Intelligent street lamp life prediction method, system, device and medium

Abstract

The application discloses a kind of intelligent street lamp life prediction method, system, device and medium, the method includes: judging the correlation of street lamp independent variable factor and light decay state and failure state, select the street lamp independent variable factor with correlation and form independent variable factor sample;Each independent variable factor sample is processed to reduce dimension, and training sample is obtained;The neural network model about light decay state and the neural network model of failure state are constructed, and the neural network model is trained using training sample;The independent variable factor to be predicted is input to neural network model, and the life prediction result of intelligent street lamp is acquired.The application constructs neural network model to the street lamp life prediction using the independent variable factor with high correlation, by the correlation of the independent variable factor with high correlation is handled, the data screened out is processed to reduce dimension, so that training data has good distinguishability in low-dimensional space, prediction rate is fast and prediction is more accurate.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data analysis, and in particular to a smart street lamp life prediction method, system, device and medium. BACKGROUND

[0002] With the development of LED street lamp technology, more and more cities begin to use LED street lamps to replace sodium light. The development of cities promotes the increase of street lamp types and the expansion of application scope. Not only on public roads, streets, tunnels, subways, parks and other urban areas need street lamps for lighting. And the wide range of street lamps needs LED street lamp operation analysis system to analyze the operation of LED street lamps. When a city or an enterprise decides to use a certain brand of LED street lamp, it needs to consider various factors to help make a choice. Establishing and improving the LED street lamp operation analysis system has become a problem that must be solved.

[0003] The existing LED street lamp is basically equipped with an adapter and a communication line, which can realize communication connection with the management center. Through the management system of the management center, the control of the LED street lamp can be realized. The service life of the LED street lamp on the market is usually 25000-50000 hours, but it is found in the use process that the service life of the LED street lamp is often lower than the marked service life, which does not have reliability, and the functions configured by the background management center are relatively single, which can only realize remote control start, and cannot predict the potential failure and life of the street lamp. Once the street lamp fails, it cannot be repaired in time, which easily affects the traffic conditions nearby. SUMMARY

[0004] The purpose of the present application is to provide a smart street lamp life prediction method, system, device and medium to solve one or more technical problems existing in the prior art, and at least to provide a beneficial choice or create conditions.

[0005] In a first aspect, a smart street lamp life prediction method is provided, comprising:

[0006] determining the correlation of the street lamp independent variable factor with the light decay state and the failure state, selecting the street lamp independent variable factors with correlation to form an independent variable factor sample;

[0007] performing dimension reduction processing on each independent variable factor sample to obtain a training sample;

[0008] constructing a neural network model about the light decay state and a neural network model about the failure state, and training the neural network model using the training sample;

[0009] inputting the to-be-predicted independent variable factor into the neural network model to obtain a smart street lamp life prediction result.

[0010] In one embodiment, determining the correlation between the street light independent variable factors and the light decay state and fault state, and selecting the street light independent variable factors with correlation to form an independent variable factor sample, specifically includes the following steps:

[0011] Pearson correlation calculations were performed on each independent variable factor with the light decay state parameter and the fault state parameter to obtain the correlation coefficient between the independent variable factor and the light decay state and the fault state.

[0012] Based on the correlation strength and correlation direction reflected by the correlation coefficient, a sample set of independent variable factors is formed by selecting several independent variable factors with correlation coefficients closest to 1 or -1 in equal proportions.

[0013] In one embodiment, the street light independent variable factors include one or more of the following: input electrical signal factor, temperature and humidity factor, operating duration factor, lighting status factor, and climate factor.

[0014] In one embodiment, the dimensionality reduction of the independent variable factor samples to obtain training samples specifically includes the following steps:

[0015] Arrange the data of the independent variable factor samples into an m×n matrix X, and perform zero-mean normalization on each row of matrix X;

[0016] Calculate the covariance matrix C of matrix X. The formula for calculating the covariance matrix C is:

[0017]

[0018] In the formula, m is the number of rows in matrix X, and X T This is the transpose of matrix X;

[0019] Calculate the eigenvalues ​​and corresponding eigenvectors of the covariance matrix C. Arrange the eigenvectors into several row matrices according to the size of their corresponding eigenvalues. Take the first K row matrices to form matrix P. Multiply matrix P by matrix X to obtain matrix Y as the training sample.

[0020] In one embodiment, the construction of the neural network model for the light decay state and the neural network model for the fault state, and the training of the neural network model using training samples, specifically includes the following steps:

[0021] A neural network model is constructed based on the support vector machine regression algorithm. The neural network model is as follows:

[0022]

[0023] In the formula, and They are Lagrange multipliers a i and The unique solution, K(x) i x) is the SVM kernel function. σ is the width of the Cauchy kernel function;

[0024] Using the Cauchy kernel width σ and the penalty factor c as independent variables, the minimum value of MAE is calculated within a preset range. The value of the Cauchy kernel width σ when the MAE is minimum is taken to reconstruct two neural network models for optical decay state or fault state.

[0025] Where MAE represents the mean absolute error.

[0026] In one embodiment, calculating the minimum value of MAE within a preset range specifically includes the following steps:

[0027] Continuously extract negatively correlated data from the training samples and input them into the neural network model. Adjust the parameters of the neural network model until the MAE reaches the preset threshold range.

[0028] Continuously extract positively correlated data from the training samples and input them into the neural network model. Adjust the parameters of the neural network model until the MAE reaches the preset threshold range.

[0029] The positive and negative correlation data from the training samples are continuously and alternately extracted and input into the neural network model. The parameters of the neural network model are adjusted until the MAE reaches the preset threshold range.

[0030] In one embodiment, the step of inputting the independent variable factor to be predicted into the neural network model to obtain the lifespan prediction result of the smart street light specifically includes the following steps:

[0031] The street light independent variable factors related to light decay to be predicted are input into a neural network model about the light decay state to obtain the prediction results of the light decay state.

[0032] The independent variable factors of the streetlights related to the fault to be predicted are input into the neural network model about the fault state to obtain the prediction results of the fault state.

[0033] Based on the prediction results of light decay state and fault state, a fuzzy evaluation is made to obtain the prediction result of street lamp life.

[0034] Secondly, a smart street light lifespan prediction system is provided, including:

[0035] The filtering module is used to determine the correlation between the independent variable factors of streetlights and the light decay state and fault state, and select the streetlight independent variable factors with correlation to form a sample of independent variable factors.

[0036] The dimensionality reduction module is used to reduce the dimensionality of the independent variable factor samples to obtain training samples;

[0037] The training module is used to build neural network models of light decay states and fault states, and to train the neural network models using training samples.

[0038] The output module is used to input the independent variable factors to be predicted into the neural network model to obtain the life prediction results of smart streetlights.

[0039] Thirdly, a computer device is provided, comprising:

[0040] Memory, which stores computer programs;

[0041] A processor that, when executing the computer program, implements the smart street light lifespan prediction method as described in the first aspect.

[0042] Fourthly, a computer storage medium is provided, on which a computer program is stored, wherein the computer program, when executed by a processor, implements the smart street light lifespan prediction method as described in the first aspect.

[0043] The beneficial effects of this invention are as follows: a neural network model is constructed using highly correlated independent variable factors to predict the lifespan of streetlights. By processing the highly correlated independent variable factors, the selected data is subjected to dimensionality reduction, which makes the training data have very good discriminative power in the low-dimensional space. This embodiment requires less training data to be stored and all of them have a high correlation with the light decay state or fault state, resulting in a faster prediction speed and more accurate prediction. Attached Figure Description

[0044] Figure 1 This is a flowchart illustrating a method for predicting the lifespan of a smart street light according to one embodiment.

[0045] Figure 2 yes Figure 1 Flowchart of Example S100.

[0046] Figure 3 yes Figure 1 Flowchart of Example S200.

[0047] Figure 4 yes Figure 1 Flowchart of Example S300.

[0048] Figure 5 yes Figure 4 The flowchart for calculating the minimum MAE value in the embodiment.

[0049] Figure 6 yes Figure 1 Flowchart of Example S400.

[0050] Figure 7 This is a structural block diagram of a smart street light life prediction system according to one embodiment.

[0051] Figure 8 This is an internal structural diagram of a computer device according to an embodiment. Detailed Implementation

[0052] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the present invention will be further described below in conjunction with the embodiments and accompanying drawings.

[0053] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numerals in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this disclosure. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this disclosure as detailed in the appended claims.

[0054] According to a first aspect of the present invention, a method for predicting the lifespan of a smart street light is provided.

[0055] See Figure 1 , Figure 1 This is a flowchart illustrating a method for predicting the lifespan of a smart street light according to one embodiment. Figure 1 As shown, the method includes the following steps:

[0056] Step S100. Determine the correlation between the street light independent variable factors and the light decay state and fault state, and select the street light independent variable factors with correlation to form a sample of independent variable factors.

[0057] Step S200. Perform dimensionality reduction on the independent variable factor samples to obtain training samples.

[0058] Step S300. Construct a neural network model for the light decay state and a neural network model for the fault state, and train the neural network model using training samples.

[0059] Step S400. Input the independent variable factors to be predicted into the neural network model to obtain the life prediction results of the smart street light.

[0060] This embodiment utilizes highly correlated independent variables to construct a neural network model for predicting street light lifespan. By filtering out highly correlated street light independent variable factors through correlation processing and dimensionality reduction, the training data exhibits excellent discriminative power in the low-dimensional space. This embodiment requires less training data storage, and all data are highly correlated with the light decay or fault state of street lights, resulting in faster prediction speed and more accurate predictions.

[0061] In this embodiment, the independent variable factors of the street light include input electrical signal factors, temperature and humidity factors, operating duration factors, lighting status factors, and climate factors. The input electrical signal factors may include input voltage, input current, the highest value of line voltage surge, and the number of surge impacts, etc. The lighting status factors may include brightness, flicker, and luminous flux, etc. The climate factors may include sandstorms, dense fog, PM10, rainfall, and salt spray data, etc.

[0062] See Figure 2 As a further improvement to this embodiment, step S100 includes the following steps:

[0063] Step S110. Perform Pearson correlation calculations on each independent variable factor with the light decay state parameter and the fault state parameter to obtain the correlation coefficient between the independent variable factor and the light decay state and the fault state.

[0064] Step S120. Based on the correlation strength and correlation direction reflected by the correlation coefficient, select several independent variable factors with correlation coefficients closest to 1 or -1 in equal proportions to form an independent variable factor sample set.

[0065] In Pearson correlation calculations, the larger the absolute value of the correlation coefficient, the stronger the correlation; the closer the correlation coefficient is to 1 or -1, the stronger the correlation; and the closer the correlation coefficient is to 0, the weaker the correlation. In this embodiment, several positively correlated independent variables and negatively correlated independent variables are selected in equal proportions to ensure a good balance in the correlation tendency of the independent variable factor sample set.

[0066] See Figure 3 As a further improvement to this embodiment, step S200 includes the following steps:

[0067] Step S210. Arrange the data of the independent variable factor sample into an m×n matrix X, and perform zero-mean normalization on each row of matrix X.

[0068] In this embodiment, the matrix X is reduced to K dimensions. The goal is to select K orthonormal bases so that after the original data is transformed onto these bases, the covariance between each pair of variables is 0, while the variance of the variables is as large as possible (under the constraint of orthogonality, the largest K variances are selected).

[0069] Step S220. Calculate the covariance matrix C of matrix X. The formula for calculating the covariance matrix C is:

[0070]

[0071] In the formula, m is the number of rows in matrix X, and X T Let X be the transpose of matrix X.

[0072] Step S230. Calculate the eigenvalues ​​and corresponding eigenvectors of the covariance matrix C. Arrange the eigenvectors into several row matrices according to the size of their corresponding eigenvalues. Take the first K row matrices to form matrix P. Multiply matrix P by matrix X to obtain matrix Y as the training sample.

[0073] The covariance matrix C obtained through step S220 above is a symmetric matrix. Matrix P is a matrix formed by arranging the eigenvectors of the covariance matrix C in rows after normalization, where each row is an eigenvector of matrix C. Multiplying the matrix formed by the first K rows of P by the original data matrix X yields the required dimensionality-reduced data matrix Y.

[0074] See Figure 4 As a further improvement to this embodiment, step S300 includes the following steps:

[0075] Step S310. Construct a neural network model based on the support vector machine regression algorithm.

[0076] The neural network model is as follows:

[0077]

[0078] In the formula, and They are Lagrange multipliers a i and The unique solution, K(x) i x) is the SVM kernel function. σ is the width of the Cauchy kernel function.

[0079] Step S320. Using the Cauchy kernel function width σ and the penalty factor c as independent variables, calculate the minimum value of MAE within a preset range, and reconstruct the two neural network models for the light decay state or the fault state by taking the value of the Cauchy kernel function width σ when the MAE is minimum.

[0080] Where MAE represents the mean absolute error.

[0081] See Figure 5 More specifically, calculating the minimum value of MAE within the preset range includes the following steps:

[0082] Step S321. Continuously extract negatively correlated data from the training samples and input them into the neural network model. Adjust the parameters of the neural network model until the MAE reaches the preset threshold range.

[0083] Step S322. Continuously extract positively correlated data from the training samples and input them into the neural network model. Adjust the parameters of the neural network model until the MAE reaches the preset threshold range.

[0084] Step S323. Continuously and alternately extract positively correlated and negatively correlated data from the training samples and input them into the neural network model, and adjust the parameters of the neural network model until the MAE reaches the preset threshold range.

[0085] This embodiment uses three methods of extracting sample data to train the neural network model. Each time data is extracted from the training samples and input into the neural network model, a MAE operation is performed. Before the MAE reaches the preset threshold range, the Cauchy kernel function width σ and the penalty factor c are adjusted before the next data extraction is performed. In the limited extraction process, the neural network model is adapted to two sample data with opposite correlations.

[0086] See Figure 6 As a further improvement to this embodiment, step S400 includes the following steps:

[0087] Step S410. Input the street lamp independent variable factors related to light decay to be predicted into the neural network model about the light decay state, and obtain the prediction result of the light decay state.

[0088] Step S420. Input the street light independent variable factors related to the fault to be predicted into the neural network model about the fault state, and obtain the prediction result of the fault state.

[0089] Step S430. Make a fuzzy evaluation based on the prediction results of light decay state and fault state to obtain the prediction results of street lamp life.

[0090] In this embodiment, light decay state prediction data and fault state prediction data obtained by neural network model are selected, corresponding weights are set, an evaluation matrix is ​​constructed using the analytic hierarchy process, a weighted evaluation function is selected for calculation, and the resulting fuzzy evaluation result is used as the prediction result of the street lamp life.

[0091] According to a second aspect of the present invention, a smart street light lifespan prediction system is provided.

[0092] See Figure 7 , Figure 7 This is a structural block diagram illustrating a smart street light lifespan prediction system according to one embodiment. Figure 7 As shown, the system includes:

[0093] The screening module 710 is used to determine the correlation between the independent variable factors of streetlights and the light decay state and fault state, and to select the streetlight independent variable factors with correlation to form a sample of independent variable factors.

[0094] Dimensionality reduction module 720 is used to reduce the dimensionality of the independent variable factor samples to obtain training samples;

[0095] Training module 730 is used to construct neural network models of light decay state and fault state, and to train the neural network models using training samples;

[0096] The output module 740 is used to input the independent variable factors to be predicted into the neural network model to obtain the life prediction results of the smart street light.

[0097] In one embodiment, determining the correlation between the street light independent variable factors and the light decay state and fault state, and selecting the street light independent variable factors with correlation to form an independent variable factor sample, specifically includes the following steps:

[0098] Pearson correlation calculations were performed on each independent variable factor with the light decay state parameter and the fault state parameter to obtain the correlation coefficient between the independent variable factor and the light decay state and the fault state.

[0099] Based on the correlation strength and correlation direction reflected by the correlation coefficient, a sample set of independent variable factors is formed by selecting several independent variable factors with correlation coefficients closest to 1 or -1 in equal proportions.

[0100] In one embodiment, the street light independent variable factors include one or more of the following: input electrical signal factor, temperature and humidity factor, operating duration factor, lighting factor, and climate factor.

[0101] In one embodiment, the dimensionality reduction of the independent variable factor samples to obtain training samples specifically includes the following steps:

[0102] Arrange the data of the independent variable factor samples into an m×n matrix X, and perform zero-mean normalization on each row of matrix X;

[0103] Calculate the covariance matrix C of matrix X. The formula for calculating the covariance matrix C is:

[0104]

[0105] In the formula, m is the number of rows in matrix X, and X T This is the transpose of matrix X;

[0106] Calculate the eigenvalues ​​and corresponding eigenvectors of the covariance matrix C. Arrange the eigenvectors into several row matrices according to the size of their corresponding eigenvalues. Take the first K row matrices to form matrix P. Multiply matrix P by matrix X to obtain matrix Y as the training sample.

[0107] In one embodiment, the construction of the neural network model for the light decay state and the neural network model for the fault state, and the training of the neural network model using training samples, specifically includes the following steps:

[0108] A neural network model is constructed based on the support vector machine regression algorithm. The neural network model is as follows:

[0109]

[0110] In the formula, and They are Lagrange multipliers a i and The unique solution, K(x) i x) is the SVM kernel function. σ is the width of the Cauchy kernel function;

[0111] Using the Cauchy kernel width σ and the penalty factor c as independent variables, the minimum value of MAE is calculated within a preset range. The value of the Cauchy kernel width σ when the MAE is minimum is taken to reconstruct two neural network models for optical decay state or fault state.

[0112] Where MAE represents the mean absolute error.

[0113] In one embodiment, calculating the minimum value of MAE within a preset range specifically includes the following steps:

[0114] Continuously extract negatively correlated data from the training samples and input them into the neural network model. Adjust the parameters of the neural network model until the MAE reaches the preset threshold range.

[0115] Continuously extract positively correlated data from the training samples and input them into the neural network model. Adjust the parameters of the neural network model until the MAE reaches the preset threshold range.

[0116] The positive and negative correlation data from the training samples are continuously and alternately extracted and input into the neural network model. The parameters of the neural network model are adjusted until the MAE reaches the preset threshold range.

[0117] In one embodiment, the step of inputting the independent variable factor to be predicted into the neural network model to obtain the lifespan prediction result of the smart street light specifically includes the following steps:

[0118] The street light independent variable factors related to light decay to be predicted are input into a neural network model about the light decay state to obtain the prediction results of the light decay state.

[0119] The independent variable factors of the streetlights related to the fault to be predicted are input into the neural network model about the fault state to obtain the prediction results of the fault state.

[0120] Based on the prediction results of light decay state and fault state, a fuzzy evaluation is made to obtain the prediction result of street lamp life.

[0121] The smart street light life prediction system implements the smart street light life prediction method of the first aspect described above. For specific limitations of the smart street light life prediction system, please refer to the limitations of the smart street light life prediction method mentioned above, which will not be repeated here.

[0122] The modules in the aforementioned smart street light lifespan prediction system can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of the computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0123] According to a third aspect of the present invention, a computer device is provided.

[0124] See Figure 8 , Figure 8 This is a diagram illustrating the internal structure of a computer device according to one embodiment. Figure 8 As shown, the computer device includes a processor, memory, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. When the computer program is executed by the processor, it implements the smart street light lifespan prediction method described in the first aspect.

[0125] The components of the memory and processor are electrically connected directly or indirectly to enable data transmission or interaction. For example, these components can be electrically connected to each other via one or more communication buses or signal lines. The processor includes at least one software functional module, which can be stored in memory or embedded in the server's operating system (OS) in the form of software or firmware. The processor is used to execute the executable module stored in memory.

[0126] The memory can be Random Access Memory (RAM), Read Only Memory (ROM), Programmable Read-Only Memory (PROM), Erasable Programmable Read-Only Memory (EPROM), Electrically Erasable Programmable Read-Only Memory (EEPROM), etc. The memory is used to store programs and voice data; the processor executes the program after receiving the execution instruction.

[0127] The processor may be an integrated circuit chip with signal processing capabilities. The aforementioned processor can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application-Specific Integrated Circuit (ASIC), an On-Premises Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this invention. The general-purpose processor can be a microprocessor or any conventional processor.

[0128] The processor couples various input / output devices to itself and memory. In some embodiments, the processor and memory can be implemented on a single chip. In other instances, they can be implemented on separate chips.

[0129] Peripheral interfaces couple various input / output devices to the processor and memory. In some embodiments, the peripheral interface, processor, and memory can be implemented on a single chip. In other instances, they can be implemented on separate chips.

[0130] According to a fourth aspect of the present invention, a computer storage medium is also provided, wherein a computer program is stored in the computer storage medium. The computer storage medium may be a magnetic random access memory, a read-only memory, a programmable read-only memory, an erasable programmable read-only memory, an electrically erasable programmable read-only memory, a flash memory, a magnetic surface memory, an optical disc, or a read-only optical disc, etc.; it may also be various devices including one or any combination of the above-mentioned memories, such as mobile phones, computers, tablet devices, personal digital assistants, etc. When the computer program is executed by a processor, it implements the smart street light lifespan prediction method described in the first aspect.

[0131] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0132] In this document, the terms “comprising,” “including,” or any other variations thereof are intended to cover non-exclusive inclusion, which includes not only the elements listed but also other elements not expressly listed.

[0133] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for predicting the lifespan of a smart street light, characterized in that, include: Determine the correlation between the independent variable factors of streetlights and their light decay and fault states, and select the streetlight independent variable factors with correlation to form a sample of independent variable factors. Dimensionality reduction is performed on the independent variable factor samples to obtain training samples; A neural network model for light decay state and a neural network model for fault state are constructed. The neural network model is trained using training samples. The neural network model is constructed based on the support vector machine regression algorithm. The neural network model is as follows: In the formula, and They are Lagrange multipliers and The only solution For SVM kernel functions, , The width of the Cauchy kernel function; MAE represents the mean absolute error. ; With Cauchy kernel width Using the penalty factor c as the independent variable, calculate the minimum value of MAE within a preset range, and take the Cauchy kernel function width when MAE is minimized. The values ​​are used to reconstruct two neural network models regarding the light decay state or the fault state; The calculation of the minimum value of MAE within the preset range specifically includes the following steps: Continuously extract negatively correlated data from the training samples and input them into the neural network model. Adjust the parameters of the neural network model until the MAE reaches the preset threshold range. Continuously extract positively correlated data from the training samples and input them into the neural network model. Adjust the parameters of the neural network model until the MAE reaches the preset threshold range. The positive and negative correlation data from the training samples are continuously and alternately extracted and input into the neural network model. The parameters of the neural network model are adjusted until the MAE reaches the preset threshold range. The street light independent variable factors related to light decay to be predicted are input into a neural network model about the light decay state to obtain the prediction results of the light decay state. The independent variable factors of the streetlights related to the fault to be predicted are input into the neural network model about the fault state to obtain the prediction results of the fault state. Based on the prediction results of light decay state and fault state, the light decay state prediction data and fault state prediction data obtained by the neural network model are selected, corresponding weights are set, the analytic hierarchy process is used to construct the evaluation matrix, the weighted evaluation function is selected for calculation, and the resulting fuzzy evaluation result is used as the prediction result of the street lamp life.

2. The method for predicting the lifespan of smart streetlights according to claim 1, characterized in that, The process of determining the correlation between the independent variable factors of streetlights and their light decay and fault states, and selecting correlated independent variable factors to form a sample of independent variable factors, specifically includes the following steps: Pearson correlation calculations were performed on each independent variable factor with the light decay state parameter and the fault state parameter to obtain the correlation coefficient between the independent variable factor and the light decay state and the fault state. Based on the correlation strength and correlation direction reflected by the correlation coefficient, a sample set of independent variable factors is formed by selecting several independent variable factors with correlation coefficients closest to 1 or -1 in equal proportions.

3. The method for predicting the lifespan of smart streetlights according to claim 1, characterized in that, The independent variable factors of the street light include one or more of the following: input electrical signal factor, temperature and humidity factor, working duration factor, lighting status factor, and climate factor.

4. The method for predicting the lifespan of smart streetlights according to claim 1, characterized in that, The step of reducing the dimensionality of the independent variable factor samples to obtain training samples specifically includes the following steps: Arrange the data of the independent variable factor samples into a matrix with m rows and N columns. For the matrix Each row is zero-mean normalized; Calculate matrix covariance matrix covariance matrix The calculation formula is: In the formula, m is a matrix. the number of rows, For matrix transpose; Calculate the covariance matrix Given the eigenvalues ​​and corresponding eigenvectors, arrange the eigenvectors into several row matrices according to the magnitude of their corresponding eigenvalues. Take the first K row matrices to form matrix P, and multiply matrix P by matrix ... The resulting matrix Y is used as a training sample.

5. A smart street light lifespan prediction system, characterized in that, include: The filtering module is used to determine the correlation between the independent variable factors of streetlights and the light decay state and fault state, and select the streetlight independent variable factors with correlation to form a sample of independent variable factors. The dimensionality reduction module is used to reduce the dimensionality of the independent variable factor samples to obtain training samples; The training module is used to construct neural network models for light decay states and fault states. The neural network models are trained using training samples. Specifically, the neural network model is constructed based on a support vector machine regression algorithm. The neural network model is as follows: In the formula, and They are Lagrange multipliers and The only solution For SVM kernel functions, , The width of the Cauchy kernel function; MAE represents the mean absolute error. ; With Cauchy kernel width Using the penalty factor c as the independent variable, calculate the minimum value of MAE within a preset range, and take the Cauchy kernel function width when MAE is minimized. The values ​​are used to reconstruct two neural network models regarding the light decay state or the fault state; The calculation of the minimum value of MAE within the preset range specifically includes the following steps: Continuously extract negatively correlated data from the training samples and input them into the neural network model. Adjust the parameters of the neural network model until the MAE reaches the preset threshold range. Continuously extract positively correlated data from the training samples and input them into the neural network model. Adjust the parameters of the neural network model until the MAE reaches the preset threshold range. The positive and negative correlation data from the training samples are continuously and alternately extracted and input into the neural network model. The parameters of the neural network model are adjusted until the MAE reaches the preset threshold range. The output module is used to input the street lamp independent variable factors related to light decay to be predicted into the neural network model about the light decay state, and obtain the prediction result of the light decay state. The independent variable factors of the streetlights related to the fault to be predicted are input into the neural network model about the fault state to obtain the prediction results of the fault state. Based on the prediction results of light decay state and fault state, the light decay state prediction data and fault state prediction data obtained by the neural network model are selected, corresponding weights are set, the analytic hierarchy process is used to construct the evaluation matrix, the weighted evaluation function is selected for calculation, and the resulting fuzzy evaluation result is used as the prediction result of the street lamp life.

6. A computer device, characterized in that, include: Memory, which stores computer programs; A processor that, when executing the computer program, implements the smart street light lifespan prediction method as described in any one of claims 1-4.

7. A computer storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the smart street light life prediction method as described in any one of claims 1-4.

Images