A street lamp intelligent operation and maintenance optimization management platform based on big data analysis

Abstract

This invention discloses a smart street light operation and maintenance optimization management platform based on big data analysis, comprising: an acquisition module for acquiring target data of smart street lights within a preset time period; an extraction module for extracting features from the target data to obtain target features; an evaluation module for inputting the target features into a pre-trained big data analysis model for evaluation to determine the operation and maintenance evaluation value of the smart street light; a determination module for comparing the operation and maintenance evaluation value with a preset operation and maintenance evaluation threshold, and designating the smart street light as a smart street light to be optimized when the operation and maintenance evaluation value is less than or equal to the preset operation and maintenance evaluation threshold; and an optimization module for querying the target optimization strategy of the smart street light to be optimized and optimizing the smart street light to be optimized based on the target optimization strategy. This enhances the data processing and analysis capabilities during street light operation and maintenance, establishes a scientific evaluation system, and improves operation and maintenance efficiency based on precise optimization strategies.

Description

Technical Field

[0001] This invention relates to the field of intelligent energy-saving street lighting, and in particular to a smart operation and maintenance optimization management platform for street lights based on big data analysis. Background Technology

[0002] With the acceleration of urbanization, the scale of urban lighting systems is constantly expanding. As an important component of urban lighting systems, smart streetlights are gradually becoming more widespread and being applied in every corner of the city. Intelligent energy-saving streetlights not only possess traditional lighting functions but also integrate various sensors and communication devices, enabling intelligent lighting control, environmental monitoring, video surveillance, and other functions, providing strong support for the intelligent management of cities.

[0003] However, the operation and maintenance management of smart energy-saving streetlights still faces many challenges:

[0004] Intelligent energy-saving streetlights generate a large amount of target data during operation, such as light intensity, current and voltage, and fault alarm information. This data contains a wealth of information, but due to a lack of effective data processing and analysis methods, most of it is not fully utilized. Existing management methods often only provide simple storage and display of data, failing to delve into the underlying problems and thus hindering the timely detection of potential streetlight malfunctions and performance degradation trends.

[0005] In traditional street light operation and maintenance management, the evaluation of smart energy-saving street lights mainly relies on manual experience and simple indicator judgments, lacking a scientific and comprehensive evaluation system. This evaluation method is highly subjective and prone to inaccurate assessments, failing to accurately reflect the actual operating status and performance level of smart street lights. For example, judging whether a street light is working properly solely based on whether it is lit, while ignoring important indicators such as light intensity and energy consumption, may lead to the overlooking of some potential problems, affecting the lifespan of the street light and its lighting effect.

[0006] Due to a lack of effective data support and a scientific evaluation system, existing street light operation and maintenance management often relies on periodic inspections and reactive repairs. This approach not only consumes significant manpower, resources, and time but also fails to promptly detect and resolve street light malfunctions, leading to untimely repairs and impacting urban lighting quality and traffic safety. Furthermore, the lack of targeted optimization strategies for different types of street light malfunctions hinders precise operation and maintenance, further reducing efficiency.

[0007] In the operation and maintenance of streetlights, when problems are discovered, a uniform approach is often adopted, lacking personalized optimization strategies for different streetlights and different types of faults. This "one-size-fits-all" optimization method fails to fully consider the actual situation and operating environment of the streetlights, which may lead to poor optimization results or even cause new problems. For example, for some streetlight faults caused by environmental factors, simply replacing streetlight components without considering the impact of environmental factors may not fundamentally solve the problem and may even increase operation and maintenance costs.

[0008] In summary, existing smart street light operation and maintenance management methods suffer from insufficient data processing and analysis capabilities, a lack of scientific evaluation systems, low operation and maintenance efficiency, and inaccurate optimization strategies, failing to meet the needs of intelligent and efficient management of urban lighting systems. Therefore, there is an urgent need for a smart street light operation and maintenance optimization management platform based on big data analytics to improve the operation and maintenance management level of smart street lights and ensure the stable operation of urban lighting systems. Summary of the Invention

[0009] This invention aims to at least partially solve one of the technical problems in the aforementioned technologies. Therefore, the purpose of this invention is to propose a smart street light operation and maintenance optimization management platform based on big data analysis, which enhances data processing and analysis capabilities during street light operation and maintenance, establishes a scientific evaluation system, and improves operation and maintenance efficiency based on precise optimization strategies.

[0010] To achieve the above objectives, this invention proposes a smart street light operation and maintenance optimization management platform based on big data analysis, comprising:

[0011] The acquisition module is used to acquire target data of smart streetlights within a preset time period;

[0012] The extraction module is used to extract features from the target data to obtain the target features;

[0013] The evaluation module is used to input the target features into a pre-trained big data analysis model for evaluation, and to determine the operation and maintenance evaluation value of the smart street light;

[0014] The determination module is used to compare the operation and maintenance evaluation value with the preset operation and maintenance evaluation threshold, and when it is determined that the operation and maintenance evaluation value is less than or equal to the preset operation and maintenance evaluation threshold, the smart street light is identified as a smart street light to be optimized.

[0015] The optimization module is used to query the target optimization strategy for the smart street light to be optimized, and then optimize the smart street light based on the target optimization strategy.

[0016] Preferably, the acquisition module includes:

[0017] The first acquisition submodule is used to acquire the operating data of the smart streetlights within a preset time period to obtain the first data;

[0018] The second acquisition submodule is used to acquire monitoring image data of smart streetlights within a preset time period to obtain the second data;

[0019] The first determining submodule is used to use the first data and the second data as the target data for the smart street light.

[0020] Preferably, it further includes a preprocessing module, used to preprocess the target data before the extraction module extracts features from the target data to obtain target features, so as to obtain preprocessed target data.

[0021] Preferably, the preprocessing module includes:

[0022] The data cleaning submodule is used to clean the first data to obtain the cleaned first data.

[0023] The noise reduction submodule is used to perform image noise reduction on the second data to obtain the noise-reduced second data.

[0024] The second determining submodule is used to take the first data after data cleaning and the second data after noise reduction as the target data after preprocessing.

[0025] Preferably, the data cleaning submodule includes:

[0026] Segmentation unit, used for:

[0027] Randomly select one type of data from the first set of data as the target category data;

[0028] The data in the target category is evenly divided into several target category sub-data based on the time series order;

[0029] The first computing unit is used for:

[0030] Randomly select one target category sub-data as the first target category sub-data;

[0031] Calculate the anomaly evaluation value corresponding to the first target category sub-data;

[0032] The first determining unit is used for:

[0033] The abnormal evaluation value is compared with a preset evaluation threshold. When it is determined that the abnormal evaluation value is greater than or equal to the preset evaluation threshold, the first target category sub-data is used as the second target category sub-data.

[0034] Iterate through all the target category sub-data to obtain several second target category sub-data;

[0035] The second calculation unit is used for:

[0036] Randomly select a sub-data point of the second target category;

[0037] The local fluctuation index of each data point in the second target category sub-data is calculated based on the first preset algorithm;

[0038] The second determining unit is used for:

[0039] The local fluctuation index of each data point is compared with a preset fluctuation threshold, and the data points whose local fluctuation index is greater than or equal to the preset fluctuation threshold are regarded as abnormal data points.

[0040] Traverse all the sub-data of the second target category to obtain several outlier data points;

[0041] The data cleaning unit is used for:

[0042] Several abnormal data points are deleted to obtain the target category data after data cleaning;

[0043] Iterate through all the data types in the first dataset to obtain the cleaned first dataset.

[0044] Preferably, the first computing unit includes:

[0045] Get a sub-unit, which is used to arbitrarily select a target category sub-data as the first target category sub-data;

[0046] The first calculation subunit is used to calculate the mean of the data values ​​corresponding to all data points in the first target category subdata to obtain the first mean.

[0047] The second computational subunit is used for:

[0048] Obtain the target category sub-data adjacent to the first target category sub-data, calculate the mean of the data values ​​corresponding to all data points in the first target category sub-data and the target category sub-data adjacent to the first target category sub-data, and obtain the second mean;

[0049] The third computational subunit is used for:

[0050] Calculate the absolute value of the difference between the first mean and the second mean to obtain the first difference value;

[0051] Calculate the range of data points corresponding to all data points in the first target category sub-data and the target category sub-data adjacent to the first target category sub-data to obtain the second difference value;

[0052] A sub-unit is defined to determine the anomaly evaluation value of the first target category sub-data based on the first difference value and the second difference value.

[0053] Preferably, the noise reduction submodule includes:

[0054] Acquisition unit, used for:

[0055] Take any monitoring image of a smart street light from the second dataset and perform grayscale processing;

[0056] The monitoring images of the smart streetlights after grayscale processing are divided into several sub-images;

[0057] Several sub-images are clustered to obtain several sub-image sets;

[0058] Choose any sub-image set as the target sub-image set;

[0059] Select any one sub-image from the target sub-image set as the target sub-image;

[0060] Take any pixel in the target sub-image as the first pixel;

[0061] The third calculation unit is used to calculate the brightness value and blur value of the first pixel respectively;

[0062] The comparison unit is used to compare the brightness value of the first pixel with a preset brightness threshold and to compare the blur value of the first pixel with a preset blur threshold.

[0063] The third determining unit is used for:

[0064] When it is determined that the brightness value of the first pixel is less than the preset brightness threshold and the blur value of the first pixel is greater than the preset blur threshold, the first pixel is taken as the pixel to be denoised.

[0065] Traverse all pixels to obtain a number of pixels to be denoised;

[0066] Noise reduction unit, used for:

[0067] The filtering algorithm is used to denoise several pixels to be denoised, and the denoised target sub-image is obtained.

[0068] Based on the target sub-image and the denoised target sub-image, determine the denoising amplitude of each pixel to be denoised in the target sub-image;

[0069] Denoising is performed on other sub-images in the target sub-image set based on the denoising amplitude of each pixel to be denoised;

[0070] Traverse all sub-image sets to obtain the noise-reduced monitoring images of the smart streetlights;

[0071] The monitoring images of all smart streetlights in the second dataset are traversed to obtain the noise-reduced second dataset.

[0072] Preferably, the third computing unit includes:

[0073] The brightness calculation subunit is used for:

[0074] The target area is determined with the first pixel as the center and a preset distance as the radius;

[0075] Calculate the average grayscale value of each pixel in the target area, and use it as the brightness value of the first pixel;

[0076] The ambiguity calculation subunit is used for:

[0077] The sum of the squares of the grayscale differences between the first pixel and all other pixels in the target region is calculated to obtain the target sum value.

[0078] Obtain the total number of pixels in the target area excluding the first pixel, and get the target value;

[0079] The ratio of the target sum to the target value is used as the blur value of the first pixel.

[0080] Preferred methods for constructing big data analytics models include:

[0081] Obtain training datasets for big data analysis;

[0082] The neural network model is trained based on the big data analysis training dataset to obtain the initial big data analysis model;

[0083] Obtain a big data analysis test dataset;

[0084] The initial big data analysis model is tested using a big data analysis test dataset. When the test results are satisfactory, a well-trained big data analysis model is obtained.

[0085] Preferred, optimized modules include:

[0086] The query submodule is used to query the smart operation and maintenance optimization database based on the target characteristics of the smart street light to be optimized, and to determine the target optimization strategy for the smart street light to be optimized.

[0087] The optimization submodule is used to optimize the smart streetlights to be optimized based on the target optimization strategy.

[0088] This invention discloses a smart street light operation and maintenance optimization management platform based on big data analysis. It acquires target data of smart street lights within a preset time period, providing a rich data foundation for comprehensively understanding the operational status of street lights. Feature extraction is performed on this target data to extract key information from massive amounts of data, yielding target features. These target features are then input into a pre-trained big data analysis model for evaluation, determining the evaluation value of the smart street lights. This evaluation of target features through the big data analysis model avoids the subjectivity and limitations of traditional evaluation methods that rely primarily on human experience and simple indicator judgments. Utilizing the powerful analytical capabilities of the big data analysis model, potential faults and performance degradation trends of street lights can be detected in a timely manner. For example, by analyzing the energy consumption data of street lights, abnormal energy consumption can be detected in advance, allowing for timely intervention and preventing further deterioration. Comparing the evaluation value with a preset evaluation threshold accurately identifies smart street lights that need optimization. This precise identification method avoids the blindness of periodic inspections and post-event maintenance in traditional operation and maintenance methods, enabling timely detection of street light faults and improving the timeliness of operation and maintenance. Meanwhile, for the optimization strategy of the smart street light query target, personalized optimization measures can be taken according to the actual situation and fault type of different street lights, so as to achieve precise operation and maintenance and improve operation and maintenance efficiency.

[0089] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description and the accompanying drawings.

[0090] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0091] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0092] Figure 1 This is a block diagram of a smart street light operation and maintenance optimization management platform based on big data analysis according to an embodiment of the present invention;

[0093] Figure 2 This is a block diagram of an acquisition module according to an embodiment of the present invention;

[0094] Figure 3 This is a block diagram of a preprocessing module according to an embodiment of the present invention. Detailed Implementation

[0095] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0096] Example 1

[0097] like Figure 1 As shown, a smart operation and maintenance optimization management platform for streetlights based on big data analytics includes:

[0098] The acquisition module is used to acquire target data of smart streetlights within a preset time period;

[0099] The extraction module is used to extract features from the target data to obtain the target features;

[0100] The evaluation module is used to input the target features into a pre-trained big data analysis model for evaluation, and to determine the operation and maintenance evaluation value of the smart street light;

[0101] The determination module is used to compare the operation and maintenance evaluation value with the preset operation and maintenance evaluation threshold, and when it is determined that the operation and maintenance evaluation value is less than or equal to the preset operation and maintenance evaluation threshold, the smart street light is identified as a smart street light to be optimized.

[0102] The optimization module is used to query the target optimization strategy for the smart street light to be optimized, and then optimize the smart street light based on the target optimization strategy.

[0103] In this embodiment, the target data includes the operation data of the smart streetlights and the monitoring image data of the smart streetlights.

[0104] In this embodiment, the target features include electrical performance-related features, lighting effect-related features, fault and anomaly-related features, and environmental-related features.

[0105] In this embodiment, the big data analysis model is a pre-trained evaluation model.

[0106] The working principle and beneficial effects of the above technical solution are as follows: It acquires target data of smart streetlights within a preset time period, providing a rich data foundation for comprehensively understanding the operational status of streetlights; it extracts features from this target data, refining key information from massive amounts of data to obtain target features; it inputs these target features into a pre-trained big data analysis model for evaluation, determining the evaluation value of the smart streetlights; by evaluating target features through a big data analysis model, it avoids the subjectivity and limitations of traditional evaluation methods that mainly rely on human experience and simple indicator judgments; and it utilizes the powerful analytical capabilities of the big data analysis model to promptly detect potential faults and performance degradation trends in streetlights. For example, by analyzing the energy consumption data of streetlights, it is possible to detect abnormal energy consumption in advance, take timely measures to address the problem, and prevent further deterioration; by comparing the evaluation value with a preset evaluation threshold, it is possible to accurately identify smart streetlights that need optimization. This precise identification method avoids the blindness of periodic inspections and post-event maintenance in traditional operation and maintenance methods, enabling timely detection of streetlight faults and improving the timeliness of operation and maintenance. Meanwhile, for the optimization strategy of the smart street light query target, personalized optimization measures can be taken according to the actual situation and fault type of different street lights, so as to achieve precise operation and maintenance and improve operation and maintenance efficiency.

[0107] Example 2

[0108] like Figure 2 As shown, the acquisition module includes:

[0109] The first acquisition submodule is used to acquire the operating data of the smart streetlights within a preset time period to obtain the first data;

[0110] The second acquisition submodule is used to acquire monitoring image data of smart streetlights within a preset time period to obtain the second data;

[0111] The first determining submodule is used to use the first data and the second data as the target data for the smart street light.

[0112] In this embodiment, the operational data includes electrical operational data of the streetlights, lighting-related data, streetlight equipment status data, and environmental perception data surrounding the streetlights.

[0113] The beneficial effects of the above technical solution are as follows: through the collaborative work of the first acquisition submodule and the second acquisition submodule, comprehensive monitoring of the operation status of smart streetlights and the surrounding environment can be achieved, ensuring the normal operation of the streetlight system and public safety; the integrated smart streetlight target dataset provides rich data support for city managers, helping them to make more scientific and reasonable decisions and improve the efficiency and scientific nature of urban management; by comprehensively acquiring and analyzing the operation data and monitoring image data of smart streetlights, rich data support is provided for city managers, which helps to achieve more scientific and efficient urban management.

[0114] Example 3

[0115] It also includes a preprocessing module, which is used to preprocess the target data before the extraction module extracts features from the target data to obtain the target features, so as to obtain preprocessed target data.

[0116] Example 4

[0117] like Figure 3 As shown, the preprocessing module includes:

[0118] The data cleaning submodule is used to clean the first data to obtain the cleaned first data.

[0119] The noise reduction submodule is used to perform image noise reduction on the second data to obtain the noise-reduced second data.

[0120] The second determining submodule is used to take the first data after data cleaning and the second data after noise reduction as the target data after preprocessing.

[0121] The beneficial effects of the above technical solution are as follows: By cleaning and denoising the data, the quality of smart street light operation data and monitoring image data is effectively improved, providing more accurate and reliable data support for subsequent data analysis and decision-making; integrating data cleaning and denoising into the preprocessing module enables rapid and efficient processing of raw data, optimizing the data processing flow; the preprocessed target data is more accurate and complete, helping to improve the overall performance of the smart street light system and enhance the efficiency and scientific nature of urban management; through the collaborative work of various sub-modules in the preprocessing module, comprehensive and efficient preprocessing of raw data is achieved, providing a solid foundation for subsequent data analysis and decision-making.

[0122] Example 5

[0123] The data cleaning submodule includes:

[0124] Segmentation unit, used for:

[0125] Randomly select one type of data from the first set of data as the target category data;

[0126] The data in the target category is evenly divided into several target category sub-data based on the time series order;

[0127] The first computing unit is used for:

[0128] Randomly select one target category sub-data as the first target category sub-data;

[0129] Calculate the anomaly evaluation value corresponding to the first target category sub-data;

[0130] The first determining unit is used for:

[0131] The abnormal evaluation value is compared with a preset evaluation threshold. When it is determined that the abnormal evaluation value is greater than or equal to the preset evaluation threshold, the first target category sub-data is used as the second target category sub-data.

[0132] Iterate through all the target category sub-data to obtain several second target category sub-data;

[0133] The second calculation unit is used for:

[0134] Randomly select a sub-data point of the second target category;

[0135] The local fluctuation index of each data point in the second target category sub-data is calculated based on the first preset algorithm;

[0136] The second determining unit is used for:

[0137] The local fluctuation index of each data point is compared with a preset fluctuation threshold, and the data points whose local fluctuation index is greater than or equal to the preset fluctuation threshold are regarded as abnormal data points.

[0138] Traverse all the sub-data of the second target category to obtain several outlier data points;

[0139] The data cleaning unit is used for:

[0140] Several abnormal data points are deleted to obtain the target category data after data cleaning;

[0141] Iterate through all the data types in the first dataset to obtain the cleaned first dataset.

[0142] In this embodiment, the first preset algorithm includes:

[0143]

[0144] Among them, A b,c Y represents the local volatility index of the c-th data point in the b-th sub-data of the second target category; b,c Y represents the data value of the c-th data point in the b-th sub-data of the second target category; b,c,x d represents the data value of the x-th data point within the neighborhood of the c-th data point in the b-th second target category sub-data; b,c,x denoted as , representing the distance between the c-th data point in the b-th second target category sub-data and the x-th data point within the neighborhood of the c-th data point; V represents the total number of data points within the neighborhood of the c-th data point; σ represents the parameter controlling weight decay; norm() represents the normalization function; exp() represents the exponential function with the natural constant as the base.

[0145] The working principle of the above technical solution is as follows: Randomly select one target category of data from the first data set, and evenly divide this target category data into several target category sub-data based on time order. At this point, the sub-data also have a sequential order. Randomly select one sub-data as the first target category sub-data, and calculate the anomaly evaluation value of the first target category sub-data. This essentially involves dividing the data into several data blocks and first judging the degree of anomaly between the data blocks, thus reducing the overall computational load of data cleaning. When the anomaly evaluation value is determined to be greater than or equal to a preset evaluation threshold, the first target category sub-data is selected as the second target category sub-data. This process is repeated for all sub-data. The first target category sub-data is processed to obtain several second target category sub-data. At this point, the filtering of the first target category sub-data is complete, resulting in several second target category sub-data. For each second target category sub-data, outlier identification is performed, and the local fluctuation index of each data point in each second target category sub-data is calculated. Because the data values ​​of data points fluctuate significantly in continuous running data, the possibility of anomalies is high, resulting in several outlier data points. These outlier data points are then deleted, resulting in the cleaned target category data. Finally, all categories of data in the first dataset are traversed to obtain the cleaned first dataset.

[0146] The beneficial effects of the above technical solution are as follows: By uniformly segmenting the target category data based on time series order through the segmentation unit, it facilitates detailed data analysis in subsequent steps. This segmentation method ensures the uniformity and continuity of the data in the time dimension, providing a solid foundation for subsequent calculation of anomaly evaluation values. The combined use of the first calculation unit and the second determination unit can accurately identify target category sub-data with high anomaly evaluation values. The second calculation unit further calculates the local fluctuation index of each data point in the second target category sub-data and compares it with a preset fluctuation threshold, thereby accurately identifying abnormal data points. The data cleaning unit removes several abnormal data points to obtain the cleaned target category data. This process not only removes noise and outliers from the data but also preserves the integrity and coherence of the data, providing a more accurate and reliable data foundation for subsequent data analysis and decision-making. Traversing all category data in the first dataset ensures that the entire dataset has undergone effective cleaning processing, improving the overall data quality.

[0147] Example 6

[0148] The first computing unit includes:

[0149] Get a sub-unit, which is used to arbitrarily select a target category sub-data as the first target category sub-data;

[0150] The first calculation subunit is used to calculate the mean of the data values ​​corresponding to all data points in the first target category subdata to obtain the first mean.

[0151] The second computational subunit is used for:

[0152] Obtain the target category sub-data adjacent to the first target category sub-data, calculate the mean of the data values ​​corresponding to all data points in the first target category sub-data and the target category sub-data adjacent to the first target category sub-data, and obtain the second mean;

[0153] The third computational subunit is used for:

[0154] Calculate the absolute value of the difference between the first mean and the second mean to obtain the first difference value;

[0155] Calculate the range of data points corresponding to all data points in the first target category sub-data and the target category sub-data adjacent to the first target category sub-data to obtain the second difference value;

[0156] A sub-unit is defined to determine the anomaly evaluation value of the first target category sub-data based on the first difference value and the second difference value.

[0157] In this embodiment, the anomaly evaluation value of the first target category sub-data includes:

[0158] T = e w +e q

[0159] Where T represents the anomaly evaluation value of the first target category sub-data; e represents the natural constant; w represents the first difference value; and q represents the second difference value.

[0160] The working principle of the above technical solution is as follows: The acquisition sub-unit arbitrarily selects a target category sub-data from the dataset as the first target category sub-data. This step is the starting point of data processing, providing the basic data for subsequent calculations. The first calculation sub-unit calculates the mean of the data values ​​corresponding to all data points in the first target category sub-data, obtaining the first mean. The mean is an important indicator for measuring the central tendency of the dataset, helping to understand the overall distribution of the data. The second calculation sub-unit further acquires the target category sub-data adjacent to the first target category sub-data and calculates the mean of the data values ​​corresponding to all data points in these two parts of data, obtaining the second mean. This step considers the continuity of the data in time series or spatial distribution, helping to capture the changing trend of the data. The third calculation sub-unit performs two key calculations: first, it calculates the absolute value of the difference between the first mean and the second mean, obtaining the first difference value; second, it calculates the range of the data values ​​corresponding to all data points in the first target category sub-data and its adjacent sub-data, obtaining the second difference value. The first difference value reflects the degree of change of the data around the mean, while the second difference value reflects the range of fluctuation of the data around the extreme values. Based on the first and second difference values, the sub-units are determined to comprehensively evaluate the degree of anomaly of the first target category sub-data and provide an anomaly evaluation value. This step is the core of anomaly detection; by comparing the difference value with a preset threshold or standard, it can be determined whether the data is abnormal.

[0161] The beneficial effects of the above technical solution are as follows: By calculating the mean of the first target category sub-data and its difference from the mean of adjacent sub-data, subtle changes in the time series or spatial distribution of the data can be captured. This mean-based difference analysis helps to accurately identify abnormal fluctuations in the data; at the same time, calculating the range further enhances the sensitivity of anomaly detection; the range reflects the difference between the maximum and minimum values ​​in the dataset and is an important indicator for measuring the range of data fluctuations; combining the first and second difference values, determining the sub-unit can more comprehensively assess the degree of data anomaly; through parallel computing or pipelined processing, the first computing unit can efficiently process large-scale datasets; the data transfer and computing processes between sub-units can be optimized to reduce computing time and resource consumption; accurate anomaly detection helps to promptly discover and correct errors or outliers in the data, thereby improving data quality and reliability. This is crucial for subsequent data analysis and decision-making processes; the design of the first computing unit provides strong support for data anomaly detection tasks by improving the accuracy of anomaly detection, enhancing the flexibility of data processing, improving data processing efficiency, promoting data quality improvement, and supporting multiple application scenarios. This design not only improves the efficiency and accuracy of data processing but also provides a reliable foundation for subsequent data analysis and decision-making.

[0162] Example 7

[0163] The noise reduction submodule includes:

[0164] Acquisition unit, used for:

[0165] Take any monitoring image of a smart street light from the second dataset and perform grayscale processing;

[0166] The monitoring images of the smart streetlights after grayscale processing are divided into several sub-images;

[0167] Several sub-images are clustered to obtain several sub-image sets;

[0168] Choose any sub-image set as the target sub-image set;

[0169] Select any one sub-image from the target sub-image set as the target sub-image;

[0170] Take any pixel in the target sub-image as the first pixel;

[0171] The third calculation unit is used to calculate the brightness value and blur value of the first pixel respectively;

[0172] The comparison unit is used to compare the brightness value of the first pixel with a preset brightness threshold and to compare the blur value of the first pixel with a preset blur threshold.

[0173] The third determining unit is used for:

[0174] When it is determined that the brightness value of the first pixel is less than the preset brightness threshold and the blur value of the first pixel is greater than the preset blur threshold, the first pixel is taken as the pixel to be denoised.

[0175] Traverse all pixels to obtain a number of pixels to be denoised;

[0176] Noise reduction unit, used for:

[0177] The filtering algorithm is used to denoise several pixels to be denoised, and the denoised target sub-image is obtained.

[0178] Based on the target sub-image and the denoised target sub-image, determine the denoising amplitude of each pixel to be denoised in the target sub-image;

[0179] Denoising is performed on other sub-images in the target sub-image set based on the denoising amplitude of each pixel to be denoised;

[0180] Traverse all sub-image sets to obtain the noise-reduced monitoring images of the smart streetlights;

[0181] The monitoring images of all smart streetlights in the second dataset are traversed to obtain the noise-reduced second dataset.

[0182] In this embodiment, the filtering algorithm includes, but is not limited to, median filtering algorithm and mean filtering algorithm.

[0183] The working principle of the above technical solution is as follows: An image of a smart street light is randomly selected from the second data set and processed to grayscale. The grayscale-processed image is divided into several sub-images. These sub-images are then clustered to obtain several sets of sub-images. The purpose of clustering is to group similar content in the monitoring image into a single category, resulting in several sets of sub-images. A noise reduction coefficient is obtained by denoising one sub-image within the set. Based on this coefficient, the same noise reduction process is applied to other sub-images within the same set. The third calculation unit calculates the brightness and blur value of the first pixel, filters it, and obtains several pixels to be denoised. Noise reduction is then performed on these pixels based on a filtering algorithm.

[0184] The beneficial effects of the above technical solution are as follows: Through grayscale processing, sub-image segmentation, and clustering, this module can efficiently process a large number of smart street light monitoring images; this processing method not only reduces computational complexity but also improves image processing efficiency; by calculating the brightness and blur values ​​of pixels and comparing them with preset thresholds, this module can accurately identify pixels that need noise reduction. This method avoids blindly performing noise reduction on all pixels, thereby improving the targeting and effectiveness of noise reduction; based on filtering algorithms to perform noise reduction on the identified pixels to be denoised, this module can intelligently remove noise from the image. Simultaneously, by determining the noise reduction amplitude and performing noise reduction on other sub-images, this module can ensure consistent noise reduction effects across the entire target sub-image set, improving the overall image quality.

[0185] Example 8

[0186] The third calculation unit includes:

[0187] The brightness calculation subunit is used for:

[0188] The target area is determined with the first pixel as the center and a preset distance as the radius;

[0189] Calculate the average grayscale value of each pixel in the target area, and use it as the brightness value of the first pixel;

[0190] The ambiguity calculation subunit is used for:

[0191] The sum of the squares of the grayscale differences between the first pixel and all other pixels in the target region is calculated to obtain the target sum value.

[0192] Obtain the total number of pixels in the target area excluding the first pixel, and get the target value;

[0193] The ratio of the target sum to the target value is used as the blur value of the first pixel.

[0194] The beneficial effects of the above technical solution are as follows: The brightness calculation subunit determines the target area with the first pixel as the center and a preset distance as the radius, and calculates the average gray value of all pixels in the area as the brightness value of the first pixel. This method can more accurately reflect the average brightness around the first pixel and avoid errors that may be caused by the brightness value of a single pixel. The blurriness calculation subunit calculates the sum of the squares of the gray value differences between the first pixel and other pixels in the target area, and then divides it by the total number of pixels in the target area, excluding the first pixel, to obtain the blurriness value of the first pixel. This calculation method can quantify the degree of gray value difference between the first pixel and its surrounding pixels, thereby effectively evaluating the blurriness of the image. Through accurate brightness calculation and effective blurriness evaluation, this technical solution can more accurately identify pixels that need noise reduction. This helps to improve the accuracy of noise reduction processing, avoid over-processing of clear pixels, and ensure that blurred pixels receive sufficient noise reduction processing.

[0195] Example 9

[0196] Methods for constructing big data analytics models include:

[0197] Obtain training datasets for big data analysis;

[0198] The neural network model is trained based on the big data analysis training dataset to obtain the initial big data analysis model;

[0199] Obtain a big data analysis test dataset;

[0200] The initial big data analysis model is tested using a big data analysis test dataset. When the test results are satisfactory, a well-trained big data analysis model is obtained.

[0201] In this embodiment,

[0202]

[0203] Where, φ i U represents the first evaluation value of the smart street light at time i; i I represents the voltage value across the smart street light at time i; i R represents the current value at both ends of the smart street light at time i; i t1 represents the internal resistance value of the smart street light at time i; t2 represents the time the smart street light has been used; t3 represents the ideal usage time of the smart street light; α represents the aging coefficient of the smart street light. δ represents the theoretical remaining service life of the smart street light; δ represents the number of failures of the smart street light.

[0204] Obtain the output power and rated power of the smart street light;

[0205] The ratio of output power to rated power is used as the lighting output index of the smart street light;

[0206] Take any monitoring image of a smart street light and obtain the moving speed of the moving objects in each lane of the monitoring image; determine the moving distance of the moving objects in each lane based on the moving speed and the unit time of the moving objects in each lane; divide each lane with the moving distance of the moving objects as the length and the width of the corresponding lane as the width, and determine the target area corresponding to each lane.

[0207] The lighting demand index for each lane is calculated based on a preset algorithm;

[0208]

[0209] Among them, C i,a denoted by , v represents the lighting demand index of lane a in the monitoring image of the smart street light at time i; v represents the speed of the moving object in lane a; θ represents the direction of the line connecting the center point of the current moving object to the center point of the current target area, and the angle between the direction of the line and the direction of travel of the moving object; l represents the distance between the center point of the current moving object and the center point of the current target area; b represents the natural constant.

[0210] The average of the lighting demand indices for all lanes is used as the lighting demand index for smart streetlights.

[0211] The ratio of the lighting demand index to the lighting output index of a smart street light is used as the second evaluation value of the smart street light.

[0212] Obtain the preset weight value;

[0213] Based on the first evaluation value, the second evaluation value, and the preset weight value of the smart street light, the operation and maintenance evaluation value of the smart street light is determined.

[0214]

[0215] Where P represents the operation and maintenance assessment value of the smart street light; φ i β1 represents the first evaluation value of the smart street light at time i; β1 represents the first preset weight; n represents the total number of times in the target data; C i,a β1 represents the lighting demand index of lane a in the monitoring image of the smart street light at time i; m represents the total number of lanes in the monitoring image; β2 represents the second preset weight, and β1+β2=1.

[0216] The beneficial effects of the above technical solution are: the construction method of big data analysis model improves the accuracy of smart street light operation and maintenance assessment values ​​through targeted training, accurate performance evaluation, improved generalization ability, optimized business processes, and promotion of a data-driven decision-making culture.

[0217] Example 10

[0218] The optimization module includes:

[0219] The query submodule is used to query the smart operation and maintenance optimization database based on the target characteristics of the smart street light to be optimized, and to determine the target optimization strategy for the smart street light to be optimized.

[0220] The optimization submodule is used to optimize the smart streetlights to be optimized based on the target optimization strategy.

[0221] The beneficial effects of the above technical solution are: the query submodule can quickly query and determine the target optimization strategy in the smart operation and maintenance optimization database based on the target characteristics of the smart street light to be optimized; this process avoids the tedious process of manual analysis and judgment of optimization strategies in traditional methods, and significantly improves optimization efficiency; with the support of the smart operation and maintenance optimization database, the query submodule can obtain optimization strategies that match the target characteristics of the smart street light to be optimized.

[0222] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A street lamp intelligent operation and maintenance optimization management platform based on big data analysis, characterized in that, The method comprises the following steps: acquiring a target data of a smart street lamp in a preset time period; extracting features from the target data to obtain target features; evaluating the target features in a pre-trained big data analysis model to determine an operation and maintenance evaluation value of the smart street lamp; comparing the operation and maintenance evaluation value with a preset operation and maintenance evaluation threshold value, and determining the smart street lamp as a to-be-optimized smart street lamp when the operation and maintenance evaluation value is less than or equal to the preset operation and maintenance evaluation threshold value; querying a target optimization strategy for the to-be-optimized smart street lamp, and optimizing the to-be-optimized smart street lamp based on the target optimization strategy; further comprising a preprocessing module for preprocessing the target data to obtain preprocessed target data before the extracting module extracts features from the target data to obtain target features; the preprocessing module comprises: a data cleaning submodule for cleaning the first data to obtain cleaned first data; a noise reduction submodule for reducing noise in the second data to obtain noise-reduced second data; a second determining submodule for determining the cleaned first data and the noise-reduced second data as preprocessed target data; the noise reduction submodule comprises: an acquisition unit for: randomly selecting a monitoring image of a smart street lamp in the second data and performing grayscale processing; dividing the grayscale-processed monitoring image of the smart street lamp into a plurality of sub-images; clustering the plurality of sub-images to obtain a plurality of sub-image sets; randomly selecting one of the sub-image sets as a target sub-image set; randomly selecting one of the sub-images in the target sub-image set as a target sub-image; randomly selecting a pixel point in the target sub-image as a first pixel point; a third calculation unit for calculating the brightness value and the blur value of the first pixel point, respectively; a comparison unit for comparing the brightness value of the first pixel point with a preset brightness threshold value and comparing the blur value of the first pixel point with a preset blur threshold value; a third determining unit for: determining the first pixel point as a to-be-noise-reduced pixel point when the brightness value of the first pixel point is less than the preset brightness threshold value and the blur value of the first pixel point is greater than the preset blur threshold value; traversing all pixel points to obtain a plurality of to-be-noise-reduced pixel points; a noise reduction unit for: noise-reducing the plurality of to-be-noise-reduced pixel points based on a filtering algorithm to obtain a noise-reduced target sub-image; determining a noise reduction amplitude of each to-be-noise-reduced pixel point in the target sub-image based on the target sub-image and the noise-reduced target sub-image; noise-reducing other sub-images in the target sub-image set based on the noise reduction amplitude of each to-be-noise-reduced pixel point; traversing all sub-image sets to obtain noise-reduced monitoring images of the smart street lamps; traversing all monitoring images of the smart street lamps in the second data to obtain noise-reduced second data; the third calculation unit comprises: a brightness calculation submodule for: determining a target region with the first pixel point as the center and a preset distance as the radius; calculating the average grayscale value of each pixel point in the target region as the brightness value of the first pixel point; a blur calculation submodule for: Calculate the sum of squares of the gray scale difference between the first pixel point and other pixel points in the target region, and obtain a target sum value; Obtain the total number of pixel points other than the first pixel point in the target region, and obtain a target numerical value; The ratio of the target sum value to the target numerical value is taken as the blur value of the first pixel point. 2.The big data analysis based street lamp intelligent operation and maintenance optimization management platform of claim 1, wherein, The acquisition module comprises: The first acquisition submodule is configured to acquire operation data of the intelligent street lamp in the preset time period to obtain first data. The second acquisition submodule is configured to acquire monitoring image data of the intelligent street lamp in the preset time period to obtain second data. The first determination submodule is configured to take the first data and the second data as target data of the intelligent street lamp. 3.The big data analysis based street lamp intelligent operation and maintenance optimization management platform of claim 1, wherein, The data cleaning submodule comprises: The segmentation unit is configured to: Take one type of data in the first data as target category data; Divide the target category data into a plurality of target category sub-data based on the time sequence order; The first calculation unit is configured to: Take one target category sub-data as a first target category sub-data; Calculate the abnormal evaluation value corresponding to the first target category sub-data; The first determination unit is configured to: Compare the abnormal evaluation value with a preset evaluation threshold, and determine that the first target category sub-data is a second target category sub-data when the abnormal evaluation value is greater than or equal to the preset evaluation threshold; Iterate through all target category sub-data to obtain a plurality of second target category sub-data; The second calculation unit is configured to: Take one second target category sub-data; Calculate the local fluctuation index of each data point in the second target category sub-data based on a first preset algorithm; The second determination unit is configured to: Compare the local fluctuation index of each data point with a preset fluctuation threshold, and take the data point whose local fluctuation index is greater than or equal to the preset fluctuation threshold as an abnormal data point; Iterate through all second target category sub-data to obtain a plurality of abnormal data points; The data cleaning unit is configured to: Delete the plurality of abnormal data points to obtain target category data after data cleaning; Iterate through all types of data in the first data to obtain first data after data cleaning. 4.The big data analysis based intelligent operation and maintenance optimization management platform of street lamps according to claim 3, characterized in that, The first calculation unit comprises: The acquisition subunit is configured to take one target category sub-data as a first target category sub-data; The first calculation subunit is configured to calculate the mean value of the data values corresponding to all data points in the first target category sub-data to obtain a first mean value; The second calculation subunit is configured to: Obtain the adjacent target category sub-data of the first target category sub-data, calculate the mean value of the data values corresponding to all data points in the first target category sub-data and the adjacent target category sub-data of the first target category sub-data to obtain a second mean value; The third calculation subunit is configured to: Calculate the absolute value of the difference between the first mean value and the second mean value to obtain a first difference value; Calculate the range of the data values corresponding to all data points in the first target category sub-data and the adjacent target category sub-data of the first target category sub-data to obtain a second difference value; The determination subunit is configured to determine the abnormal evaluation value of the first target category sub-data based on the first difference value and the second difference value. 5.The big data analysis based street lamp intelligent operation and maintenance optimization management platform of claim 1, wherein, The method for constructing a big data analysis model comprises: Obtaining a big data analysis training data set; Training a neural network model based on the big data analysis training data set to obtain an initial big data analysis model; Obtaining a big data analysis test data set; Testing the initial big data analysis model based on the big data analysis test data set, and obtaining a trained big data analysis model when the test result is qualified. 6.The big data analysis based street lamp intelligent operation and maintenance optimization management platform of claim 1, wherein, The optimization module comprises: The query submodule is configured to query the intelligent operation and maintenance optimization database based on the target characteristics of the to-be-optimized intelligent street lamp, and determine the target optimization strategy of the to-be-optimized intelligent street lamp. The optimization submodule is configured to optimize the to-be-optimized intelligent street lamp based on the target optimization strategy.

Images