Search

CN-122020385-A - Lamp strip service life prediction method and system based on multi-source data fusion

CN122020385ACN 122020385 ACN122020385 ACN 122020385ACN-122020385-A

Abstract

The invention relates to the technical field of lamp strip life prediction, in particular to a lamp strip life prediction method and system based on multi-source data fusion, comprising the following steps: and obtaining a plurality of test lamp strip sections based on the lamp strip to be predicted, constructing a lamp strip data fusion model, carrying out life test on the plurality of test lamp strip sections by utilizing a lamp strip monitoring index set to obtain a target life matching database, carrying out life prediction on the lamp strip to be predicted according to the current monitoring time, the target life matching database and the lamp strip data fusion model to obtain a current predicted lamp strip life value and a historical predicted lamp strip life value, and carrying out historical environment correction on the current predicted lamp strip life value by utilizing the historical predicted lamp strip life value to obtain a target predicted lamp strip life value. The invention can improve the stability and environmental adaptability of the life prediction of the lamp strip and reduce the prediction deviation caused by environmental mutation.

Inventors

  • LUO HAIHUA

Assignees

  • 惠州市路森照明有限公司

Dates

Publication Date
20260512
Application Date
20260203

Claims (10)

  1. 1. A lamp strip life prediction method based on multi-source data fusion, the method comprising: Determining a lamp band to be predicted and a lamp band monitoring index set, and acquiring a plurality of test lamp band segments based on the lamp band to be predicted, wherein the lamp band monitoring index set comprises a plurality of lamp band monitoring indexes; constructing a lamp strip data fusion model, wherein the lamp strip data fusion model comprises a data monitoring unit, a characteristic statistics unit and a data storage unit; Performing life test on the plurality of test lamp strip sections by using the lamp strip monitoring index set to obtain a target life matching database, wherein the target life matching database comprises a plurality of target life matching data sets, and the target life matching data sets are in one-to-one correspondence with the test lamp strip sections; Receiving a life prediction instruction, determining a current monitoring time based on the life prediction instruction, and predicting the life of the lamp strip to be predicted according to the current monitoring time, a target life matching database and a lamp strip data fusion model to obtain a current predicted lamp strip life value and a historical predicted lamp strip life value; carrying out historical environment correction on the current predicted lamp strip life value by utilizing the historical predicted lamp strip life value to obtain a target predicted lamp strip life value; And completing the lamp band life prediction based on multi-source data fusion based on the target predicted lamp band life value.
  2. 2. The method for predicting life of a lamp strip based on multi-source data fusion of claim 1, wherein the performing life testing on the plurality of test lamp strip segments using the lamp strip monitoring index set to obtain the target life matching database comprises: The following is performed for each of the plurality of test light strip segments: determining a test lamp strip working environment of the test lamp strip section, and constructing a test lamp strip environment vector according to the test lamp strip working environment; Carrying out life data analysis on the test lamp strip section according to the lamp strip monitoring index set to obtain a lamp strip life matching data set; marking the lamp strip service life matching data set by using the test lamp strip environment vector to obtain a target service life matching data set; And summarizing the target life matching data set corresponding to each test lamp band segment to obtain a target life matching database.
  3. 3. The method for predicting life of a lamp strip based on multi-source data fusion of claim 2, wherein the analyzing life data of the test lamp strip segment according to the lamp strip monitoring index set to obtain a lamp strip life matching data set comprises: performing data monitoring on the test lamp strip segments based on the lamp strip monitoring index set to obtain a plurality of test lamp strip parameter sequences, wherein the test lamp strip parameter sequences correspond to the lamp strip monitoring indexes one by one; Performing feature analysis according to the plurality of test lamp band parameter sequences to obtain a plurality of test lamp band operation feature groups, wherein the test lamp band operation feature groups are in one-to-one correspondence with the test lamp band parameter sequences; Detecting the service life of the test lamp strip section to obtain a test lamp strip service life value; performing data matching on a plurality of test lamp band operation characteristic groups by using the test lamp band life values to obtain lamp band life matching data; If the service life value of the test lamp strip is not less than the preset service life threshold value of the lamp strip, returning to the step of monitoring the data of the test lamp strip section based on the lamp strip monitoring index set until the service life value of the test lamp strip is less than the service life threshold value of the lamp strip; and if the service life value of the tested lamp strip is smaller than the service life threshold value of the lamp strip, summarizing the service life matching data of the lamp strip to obtain a service life matching data set of the lamp strip.
  4. 4. The method for predicting life of a lamp strip based on multi-source data fusion of claim 3, wherein the performing feature analysis according to a plurality of test lamp strip parameter sequences to obtain a plurality of test lamp strip operation feature sets comprises: The following is performed for each of the plurality of test strip parameter sequences: Carrying out mathematical statistical feature analysis on the test lamp strip parameter sequence to obtain a test short-term lamp strip operation feature set; Determining a historical light strip parameter total sequence based on the test light strip parameter sequence; Supplementing the test lamp strip parameter sequence to the history lamp strip parameter total sequence to obtain the test lamp strip parameter total sequence; carrying out mathematical statistical feature analysis on the total sequence of the parameters of the test lamp strip to obtain a test long-term lamp strip operation feature set; Combining the test short-term lamp band operation characteristic set and the test long-term lamp band operation characteristic set to obtain a test lamp band operation characteristic set; Summarizing the test lamp band operation characteristic groups corresponding to each test lamp band parameter sequence to obtain a plurality of test lamp band operation characteristic groups.
  5. 5. The method for predicting the lifetime of a lamp strip based on multi-source data fusion according to claim 4, wherein predicting the lifetime of the lamp strip to be predicted according to the current monitoring time, the target lifetime matching database and the lamp strip data fusion model to obtain the current predicted lifetime value and the historical predicted lifetime value of the lamp strip comprises: Determining a target lamp strip environment of a lamp strip to be predicted; Based on a data monitoring unit and a lamp band monitoring index set in the lamp band data fusion model, carrying out data monitoring on a lamp band to be predicted to obtain a plurality of target lamp band parameter sequences, wherein the target lamp band parameter sequences in the plurality of target lamp band parameter sequences are in one-to-one correspondence with the lamp band monitoring indexes; Carrying out feature statistics on the multiple target lamp band parameter sequences by using a feature statistics unit to obtain multiple target short-term operation feature groups; Reading a plurality of historical monitoring parameter total sequences and historical predicted lamp strip life values from a data storage unit based on the current monitoring moment; Correspondingly supplementing a plurality of historical monitoring parameter total sequences by a plurality of target lamp band parameter sequences to obtain a plurality of current monitoring parameter total sequences; carrying out feature statistics on the total sequence of the plurality of current monitoring parameters by utilizing a feature statistics unit to obtain a plurality of target long-term operation feature groups; combining the plurality of target short-term operation characteristic groups and the plurality of target long-term operation characteristic groups to obtain a plurality of current lamp band operation characteristic groups; Acquiring environmental parameters of the target lamp strip environment based on the current monitoring moment to obtain a current lamp strip environment vector; And calculating a current predicted lamp band life value based on the plurality of current lamp band operation characteristic groups, the current lamp band environment vector and the target life matching database.
  6. 6. The method for predicting lamp strip life based on multi-source data fusion of claim 5, wherein calculating the current predicted lamp strip life value based on the plurality of current lamp strip operational feature sets, the current lamp strip environment vector, and the target life matching database comprises: the following is performed for each target lifetime matching dataset in the target lifetime matching database: Obtaining a target lamp strip environment vector corresponding to the target life matching data set; Sequentially extracting target life matching data in a target life matching data set, and recording the extracted target life matching data as comparison life matching data; Confirming a comparison lamp strip service life value and a plurality of comparison lamp strip operation characteristic groups in the comparison service life matching data; Calculating comparison feature similarity according to the current lamp band operation feature groups and the comparison lamp band operation feature groups; summarizing the contrast characteristic similarity corresponding to each contrast life matching data to obtain a contrast characteristic similarity set; identifying optimal feature similarity in the comparison feature similarity set, and marking a comparison lamp strip life value in target life matching data corresponding to the optimal feature similarity as an optimal lamp strip life value; Combining the optimal lamp band life value and the target lamp band environment vector to obtain optimal lamp band fusion data; summarizing the optimal lamp band fusion data corresponding to each target life matching data set to obtain an optimal lamp band fusion data set; And carrying out weighted calculation according to the optimal lamp band fusion data set and the current lamp band environment vector to obtain the current predicted lamp band life value.
  7. 7. The method for predicting lamp strip life based on multi-source data fusion of claim 6, wherein calculating a comparison feature similarity from a plurality of current lamp strip operating feature sets and a plurality of comparison lamp strip operating feature sets comprises: constructing a current lamp band operation feature matrix based on the plurality of current lamp band operation feature sets, and constructing a comparison lamp band operation feature matrix based on the plurality of comparison lamp band operation feature sets; Calculating contrast characteristic similarity according to the current lamp band operation characteristic matrix and the contrast lamp band operation characteristic matrix, wherein the contrast characteristic similarity is expressed as: , Wherein, the The degree of similarity of the contrast features is indicated, Representing the matrix number of the current lamp band operation feature matrix or the matrix number of the comparison lamp band operation feature matrix, Representing the number of matrix columns of the current lamp band operation feature matrix or the number of matrix columns of the comparison lamp band operation feature matrix, An exponential function that represents the base of the natural constant, The representation takes the absolute value of the value, Reference lamp strip run feature matrix Line 1 The matrix elements of the columns are arranged such that, Representing the first of the current lamp band run feature matrices Line 1 Matrix elements of columns.
  8. 8. The method for predicting lamp life based on multi-source data fusion of claim 7, wherein the performing historical environmental correction on the current predicted lamp life value using the historical predicted lamp life value to obtain the target predicted lamp life value comprises: Acquiring a historical average lamp band environment vector; calculating a historical continuous lamp band life value based on a plurality of current lamp band operation feature groups, historical average lamp band environment vectors and a target life matching database; acquiring a historical monitoring total time length and a current monitoring time length; And calculating a target predicted lamp band life value according to the historical monitoring total time length, the current monitoring time length, the historical predicted lamp band life value, the historical continuous lamp band life value and the current predicted lamp band life value.
  9. 9. The multi-source data fusion based lamp strip life prediction method of claim 8, wherein the target predicted lamp strip life value is expressed as: , Wherein, the Representing a target predicted lamp strip life value, Representing a historical predicted lamp strip life value, Indicating the total duration of the historical monitoring, Indicating the current duration of the monitoring, Indicating a historical extended lamp strip life value, Indicating the current predicted lamp strip life value.
  10. 10. A lamp strip life prediction system based on multi-source data fusion, the system comprising: The system comprises a fusion model construction module, a light band data fusion module and a test module, wherein the fusion model construction module is used for determining a light band to be predicted and a light band monitoring index set, acquiring a plurality of test light band segments based on the light band to be predicted, the light band monitoring index set comprises a plurality of light band monitoring indexes, and constructing the light band data fusion model, and the light band data fusion model comprises a data monitoring unit, a characteristic statistics unit and a data storage unit; the system comprises a matching data acquisition module, a target life matching database and a test module, wherein the matching data acquisition module is used for carrying out life test on a plurality of test lamp strip sections by utilizing a lamp strip monitoring index set to obtain the target life matching database, the target life matching database comprises a plurality of target life matching data sets, and the target life matching data sets are in one-to-one correspondence with the test lamp strip sections; The current life prediction module is used for receiving a life prediction instruction, determining a current monitoring time based on the life prediction instruction, and predicting the life of the lamp strip to be predicted according to the current monitoring time, the target life matching database and the lamp strip data fusion model to obtain a current predicted lamp strip life value and a historical predicted lamp strip life value; and the historical environment correction module is used for carrying out historical environment correction on the current predicted lamp strip life value by utilizing the historical predicted lamp strip life value to obtain the target predicted lamp strip life value.

Description

Lamp strip service life prediction method and system based on multi-source data fusion Technical Field The invention relates to the technical field of lamp strip life prediction, in particular to a lamp strip life prediction method and system based on multi-source data fusion. Background The reliability of the lamp strip serving as a key component in an industrial illumination and automation system directly influences the stability and safety of a production environment, the accurate life prediction can realize preventive maintenance, the illumination interruption or production loss caused by the sudden failure of the lamp strip is avoided, and particularly in a continuous operation factory or precision manufacturing scene, the life prediction technology plays an important role in guaranteeing the seamless operation of the system and improving the efficiency of the whole equipment. Conventional lamp strip life prediction methods are typically based on static threshold decisions for a single operating parameter. The method has the defects that the prediction result is easy to be interfered by transient fluctuation and cannot adapt to long-term performance evolution due to the fact that the method ignores the dynamic influence of multiple factors such as temperature, humidity and the like in an actual working environment, and the prediction accuracy is low. Disclosure of Invention The invention provides a lamp strip life prediction method based on multi-source data fusion and a computer readable storage medium, which mainly aim to improve the stability and environmental adaptability of lamp strip life prediction and reduce prediction deviation caused by environmental mutation. In order to achieve the above object, the present invention provides a lamp strip life prediction method based on multi-source data fusion, including: Determining a lamp band to be predicted and a lamp band monitoring index set, and acquiring a plurality of test lamp band segments based on the lamp band to be predicted, wherein the lamp band monitoring index set comprises a plurality of lamp band monitoring indexes; constructing a lamp strip data fusion model, wherein the lamp strip data fusion model comprises a data monitoring unit, a characteristic statistics unit and a data storage unit; Performing life test on the plurality of test lamp strip sections by using the lamp strip monitoring index set to obtain a target life matching database, wherein the target life matching database comprises a plurality of target life matching data sets, and the target life matching data sets are in one-to-one correspondence with the test lamp strip sections; Receiving a life prediction instruction, determining a current monitoring time based on the life prediction instruction, and predicting the life of the lamp strip to be predicted according to the current monitoring time, a target life matching database and a lamp strip data fusion model to obtain a current predicted lamp strip life value and a historical predicted lamp strip life value; carrying out historical environment correction on the current predicted lamp strip life value by utilizing the historical predicted lamp strip life value to obtain a target predicted lamp strip life value; And completing the lamp band life prediction based on multi-source data fusion based on the target predicted lamp band life value. Optionally, the performing life test on the plurality of test light band segments by using the light band monitoring index set to obtain a target life matching database includes: The following is performed for each of the plurality of test light strip segments: determining a test lamp strip working environment of the test lamp strip section, and constructing a test lamp strip environment vector according to the test lamp strip working environment; Carrying out life data analysis on the test lamp strip section according to the lamp strip monitoring index set to obtain a lamp strip life matching data set; marking the lamp strip service life matching data set by using the test lamp strip environment vector to obtain a target service life matching data set; And summarizing the target life matching data set corresponding to each test lamp band segment to obtain a target life matching database. Optionally, the analyzing the life data of the test lamp strip section according to the lamp strip monitoring index set to obtain a lamp strip life matching data set includes: performing data monitoring on the test lamp strip segments based on the lamp strip monitoring index set to obtain a plurality of test lamp strip parameter sequences, wherein the test lamp strip parameter sequences correspond to the lamp strip monitoring indexes one by one; Performing feature analysis according to the plurality of test lamp band parameter sequences to obtain a plurality of test lamp band operation feature groups, wherein the test lamp band operation feature groups are in one-to-one correspondence with the test l