An intelligent adjustment method and system for street lamps

Abstract

The present application provides a method and system for intelligent adjustment of street lights, relating to the technical field of intelligent control of street lights. The method includes: performing brightness trend analysis on Q sets of park images respectively to obtain Q brightness trend analysis results; performing regional grid fusion analysis to obtain the park brightness trend distribution result; sending the multiple brightness fusion central values of multiple grid fusion regions in the distribution result to the street light intelligent adjustment platform to obtain multiple first template adjustment scheme sets; respectively performing fine-tuning and optimization on the multiple first template adjustment scheme sets to obtain the first fine-tuning and optimization scheme set for brightness adjustment. Through the present application, the technical problem that unnecessary waste of lighting resources exists and the energy consumption of park street lights is relatively high due to the inability to timely adaptively adjust the brightness of street lights according to the actual lighting needs in the park can be solved, the technical goal of intelligently adjusting the brightness of park street lights can be achieved, and the effect of effectively reducing the energy consumption of park street lights can be achieved.

Description

Technical Field

[0001] This application relates to the technical field of intelligent control of street lamps, and particularly to an intelligent adjustment method and system for street lamps. Background Art

[0002] With the large-scale construction of industrial parks and commercial parks, the number of lighting street lamps has increased sharply. Although the energy consumption of a single street lamp is not high, the cumulative energy consumption of multiple street lamps in the park will cause a large amount of energy waste.

[0003] Currently, most parks are still using traditional methods for lighting, and the total control of park street lamps is carried out according to fixed time periods. This method cannot adaptively adjust the lighting state of park street lamps according to the actual situation in the park, resulting in a low matching degree between the lighting state of park street lamps and the actual lighting needs, and the situation of lighting resource waste often occurs.

[0004] In summary, the existing park street lamp adjustment method cannot adaptively adjust the brightness of park street lamps in a timely manner according to the actual lighting needs in the park, resulting in unnecessary waste of lighting resources and the technical problem of high energy consumption of park street lamps. Summary of the Invention

[0005] The purpose of this application is to provide an intelligent adjustment method and system for street lamps to solve the technical problem that the existing park street lamp adjustment method cannot adaptively adjust the brightness of park street lamps in a timely manner according to the actual lighting needs in the park, resulting in unnecessary waste of lighting resources and high energy consumption of park street lamps.

[0006] In view of the above problems, this application provides an intelligent adjustment method and system for street lamps.

[0007] First aspect, the present application provides a method for intelligent adjustment of street lights. The method is implemented through a street light intelligent adjustment system. Among them, the method is applied to a street light intelligent adjustment platform, and the street light intelligent adjustment platform is communicatively connected to an array of street light control units and camera component layouts. The method includes: using the camera component layout array to collect images of a target park within a preset monitoring window to obtain Q park image sets, where the camera component layout array has Q camera components; performing brightness trend analysis on the Q park image sets respectively to obtain Q brightness trend analysis results, where each brightness trend analysis result includes multiple grid regions, and each grid region has a brightness concentration value and a trend identifier, and the trend identifier includes a positive brightness identifier, a steady brightness identifier, and a negative brightness identifier; performing regional grid fusion analysis on the Q brightness trend analysis results to obtain a target park brightness trend distribution result, where the target park brightness trend distribution result has multiple grid fusion regions, and the multiple grid fusion regions have multiple brightness fusion concentration values and multiple trend identifiers; sending the multiple brightness fusion concentration values of the multiple grid fusion regions in the target park brightness trend distribution result to the street light intelligent adjustment platform, and respectively calling adjustment plan templates for the multiple grid fusion regions to obtain multiple first template adjustment plan sets; respectively performing fine-tuning optimization on the multiple first template adjustment plan sets according to the multiple brightness fusion concentration values and the multiple trend identifiers to obtain a first fine-tuning optimization plan set; the street light control unit adjusts the brightness of the street lights in the target park according to the first fine-tuning optimization plan set.

[0008] Second aspect, the present application also provides a street lamp intelligent adjustment system for implementing a street lamp intelligent adjustment method as described in the first aspect. Among them, the system is communicatively connected to a street lamp intelligent adjustment platform, and the street lamp intelligent adjustment platform is communicatively connected to a street lamp control unit and an array of camera components. The system includes: a park image acquisition module, which is used to use the array of camera components to collect images of a target park within a preset monitoring window, obtaining Q park image sets. Among them, the array of camera components has Q camera components; a brightness trend analysis module, which is used to perform brightness trend analysis on the Q park image sets respectively, obtaining Q brightness trend analysis results. Among them, each brightness trend analysis result includes multiple grid regions, and each grid region has a brightness concentration value and a trend identifier. The trend identifier includes a positive brightness identifier, a steady brightness identifier, and a negative brightness identifier; a regional grid fusion analysis module, which is used to perform regional grid fusion analysis on the Q brightness trend analysis results, obtaining a target park brightness trend distribution result. Among them, the target park brightness trend distribution result has multiple grid fusion regions, and the multiple grid fusion regions have multiple brightness fusion concentration values and multiple trend identifiers; an adjustment plan template calling module, which is used to send the multiple brightness fusion concentration values of the multiple grid fusion regions in the target park brightness trend distribution result to the street lamp intelligent adjustment platform, respectively perform adjustment plan template calling on the multiple grid fusion regions, obtaining multiple first template adjustment plan sets; a template adjustment plan optimization module, which is used to fine-tune and optimize the multiple first template adjustment plan sets respectively according to the multiple brightness fusion concentration values and the multiple trend identifiers, obtaining a first fine-tuning and optimization plan set; a street lamp brightness adjustment module, which is used to adjust the brightness of the street lamps in the target park by the street lamp control unit according to the first fine-tuning and optimization plan set.

[0009] One or more technical solutions provided in the present application have at least the following technical effects or advantages:

[0010] 1. By using the camera component layout array to collect images of the target park within the preset monitoring window, Q sets of park images are obtained, where the camera component layout array has Q camera components; performing brightness trend analysis on the Q sets of park images respectively to obtain Q brightness trend analysis results, where each brightness trend analysis result includes multiple grid regions, each grid region has a brightness concentration value and a trend identifier, and the trend identifier includes a positive brightness identifier, a steady brightness identifier, and a negative brightness identifier; performing regional grid fusion analysis on the Q brightness trend analysis results to obtain the target park brightness trend distribution result, where the target park brightness trend distribution result has multiple grid fusion regions, and the multiple grid fusion regions have multiple brightness fusion concentration values and multiple trend identifiers; sending the multiple brightness fusion concentration values of the multiple grid fusion regions in the target park brightness trend distribution result to the street lamp intelligent adjustment platform, and respectively calling the adjustment plan templates for the multiple grid fusion regions to obtain multiple first template adjustment plan sets; fine-tuning and optimizing the multiple first template adjustment plan sets respectively according to the multiple brightness fusion concentration values and the multiple trend identifiers to obtain the first fine-tuning and optimization plan set; adjusting the brightness of the street lamps in the target park by the street lamp control unit according to the first fine-tuning and optimization plan set. That is to say, by performing real-time brightness analysis and brightness trend analysis based on the park image set, multiple brightness trend analysis results are determined, and further integrated to obtain the overall park brightness trend distribution result; based on the brightness trend distribution result, the adjustment plan template is called, and the adjustment plan is fine-tuned and optimized according to the brightness fusion concentration value and the brightness trend identifier, and the brightness of the park street lamps is adjusted based on the optimized plan, which can reduce or avoid waste of additional lighting resources on the premise of meeting the lighting requirements, improve the matching degree of lighting requirements and the brightness of park street lamps, achieve the technical goal of intelligently adjusting the brightness of park street lamps, and achieve the technical effect of effectively reducing the energy consumption of park street lamps.

[0011] 2. By calling the adjustment plan template based on the street lamp intelligent adjustment platform according to the brightness fusion concentration value, the efficiency and accuracy of obtaining the adjustment plan can be improved; at the same time, fine-tuning and optimizing the adjustment plan according to the brightness fusion concentration value and the brightness trend identifier can further improve the adaptability of the adjustment plan to the actual lighting requirements of the park, improve the accuracy of the adjustment plan setting, and thus improve the accuracy, efficiency and intelligence of the park street lamp brightness adjustment.

[0012] The above description is only an overview of the technical solution of the present application. In order to be able to more clearly understand the technical means of the present application, it can be implemented in accordance with the content of the specification. And in order to make the above and other purposes, features and advantages of the present application more obvious and understandable, the specific embodiments of the present application are specifically given below. It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of the present application, nor is it used to limit the scope of the present application. Other features of the present application will become easily understood through the following description. Description of the Drawings

[0013] In order to more clearly illustrate the technical solutions in the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings described below are only exemplary. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained according to the provided drawings.

[0014] Figure 1 It is a schematic flowchart of a method for intelligent adjustment of street lights in the present application;

[0015] Figure 2 It is a schematic flowchart of obtaining Q brightness trend analysis results in a method for intelligent adjustment of street lights in the present application;

[0016] Figure 3 It is a schematic structural diagram of a system for intelligent adjustment of street lights in the present application.

[0017] Description of the reference numerals:

[0018] Park image acquisition module 11, brightness trend analysis module 12, regional grid fusion analysis module 13, adjustment plan template call module 14, template adjustment plan optimization module 15, street light brightness adjustment module 16. Detailed Embodiments

[0019] By providing a method and system for intelligent adjustment of street lights, the present application solves the technical problem that the existing method for adjusting street lights in a park causes unnecessary waste of lighting resources and high energy consumption of park street lights because it cannot adaptively adjust the brightness of park street lights in a timely manner according to the actual lighting needs in the park. It can reduce or avoid additional waste of lighting resources on the premise of meeting the lighting needs, improve the matching degree between the lighting needs and the brightness of park street lights, achieve the technical goal of intelligent adjustment of the brightness of park street lights, and achieve the technical effect of effectively reducing the energy consumption of park street lights.

[0020] Next, the technical solutions in the present application will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. It should be understood that the present application is not limited by the example embodiments described herein. Based on the embodiments of the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the scope of protection of the present application. Additionally, it should be noted that for the convenience of description, only the parts related to the present application are shown in the drawings rather than all of them.

[0021] Embodiment 1

[0022] Please refer to the attached Figure 1 , the present application provides a method for intelligent adjustment of street lights. Among them, the method is applied to a street light intelligent adjustment system, the system is communicatively connected to a street light intelligent adjustment platform, and the street light intelligent adjustment platform is communicatively connected to a street light control unit and an array of camera components. The method specifically includes the following steps:

[0023] Step 1: Use the array of camera components to collect images of the target park within a preset monitoring window, and obtain Q sets of park images, where the array of camera components has Q camera components;

[0024] Specifically, the method provided by the present application is used for intelligent adjustment of park street lights. The method is specifically implemented in a street light intelligent adjustment system, and the system and the street light intelligent adjustment platform can realize data transmission and information interaction. The street light intelligent adjustment platform inputs the adjustment scheme into the street light control unit through signal transmission, thereby realizing intelligent adjustment of park street lights; the array of camera components includes multiple camera components, which are used to collect images of different areas of the park and input the image collection results into the street light intelligent adjustment platform through signal transmission.

[0025] Collect images of the target park within a preset monitoring window through the array of camera components to obtain Q sets of park images. The preset monitoring window is a time period, and the preset monitoring window includes multiple monitoring time nodes. Those skilled in the art can set the monitoring time nodes based on the street light adjustment accuracy requirements. The higher the adjustment accuracy requirements, the smaller the interval of the monitoring time nodes. For example, the preset monitoring window can be set to 1 hour, and the monitoring time nodes can be set to 3 minutes. That is, within the preset monitoring window, the target park is imaged every 3 minutes; the array of camera components includes Q camera components, Q is an integer greater than 1, and the specific value of Q can be calculated according to the scale of the target park and the coverage area of each camera component.

[0026] By obtaining Q sets of park images, it provides data support for the next step of park brightness recognition and brightness change trend analysis.

[0027] Step 2: Perform brightness trend analysis on each of the Q sets of park images to obtain Q brightness trend analysis results. Each brightness trend analysis result includes multiple grid regions, and each grid region has a brightness concentration value and a trend identifier. The trend identifier includes a positive brightness identifier, a steady brightness identifier, and a negative brightness identifier.

[0028] Specifically, perform brightness trend analysis on each of the Q sets of park images. First, divide the park images into grids. Each park image includes multiple grid regions. Then, perform brightness trend analysis on the grid regions of each image set respectively. The brightness trend analysis includes brightness recognition and brightness change trend analysis, obtaining Q brightness trend analysis results. Each brightness trend analysis result includes multiple grid regions, and each grid region has a brightness concentration value and a trend identifier. The brightness concentration value represents the average brightness state of the region. The trend identifier includes a positive brightness identifier, a steady brightness identifier, and a negative brightness identifier. The positive brightness identifier indicates that the brightness in the region shows an increasing trend over time. The steady brightness identifier indicates that the brightness in the region is hardly affected by time. The negative brightness identifier indicates that the brightness in the region shows a decreasing trend over time.

[0029] By performing brightness trend analysis and obtaining Q brightness trend analysis results, it can intuitively display the brightness state and brightness change trend of each region in the park, and at the same time provide data support for the overall park brightness analysis and brightness change trend analysis.

[0030] Step 3: Perform regional grid fusion analysis on the Q brightness trend analysis results to obtain the target park brightness trend distribution result. The target park brightness trend distribution result has multiple grid fusion regions, and the multiple grid fusion regions have multiple brightness fusion concentration values and multiple trend identifiers.

[0031] Specifically, perform regional grid fusion analysis on the Q brightness trend analysis results, that is, splice and fuse the park image sets according to the position coordinates of each camera component in the park, and then realize the splicing and fusion of the Q brightness trend analysis results to obtain the target park brightness trend distribution result of the entire park. The target park brightness trend distribution result includes multiple grid fusion regions, and each grid fusion region has a brightness fusion concentration value and a trend identifier. By obtaining the target park brightness trend distribution result, it provides a basis for the next step of calling the street lamp adjustment plan template and intelligently generating the street lamp adjustment plan.

[0032] Step 4: Send the multiple brightness fusion central values of multiple grid fusion regions in the target park brightness trend distribution result to the street lamp intelligent adjustment platform, and respectively call the adjustment plan templates for multiple grid fusion regions to obtain multiple first template adjustment plan sets;

[0033] Specifically, send the multiple brightness fusion central values of multiple grid fusion regions in the target park brightness trend distribution result to the street lamp intelligent adjustment platform, where the street lamp intelligent adjustment platform is embedded with an adjustment plan matching model. The adjustment plan matching model is constructed based on a BP neural network, which is a neural network model in machine learning that can be iteratively optimized and is obtained through supervised training with sample data; further, use the adjustment plan matching model to respectively perform adjustment plan template matching on the multiple brightness fusion central values of the multiple grid fusion regions. Since the brightness distribution results within each grid fusion region are not completely the same, in order to select a better adjustment plan subsequently, the plan matching can be performed multiple times to obtain multiple first template adjustment plan sets, where each first template adjustment plan set corresponds to a grid fusion region.

[0034] The adjustment plan matching model includes an input layer, multiple hidden layers, and an output layer. The input data of the input layer is the brightness fusion central value, and the output data of the output layer is the template adjustment plan. Retrieve sample data based on big data technology. Big data technology refers to a series of technologies and methods that enable the value contained in massive data to be mined and presented, including technical means such as data planning, data collection, analysis and mining, etc. The sample data includes multiple sample brightness fusion central values and multiple sample template adjustment plans to obtain a sample data set; then use the sample data set to perform supervised training on the adjustment plan matching model to obtain an adjustment plan matching model in a converging state.

[0035] By constructing an adjustment plan matching model based on a BP neural network and performing intelligent plan matching on multiple brightness fusion central values through the adjustment plan matching model, the intelligence of generating template adjustment plans can be improved, and at the same time, the accuracy and efficiency of obtaining template adjustment plans can be improved.

[0036] Step 5: According to the multiple brightness fusion central values and the multiple trend identifiers, respectively perform fine-tuning and optimization on the multiple first template adjustment plan sets to obtain a first fine-tuning and optimization plan set;

[0037] Specifically, a plurality of fine-tuning directions and a plurality of fine-tuning step sizes are determined based on the plurality of luminance fusion concentration values and the plurality of tendency identifiers, and the plurality of first template adjustment scheme sets are fine-tuned and optimized respectively based on the plurality of fine-tuning directions and the plurality of fine-tuning step sizes to obtain a first fine-tuning and optimization scheme set. Among them, the fine-tuning direction is the direction of luminance adjustment, and the fine-tuning step size is the luminance value adjusted during a single luminance adjustment.

[0038] Since the template adjustment scheme obtained through intelligent matching only considers the luminance distribution in the park and does not take into account the impact of the luminance change trend on the adjustment scheme, fine-tuning and optimizing the template adjustment scheme by combining the luminance fusion concentration value and the luminance tendency identifier can further improve the accuracy and rationality of the fine-tuning and optimization scheme setting, thereby improving the accuracy of the park street lamp luminance adjustment.

[0039] Step Six: The street lamp control unit adjusts the luminance of the street lamps in the target park according to the first fine-tuning and optimization scheme set.

[0040] Specifically, the first fine-tuning and optimization scheme set is transmitted to the street lamp control unit, and the street lamp control unit adjusts the luminance of the target park street lamps according to the first fine-tuning and optimization scheme set. By adjusting the luminance of the park street lamps according to the fine-tuning and optimization scheme, the automation and intelligence level of the park street lamp luminance adjustment can be improved.

[0041] The described street lamp intelligent adjustment method is applied to a street lamp intelligent adjustment system. The system is communicatively connected to a street lamp intelligent adjustment platform, and the street lamp intelligent adjustment platform is communicatively connected to a street lamp control unit and an imaging component layout array, which can solve the technical problem that the existing park street lamp adjustment method cannot adaptively adjust the luminance of the park street lamps in a timely manner according to the actual lighting requirements in the park, resulting in unnecessary waste of lighting resources and high energy consumption of the park street lamps.

[0042] First, use the camera component layout array to collect images of the target park within a preset monitoring window, obtaining Q sets of park images. Among them, the camera component layout array has Q camera components. Then, perform brightness trend analysis on the Q sets of park images respectively to obtain Q brightness trend analysis results. Each brightness trend analysis result includes multiple grid regions, and each grid region has a brightness concentration value and a trend identifier. The trend identifier includes a positive brightness identifier, a steady brightness identifier, and a negative brightness identifier. Next, perform regional grid fusion analysis on the Q brightness trend analysis results to obtain the brightness trend distribution result of the target park. The brightness trend distribution result of the target park has multiple grid fusion regions, and the multiple grid fusion regions have multiple brightness fusion concentration values and multiple trend identifiers. Next, send the multiple brightness fusion concentration values of the multiple grid fusion regions in the brightness trend distribution result of the target park to the street lamp intelligent adjustment platform, and respectively call the adjustment plan templates for the multiple grid fusion regions to obtain multiple first template adjustment plan sets. In addition, perform fine-tuning optimization on the multiple first template adjustment plan sets according to the multiple brightness fusion concentration values and the multiple trend identifiers respectively to obtain the first fine-tuning optimization plan set. Finally, the street lamp control unit adjusts the brightness of the street lamps in the target park according to the first fine-tuning optimization plan set. By performing real-time brightness analysis and brightness trend analysis based on the park image set, determining multiple brightness trend analysis results, and further fusing to obtain the overall park brightness trend distribution result; calling the adjustment plan template based on the brightness trend distribution result, and performing fine-tuning optimization on the adjustment plan according to the brightness fusion concentration value and the brightness trend identifier, and adjusting the brightness of the park street lamps based on the optimized plan, it is possible to reduce or avoid waste of additional lighting resources while meeting the lighting requirements, improve the matching degree of lighting requirements and the brightness of park street lamps, achieve the technical goal of intelligently adjusting the brightness of park street lamps, and achieve the technical effect of effectively reducing the energy consumption of park street lamps.

[0043] Further, perform brightness trend analysis on the Q sets of park images respectively to obtain Q brightness trend analysis results. As shown in the appendix Figure 2 This application's step two includes:

[0044] Extract the first set of park images from the Q sets of park images. Each park image in the first set of park images has M pixel points;

[0045] Using the M pixel points as indexes, extract the brightness values of the first set of park images in chronological order to obtain M brightness value sequences;

[0046] Calculate the mean values of the M brightness value sequences respectively to determine M brightness means;

[0047] Using the first park image set as the simulated terrain area and the M brightness means as the terrain surface height of the simulated terrain area, perform raster area division on the first park image set to obtain a first raster area set;

[0048] Specifically, first, randomly extract a first park image set from the Q park image sets. The first park image set is any one of the Q park image sets, and each park image in the first park image set has M pixel points, where the value of M can be set based on the specifications and acquisition accuracy of the park image. Further, using the M pixel points as indexes, extract the brightness values of the first park image set in the order of the image acquisition time nodes to obtain M brightness value sequences. Then, calculate the mean values of the multiple brightness values in the M brightness value sequences respectively to obtain M brightness means.

[0049] Take the park area corresponding to the first park image set as the simulated terrain area, take the M brightness means as the terrain surface height of the simulated terrain area, perform raster area division on the first park image set, and obtain a first raster area set according to the division result.

[0050] Further, using the first park image set as the simulated terrain area and the M brightness means as the terrain surface height of the simulated terrain area, perform raster area division on the first park image set to obtain a first raster area set. This application further includes the following steps:

[0051] Take the minimum value of the M brightness means as the starting point for filling water into the simulated terrain area;

[0052] Take the first brightness mean that the water surface rises and submerges after filling water into the simulated terrain area as the first-stage brightness mean;

[0053] Calculate the brightness difference between the brightness mean at the starting point and the first-stage brightness mean, and determine whether the calculation result meets the preset brightness difference. If so, continue to fill water into the simulated terrain area;

[0054] If not, generate a first dividing ridge line at the first-stage brightness mean, and then continue to fill water into the simulated terrain area, where the first dividing ridge line is used to separate the first-stage brightness mean and the starting point;

[0055] When the water surface submerges the maximum value of the M brightness means, stop filling water to obtain multiple dividing ridge lines;

[0056] Use the multiple dividing ridge lines to divide the first park image set into multiple first raster areas, thereby obtaining the first raster area set.

[0057] Specifically, first, take the minimum value of the M brightness means as the starting point for filling water into the simulated terrain area, perform filling of the simulated terrain area, and take the first brightness mean that the water surface rises above after filling the simulated terrain area as the first-stage brightness mean, where the first brightness mean has the smallest deviation from the minimum brightness mean. Further calculate the brightness difference between the brightness mean at the starting point and the first-stage brightness mean to obtain the first brightness deviation, and then judge the first brightness deviation according to a preset brightness difference. The preset brightness difference is the maximum brightness difference at which two pixel points can be divided into the same area, and those skilled in the art can set it according to the actual situation.

[0058] If the first brightness difference is less than the preset brightness difference, continue to fill water into the simulated terrain area; if the first brightness difference is greater than or equal to the preset brightness difference, generate a first dividing ridge line at the first-stage brightness mean, and then continue to fill water into the simulated terrain area, where the first dividing ridge line is used to separate the first-stage brightness mean and the starting point. Thus, the goal of dividing pixel points with similar brightness into the same area and pixel points with large brightness differences into different areas, and performing grid area division on the park area corresponding to the first park image set is achieved. Furthermore, continuously perform iterative simulated water filling until the water surface rises above the maximum value of the M brightness means, then stop water filling to obtain multiple dividing ridge lines. Then use the multiple dividing ridge lines to divide the first park image set into multiple first grid areas to obtain the first grid area set.

[0059] By using the simulated water filling algorithm to perform grid area division on the first park image set, the accuracy and efficiency of grid area division can be improved, thereby achieving the technical goal of intelligent, efficient, and accurate area division.

[0060] Calculate the brightness concentration value and tendency identification for the first grid area set according to the M brightness value sequences, and combine the first grid area set to obtain the first brightness tendency analysis result;

[0061] Specifically, calculate the brightness concentration value and tendency identification for the first grid area set according to the M brightness value sequences, where the tendency identification includes positive brightness identification, steady brightness identification, and negative brightness identification, to obtain the brightness concentration value calculation result and the tendency identification result. Based on the brightness concentration value calculation result, the tendency identification result, and the first grid area set, obtain the first brightness tendency analysis result. The first brightness tendency analysis result includes multiple first grid areas, and each grid area has a brightness concentration value and a tendency identification.

[0062] Further, perform brightness concentration value calculation and trend identification on the first grid region set according to the M brightness value sequences, and obtain a first brightness trend analysis result in combination with the first grid region set. The present application further includes the following steps:

[0063] Match multiple first grid regions in the first grid region set with the M brightness means respectively to obtain multiple brightness mean sets, where the multiple brightness mean sets and the multiple first grid regions correspond one by one;

[0064] Take the mode of each of the multiple brightness mean sets as the multiple brightness concentration values of the multiple first grid regions;

[0065] Use the linear regression method to fit the multiple brightness value sequence sets in the multiple first grid regions respectively to determine the brightness development trend and obtain multiple pixel trend identification sets;

[0066] Take the pixel trend identification with a relatively large proportion in the multiple pixel trend identification sets as the trend identification.

[0067] Specifically, match the multiple first grid regions in the first grid region set with the M brightness means to obtain multiple brightness mean sets, where each brightness mean set corresponds to a first grid region. Further extract the mode of the brightness means in the brightness mean set as the brightness concentration value of the corresponding first grid region, and obtain multiple brightness concentration values corresponding to the multiple first grid regions.

[0068] Linear regression is a commonly used predictive modeling technique in statistics. It studies the linear relationship between the dependent variable and the independent variable, that is, the dependent variable is a linear combination of the independent variables. The goal of linear regression is to find the best-fitting line that can minimize the sum of the squares of the vertical distances between the data points and the line. Then use the linear regression method to fit the multiple brightness value sequence sets in the multiple first grid regions respectively, and determine the brightness development trend according to the fitting results. The brightness development trend includes an increasing trend, a stable trend, and a decreasing trend. Determine the pixel trend identification based on the brightness development trend, where the pixel trend identification corresponding to the increasing trend is a positive identification, the pixel trend identification corresponding to the stable trend is a steady-state identification, and the pixel trend identification corresponding to the decreasing trend is a negative identification. Each first grid region includes multiple pixel points, and each pixel point corresponds to a brightness value sequence, and multiple pixel trend identification sets are obtained. Finally, take the pixel trend identification with a relatively large proportion in the multiple pixel trend identification sets as the trend identification to obtain a trend identification result.

[0069] Obtain the Q brightness trend analysis results according to the Q park image sets.

[0070] Specifically, the Q sets of park images are analyzed in sequence by using the method for obtaining the first luminance trend analysis result, and Q luminance trend analysis results are obtained, where each set of park images corresponds to a luminance trend analysis result.

[0071] Furthermore, a regional grid fusion analysis is performed on the Q luminance trend analysis results to obtain a target park luminance trend distribution result. Step three of this application includes:

[0072] According to the distribution positions of the Q camera components, regional grid splicing is performed on the Q luminance trend analysis results to obtain an initial park luminance trend distribution result, where the initial park luminance trend distribution result has multiple grid fusion regions, and the multiple grid fusion regions include multiple overlapping regions and multiple non-overlapping regions;

[0073] Mean processing is performed on the multiple luminance concentration values of the multiple overlapping regions to obtain multiple overlapping luminance fusion concentration values of the overlapping regions;

[0074] The luminance concentration values of the multiple non-overlapping regions are used as multiple non-overlapping luminance fusion concentration values;

[0075] Multiple luminance fusion concentration values of the multiple grid fusion regions are obtained according to the multiple overlapping luminance fusion concentration values and the multiple non-overlapping luminance fusion concentration values.

[0076] Specifically, according to the distribution positions of the Q camera components in the park, regional grid splicing is performed on the Q luminance trend analysis results corresponding to the Q sets of park images to obtain an initial park luminance trend distribution result, where the initial park luminance trend distribution result includes multiple grid fusion regions, and each grid fusion region includes an overlapping region and a non-overlapping region, and the overlapping region is the region where there is image overlap when adjacent two camera components perform image acquisition.

[0077] Mean processing is performed on the multiple luminance concentration values of the multiple overlapping regions, and the mean calculation result is used as the overlapping luminance fusion concentration value corresponding to the corresponding overlapping region, to obtain multiple overlapping luminance fusion concentration values corresponding to the multiple overlapping regions; the luminance concentration values of the non-overlapping regions are used as non-overlapping luminance fusion concentration values, to obtain multiple non-overlapping luminance fusion concentration values corresponding to the multiple non-overlapping regions. Finally, multiple luminance fusion concentration values of the multiple grid fusion regions are obtained based on the multiple overlapping luminance fusion concentration values and the multiple non-overlapping luminance fusion concentration values, where the luminance fusion concentration value is the sum of the product of the overlapping luminance fusion concentration value in the grid fusion region and the proportion of the overlapping region and the product of the non-overlapping luminance fusion concentration value and the proportion of the non-overlapping region.

[0078] By dividing the raster fusion area into overlapping and non - overlapping areas, and calculating the luminance fusion central value of the raster fusion area based on the overlapping luminance fusion central value of the overlapping area and the non - overlapping luminance fusion central value of the non - overlapping area, the accuracy of setting the luminance fusion central value can be improved, thereby improving the accuracy and precision of subsequent street lamp adjustment scheme matching and optimization.

[0079] Further, fine - tuning and optimizing the multiple first - template adjustment scheme sets according to the multiple luminance fusion central values and the multiple tendency identifiers respectively. Step five of the present application includes:

[0080] Determining multiple fine - tuning directions based on the multiple tendency identifiers, and determining multiple fine - tuning step sizes according to the multiple luminance fusion central values;

[0081] Fine - tuning the multiple first - template adjustment scheme sets respectively according to the multiple fine - tuning step sizes and in accordance with the multiple fine - tuning directions to obtain multiple first - template fine - tuning scheme sets;

[0082] Performing fitness analysis on the multiple first - template fine - tuning scheme sets by using the expert investigation method to obtain multiple first - fitness sets;

[0083] Respectively selecting the multiple first - template fine - tuning schemes corresponding to the maximum fitness values in the multiple first - fitness sets as the updated fine - tuning directions, and performing fine - tuning according to the multiple fine - tuning step sizes to obtain multiple second - template fine - tuning scheme sets;

[0084] Performing fitness analysis on the multiple second - template fine - tuning scheme sets again to obtain multiple second - fitness sets.

[0085] Specifically, first, determine multiple fine - tuning directions based on the multiple tendency identifiers. Among them, the fine - tuning direction of the luminance positive identifier is to lower the luminance, the fine - tuning direction of the luminance steady - state identifier is to lower or raise the luminance, and the fine - tuning direction of the luminance negative identifier is to raise the luminance. Determine multiple fine - tuning step sizes according to the multiple luminance fusion central values. Among them, the luminance deviation between the luminance fusion central value and the standard luminance is proportional to the fine - tuning step size, that is, the larger the luminance deviation, the larger the fine - tuning step size. The fine - tuning step size matching can be performed by constructing a central - value - step - size mapping channel, and then multiple fine - tuning step sizes can be obtained.

[0086] The method for constructing the central value-step size mapping channel is as follows. First, set multiple sample brightness fusion central values, and use an expert system to analyze the fine-tuning step sizes corresponding to the multiple sample brightness fusion central values to determine multiple qualified fine-tuning step sizes. There is a one-to-one mapping relationship between the sample brightness fusion central values and the qualified fine-tuning step sizes. Then, based on the decision tree principle, use the sample brightness fusion central values as sub-nodes and the qualified fine-tuning step sizes corresponding to the sample brightness fusion central values as the leaf nodes of the sub-nodes to generate the central value-step size mapping channel.

[0087] Further, based on the multiple fine-tuning step sizes, fine-tune the multiple first template adjustment plan sets according to the multiple fine-tuning directions respectively to obtain multiple first template fine-tuning plan sets. Then, use the expert investigation method to perform fitness analysis on the multiple first template fine-tuning plan sets to obtain multiple first fitness sets. The expert investigation method, also known as the Delphi method, is a survey method that solicits the opinions and views of relevant experts or authorities around a certain theme or issue. The survey objects of this method are limited to the expert level and usually carried out through multiple rounds of surveys to obtain relatively accurate results after multiple rounds of survey analysis. Then, respectively select the multiple first template fine-tuning plans corresponding to the maximum fitness values in the multiple first fitness sets as the updated fine-tuning directions, and perform fine-tuning according to the multiple fine-tuning step sizes to obtain multiple second template fine-tuning plan sets. Then, use the expert investigation method again to perform fitness analysis on the multiple second template fine-tuning plan sets to obtain multiple second fitness sets.

[0088] Further, after obtaining the multiple second fitness sets, the present application further includes the following steps:

[0089] Use the multiple second template fine-tuning plans corresponding to the maximum fitness values in the multiple second fitness sets as the updated fine-tuning directions, and perform fine-tuning according to the multiple fine-tuning step sizes to obtain multiple third template fine-tuning plan sets;

[0090] After preset times of updated fine-tuning, obtain multiple target fine-tuning optimization plans, where the multiple target fine-tuning optimization plans are the template fine-tuning plans corresponding to the maximum fitness values in the multiple fine-tuning processes;

[0091] Generate the first fine-tuning optimization plan set according to the multiple target fine-tuning optimization plans.

[0092] Specifically, after obtaining multiple second fitness sets, multiple second template fine-tuning schemes corresponding to the maximum fitness values in the multiple second fitness sets are used as update fine-tuning directions, and fine-tuning is performed according to the multiple fine-tuning steps to obtain multiple third template fine-tuning scheme sets; then iterative fine-tuning is continuously performed until a preset number of iterative fine-tuning times is reached, and the preset number of iterative fine-tuning times can be set based on the optimization accuracy requirement, wherein the higher the required accuracy, the larger the preset number of iterative fine-tuning times; then the template fine-tuning scheme corresponding to the maximum fitness value in the multiple fine-tuning processes is set as the target fine-tuning optimization scheme, and multiple target fine-tuning optimization schemes are obtained; finally, the first fine-tuning optimization scheme set is generated according to the multiple target fine-tuning optimization schemes.

[0093] In summary, the intelligent street lamp adjustment method provided by this application has the following technical effects:

[0094] 1. Through real-time brightness analysis and brightness trend analysis based on the park image set, multiple brightness trend analysis results are determined, and further integrated to obtain the overall park brightness trend distribution results; based on the brightness trend distribution results, the adjustment plan template is called, and the adjustment plan is fine-tuned and optimized according to the brightness fusion concentration value and the brightness trend mark. The brightness of the park street lights is adjusted based on the optimization plan. On the premise of meeting the lighting needs, it can reduce or avoid additional lighting resource waste, improve the matching degree between the lighting needs and the brightness of the park street lights, achieve the technical goal of intelligently adjusting the brightness of the park street lights, and achieve the technical effect of effectively reducing the energy consumption of the park street lights.

[0095] 2. Based on the intelligent street light adjustment platform, the adjustment plan template is called according to the brightness fusion concentration value, which can improve the efficiency and accuracy of the adjustment plan. At the same time, the adjustment plan is fine-tuned and optimized according to the brightness fusion concentration value and the brightness trend mark, which can further improve the adaptability of the adjustment plan to the actual lighting needs of the park and improve the accuracy of the adjustment plan setting, thereby improving the accuracy, efficiency and intelligence of the brightness adjustment of the park street lights.

[0096] 3. By using the simulated watering algorithm to divide the first park image set into grid areas, the accuracy and efficiency of the grid area division can be improved, thereby achieving the technical goal of intelligent, efficient and accurate area division.

[0097] Embodiment 2

[0098] Based on the same inventive concept as the street lamp intelligent adjustment method in the aforementioned embodiment, the present application also provides a street lamp intelligent adjustment system, the system is connected to the street lamp intelligent adjustment platform, and the street lamp intelligent adjustment platform is connected to the street lamp control unit and the camera assembly array. Figure 3 , the system comprising:

[0099] The park image acquisition module 11 is configured to use the camera components arranged in an array within a preset monitoring window to acquire images of the target park, obtaining Q sets of park images. Among them, the array of camera components has Q camera components;

[0100] The brightness trend analysis module 12 is configured to perform brightness trend analysis on the Q sets of park images respectively, obtaining Q brightness trend analysis results. Among them, each brightness trend analysis result includes multiple grid regions, and each grid region has a brightness concentration value and a trend identifier. The trend identifier includes a positive brightness identifier, a steady brightness identifier, and a negative brightness identifier;

[0101] The regional grid fusion analysis module 13 is configured to perform regional grid fusion analysis on the Q brightness trend analysis results, obtaining the target park brightness trend distribution result. Among them, the target park brightness trend distribution result has multiple grid fusion regions, and the multiple grid fusion regions have multiple brightness fusion concentration values and multiple trend identifiers;

[0102] The adjustment scheme template calling module 14 is configured to send the multiple brightness fusion concentration values of the multiple grid fusion regions in the target park brightness trend distribution result to the street lamp intelligent adjustment platform, and respectively call adjustment scheme templates for the multiple grid fusion regions, obtaining multiple first template adjustment scheme sets;

[0103] The template adjustment scheme optimization module 15 is configured to perform fine-tuning optimization on the multiple first template adjustment scheme sets respectively according to the multiple brightness fusion concentration values and the multiple trend identifiers, obtaining the first fine-tuning optimization scheme set;

[0104] The street lamp brightness adjustment module 16 is configured to use the street lamp control unit to adjust the brightness of the street lamps in the target park according to the first fine-tuning optimization scheme set.

[0105] Furthermore, the brightness trend analysis module 12 in the system is further configured to:

[0106] Extract the first set of park images from the Q sets of park images. Each park image in the first set of park images has M pixel points;

[0107] Taking the M pixel points as indexes, extract the brightness values of the first set of park images in chronological order, obtaining M brightness value sequences;

[0108] Calculate the mean value of each of the M luminance value sequences to determine M luminance means;

[0109] Using the first park image set as the simulated terrain area and the M luminance means as the terrain surface heights of the simulated terrain area, divide the first park image set into grid areas to obtain a first grid area set;

[0110] Calculate the luminance concentration value and tendency identification for the first grid area set according to the M luminance value sequences, and combine with the first grid area set to obtain a first luminance tendency analysis result;

[0111] Obtain the Q luminance tendency analysis results according to the Q park image sets.

[0112] Furthermore, the luminance tendency analysis module 12 in the system is further configured to:

[0113] Take the minimum value of the M luminance means as the starting point for filling water into the simulated terrain area;

[0114] Take the first luminance mean that the water surface rises above after filling water into the simulated terrain area as the first-stage luminance mean;

[0115] Calculate the luminance difference between the luminance mean at the starting point and the first-stage luminance mean, and determine whether the calculation result meets a preset luminance difference. If so, continue to fill water into the simulated terrain area;

[0116] If not, generate a first dividing ridge line at the first-stage luminance mean, and then continue to fill water into the simulated terrain area, where the first dividing ridge line is used to separate the first-stage luminance mean and the starting point;

[0117] When the water surface exceeds the maximum value of the M luminance means, stop filling water to obtain multiple dividing ridge lines;

[0118] Use the multiple dividing ridge lines to divide the first park image set into multiple first grid areas, thereby obtaining the first grid area set.

[0119] Furthermore, the luminance tendency analysis module 12 in the system is further configured to:

[0120] Match each of the multiple first grid areas in the first grid area set with the M luminance means respectively to obtain multiple luminance mean sets, where the multiple luminance mean sets and the multiple first grid areas correspond one by one;

[0121] Take the mode of each of the multiple luminance mean sets as the multiple luminance concentration values of the multiple first grid areas;

[0122] Using the linear regression method, fit the multiple sets of brightness value sequences in the multiple first grid regions respectively to determine the brightness development trend and obtain multiple sets of pixel tendency identifiers;

[0123] Take the pixel tendency identifier with a relatively large proportion in the multiple sets of pixel tendency identifiers as the tendency identifier.

[0124] Furthermore, the area grid fusion analysis module 13 in the system is also used for:

[0125] According to the distribution positions of the Q camera components, perform area grid splicing on the Q brightness tendency analysis results to obtain an initial park brightness tendency distribution result, where the initial park brightness tendency distribution result has multiple grid fusion regions, and the multiple grid fusion regions include multiple overlapping regions and multiple non-overlapping regions;

[0126] Perform mean processing on the multiple brightness concentration values of the multiple overlapping regions to obtain multiple overlapping brightness fusion concentration values of the overlapping regions;

[0127] Take the brightness concentration values of the multiple non-overlapping regions as multiple non-overlapping brightness fusion concentration values;

[0128] Obtain multiple brightness fusion concentration values of the multiple grid fusion regions according to the multiple overlapping brightness fusion concentration values and the multiple non-overlapping brightness fusion concentration values.

[0129] Furthermore, the template adjustment scheme optimization module 15 in the system is also used for:

[0130] Determine multiple fine-tuning directions based on the multiple tendency identifiers, and determine multiple fine-tuning steps according to the multiple brightness fusion concentration values;

[0131] According to the multiple fine-tuning steps, fine-tune the multiple first template adjustment scheme sets respectively in the multiple fine-tuning directions to obtain multiple first template fine-tuning scheme sets;

[0132] Use the expert investigation method to perform fitness analysis on the multiple first template fine-tuning scheme sets to obtain multiple first fitness sets;

[0133] Select the multiple first template fine-tuning schemes corresponding to the maximum fitness values in the multiple first fitness sets respectively as the updated fine-tuning directions, and perform fine-tuning according to the multiple fine-tuning steps to obtain multiple second template fine-tuning scheme sets;

[0134] Perform fitness analysis on the multiple second template fine-tuning scheme sets again to obtain multiple second fitness sets.

[0135] Further, the template adjustment scheme optimization module 15 in the system is further configured to:

[0136] Using the multiple second template fine-tuning schemes corresponding to the maximum fitness value in the multiple second fitness sets as the updated fine-tuning directions, and performing fine-tuning according to the multiple fine-tuning step lengths to obtain multiple third template fine-tuning scheme sets;

[0137] After performing updated fine-tuning for a preset number of times, multiple target fine-tuning optimization schemes are obtained, where the multiple target fine-tuning optimization schemes are the template fine-tuning schemes corresponding to the maximum fitness value during multiple fine-tuning processes;

[0138] Generating the first fine-tuning optimization scheme set according to the multiple target fine-tuning optimization schemes.

[0139] The various embodiments in this specification are described in a progressive manner. The key point of each embodiment is the difference from other embodiments. The intelligent street lamp adjustment method and specific examples in the foregoing Embodiment 1 are equally applicable to the intelligent street lamp adjustment system in this embodiment. Through the foregoing detailed description of the intelligent street lamp adjustment method, those skilled in the art can clearly know the intelligent street lamp adjustment system in this embodiment. Therefore, for the sake of brevity of the specification, it will not be elaborated here. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the description of the method part.

[0140] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0141] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalent technologies, the present application is also intended to include these changes and modifications.

Claims

1. An intelligent adjustment method for street lamps, characterized in that, The method is applied to a street lamp intelligent regulation platform, which is communicatively connected in an array layout with a street lamp control unit and a camera component layout array. The method includes: Using the camera component layout array to collect images of a target park within a preset monitoring window, obtaining Q sets of park images, where the camera component layout array has Q camera components; Performing brightness trend analysis on the Q sets of park images respectively to obtain Q brightness trend analysis results. Each brightness trend analysis result includes multiple grid regions, and each grid region has a brightness concentration value and a trend identifier. The trend identifier includes a positive brightness identifier, a steady brightness identifier, and a negative brightness identifier; Performing regional grid fusion analysis on the Q brightness trend analysis results to obtain a brightness trend distribution result of the target park. The brightness trend distribution result of the target park has multiple grid fusion regions, and the multiple grid fusion regions have multiple brightness fusion concentration values and multiple trend identifiers; Sending the multiple brightness fusion concentration values of the multiple grid fusion regions in the brightness trend distribution result of the target park to the street lamp intelligent regulation platform, and respectively invoking adjustment plan templates for the multiple grid fusion regions to obtain multiple first template adjustment plan sets; Performing fine-tuning and optimization on the multiple first template adjustment plan sets respectively according to the multiple brightness fusion concentration values and the multiple trend identifiers to obtain a first fine-tuning and optimization plan set; The street lamp control unit adjusts the brightness of the street lamps in the target park according to the first fine-tuning and optimization plan set; Performing brightness trend analysis on the Q sets of park images respectively to obtain Q brightness trend analysis results. The method includes: Extracting a first set of park images from the Q sets of park images. Each park image in the first set of park images has M pixel points; Taking the M pixel points as indexes, extracting brightness values from the first set of park images in time sequence to obtain M brightness value sequences; Calculating the mean values of the M brightness value sequences respectively to determine M brightness means; Taking the first set of park images as a simulated terrain area and the M brightness means as the terrain surface height of the simulated terrain area, dividing the first set of park images into grid regions to obtain a first set of grid regions; Calculating the brightness concentration values and trend identifiers for the first set of grid regions according to the M brightness value sequences, and combining the first set of grid regions to obtain a first brightness trend analysis result; Obtaining the Q brightness trend analysis results according to the Q sets of park images; Taking the first set of park images as a simulated terrain area and the M brightness means as the terrain surface height of the simulated terrain area, dividing the first set of park images into grid regions to obtain a first set of grid regions. The method includes: Taking the minimum value of the M brightness means as the starting point for filling water into the simulated terrain area; Taking the first brightness mean that the water surface rises above after filling water into the simulated terrain area as the first-stage brightness mean; Calculate the brightness difference between the brightness mean at the starting point and the brightness mean in the first stage, and determine whether the calculation result meets the preset brightness difference. If so, continue to fill water into the simulated terrain area; If not, generate a first dividing ridge line at the brightness mean in the first stage, and then continue to fill water into the simulated terrain area, where the first dividing ridge line is used to separate the brightness mean in the first stage and the starting point; When the water surface exceeds the maximum value among the M brightness means, stop filling water to obtain multiple dividing ridge lines; Use the multiple dividing ridge lines to divide the first park image set into multiple first grid regions, thereby obtaining the first grid region set; 2. The method according to claim 1, wherein Perform brightness concentration value calculation and trend identification on the first grid region set according to the M brightness value sequences, and combine the first grid region set to obtain a first brightness trend analysis result. The method includes: Match multiple first grid regions in the first grid region set with the M brightness means respectively to obtain multiple brightness mean sets, where the multiple brightness mean sets and the multiple first grid regions correspond one by one; Respectively use the modes of the multiple brightness mean sets as the multiple brightness concentration values of the multiple first grid regions; Use the linear regression method to fit the multiple brightness value sequence sets in the multiple first grid regions respectively to determine the brightness development trend and obtain multiple pixel trend identification sets; Use the pixel trend identification with a relatively large proportion in the multiple pixel trend identification sets as the trend identification; 3. The method according to claim 1, wherein Perform regional grid fusion analysis on the Q brightness trend analysis results to obtain the target park brightness trend distribution result. The method includes: According to the distribution positions of the Q camera components, perform regional grid splicing on the Q brightness trend analysis results to obtain the initial park brightness trend distribution result, where the initial park brightness trend distribution result has multiple grid fusion regions, and the multiple grid fusion regions include multiple overlapping regions and multiple non-overlapping regions; Perform mean processing on the multiple brightness concentration values of the multiple overlapping regions to obtain multiple overlapping brightness fusion concentration values of the overlapping regions; Use the brightness concentration values of the multiple non-overlapping regions as multiple non-overlapping brightness fusion concentration values; Obtain the multiple brightness fusion concentration values of the multiple grid fusion regions according to the multiple overlapping brightness fusion concentration values and the multiple non-overlapping brightness fusion concentration values; 4. The method according to claim 1, wherein Fine-tune and optimize the multiple first template adjustment plan sets according to the multiple brightness fusion concentration values and the multiple trend identifications respectively. The method includes: Determine multiple fine-tuning directions based on the multiple trend identifications, and determine multiple fine-tuning steps according to the multiple brightness fusion concentration values; According to the multiple fine-tuning steps, fine-tune the multiple first template adjustment plan sets respectively in the multiple fine-tuning directions to obtain multiple first template fine-tuning plan sets; Use the expert investigation method to perform fitness analysis on the multiple first template fine-tuning plan sets to obtain multiple first fitness sets; Respectively select multiple first template fine-tuning schemes corresponding to the maximum fitness values in the multiple first fitness sets as the updated fine-tuning directions, and perform fine-tuning according to the multiple fine-tuning step sizes to obtain multiple second template fine-tuning scheme sets; Perform fitness analysis on the multiple second template fine-tuning scheme sets again to obtain multiple second fitness sets.

5. The method according to claim 4, wherein After obtaining the multiple second fitness sets, the method includes: Use multiple second template fine-tuning schemes corresponding to the maximum fitness values in the multiple second fitness sets as the updated fine-tuning directions, and perform fine-tuning according to the multiple fine-tuning step sizes to obtain multiple third template fine-tuning scheme sets; After a preset number of updated fine-tuning operations, obtain multiple target fine-tuning optimization schemes, where the multiple target fine-tuning optimization schemes are template fine-tuning schemes corresponding to the maximum fitness values during multiple fine-tuning processes; Generate the first fine-tuning optimization scheme set according to the multiple target fine-tuning optimization schemes.

6. An intelligent street lamp adjustment system, characterized in that, For implementing the steps of the method according to any one of claims 1 to 5, the system is communicatively connected to the street lamp intelligent adjustment platform, and the street lamp intelligent adjustment platform is communicatively connected to the street lamp control unit and the camera component layout array. The method includes: A park image acquisition module, which is used to use the camera component layout array to collect images of the target park within a preset monitoring window to obtain Q park image sets, where the camera component layout array has Q camera components; A brightness trend analysis module, which is used to perform brightness trend analysis on the Q park image sets respectively to obtain Q brightness trend analysis results. Each brightness trend analysis result includes multiple grid regions, and each grid region has a brightness concentration value and a trend identifier. The trend identifier includes a positive brightness identifier, a brightness steady state identifier, and a negative brightness identifier; A regional grid fusion analysis module, which is used to perform regional grid fusion analysis on the Q brightness trend analysis results to obtain a target park brightness trend distribution result. The target park brightness trend distribution result has multiple grid fusion regions, and the multiple grid fusion regions have multiple brightness fusion concentration values and multiple trend identifiers; An adjustment scheme template call module, which is used to send the multiple brightness fusion concentration values of the multiple grid fusion regions in the target park brightness trend distribution result to the street lamp intelligent adjustment platform, and respectively perform adjustment scheme template calls on the multiple grid fusion regions to obtain multiple first template adjustment scheme sets; A template adjustment scheme optimization module, which is used to fine-tune and optimize the multiple first template adjustment scheme sets according to the multiple brightness fusion concentration values and the multiple trend identifiers respectively to obtain a first fine-tuning optimization scheme set; A street lamp brightness adjustment module, which is used to adjust the brightness of the street lamps in the target park by the street lamp control unit according to the first fine-tuning optimization scheme set.

Images