintenance strategy according to the prediction result; Step S104: Perform simulation operation and maintenance on the corresponding monitoring points according to the first operation and maintenance strategy, perform performance testing on the monitoring points after the simulation operation and maintenance, generate a simulation application coefficient according to the test result, and judge whether to correct the first operation and maintenance strategy. If so, obtain a second operation and maintenance strategy and issue an operation and maintenance instruction.

[0021] In this embodiment, several monitoring points are set according to the high-frequency operation and maintenance points in the historical operation and maintenance logs. By dividing the organic electroluminescent device into several monitoring points and evaluating the risk coefficients of each monitoring point, the accuracy of the state judgment of the organic electroluminescent device is improved, abnormal points are found in time, and an operation and maintenance strategy is configured, thereby improving the operation and maintenance efficiency.

[0022] In some embodiments of the present application, determining the characteristic monitoring data and monitoring time nodes corresponding to the monitoring points according to the historical fault characteristics includes: Obtain several historical operation and maintenance logs of each monitoring point, and extract the historical fault characteristics in each historical operation and maintenance log. The historical fault characteristics include historical fault types, historical fault data, historical fault time periods, historical fault coefficients, and historical fault impact coefficients; Compare the historical fault data involved in the same historical fault type of each monitoring point to obtain the historical fault frequency of each historical fault data; Calculate the historical data difference between the same historical fault data and the corresponding standard data interval, and generate the historical influence degree of the corresponding historical fault data on the corresponding historical fault coefficient according to multiple historical data differences and the corresponding historical fault coefficients; Set the historical fault data with the historical fault frequency of each historical fault type greater than the preset fault frequency threshold and the historical influence degree greater than the preset influence degree threshold as the characteristic monitoring data of the corresponding monitoring point, and generate the first attention evaluation value of the corresponding characteristic monitoring data according to the historical fault frequency and the historical influence degree; Set the weight coefficient of the corresponding characteristic monitoring data according to the historical influence degree of the characteristic monitoring data on the historical fault coefficient, and generate the second attention evaluation value of the corresponding characteristic monitoring data in combination with the historical fault influence coefficient of the corresponding historical fault type of the characteristic monitoring data; Obtain multiple historical fluctuation degrees of the same characteristic monitoring data in multiple historical fault time periods of the corresponding historical fault type, and generate the average value of the historical fluctuation degrees of the corresponding characteristic monitoring data according to the multiple historical fluctuation degrees; Generate the third attention evaluation value of the corresponding characteristic monitoring data according to the average value of the historical fluctuation degrees; Generate the comprehensive attention evaluation value of the corresponding characteristic monitoring data according to the first attention evaluation value, the second attention evaluation value, and the third attention evaluation value, and set the time interval between adjacent monitoring time nodes of the corresponding characteristic monitoring data according to the comprehensive attention evaluation value.

[0023] In this embodiment, the historical fault coefficient refers to the fault level of each historical fault type. When the historical fault coefficient is larger, the fault level is larger, and vice versa. The historical fault influence coefficient refers to the size of the harm caused by each historical fault type. When the historical fault influence coefficient is larger, the corresponding harm is larger, and vice versa.

[0024] In this embodiment, perform attention evaluation conversion according to the historical fault frequency difference between the historical fault frequency and the preset fault frequency threshold and the historical influence degree difference between the historical influence degree and the preset influence degree threshold, and generate the first attention evaluation value. When the historical fault frequency difference and the historical influence degree difference are larger, the corresponding first attention evaluation value is larger. The first attention evaluation value is used to evaluate the influence size of the corresponding characteristic monitoring data on the fault level of the corresponding historical fault type and the evaluation accuracy when the corresponding historical fault type appears.

[0025] In this embodiment, when the historical influence degree is larger, the weight coefficient of the corresponding characteristic monitoring data is larger. Calculate the average value of the historical fault influence coefficients of multiple historical fault influence coefficients, and combine the weight coefficient of the corresponding characteristic monitoring data to obtain the second attention evaluation value. The second attention evaluation value is used to evaluate the size of the harm caused by the corresponding characteristic monitoring data and the corresponding historical fault type.

[0026] In this embodiment, when the average value of the historical fluctuation degree is larger, the corresponding third attention evaluation value is larger, and vice versa. The third attention evaluation value is used to evaluate the degree of data change of the feature monitoring data during the historical fault period.

[0027] In this embodiment, when the first attention evaluation value, the second attention evaluation value, and the third attention evaluation value are larger, it indicates that the corresponding feature monitoring data is more important, the data changes more frequently, and it is more accurate for evaluating the fault type.

[0028] In some embodiments of the present application, setting the time interval between adjacent monitoring time nodes of the corresponding feature monitoring data according to the comprehensive attention evaluation value includes: Presetting a first preset comprehensive attention evaluation value interval, a second preset comprehensive attention evaluation value interval, a third preset comprehensive attention evaluation value interval, and a fourth preset comprehensive attention evaluation value interval; When the comprehensive attention evaluation value of the feature monitoring data is within the first preset comprehensive attention evaluation value interval, select the fourth preset time interval as the time interval between adjacent monitoring time nodes of the corresponding feature monitoring data; When the comprehensive attention evaluation value of the feature monitoring data is within the second preset comprehensive attention evaluation value interval, select the third preset time interval as the time interval between adjacent monitoring time nodes of the corresponding feature monitoring data; When the comprehensive attention evaluation value of the feature monitoring data is within the third preset comprehensive attention evaluation value interval, select the second preset time interval as the time interval between adjacent monitoring time nodes of the corresponding feature monitoring data; When the comprehensive attention evaluation value of the feature monitoring data is within the fourth preset comprehensive attention evaluation value interval, select the first preset time interval as the time interval between adjacent monitoring time nodes of the corresponding feature monitoring data.

[0029] In this embodiment, the first preset comprehensive attention evaluation value interval < the second preset comprehensive attention evaluation value interval < the third preset comprehensive attention evaluation value interval < the fourth preset comprehensive attention evaluation value interval, and the first preset time interval < the second preset time interval < the third preset time interval < the fourth preset time interval.

[0030] In this embodiment, when the preset comprehensive attention evaluation value interval where it is located is larger, it indicates that the corresponding feature monitoring data is more important, the data changes more frequently, and it is more accurate for evaluating the fault type. A short time interval should be selected to monitor the corresponding feature monitoring data more frequently to ensure that abnormal situations can be detected in time so that a rapid response and processing can be made when the data is abnormal.

[0031] In some embodiments of the present application, generating a risk coefficient for each monitoring point according to the analysis result includes: Set multiple monitoring time nodes for several characteristic monitoring data of each monitoring point at preset time intervals, and collect the real-time characteristic monitoring data of each monitoring point according to the monitoring time nodes; Compare the real-time characteristic monitoring data with the corresponding standard data interval to obtain the difference of the real-time characteristic monitoring data; Divide the real-time characteristic monitoring data into abnormal characteristic monitoring data, suspected abnormal characteristic monitoring data and normal characteristic monitoring data according to the difference of the real-time characteristic monitoring data, and divide them according to the corresponding fault types to obtain the set of real-time characteristic monitoring data of each fault type; Set the data sorting value of the real-time characteristic monitoring data according to the data status and data weight, and sort the real-time characteristic monitoring data in the set of the real-time characteristic monitoring data according to the data sorting value; Generate the fault probability of the corresponding fault type according to the number of abnormal data, the number of suspected abnormal data, the number of normal data, the corresponding difference of the real-time characteristic monitoring data and the weight coefficient of the corresponding real-time characteristic monitoring data in the set of the real-time characteristic monitoring data of each fault type; The calculation formula of the fault probability is: ; Where P is the fault probability, m1, m2, and m3 are respectively the number of abnormal data, the number of suspected abnormal data, and the number of normal data in the set of the real-time characteristic monitoring data of the corresponding fault type, is the difference of the v1-th abnormal real-time characteristic monitoring data, is the weight coefficient of the difference of the v1-th abnormal real-time characteristic monitoring data, is the difference of the v2-th suspected abnormal real-time characteristic monitoring data, is the weight coefficient of the difference of the v2-th suspected abnormal real-time characteristic monitoring data, is the difference of the v3-th normal real-time characteristic monitoring data, is the weight coefficient of the difference of the v3-th normal real-time characteristic monitoring data, x1 is the first fault probability conversion coefficient, x2 is the second fault probability conversion coefficient, and x3 is the third fault probability conversion coefficient; Preset the fault probability threshold of each fault type, and calculate the fault probability difference of each fault type; Generate the risk coefficient of the corresponding monitoring point according to the fault probability differences of multiple fault types of the same monitoring point and the corresponding weight coefficients.

[0032] In this embodiment, the feature monitoring data in the real-time feature monitoring data set is sorted in the state order of abnormal state, suspected abnormal state, and normal state. Moreover, the state coefficient corresponding to the real-time feature monitoring data is set according to the difference of the real-time feature monitoring data. The data sorting value is generated based on the state coefficient and the data weight. When the state coefficient is smaller and the data weight is larger, it indicates that the corresponding real-time feature monitoring data is more abnormal and more important, and then the data sorting value is more forward. On the contrary, it is more backward, laying a foundation for subsequent data extraction and improving the accuracy of fault feature prediction.

[0033] In this embodiment, a first data difference interval, a second data difference interval, and a third data difference interval are preset. When the difference of the real-time feature monitoring data is respectively in the first data difference interval, the second data difference interval, or the third data difference interval, it is abnormal, suspected abnormal, or normal. Among them, both the first data difference interval and the second data difference interval are less than 0, and the third data difference interval is greater than 0.

[0034] In this embodiment, the risk coefficient of the corresponding monitoring point is obtained by calculating the failure probability of each failure type, accurately evaluating whether there are failures at several monitoring points of the organic electroluminescent device, and timely discovering and performing operation and maintenance to ensure the operation and maintenance efficiency and the operation and maintenance accuracy rate of the organic electroluminescent device.

[0035] In some embodiments of the present application, the risk factors corresponding to the monitoring points are screened out and the fault feature prediction is performed, including: Presetting a risk coefficient threshold; If the risk coefficient is less than the risk coefficient threshold, no operation and maintenance is required; If the risk coefficient is greater than the risk coefficient threshold, the failure types with a failure probability greater than the corresponding failure probability threshold are screened out; According to the failure probability difference of the screened-out failure types and the corresponding weight coefficients, the number of data extractions of the real-time feature monitoring data set corresponding to the failure types is set; According to the number of data extractions and the sorting result, the feature monitoring data in the real-time feature monitoring data set corresponding to the corresponding failure type is extracted. According to the data extraction result in the screened-out failure types at the same monitoring point, the risk factor corresponding to the monitoring point is set. The risk factor is several real-time feature monitoring data extracted; The risk factors of the same failure type at the current monitoring point are compared with the feature monitoring data of the corresponding failure type, and the credibility of the corresponding failure type is generated according to the comparison result; If the credibility is less than the preset credibility threshold, the corresponding failure type and the risk factors involved in the corresponding failure type are excluded; If the confidence level is greater than the preset confidence threshold, retain the risk factors involved and set the corresponding fault type as the predicted fault type, and determine the predicted fault characteristics of the corresponding predicted fault type according to the data status and data change characteristics of the risk factors involved in the current predicted fault type; Among them, when the data status of the risk factors involved in the predicted fault type is all abnormal, determine the predicted fault characteristics according to the sorting result of the risk factors involved and the difference of real-time feature monitoring data; When the data status of the risk factors involved in the predicted fault type has a suspected abnormal status or a normal status, construct a real-time data change curve of the corresponding risk factor and obtain the data change characteristics, where the data change characteristics include the change trend, change rate, and change magnitude; Perform curve extrapolation according to the data change characteristics and the real-time data change curve to obtain the predicted data change curve of the corresponding risk factor in the preset time period, and determine the predicted characteristic monitoring data difference, the real-time characteristic monitoring data difference of the risk factor in the abnormal state, and the sorting result to determine the predicted fault characteristics. The predicted fault characteristics include predicted fault factors, predicted fault time periods, predicted fault coefficients, and predicted fault impact coefficients; Generate the prediction result of the current monitoring point. The prediction result includes a predicted fault sequence, where the predicted fault sequence includes several predicted fault types and their corresponding predicted fault characteristics, and sort the several predicted fault types according to the confidence level of the predicted fault type.

[0036] In this embodiment, when the fault probability difference of the selected fault type is larger and the corresponding weight coefficient is larger, the corresponding number of data extractions is more, and the number of data extractions is extracted according to the sorting result of the feature monitoring data in the real-time feature monitoring data set. For example, if the number of data extractions is 10, then the first 10 feature monitoring data in the real-time feature monitoring data set are extracted.

[0037] In this embodiment, when the data status of the risk factors involved in the same predicted fault type is all abnormal, set all the risk factors involved as predicted fault factors, determine the predicted fault time period according to the difference of the real-time feature monitoring data of the predicted fault factors, and generate several predicted fault coefficients in turn according to the sorting result of the predicted fault factors and the historical influence degree of the predicted fault factors on the predicted fault type, and perform weighted average processing to obtain the final predicted fault coefficient, and obtain the corresponding predicted fault impact coefficient according to the final predicted fault coefficient and the corresponding predicted fault type.

[0038] In this embodiment, the preset time period refers to a future time period. A prediction data change curve of risk factors for a suspected abnormal state or a normal state is constructed to obtain the difference in predicted characteristic monitoring data of the corresponding risk factors, so as to determine whether the corresponding risk factors are predicted fault factors. If not, the corresponding risk factors are excluded. If so, the predicted fault time period is determined according to the difference in predicted characteristic monitoring data of the predicted fault factors, and the predicted fault coefficient and the predicted fault impact coefficient are determined in combination with the sorting result and the historical impact degree.

[0039] In some embodiments of the present application, configuring a first operation and maintenance strategy according to the prediction result includes: Constructing an operation and maintenance strategy reference library for the corresponding monitoring points, where the operation and maintenance strategy reference library includes several preset fault types, each fault type includes several preset fault characteristics, and each preset fault characteristic is mapped to a corresponding preset operation and maintenance strategy; Performing a similarity analysis on the several preset fault types and the corresponding several preset fault characteristics in the operation and maintenance strategy reference library and the several predicted fault types and the corresponding predicted fault characteristics in the predicted fault sequence of the corresponding monitoring points to obtain the similarity between the preset fault characteristics and the predicted fault characteristics of the same fault type; Setting the preset operation and maintenance strategy mapped by the preset fault characteristic with the largest similarity as the operation and maintenance sub-strategy for the corresponding predicted fault characteristic; Generating operation and maintenance sub-strategies for the predicted fault characteristics of each predicted fault type in sequence; Judging whether several operation and maintenance sub-strategies need to be optimized. If so, constructing a first operation and maintenance strategy for the corresponding monitoring points according to the optimized operation and maintenance sub-strategies.

[0040] In this embodiment, the optimization includes conflict optimization and merging optimization. The conflict optimization includes policy conflict and execution conflict. The operation and maintenance sub-strategies that need to be optimized are optimized according to the sorting priority principle and constitute the first operation and maintenance strategy.

[0041] In some embodiments of the present application, before performing a performance test on the monitoring points after simulation operation and maintenance, it includes: Presetting several performance evaluation indicators, and constructing a performance evaluation index tree based on the importance of each performance evaluation indicator, the correlation relationship and the degree of correlation between different performance evaluation indicators; Among them, the performance evaluation index tree includes a preset main trunk. The connection relationship and the connection distance between the corresponding performance evaluation indicator and the preset main trunk are set according to the importance of the performance evaluation indicator, and the connection relationship and the connection distance between different performance evaluation indicators are set according to the correlation relationship and the degree of correlation between different performance evaluation indicators; Setting a first weight coefficient for each performance evaluation indicator according to the connection information and the position information of each performance evaluation indicator in the performance evaluation index tree; Determine the predicted failure type corresponding to the monitoring point and the predicted negative impact degree of the corresponding predicted failure characteristics on several performance evaluation indicators based on the historical operation and maintenance logs of the monitoring points, and set the second weight coefficient of each performance evaluation indicator according to the predicted negative impact degree; Generate the comprehensive weight coefficient of the corresponding performance evaluation indicator according to the first weight coefficient and the second weight coefficient, set the performance evaluation indicators with the comprehensive weight coefficient greater than the preset weight coefficient threshold as the concerned performance evaluation indicators, and sort the concerned performance evaluation indicators according to the comprehensive weight coefficient; Generate the performance test characteristics of each concerned performance evaluation indicator in turn according to the sorting result, set the standard test response characteristics corresponding to the performance test characteristics, and generate the performance test instruction according to the sorting result and the performance test characteristics of each concerned performance evaluation indicator; Among them, the performance test characteristics include test behavior, test frequency and test intensity, and the standard test response characteristics include standard test response data, standard test response frequency and standard test response intensity.

[0042] In this embodiment, several performance evaluation indicators include optoelectronic characteristics, efficiency, lifespan, chromaticity, response time, etc. The concerned performance evaluation indicator refers to the evaluation indicator that needs to be subjected to performance testing at the current monitoring point, and the application effect of the first operation and maintenance strategy is evaluated according to the performance test result.

[0043] In this embodiment, by setting the performance test instruction, the application effect of the first operation and maintenance strategy on the performance of the monitoring point is realized, and combined with the operation effect, the comprehensive evaluation of the first operation and maintenance strategy is completed, ensuring the operation and maintenance efficiency and the operation and maintenance effect, so as to maximize the application of the lifespan and performance of the organic electroluminescent device.

[0044] In some embodiments of the present application, generating a simulation application coefficient according to the test result includes: Obtain the location information, structure information and environmental information of the monitoring points that need to be operated and maintained, and build a simulation operation model of the corresponding monitoring points in combination with several predicted failure types and corresponding predicted failure characteristics of the corresponding monitoring points; Obtain the operation and maintenance information to be set according to the first operation and maintenance strategy, obtain the operation relationship between the operation and maintenance information to be set and the respective characteristic monitoring data of the corresponding monitoring points, and import it into the simulation operation model of the corresponding monitoring points to obtain the simulated monitoring points after simulation operation and maintenance; Obtain the simulated characteristic monitoring data of the simulated monitoring points, calculate the difference of the simulated characteristic monitoring data of each simulated characteristic monitoring data, and generate the simulated operation evaluation value of the simulated monitoring points according to several differences of the simulated characteristic monitoring data; Generate the predicted fault types corresponding to the monitoring points and the simulation change characteristics of the corresponding predicted fault characteristics according to the simulation feature monitoring data, and generate the compensation coefficient of the simulation operation evaluation value according to the simulation change characteristics; Perform performance tests on the simulation monitoring points according to the performance test instructions to obtain the simulation test response characteristics corresponding to the performance test characteristics of each concerned performance evaluation index. The simulation test characteristics include simulation test response data, simulation test response frequency, and simulation test response intensity; Compare the simulation test response characteristics with the standard test response characteristics of the corresponding performance test characteristics to obtain the simulation test response characteristic differences; Generate the simulation performance evaluation value of the simulation monitoring point according to the simulation test response characteristic differences of several concerned performance evaluation indexes and the corresponding comprehensive weight coefficients; Generate the simulation application coefficient of the first operation and maintenance strategy corresponding to the monitoring point according to the simulation operation evaluation value and the simulation performance evaluation value; The calculation formula of the simulation application coefficient is: ; Among them, Y is the simulation application coefficient, a1 is the weight coefficient of the simulation operation evaluation value, y1 is the first simulation application conversion coefficient, r is the compensation coefficient, n1 is the total number of the feature monitoring data of the current monitoring point, is the simulation feature monitoring data difference of the i-th simulation feature monitoring data, q1i is the weight coefficient of the i-th simulation feature monitoring data, a2 is the simulation performance evaluation value, y2 is the second simulation application conversion coefficient, n2 is the total number of concerned performance evaluation indexes, is the simulation test response characteristic difference of the s-th concerned performance evaluation index, and q2s is the comprehensive weight coefficient of the s-th concerned performance evaluation index.

[0045] In this embodiment, the simulated fault types and the corresponding simulated fault characteristics after the simulation operation and maintenance are obtained according to the simulation feature monitoring data, and compared with the predicted fault types and the corresponding predicted fault characteristics before the operation and maintenance to obtain the simulation change characteristics. The simulation change characteristics include the number reduction of the predicted fault types after the simulation operation and maintenance, the number reduction of the predicted fault factors in the corresponding predicted fault characteristics, the time period change of the predicted fault time period, the reduction change of the predicted fault coefficient and the predicted fault impact coefficient, etc. When the number of predicted fault types decreases or there is no change, and the predicted fault characteristics show a normal change trend, the corresponding compensation coefficient is larger, and vice versa. The value range of the compensation coefficient is (0.8, 1.2).

[0046] In this embodiment, the differences in simulation test response characteristics include the differences in simulation test response data, the differences in simulation test response frequencies, and the differences in simulation test response intensities, and they are uniformly quantified to obtain a simulation performance evaluation value. When the differences in simulation test response data, the differences in simulation test response frequencies, and the differences in simulation test response intensities are larger, the corresponding simulation performance evaluation value is smaller, and vice versa.

[0047] In some embodiments of the present application, to determine whether to modify the first operation and maintenance strategy, if so, obtain the second operation and maintenance strategy and issue an operation and maintenance instruction, including: Preset a simulation application coefficient threshold in advance; If the simulation application coefficient is greater than the simulation application coefficient threshold, do not modify the first operation and maintenance strategy and issue an operation and maintenance instruction; If the simulation application coefficient is less than the simulation application coefficient threshold, screen out the feature monitoring data to be optimized according to the simulation operation evaluation value, screen out the performance evaluation index to be optimized according to the simulation performance evaluation index, and modify the corresponding operation and maintenance information in the first operation strategy according to the feature monitoring data to be optimized and the performance evaluation index to be optimized to obtain the second operation and maintenance strategy, and issue an operation and maintenance instruction.

[0048] In some embodiments of the present application, as Figure 2 shown, there is also an operation and maintenance system for an organic electroluminescent device: A determination module, configured to preset multiple monitoring points of the organic electroluminescent device in advance, extract the historical fault characteristics of each monitoring point, and determine the feature monitoring data and the monitoring time node corresponding to each monitoring point according to the historical fault characteristics; An analysis module, configured to obtain the real-time feature monitoring data of each monitoring point according to the monitoring time node, analyze the real-time feature monitoring data, and generate a risk coefficient for each monitoring point according to the analysis result; A judgment module, configured to judge whether operation and maintenance are required according to the risk coefficient. If so, screen out the risk factors of the corresponding monitoring points and perform fault feature prediction, and configure the first operation and maintenance strategy according to the prediction result; An operation and maintenance module, configured to perform simulation operation and maintenance on the corresponding monitoring points according to the first operation and maintenance strategy, perform performance testing on the monitoring points after the simulation operation and maintenance, generate a simulation application coefficient according to the test result, and judge whether to modify the first operation and maintenance strategy. If so, obtain the second operation and maintenance strategy and issue an operation and maintenance instruction.

[0049] The above are only the preferred embodiments of the present application. It should be noted that for those of ordinary skill in the art in the technical field of the present application, without departing from the technical principle of the present application, several improvements and replacements can still be made, and these improvements and replacements should also be regarded as the protection scope of the present application.

Claims

1. A method for operation and maintenance of an organic electroluminescent device, characterized in that, Including: Preset multiple monitoring points of the organic electroluminescent device, extract the historical fault characteristics of each monitoring point, and determine the characteristic monitoring data and monitoring time nodes of the corresponding monitoring points according to the historical fault characteristics; Obtain the real-time characteristic monitoring data of each monitoring point according to the monitoring time node, analyze the real-time characteristic monitoring data, and generate the risk coefficient of each monitoring point according to the analysis result; Judge whether operation and maintenance are needed according to the risk coefficient. If so, screen out the risk factors of the corresponding monitoring points and conduct fault characteristic prediction, and configure the first operation and maintenance strategy according to the prediction result; Carry out simulation operation and maintenance on the corresponding monitoring points according to the first operation and maintenance strategy, conduct performance tests on the monitoring points after simulation operation and maintenance, generate the simulation application coefficient according to the test result, and judge whether to correct the first operation and maintenance strategy. If so, obtain the second operation and maintenance strategy and issue the operation and maintenance instruction.

2. The operation and maintenance method for an organic electroluminescent device according to claim 1, characterized in that, Determine the characteristic monitoring data and monitoring time nodes of the corresponding monitoring points according to the historical fault characteristics, including: Obtain a number of historical operation and maintenance logs of each monitoring point, and extract the historical fault characteristics in each historical operation and maintenance log. The historical fault characteristics include historical fault types, historical fault data, historical fault time periods, historical fault coefficients, and historical fault impact coefficients; Compare the historical fault data involved in the same historical fault type of each monitoring point to obtain the historical fault frequency of each historical fault data; Calculate the historical data difference between the same historical fault data and the corresponding standard data interval, and generate the historical influence degree of the corresponding historical fault data on the corresponding historical fault coefficient according to multiple historical data differences and the corresponding historical fault coefficients; Set the historical fault data with the historical fault frequency of each historical fault type greater than the preset fault frequency threshold and the historical influence degree greater than the preset influence degree threshold as the characteristic monitoring data of the corresponding monitoring point, and generate the first attention evaluation value of the corresponding characteristic monitoring data according to the historical fault frequency and the historical influence degree; Set the weight coefficient of the corresponding characteristic monitoring data according to the historical influence degree of the characteristic monitoring data on the historical fault coefficient, and generate the second attention evaluation value of the corresponding characteristic monitoring data in combination with the historical fault impact coefficient of the corresponding historical fault type of the characteristic monitoring data; Obtain multiple historical fluctuation degrees of the same characteristic monitoring data in multiple historical fault time periods of the corresponding historical fault type, and generate the average historical fluctuation degree of the corresponding characteristic monitoring data according to the multiple historical fluctuation degrees; Generate the third attention evaluation value of the corresponding characteristic monitoring data according to the average historical fluctuation degree; Generate the comprehensive attention evaluation value of the corresponding characteristic monitoring data according to the first attention evaluation value, the second attention evaluation value, and the third attention evaluation value, and set the time interval between adjacent monitoring time nodes of the corresponding characteristic monitoring data according to the comprehensive attention evaluation value.

3. The operation and maintenance method for an organic electroluminescent device according to claim 2, characterized in that, Set the time interval between adjacent monitoring time nodes of the corresponding characteristic monitoring data, including: Preset the first preset comprehensive attention evaluation value interval, the second preset comprehensive attention evaluation value interval, the third preset comprehensive attention evaluation value interval, and the fourth preset comprehensive attention evaluation value interval; When the comprehensive attention evaluation value of the feature monitoring data is within the first preset comprehensive attention evaluation value interval, the fourth preset time interval is selected as the time interval between adjacent monitoring time nodes of the corresponding feature monitoring data; When the comprehensive attention evaluation value of the feature monitoring data is within the second preset comprehensive attention evaluation value interval, the third preset time interval is selected as the time interval between adjacent monitoring time nodes of the corresponding feature monitoring data; When the comprehensive attention evaluation value of the feature monitoring data is within the third preset comprehensive attention evaluation value interval, the second preset time interval is selected as the time interval between adjacent monitoring time nodes of the corresponding feature monitoring data; When the comprehensive attention evaluation value of the feature monitoring data is within the fourth preset comprehensive attention evaluation value interval, the first preset time interval is selected as the time interval between adjacent monitoring time nodes of the corresponding feature monitoring data.

4. The operation and maintenance method for an organic electroluminescent device according to claim 3, characterized in that, Generate the risk coefficient for each monitoring point according to the analysis result, including: Set multiple monitoring time nodes for several feature monitoring data of each monitoring point at preset time intervals, and collect the real-time feature monitoring data of each monitoring point according to the monitoring time nodes; Compare the real-time feature monitoring data with the corresponding standard data interval to obtain the difference of the real-time feature monitoring data; Divide the real-time feature monitoring data into abnormal feature monitoring data, suspected abnormal feature monitoring data and normal feature monitoring data according to the difference of the real-time feature monitoring data, and divide them according to the corresponding fault types to obtain the set of real-time feature monitoring data of each fault type; Set the data sorting value of the real-time feature monitoring data according to the data status and data weight, and sort the real-time feature monitoring data in the set of real-time feature monitoring data according to the data sorting value; Generate the fault probability of the corresponding fault type according to the number of abnormal data, the number of suspected abnormal data, the number of normal data, the corresponding difference of the real-time feature monitoring data and the weight coefficient of the corresponding real-time feature monitoring data in the set of real-time feature monitoring data of each fault type; The calculation formula of the fault probability is: ； Among them, P is the failure probability, m1, m2, and m3 are the numbers of abnormal data, suspected abnormal data, and normal data in the real-time feature monitoring data sets corresponding to the failure types respectively. is the difference of the real-time feature monitoring data of the v1-th abnormality. is the weight coefficient of the difference of the real-time feature monitoring data of the v1-th abnormality. is the difference of the real-time feature monitoring data of the v2-th suspected abnormality. is the weight coefficient of the difference of the real-time feature monitoring data of the v2-th suspected abnormality. is the difference of the real-time feature monitoring data of the v3-th normality. is the weight coefficient of the difference of the real-time feature monitoring data of the v3-th normality, x1 is the first failure probability conversion coefficient, x2 is the second failure probability conversion coefficient, and x3 is the third failure probability conversion coefficient. Preset the fault probability threshold of each fault type and calculate the difference of the fault probability of each fault type; Generate the risk coefficient of the corresponding monitoring point according to the difference of the fault probabilities of multiple fault types at the same monitoring point and the corresponding weight coefficients.

5. The operation and maintenance method for an organic electroluminescent device according to claim 4, characterized in that, Screen out the risk factors of the corresponding monitoring point and conduct fault feature prediction, including: Preset the risk coefficient threshold; If the risk coefficient is less than the risk coefficient threshold, no operation and maintenance is required; If the risk coefficient is greater than the risk coefficient threshold, screen out the fault types with the fault probability greater than the corresponding fault probability threshold; Set the number of data extractions of the set of real-time feature monitoring data of the corresponding fault type according to the difference of the fault probabilities of the screened fault types and the corresponding weight coefficients; Extract the feature monitoring data in the set of real-time feature monitoring data of the corresponding fault type according to the number of data extractions and the sorting result, and set the risk factors of the corresponding monitoring point according to the data extraction result in the screened fault types at the same monitoring point, and the risk factors are several real-time feature monitoring data extracted; Compare the risk factors of the same fault type at the current monitoring point with the characteristic monitoring data of the corresponding fault type, and generate the credibility of the corresponding fault type according to the comparison result; If the credibility is less than the preset credibility threshold, eliminate the corresponding fault type and the risk factors involved in the corresponding fault type; If the credibility is greater than the preset credibility threshold, retain the involved risk factors and set the corresponding fault type as the predicted fault type, and determine the predicted fault characteristics of the corresponding predicted fault type according to the data status and data change characteristics of the risk factors involved in the current predicted fault type; Among them, when the data status of the risk factors involved in the predicted fault type is in an abnormal state, determine the predicted fault characteristics according to the sorting result of the involved risk factors and the difference of real-time characteristic monitoring data; When the data status of the risk factors involved in the predicted fault type is in a suspected abnormal state or a normal state, construct a real-time data change curve of the corresponding risk factor, and obtain the data change characteristics, where the data change characteristics include the change trend, the change rate, and the change magnitude; Perform curve extrapolation according to the data change characteristics and the real-time data change curve to obtain the predicted data change curve of the corresponding risk factor in the preset time period, and determine the predicted fault characteristics according to the predicted data change curve, the difference of the real-time characteristic monitoring data of the risk factor in the abnormal state, and the sorting result, where the predicted fault characteristics include the predicted fault factor, the predicted fault time period, the predicted fault coefficient, and the predicted fault impact coefficient; Generate the prediction result of the current monitoring point, where the prediction result includes a predicted fault sequence, the predicted fault sequence includes several predicted fault types and the corresponding predicted fault characteristics, and sort the several predicted fault types according to the credibility of the predicted fault type; 6. The operation and maintenance method for an organic electroluminescent device according to claim 5, characterized in that, Configure the first operation and maintenance strategy according to the prediction result, including: Construct an operation and maintenance strategy reference library for the corresponding monitoring point, where the operation and maintenance strategy reference library includes several preset fault types, each fault type includes several preset fault characteristics, and each preset fault characteristic is mapped with a corresponding preset operation and maintenance strategy; Perform a similarity analysis on the several preset fault types and the corresponding several preset fault characteristics in the operation and maintenance strategy reference library and the several predicted fault types and the corresponding predicted fault characteristics in the predicted fault sequence of the corresponding monitoring point to obtain the similarity between the preset fault characteristics and the predicted fault characteristics of the same fault type; Set the preset operation and maintenance strategy mapped by the preset fault characteristic with the largest similarity as the operation and maintenance sub-strategy of the corresponding predicted fault characteristic; Generate the operation and maintenance sub-strategies of the predicted fault characteristics of each predicted fault type in turn; Judge whether several operation and maintenance sub-strategies need to be optimized. If so, construct the first operation and maintenance strategy for the corresponding monitoring point according to the optimized operation and maintenance sub-strategy; 7. The operation and maintenance method for an organic electroluminescent device according to claim 6, characterized in that Before performing a performance test on the monitored point after simulation operation and maintenance, including: Preset several performance evaluation indicators, and construct a performance evaluation indicator tree based on the importance of each performance evaluation indicator, the correlation relationship and the degree of correlation between different performance evaluation indicators; Among them, the performance evaluation index tree includes a preset main trunk. The connection relationship and connection distance between the corresponding performance evaluation index and the preset main trunk are set according to the importance degree of the performance evaluation index. The connection relationship and connection distance between different performance evaluation indexes are set according to the association relationship and association degree between different performance evaluation indexes; The first weight coefficient of each performance evaluation index is set according to the connection information and position information of each performance evaluation index in the performance evaluation index tree; Based on the historical operation and maintenance logs of the monitoring points, determine the predicted failure type of the corresponding monitoring points and the predicted negative impact degree of the corresponding predicted failure characteristics on several performance evaluation indexes. The second weight coefficient of each performance evaluation index is set according to the predicted negative impact degree; Generate the comprehensive weight coefficient of the corresponding performance evaluation index according to the first weight coefficient and the second weight coefficient. Set the performance evaluation index with the comprehensive weight coefficient greater than the preset weight coefficient threshold as the concerned performance evaluation index, and sort the concerned performance evaluation indexes according to the comprehensive weight coefficient; Generate the performance test characteristics of each concerned performance evaluation index in turn according to the sorting result, and set the standard test response characteristics corresponding to the performance test characteristics. Generate the performance test instruction according to the sorting result and the performance test characteristics of each concerned performance evaluation index; Among them, the performance test characteristics include test behavior, test frequency and test intensity, and the standard test response characteristics include standard test response data, standard test response frequency and standard test response intensity.

8. The operation and maintenance method for an organic electroluminescent device according to claim 7, wherein Generate the simulation application coefficient according to the test result, including: Obtain the position information, structure information and environment information of the monitoring points that need operation and maintenance, and build the simulation operation model of the corresponding monitoring points in combination with several predicted failure types and corresponding predicted failure characteristics of the corresponding monitoring points; Obtain the operation and maintenance information to be set according to the first operation and maintenance strategy, obtain the operation relationship of the operation and maintenance information to be set for each characteristic monitoring data of the corresponding monitoring points, and import it into the simulation operation model of the corresponding monitoring points to obtain the simulated monitoring points after simulation operation and maintenance; Obtain the simulated characteristic monitoring data of the simulated monitoring points, calculate the difference of the simulated characteristic monitoring data of each simulated characteristic monitoring data, and generate the simulated operation evaluation value of the simulated monitoring points according to several differences of the simulated characteristic monitoring data; Generate the simulated change characteristics of the predicted failure type and corresponding predicted failure characteristics of the corresponding monitoring points according to the simulated characteristic monitoring data, and generate the compensation coefficient of the simulated operation evaluation value according to the simulated change characteristics; Conduct performance tests on the simulated monitoring points according to the performance test instructions to obtain the simulated test response characteristics corresponding to the performance test characteristics of each concerned performance evaluation index. The simulated test characteristics include simulated test response data, simulated test response frequency and simulated test response intensity; Compare the simulated test response characteristics with the standard test response characteristics of the corresponding performance test characteristics to obtain the difference of the simulated test response characteristics; Generate the simulated performance evaluation value of the simulated monitoring points according to the differences of the simulated test response characteristics of several concerned performance evaluation indexes and the corresponding comprehensive weight coefficients; Generate the simulation application coefficient of the first operation and maintenance strategy for the corresponding monitoring points according to the simulation operation evaluation value and the simulation performance evaluation value; The calculation formula of the simulation application coefficient is as follows: ； Among them, Y is the simulation application coefficient, a1 is the weight coefficient of the simulation operation evaluation value, y1 is the first simulation application conversion coefficient, r is the compensation coefficient, and n1 is the total number of characteristic monitoring data at the current monitoring point. is the simulation characteristic monitoring data difference of the i-th simulation characteristic monitoring data, q1i is the weight coefficient of the i-th simulation characteristic monitoring data, a2 is the simulation performance evaluation value, y2 is the second simulation application conversion coefficient, and n2 is the total number of concerned performance evaluation indicators. is the simulation test response characteristic difference of the s-th concerned performance evaluation indicator, and q2s is the comprehensive weight coefficient of the s-th concerned performance evaluation indicator.

9. The operation and maintenance method for an organic electroluminescent device according to claim 8, wherein, Judge whether to correct the first operation and maintenance strategy. If so, obtain the second operation and maintenance strategy and issue an operation and maintenance instruction, including: Preset the simulation application coefficient threshold in advance; If the simulation application coefficient is greater than the simulation application coefficient threshold, do not correct the first operation and maintenance strategy and issue an operation and maintenance instruction; If the simulation application coefficient is less than the simulation application coefficient threshold, screen out the to-be-optimized feature monitoring data according to the simulation operation evaluation value, screen out the to-be-optimized performance evaluation indicators according to the simulation performance evaluation indicators, correct the corresponding operation and maintenance information in the first operation strategy according to the to-be-optimized feature monitoring data and the to-be-optimized performance evaluation indicators, obtain the second operation and maintenance strategy, and issue an operation and maintenance instruction.

10. An operation and maintenance system for an organic electroluminescent device, characterized in that, Including: A determination module for presetting multiple monitoring points of the organic electroluminescent device in advance, extracting the historical fault characteristics of each monitoring point, and determining the feature monitoring data and the monitoring time node of the corresponding monitoring point according to the historical fault characteristics; An analysis module for obtaining the real-time feature monitoring data of each monitoring point according to the monitoring time node, analyzing the real-time feature monitoring data, and generating the risk coefficient of each monitoring point according to the analysis result; A judgment module for judging whether operation and maintenance are required according to the risk coefficient. If so, screen out the risk factors of the corresponding monitoring point and perform fault feature prediction, and configure the first operation and maintenance strategy according to the prediction result; An operation and maintenance module for performing simulation operation and maintenance on the corresponding monitoring points according to the first operation and maintenance strategy, performing performance testing on the monitoring points after simulation operation and maintenance, generating a simulation application coefficient according to the test result, and judging whether to correct the first operation and maintenance strategy. If so, obtain the second operation and maintenance strategy and issue an operation and maintenance instruction.

Images