A life prediction system for organic electroluminescent devices

Abstract

The present invention provides an organic electroluminescent device life prediction system, which belongs to the field of data processing technology, and includes: acquiring and analyzing the real-time operation data of the target organic electroluminescent device, determining the operation vector of the target organic electroluminescent device; collecting the historical data of multiple organic electroluminescent devices, determining the characteristic vector of each historical data and the failure mode vector of all historical data; determining the termination cluster group and the actual failure mode of each cluster group in the termination cluster group; determining the belonging cluster group of the target organic electroluminescent device; predicting the remaining life of the target organic electroluminescent device. The target organic electroluminescent device can be analyzed in a personalized manner, targeted prediction can be achieved, prediction efficiency can be improved, the accuracy and reliability of prediction can be improved, device problems can be discovered in a timely manner, maintenance costs can be reduced, serious consequences caused by device failures can be prevented, and the reliability and adaptability of the system can be improved.

Description

Technical Field

[0001] The invention relates to the technical field of data processing, and in particular to a life prediction system for an organic electroluminescent device. Background Art

[0002] Organic electroluminescent devices are a technology that emits light through organic materials. They have a wide range of applications, covering display technology, lighting, sensing and other aspects. In display technology, they are used in smartphones, TVs, tablets, etc. In lighting technology, they are used in indoor lighting. In sensing technology, they are used in health monitoring, biosensing and other fields. In wearable medical devices, they can be well integrated into wearable devices, such as heart rate monitors, blood oximeters and other medical and health equipment. They are also used in virtual reality, augmented reality, artworks, decorations and creative design products. The life of organic electroluminescent devices directly determines their long-term reliability, stability and user experience in practical applications.

[0003] Therefore, the present invention provides a life prediction system for an organic electroluminescent device. Summary of the invention

[0004] The present invention provides an organic electroluminescent device life prediction system. By analyzing the real-time operation data and multiple historical data of a target organic electroluminescent device, the operation vector, failure mode vector and characteristic vector of each historical data of the organic electroluminescent device are determined, and the termination cluster group and the actual failure mode of each cluster group in the termination cluster group are determined, so as to determine the belonging cluster group of the target organic electroluminescent device and perform failure mode prediction and remaining life prediction. The target organic electroluminescent device can be analyzed in a personalized manner, targeted prediction can be achieved, prediction efficiency can be improved, and the accuracy and reliability of prediction can be improved. Device problems can be discovered in time, maintenance costs can be reduced, serious consequences caused by device failures can be prevented, and the reliability and adaptability of the system can be improved.

[0005] The present invention provides a life prediction system for an organic electroluminescent device, comprising:

[0006] Acquisition module: acquires and analyzes the real-time operation data of the target organic electroluminescent device, and determines the operation vector of the target organic electroluminescent device;

[0007] Acquisition module: collects historical data of multiple target organic electroluminescent devices, determines the characteristic vector of each historical data, divides all historical data into an experimental group and a verification group, and determines the failure mode vector of all historical data in the experimental group;

[0008] Determination module: determining the termination cluster and the actual failure mode of each cluster in the termination cluster based on all feature vectors and failure mode vectors;

[0009] Affiliation module: determining the affiliation cluster of the target organic electroluminescent device based on the operation vector of the target organic electroluminescent device and the cluster vectors of all clusters in the termination cluster;

[0010] Prediction module: predicts the remaining life of the target organic electroluminescent device based on the operation vector and the belonging cluster of the target organic electroluminescent device.

[0011] According to an organic electroluminescent device life prediction system provided by the present invention, the acquisition module comprises:

[0012] An operation vector unit: preprocesses the real-time operation data, and determines the operation vector of the target organic electroluminescent device based on the preprocessed real-time operation data, wherein the real-time operation data includes all data related to the life of the target organic electroluminescent device.

[0013] According to an organic electroluminescent device life prediction system provided by the present invention, the acquisition module comprises:

[0014] Feature vector unit: collects historical data of multiple target organic electroluminescent devices within a specified time period before the actual failure time, preprocesses all the historical data, and determines the corresponding feature vector based on each preprocessed historical data;

[0015] Division unit: divide all preprocessed historical data into experimental group and verification group;

[0016] Failure mode vector unit: extracts actual failure modes from all historical data in the experimental group and determines a failure mode vector, wherein the failure mode vector includes all actual failure modes that appear in all historical data in the experimental group.

[0017] According to an organic electroluminescent device life prediction system provided by the present invention, the determination module comprises:

[0018] Cluster unit: standardize the feature vectors of all historical data in the experimental group, and determine each feature vector after standardization as a cluster;

[0019] Stop cluster number unit: determines the stop cluster number of cluster analysis based on the actual failure mode number in the failure mode vector;

[0020] Clustering unit: cluster all clusters, determine the current number of cluster groups and all current cluster groups for each clustering result;

[0021] Comparison unit: compares the current number of cluster groups and the number of stop cluster families. If the current number of cluster groups is greater than the number of stop cluster families, iterative clustering is performed based on all current cluster groups of the current clustering result. Otherwise, all current cluster families of the current clustering result are determined to be stop cluster families.

[0022] Cluster vector unit: statistics the failure characteristics of all characteristic vectors included in each cluster in the terminated cluster, and determine the corresponding cluster vector;

[0023] Actual failure mode unit: based on the cluster vector and failure mode vector of each cluster in the terminated cluster, determines the actual failure mode corresponding to each cluster in the terminated cluster.

[0024] According to an organic electroluminescent device life prediction system provided by the present invention, the attribution module comprises:

[0025] A scoring value unit: based on the running vector of the target organic electroluminescent device and the cluster vectors of all clusters in the terminated cluster, determining the scoring value of the target organic electroluminescent device based on each cluster in the terminated cluster;

[0026] in, 2, N2 represents the actual number of cluster groups in the terminated cluster group, represents the consistency score value of the i-th cluster in the terminated cluster, represents the jth eigenvector of the ith cluster in the terminated cluster, represents the cluster vector of the ith cluster in the terminated cluster, represents the cluster center feature vector of the ith cluster in the terminated cluster, represents the cluster vector weight of the ith cluster in the termination cluster, represents the number of eigenvectors of the ith cluster in the terminated cluster, represents the separation score value of the ith cluster in the termination cluster, represents the cluster vector of the kth cluster in the terminated cluster, represents the cluster center feature vector of the kth cluster in the termination cluster, represents the cluster vector weight of the kth cluster in the termination cluster, The operation vector of the target organic electroluminescent device is based on the score value of the i-th cluster in the termination cluster. represents the operation vector of the target organic electroluminescent device, represents the similarity score value between the running vector of the target organic electroluminescent device and the cluster vector of the i-th cluster in the termination cluster, represents the running vector of the target organic electroluminescent device and the fitting score value of the i-th cluster in the termination cluster, Represents the first weight of the similarity score value, The second weight representing the consistency score, The third weight representing the separation score, The fourth weight representing the fitting score value;

[0027] The attribution unit: based on the score values ​​of all clusters in the termination cluster of the target organic electroluminescent device, selects the cluster with the largest score value as the attribution cluster of the target organic electroluminescent device.

[0028] According to an organic electroluminescent device life prediction system provided by the present invention, the prediction module comprises:

[0029] First failure mode unit: based on the cluster vector of each cluster in the terminated cluster, predict the first failure mode of each feature vector of the historical data in the verification group;

[0030] Reasonableness unit: based on the actual failure mode in all historical data in the verification group and the consistency of the first failure mode, determine the rationality of the actual failure mode of each cluster group in the termination cluster group;

[0031] A second failure mode unit: determining an actual failure mode of the cluster to which the target organic electroluminescent device belongs, as a second failure mode of the target organic electroluminescent device;

[0032] Life prediction unit: predicts the remaining life of the target organic electroluminescent device based on the operation vector of the target organic electroluminescent device and the rationality of the actual failure mode of the belonging cluster;

[0033] Early warning unit: based on the remaining life and the second failure mode of the target organic electroluminescent device, sends early warning information to the user.

[0034] According to an organic electroluminescent device life prediction system provided by the present invention, the life prediction unit comprises:

[0035] The first extraction subunit is used to extract the start time of use in the historical data of each organic electroluminescent device, and determine the fitting life of each cluster group based on the start time of use and the actual failure time of all historical data of each cluster group in the terminated cluster group;

[0036] A second extraction subunit: extracting the current usage time from the real-time operation data of the target organic electroluminescent device;

[0037] Calculation subunit: Calculate the remaining life of the target organic electroluminescent device based on the current usage time of the target organic electroluminescent device, the fitted life of the cluster group to which the target organic electroluminescent device belongs, the score value of the cluster group to which the target organic electroluminescent device belongs, and the rationality of the actual failure mode of the cluster group to which the target organic electroluminescent device belongs.

[0038] According to an organic electroluminescent device life prediction system provided by the present invention, the calculation subunit comprises:

[0039] in, represents the remaining life of the target organic electroluminescent device, represents the fitted lifetime of the target organic electroluminescent device to which the cluster belongs, Indicates the current usage time of the target organic electroluminescent device, represents the score value of the cluster to which the target organic electroluminescent device belongs, represents the life attenuation coefficient, represents the reasonableness adjustment coefficient of the actual failure mode, Indicates the actual failure mode rationality of the target organic electroluminescent device's belonging cluster family.

[0040] Compared with the prior art, the present invention has the following beneficial effects:

[0041] By analyzing the real-time operation data and multiple historical data of the target organic electroluminescent device, the operation vector, failure mode vector and characteristic vector of each historical data of the organic electroluminescent device are determined, and the termination cluster and the actual failure mode of each cluster in the termination cluster are determined, so as to determine the cluster to which the target organic electroluminescent device belongs and perform failure mode prediction and remaining life prediction. The target organic electroluminescent device can be analyzed in a personalized manner to achieve targeted prediction, improve prediction efficiency, improve prediction accuracy and reliability, discover device problems in a timely manner, reduce maintenance costs, prevent serious consequences caused by device failures, and improve system reliability and adaptability. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0043] Figure 1 It is a structural schematic diagram of a life prediction system for an organic electroluminescent device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0044] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are 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.

[0045] Embodiment 1:

[0046] The embodiment of the present invention provides a system for predicting the life of an organic electroluminescent device, such as Figure 1 As shown, including:

[0047] Acquisition module: acquires and analyzes the real-time operation data of the target organic electroluminescent device, and determines the operation vector of the target organic electroluminescent device;

[0048] Acquisition module: collects historical data of multiple target organic electroluminescent devices, determines the characteristic vector of each historical data, divides all historical data into an experimental group and a verification group, and determines the failure mode vector of all historical data in the experimental group;

[0049] Determination module: determining the termination cluster and the actual failure mode of each cluster in the termination cluster based on all feature vectors and failure mode vectors;

[0050] Affiliation module: determining the affiliation cluster of the target organic electroluminescent device based on the operation vector of the target organic electroluminescent device and the cluster vectors of all clusters in the termination cluster;

[0051] Prediction module: predicts the remaining life of the target organic electroluminescent device based on the operation vector and the belonging cluster of the target organic electroluminescent device.

[0052] In this embodiment, by monitoring the real-time operation data of the target organic electroluminescent device, the current state of the device can be accurately reflected, and these data can be converted into an operation vector.

[0053] In this embodiment, the operation vector may include a current fluctuation range, a voltage increase rate, a brightness decay rate, a temperature change rate, a color drift amount, etc. of the target organic electroluminescent device.

[0054] In this embodiment, each cluster in the terminated cluster corresponds to an actual failure mode and a cluster vector.

[0055] The beneficial effects of the above technical solution are as follows: by analyzing the real-time operation data and multiple historical data of the target organic electroluminescent device, the operation vector, failure mode vector and characteristic vector of each historical data of the organic electroluminescent device are determined, and the termination cluster and the actual failure mode of each cluster in the termination cluster are determined, so as to determine the cluster to which the target organic electroluminescent device belongs and perform failure mode prediction and remaining life prediction. The target organic electroluminescent device can be analyzed in a personalized manner to achieve targeted prediction, improve prediction efficiency, improve prediction accuracy and reliability, timely discover device problems, reduce maintenance costs, prevent serious consequences caused by device failures, and improve system reliability and adaptability.

[0056] Embodiment 2:

[0057] An embodiment of the present invention provides a system for predicting the life of an organic electroluminescent device, wherein the acquisition module comprises:

[0058] An operation vector unit: preprocesses the real-time operation data, and determines the operation vector of the target organic electroluminescent device based on the preprocessed real-time operation data, wherein the real-time operation data includes all data related to the life of the target organic electroluminescent device.

[0059] In this embodiment, all data related to the life span of the target organic electroluminescent device may include current data, voltage data, light emitting brightness data, temperature data, color data, and operating time.

[0060] In this embodiment, the current data reflects the operating current of the target organic electroluminescent device, which is directly related to the luminous efficiency and power consumption of the device. Abnormal current fluctuations may be an early sign of device failure.

[0061] In this embodiment, the voltage data reflects the driving voltage required by the target organic electroluminescent device when it is working. An increase in voltage may indicate device aging or a decrease in luminous efficiency. The voltage data is usually combined with the current data to analyze power consumption and device health status.

[0062] In this embodiment, the brightness data reflects the luminous intensity of the target organic electroluminescent device. The attenuation of brightness is usually an important indicator of device aging. The brightness gradually decreases over time and is one of the key factors in life prediction.

[0063] In this embodiment, the temperature data reflects the operating temperature of the target organic electroluminescent device. Too high a temperature will accelerate the aging of the device material, resulting in a shortened device lifespan. Controlling the temperature can effectively extend the device's service life.

[0064] In this embodiment, the color data reflects the color stability of the light emitted by the target organic electroluminescent device. Over time, the color may drift, which is another important feature of device aging. The change in color can indicate the degradation of the material.

[0065] The beneficial effects of the above technical solution are as follows: by acquiring and analyzing the real-time operating data of the target organic electroluminescent device and determining the operating vector of the target organic electroluminescent device, the real-time state and performance of the target organic electroluminescent device can be determined, providing a data basis for determining the cluster to which the target organic electroluminescent device belongs, so as to predict the remaining life of the target organic electroluminescent device.

[0066] Embodiment 3:

[0067] The embodiment of the present invention provides a system for predicting the life of an organic electroluminescent device, wherein the acquisition module comprises:

[0068] Feature vector unit: collects historical data of multiple target organic electroluminescent devices within a specified time period before the actual failure time, preprocesses all the historical data, and determines the corresponding feature vector based on each preprocessed historical data;

[0069] Division unit: divide all preprocessed historical data into experimental group and verification group;

[0070] Failure mode vector unit: extracts actual failure modes from all historical data in the experimental group and determines a failure mode vector, wherein the failure mode vector includes all actual failure modes that appear in all historical data in the experimental group.

[0071] In this embodiment, historical data of the organic electroluminescent device is collected within a specified time period (eg, 80% or 90% of the total life) before the organic electroluminescent device completely fails.

[0072] In this embodiment, the historical data may include the start time of use of the organic electroluminescent device, actual failure mode, historical current data, historical voltage data, historical luminous brightness data, historical temperature data, and historical color data.

[0073] In this embodiment, preprocessing of all historical data includes cleaning data (such as processing noise, missing data, outliers, etc.), normalizing (making the data have a uniform scale), and extracting key features that can reflect device aging and failure to determine feature vectors.

[0074] In this embodiment, the feature vector may include current fluctuation range, voltage rise rate, brightness decay rate, temperature change rate, color drift amount, etc.

[0075] In this embodiment, the historical data are divided into an experimental group (for cluster analysis) and a validation group (for testing and validating the termination of clusters). A common method is random allocation or time splitting. The experimental group usually accounts for 70%-80% of the multiple historical data, while the validation group accounts for 20%-30%.

[0076] In this embodiment, in the experimental group data, the actual failure mode of each device is extracted, and these modes are vectorized to form a failure mode vector representing all failure modes in the experimental group.

[0077] The beneficial effects of the above technical solution are as follows: by collecting historical data of multiple organic electroluminescent devices, determining the characteristic vector of each historical data, and determining the failure mode vector of all historical data in the experimental group, the data quality can be improved, and a data basis can be provided for determining the actual failure mode of the termination cluster group and each cluster group in the termination cluster group, so as to improve the accuracy of the prediction.

[0078] Embodiment 4:

[0079] An embodiment of the present invention provides a system for predicting the life of an organic electroluminescent device, wherein the determination module comprises:

[0080] Cluster unit: standardize the feature vectors of all historical data in the experimental group, and determine each feature vector after standardization as a cluster;

[0081] Stop cluster number unit: determines the stop cluster number of cluster analysis based on the actual failure mode number in the failure mode vector;

[0082] Clustering unit: cluster all clusters, determine the current number of cluster groups and all current cluster groups for each clustering result;

[0083] Comparison unit: compares the current number of cluster groups and the number of stop cluster families. If the current number of cluster groups is greater than the number of stop cluster families, iterative clustering is performed based on all current cluster groups of the current clustering result. Otherwise, all current cluster families of the current clustering result are determined to be stop cluster families.

[0084] Cluster vector unit: statistics the failure characteristics of all characteristic vectors included in each cluster in the terminated cluster, and determine the corresponding cluster vector;

[0085] Actual failure mode unit: based on the cluster vector and failure mode vector of each cluster in the terminated cluster, determines the actual failure mode corresponding to each cluster in the terminated cluster.

[0086] In this embodiment, all historical data in the experimental group are standardized to make the numerical ranges of different features consistent, so as to eliminate the differences in the magnitudes of the various features and ensure that their contributions to clustering are relatively balanced. Each feature vector after standardization is a cluster.

[0087] In this embodiment, the termination condition of the clustering analysis, that is, the number of stopping clusters, is determined based on the actual number of failure modes in the failure mode vector. This number determines when the clustering process stops, ensuring that the clustering result is consistent with the actual number of failure modes. For example, the larger the actual number of failure modes in the failure mode vector, the larger the number of stopping clusters; conversely, the smaller the actual number of failure modes in the failure mode vector, the smaller the number of stopping clusters.

[0088] In this embodiment, it is assumed that the actual number of failure modes in the failure mode vector is 100, and the number of determined stop clusters is 20. If the clustering result generated by the current clustering analysis is

[0089] In this embodiment, cluster analysis is performed based on all feature vectors in the standardized experimental group, and each clustering generates a clustering result, including the number of cluster groups and detailed information of each cluster group. Clustering can identify historical data with similar feature vectors and classify them into the same cluster family.

[0090] In this embodiment, if the current number of cluster groups is greater than the number of stopping cluster families, iterative clustering of the current cluster group continues until the actual number of cluster groups is equal to or less than the number of stopping cluster families, thereby ensuring that the clustering process will not terminate early or late. For example, if the actual number of failure modes in the failure mode vector is 100, and the determined number of stopping cluster families is 20, if the current number of cluster groups in the clustering result generated by the current clustering analysis is less than 100 and greater than 21, iterative clustering of the current cluster group continues. If the current number of cluster groups in the clustering result generated by the current clustering analysis is equal to or less than 20, all current cluster families of the current clustering result are determined to be stopping cluster families.

[0091] In this embodiment, for each cluster in the terminated clusters, all features of all feature vectors therein are counted to generate a cluster vector, which represents the overall features of the cluster related to failure.

[0092] In this embodiment, the cluster vector represents a vector composed of multiple failure features representing the corresponding cluster, and is determined based on the statistical results of the failure features of all feature vectors included in the corresponding cluster. The statistical results can be the number of times each failure feature appears in all feature vectors, the average, etc.

[0093] The beneficial effects of the above technical solution are as follows: the actual failure mode of the termination cluster family and each cluster family in the termination cluster family is determined according to the characteristic vector and the failure mode vector, and the cluster family vector of the cluster family corresponding to each failure mode in the failure mode vector can be determined, thereby improving the recognition ability of the actual failure mode and the accuracy of life prediction.

[0094] Embodiment 5:

[0095] The embodiment of the present invention provides a system for predicting the life of an organic electroluminescent device, wherein the attribution module comprises:

[0096] A scoring value unit: based on the running vector of the target organic electroluminescent device and the cluster vectors of all clusters in the terminated cluster, determining the scoring value of the target organic electroluminescent device based on each cluster in the terminated cluster;

[0097] in, 2, N2 represents the actual number of cluster groups in the terminated cluster group, represents the consistency score of the ith cluster in the terminated cluster, represents the jth eigenvector of the ith cluster in the terminated cluster, represents the cluster vector of the ith cluster in the terminated cluster, represents the cluster center feature vector of the ith cluster in the terminated cluster, represents the cluster vector weight of the ith cluster in the termination cluster, represents the number of eigenvectors of the ith cluster in the terminated cluster, represents the separation score value of the ith cluster in the termination cluster, represents the cluster vector of the kth cluster in the terminated cluster, represents the cluster center feature vector of the kth cluster in the termination cluster, represents the cluster vector weight of the kth cluster in the termination cluster, The operation vector of the target organic electroluminescent device is based on the score value of the i-th cluster in the termination cluster. represents the operation vector of the target organic electroluminescent device, represents the similarity score value between the running vector of the target organic electroluminescent device and the cluster vector of the i-th cluster in the termination cluster, represents the running vector of the target organic electroluminescent device and the fitting score value of the i-th cluster in the termination cluster, Represents the first weight of the similarity score value, The second weight representing the consistency score, The third weight representing the separation score, The fourth weight representing the fitting score value;

[0098] The attribution unit: based on the score values ​​of all clusters in the termination cluster of the target organic electroluminescent device, selects the cluster with the largest score value as the attribution cluster of the target organic electroluminescent device.

[0099] In this embodiment, the consistency score value represents a score value calculated based on the consistency between all feature vectors included in the corresponding cluster and the cluster vector, and the consistency between all feature vectors in the cluster except the cluster center and the cluster center feature vector.

[0100] In this embodiment, the separability score value represents a score value calculated based on the separability between the corresponding cluster family vector and the cluster family vectors of all other cluster families in the terminated cluster family except the cluster family vector, and the separability between the corresponding cluster center feature vector and the cluster center feature vectors of all other cluster families in the terminated cluster family except the changed cluster family.

[0101] In this embodiment, the fitting score value represents the regularization result based on the running vector of the target organic electroluminescent device, the cluster vector of the corresponding cluster family and the cluster center feature vector, so as to prevent overfitting and balance the influence of the length of the running vector, the cluster center feature vector and the cluster vector on the score value.

[0102] In this embodiment, the score value represents a comprehensive score value calculated based on the consistency score value, separability score value, similarity score value of the running vector of the target organic electroluminescent device and the cluster vector of the cluster in the termination cluster, and fitting score value of the corresponding cluster.

[0103] In this embodiment, after score comparison, the target organic electroluminescent device is finally classified into a specific cluster group, which is the cluster group that is most similar to the operation vector of the target device, that is, the cluster group with the largest score value.

[0104] The beneficial effects of the above technical solution are as follows: according to the running vector of the target organic electroluminescent device and the cluster vectors of all clusters in the terminated cluster, the cluster to which the target organic electroluminescent device belongs is determined, and the target organic electroluminescent device can be personalized identified, which significantly improves the accuracy and adaptability of life prediction.

[0105] Embodiment 6:

[0106] An embodiment of the present invention provides a system for predicting the life of an organic electroluminescent device, wherein the prediction module comprises:

[0107] The first failure mode unit: based on the cluster vector of each cluster in the termination cluster, the first failure mode prediction is performed on the feature vector of each historical data in the verification group;

[0108] Reasonableness unit: based on the actual failure mode in all historical data in the verification group and the consistency of the first failure mode, determine the rationality of the actual failure mode of each cluster group in the termination cluster group;

[0109] A second failure mode unit: determining an actual failure mode of the cluster to which the target organic electroluminescent device belongs, as a second failure mode of the target organic electroluminescent device;

[0110] Life prediction unit: predicts the remaining life of the target organic electroluminescent device based on the operation vector of the target organic electroluminescent device and the rationality of the actual failure mode of the belonging cluster group;

[0111] Early warning unit: based on the remaining life and the second failure mode of the target organic electroluminescent device, sends early warning information to the user.

[0112] In this embodiment, the first failure mode prediction of the feature vector of each historical data in the verification group can be performed by calculating the feature vector of the historical data based on the score value of each cluster group in the termination cluster group, determining the cluster group to which the feature vector of the historical data belongs, so as to predict the first failure mode of the feature vector of the historical data.

[0113] In this embodiment, the rationality of the actual failure mode of each cluster group in the terminated cluster group is determined by extracting the feature vectors of the historical data of each cluster group belonging to the cluster group in the verification group respectively, and the rationality of the actual failure mode of the cluster group is determined based on the actual failure mode of all the extracted feature vectors and the consistency of the first failure mode.

[0114] In this embodiment, the actual failure mode of the cluster to which the target organic electroluminescent device belongs is the second failure mode of the target organic electroluminescent device.

[0115] In this embodiment, when it is predicted that the remaining life of the target organic electroluminescent device is about to reach a critical point, the system will send a warning message to the user according to the second predicted failure mode to prompt the possible failure risk.

[0116] The beneficial effects of the above technical solution are as follows: according to the operating vector of the target organic electroluminescent device and the belonging cluster group, the second failure mode is determined, and the remaining life of the target organic electroluminescent device is predicted. The rationality of the actual failure mode of each cluster group in the termination cluster group can be verified, and the targeted prediction of the target organic electroluminescent device can be achieved, thereby improving the accuracy and credibility of the prediction.

[0117] Embodiment 7:

[0118] An embodiment of the present invention provides a life prediction system for an organic electroluminescent device, wherein the life prediction unit comprises:

[0119] The first extraction subunit is used to extract the start time of use in the historical data of each organic electroluminescent device, and determine the fitting life of each cluster group based on the start time of use and the actual failure time of all historical data of each cluster group in the terminated cluster group;

[0120] A second extraction subunit: extracting the current usage time from the real-time operation data of the target organic electroluminescent device;

[0121] Calculation subunit: Calculate the remaining life of the target organic electroluminescent device based on the current usage time of the target organic electroluminescent device, the fitted life of the cluster group to which the target organic electroluminescent device belongs, the score value of the cluster group to which the target organic electroluminescent device belongs, and the rationality of the actual failure mode of the cluster group to which the target organic electroluminescent device belongs.

[0122] In this embodiment, the fitting life of each cluster group is determined by analyzing the start-of-use time and actual failure time of all devices with similar historical data in the terminated cluster group.

[0123] In this embodiment, the fitting life of each cluster group can be determined by performing distribution fitting, regression analysis, model fitting, etc. on the start-of-use time and actual failure time of all historical data of the corresponding cluster group.

[0124] In this embodiment, the current usage time is a dynamic indicator, which can reflect the real-time usage of the organic electroluminescent device and ensure the timeliness and accuracy of the life prediction.

[0125] The beneficial effects of the above technical solution are as follows: the remaining life of the target organic electroluminescent device is predicted based on the operating vector of the target organic electroluminescent device and the rationality of the actual failure mode of the belonging cluster group, which can improve the prediction accuracy of the life of each cluster group in the terminated cluster group, further improve the accuracy of the life prediction of the target organic electroluminescent device, timely discover device problems, and reduce maintenance costs.

[0126] Example 8

[0127] An embodiment of the present invention provides a life prediction system for an organic electroluminescent device, wherein the calculation subunit comprises:

[0128] in, represents the remaining life of the target organic electroluminescent device, represents the fitted lifetime of the target organic electroluminescent device to which the cluster belongs, Indicates the current usage time of the target organic electroluminescent device, represents the score value of the cluster to which the target organic electroluminescent device belongs, represents the life attenuation coefficient, represents the reasonableness adjustment coefficient of the actual failure mode, Indicates the actual failure mode rationality of the target organic electroluminescent device's belonging cluster family.

[0129] In this embodiment, the remaining life of the target organic electroluminescent device represents the remaining usage time calculated according to the fitted life of the belonging cluster, the score value, the rationality of the actual failure mode and the current usage time of the target organic electroluminescent device.

[0130] The beneficial effects of the above technical solution are as follows: the remaining life of the target organic electroluminescent device is calculated according to the current usage time of the target organic electroluminescent device, the fitted life of the cluster to which it belongs, the score value and the rationality of the actual failure mode, and targeted prediction can be made in real time to accurately predict the remaining life of the target organic electroluminescent device, thereby improving the accuracy and efficiency of the prediction.

[0131] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0132] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0133] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An organic electroluminescent device life prediction system, characterized in that: include: Acquisition module: acquires and analyzes the real-time operation data of the target organic electroluminescent device, and determines the operation vector of the target organic electroluminescent device; Acquisition module: collects historical data of multiple target organic electroluminescent devices, determines the characteristic vector of each historical data, divides all historical data into an experimental group and a verification group, and determines the failure mode vector of all historical data in the experimental group; Determination module: determining the termination cluster and the actual failure mode of each cluster in the termination cluster based on all feature vectors and failure mode vectors; Affiliation module: determining the affiliation cluster of the target organic electroluminescent device based on the operation vector of the target organic electroluminescent device and the cluster vectors of all clusters in the termination cluster; Prediction module: predicting the remaining life of the target organic electroluminescent device based on the operation vector and the belonging cluster of the target organic electroluminescent device; Wherein, the determination module includes: Cluster unit: standardize the feature vectors of all historical data in the experimental group, and determine each feature vector after standardization as a cluster; Stop cluster number unit: determines the stop cluster number of cluster analysis based on the actual failure mode number in the failure mode vector; Clustering unit: cluster all clusters, determine the current number of cluster groups and all current cluster groups for each clustering result; Comparison unit: compares the current number of cluster groups and the number of stop cluster families. If the current number of cluster groups is greater than the number of stop cluster families, iterative clustering is performed based on all current cluster groups of the current clustering result. Otherwise, all current cluster families of the current clustering result are determined to be stop cluster families. Cluster vector unit: statistics the failure characteristics of all characteristic vectors included in each cluster in the terminated cluster, and determine the corresponding cluster vector; Actual failure mode unit: determining an actual failure mode corresponding to each cluster group in the terminated cluster group based on the cluster group vector and the failure mode vector of each cluster group in the terminated cluster group; The attribution module comprises: A scoring value unit: based on the running vector of the target organic electroluminescent device and the cluster vectors of all clusters in the terminated cluster, determining the scoring value of the target organic electroluminescent device based on each cluster in the terminated cluster; in, 2, N2 represents the actual number of cluster groups in the terminated cluster group, represents the consistency score value of the i-th cluster in the terminated cluster, represents the jth eigenvector of the ith cluster in the terminated cluster, represents the cluster vector of the ith cluster in the terminated cluster, represents the cluster center feature vector of the ith cluster in the terminated cluster, represents the cluster vector weight of the ith cluster in the termination cluster, represents the number of eigenvectors of the ith cluster in the terminated cluster, represents the separation score value of the ith cluster in the termination cluster, represents the cluster vector of the kth cluster in the terminated cluster, represents the cluster center feature vector of the kth cluster in the termination cluster, represents the cluster vector weight of the kth cluster in the termination cluster, The operation vector of the target organic electroluminescent device is based on the score value of the i-th cluster in the termination cluster. represents the operation vector of the target organic electroluminescent device, represents the similarity score value between the running vector of the target organic electroluminescent device and the cluster vector of the i-th cluster in the termination cluster, represents the running vector of the target organic electroluminescent device and the fitting score value of the i-th cluster in the termination cluster, Represents the first weight of the similarity score value, The second weight representing the consistency score, The third weight representing the separation score, The fourth weight representing the fit score value; The attribution unit: based on the score values ​​of all clusters in the termination cluster of the target organic electroluminescent device, selects the cluster with the largest score value as the attribution cluster of the target organic electroluminescent device.

2. The organic electroluminescent device life prediction system according to claim 1, characterized in that: The acquisition module comprises: An operation vector unit: preprocesses the real-time operation data, and determines the operation vector of the target organic electroluminescent device based on the preprocessed real-time operation data, wherein the real-time operation data includes all data related to the life of the target organic electroluminescent device.

3. The organic electroluminescent device life prediction system according to claim 1, characterized in that: The acquisition module comprises: Feature vector unit: collects historical data of multiple target organic electroluminescent devices within a specified time period before the actual failure time, preprocesses all the historical data, and determines the corresponding feature vector based on each preprocessed historical data; Division unit: divide all preprocessed historical data into experimental group and verification group; Failure mode vector unit: extracts actual failure modes from all historical data in the experimental group and determines a failure mode vector, wherein the failure mode vector includes all actual failure modes that appear in all historical data in the experimental group.

4. The organic electroluminescent device life prediction system according to claim 1, characterized in that: The prediction module comprises: The first failure mode unit: based on the cluster vector of each cluster in the termination cluster, the first failure mode prediction is performed on the feature vector of each historical data in the verification group; Reasonableness unit: based on the actual failure mode in all historical data in the verification group and the consistency of the first failure mode, determine the rationality of the actual failure mode of each cluster group in the termination cluster group; A second failure mode unit: determining an actual failure mode of the cluster to which the target organic electroluminescent device belongs, as a second failure mode of the target organic electroluminescent device; Life prediction unit: predicts the remaining life of the target organic electroluminescent device based on the operation vector of the target organic electroluminescent device and the rationality of the actual failure mode of the belonging cluster group; Early warning unit: based on the remaining life and the second failure mode of the target organic electroluminescent device, sends early warning information to the user.

5. The organic electroluminescent device life prediction system according to claim 4, characterized in that: The life prediction unit comprises: The first extraction subunit is used to extract the start time of use in the historical data of each organic electroluminescent device, and determine the fitting life of each cluster group based on the start time of use and the actual failure time of all historical data of each cluster group in the terminated cluster group; A second extraction subunit: extracting the current usage time from the real-time operation data of the target organic electroluminescent device; Calculation subunit: Calculate the remaining life of the target organic electroluminescent device based on the current usage time of the target organic electroluminescent device, the fitted life of the cluster group to which the target organic electroluminescent device belongs, the score value of the cluster group to which the target organic electroluminescent device belongs, and the rationality of the actual failure mode of the cluster group to which the target organic electroluminescent device belongs.

6. The organic electroluminescent device life prediction system according to claim 5, characterized in that: The computing subunit comprises: in, represents the remaining life of the target organic electroluminescent device, represents the fitted lifetime of the target organic electroluminescent device to which the cluster belongs, Indicates the current usage time of the target organic electroluminescent device, represents the score value of the cluster to which the target organic electroluminescent device belongs, represents the life attenuation coefficient, represents the reasonableness adjustment coefficient of the actual failure mode, Indicates the actual failure mode rationality of the target organic electroluminescent device's belonging cluster family.

Images