Method for analyzing terminal position distribution based on public network base station data

Abstract

The invention relates to the technical field of big data analysis, and provides a method for analyzing terminal position distribution based on public network base station data. The method comprises the following steps: acquiring air quality data of a target area based on public network base station data, and calculating a pollution index mean value and a fluctuation degree; constructing a pollution topological graph and a fluctuation characteristic graph according to the pollution index mean value and the fluctuation degree; acquiring weather change and people flow density prediction information, adjusting the pollution topological graph and the fluctuation characteristic graph, and outputting a corrected pollution distribution topological graph and a corrected index fluctuation characteristic graph; and optimizing a sensor layout scheme and adjusting the position according to the correction graph. The technical problem that in the prior art, the PM2.5 sensor layout position lacks dynamic and accurate layout, and consequently the air quality monitoring precision is insufficient is solved, and the technical effects that terminal position dynamic analysis is conducted on public network base station data through the big data analysis technology, the sensor layout position is accurately optimized, and the air quality monitoring precision is improved are achieved.

Description

Technical Field

[0001] The present application relates to the field of big data analysis technology, and specifically to a method for analyzing terminal location distribution based on public network base station data. Background Art

[0002] Due to the accelerated urbanization process and the continued expansion of industrial activities, air pollution has become increasingly prominent, with PM2.5 pollution posing a particularly significant threat to residents' health. PM2.5 refers to particulate matter in the air with a diameter of less than or equal to 2.5 microns, also known as fine particulate matter. It is typically produced by motor vehicle exhaust, industrial emissions, coal-fired power generation, construction dust, and biomass combustion. Currently, common PM2.5 monitoring methods are mostly based on fixed-point sensor deployments, collecting air pollution data through a limited number of monitoring points. This approach struggles to dynamically adapt to changes in urban population density, weather factors, and sudden pollution events, resulting in insufficient accuracy and real-time monitoring data. Furthermore, existing technologies often overlook the potential application value of public network base station data in terminal location analysis and crowd activity monitoring. This results in sensor deployment solutions lacking the ability to respond to dynamic changes in terminal location based on big data analysis, making it impossible to mine the correlation between population migration patterns and pollution exposure risks through massive communication signal data, further reducing the spatial coverage and accuracy of air quality monitoring. Therefore, there is an urgent need for a method that can combine public network base station data, analyze terminal location changes in real time through big data analysis technology, and accurately guide the dynamic layout of PM2.5 sensors to improve the accuracy and reliability of air quality monitoring. Summary of the Invention

[0003] This application provides a method for analyzing terminal location distribution based on public network base station data, aiming to solve the technical problem in the existing technology that the PM2.5 sensor deployment positions lack dynamic and precise layout, resulting in insufficient air quality monitoring accuracy.

[0004] The present application provides a method for analyzing terminal location distribution based on public network base station data, the method comprising: obtaining air quality monitoring data of a target area in a historical time zone based on the public network base station data, and calculating a plurality of air pollution index means and a plurality of index fluctuations; constructing a pollution distribution topology map and an index fluctuation characteristic map based on the plurality of air pollution index means and the plurality of index fluctuations; obtaining predicted weather change information and predicted crowd density distribution of the target area in a preset future time period, adjusting the pollution distribution topology map and the index fluctuation characteristic map, and outputting a corrected pollution distribution topology map and a corrected index fluctuation characteristic map; optimizing a sensor deployment plan based on the corrected pollution distribution topology map and the corrected index fluctuation characteristic map to obtain an adapted deployment plan, and adjusting the sensor deployment position in the preset future time period.

[0005] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0006] The aforementioned method for analyzing terminal location distribution based on public network base station data first uses public network base station data to collect historical air quality data for the target area and calculate the average and fluctuation levels of the air pollution index. Subsequently, based on this data, a topological map of the pollution distribution and a fluctuation characteristic map are constructed. The original map is then corrected and optimized by obtaining forecasts of future weather changes and pedestrian density. Finally, based on the optimized map, a more appropriate PM2.5 sensor layout plan is determined, and sensor positions are adjusted during predetermined future periods to improve air quality monitoring accuracy.

[0007] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0009] Figure 1 Schematic diagram of a flow chart of a method for analyzing terminal location distribution based on public network base station data in one embodiment.

[0010] Figure 2 Schematic diagram of a flow chart of calculating the mean value and index volatility of the air pollution index in a method for analyzing terminal location distribution based on public network base station data in one embodiment. DETAILED DESCRIPTION

[0011] The embodiments of the present application provide a method for analyzing terminal location distribution based on public network base station data to solve the technical problem in the prior art that the PM2.5 sensor deployment positions lack dynamic and precise layout, resulting in insufficient air quality monitoring accuracy.

[0012] The following will be combined with the accompanying drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only some of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0013] It should be noted that the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or are inherent to these processes, methods, products or devices.

[0014] Examples, such as Figure 1 As shown, the present application provides a method for analyzing terminal location distribution based on public network base station data, the method comprising:

[0015] Based on public network base station data, air quality monitoring data of the target area in the historical time zone is obtained, and several air pollution index means and several index fluctuations are calculated.

[0016] In an embodiment of the present application, historical air quality monitoring data within the target area is first obtained from public network base station data to form multiple time series of air pollution index records. Subsequently, mathematical statistical processing is performed on the pollution index time series corresponding to each location point, and its numerical mean is calculated to represent the average pollution level of the point during the period. Then, a volatility analysis is performed on each time series to obtain the index volatility, which is used to reflect the stability and variation range of the pollution index in the area. After all time series are calculated, several pollution index means and several index volatility values ​​are formed for subsequent pollution feature map construction.

[0017] Further, if Figure 2 As shown, this application provides a method for obtaining air quality monitoring data of a target area in a historical time zone based on public network base station data, and calculating several air pollution index means and several index fluctuations, including:

[0018] Based on public network base station data, air quality monitoring data of a target area in a historical time zone is obtained, wherein the air quality monitoring data is obtained based on an initial sensor deployment plan, and includes several air pollution index sequences, and the initial sensor deployment plan includes several PM2.5 sensors at different positions; mean calculations are performed on the several air pollution index sequences respectively to obtain several air pollution index means; volatility analysis is performed on the several air pollution index sequences respectively, and several index volatility is output, wherein the index volatility is the ratio of the index standard deviation to the index mean.

[0019] Preferably, first, the air quality monitoring data of the target area in the historical time zone is located and extracted from the public network base station data according to the geographical identification. These historical air quality monitoring data are collected based on multiple PM2.5 sensors deployed in different locations in the area through the initial sensor deployment plan. Each air pollution index sequence in the historical air quality monitoring data corresponds to a specific sensor point. Subsequently, a mean calculation is performed on each pollution index sequence to obtain the air pollution index mean of each pollution index sequence, which is used to reflect the average pollution level of each point in the selected time period. In addition, a volatility analysis is performed on each air pollution index sequence, the standard deviation is calculated, and the standard deviation is divided by the mean to obtain the corresponding index volatility, which is used to describe the stability or drastic change of the pollution data at the point. Finally, the air pollution index mean and index volatility of all monitoring points are output to provide basic input for constructing a pollution map and optimizing the deployment plan.

[0020] A pollution distribution topology map and an index fluctuation characteristic map are constructed based on the several air pollution index means and the several index fluctuation degrees.

[0021] In one embodiment, the geographic coordinates of each PM2.5 sensor in the initial deployment plan are first matched with the corresponding air pollution index mean and index volatility, constructing a pollution index mean array and a pollution index volatility array, with each data point corresponding to each sensor location. Subsequently, a pre-trained pollution distribution fitter and a fluctuation feature fitter are invoked, inputting the pollution index mean array into the pollution distribution fitter and the index volatility array into the fluctuation feature fitter, respectively. Based on existing pollution evolution patterns and geographic spatial characteristics, the fitter performs continuous spatial interpolation and topological modeling of the pollution level and fluctuation at each point, thereby outputting a pollution distribution topology map and an index fluctuation feature map covering the entire target area. The pollution distribution topology map describes the spatial diffusion structure of pollution concentration, while the index fluctuation feature map reflects the sensitivity and uncertainty of pollution changes within the region. These two maps can comprehensively reflect the spatial distribution of regional pollution and the distribution of fluctuation risks, providing a quantifiable spatial basis for subsequent dynamic adjustments that integrate factors such as weather and human traffic, thereby significantly improving the scientific and targeted nature of the sensor deployment plan.

[0022] Furthermore, the present application provides a method for constructing a pollution distribution topology map and an index fluctuation characteristic map based on the several air pollution index means and several index fluctuation degrees, including:

[0023] According to the position coordinates of several sensors in the initial sensor layout plan, the several air pollution index means and several index fluctuations are arranged respectively to construct an air pollution index mean array and an index fluctuation array; a pollution distribution fitter and an index fluctuation fitter are pre-trained; the pollution distribution fitter and the index fluctuation fitter are used to fit the air pollution index mean array and the index fluctuation array, and a pollution distribution topology map and an index fluctuation characteristic map are output.

[0024] Optionally, first, based on the spatial coordinates of each PM2.5 sensor in the initial sensor deployment plan, all sensors are numbered in geographic order or grid numbering, and the corresponding air pollution index mean values ​​for each sensor are arranged sequentially to construct an air pollution index mean array. Similarly, the index fluctuations corresponding to each sensor point are arranged sequentially to construct an index fluctuation array. Together, these two represent the spatial pollution level and fluctuation characteristics. Subsequently, a pollution distribution fitter and an index fluctuation fitter, which have been pre-trained based on historical environmental monitoring records, are called. These two fitters typically use generative adversarial networks (GANs) or other deep learning structures and can learn and restore pollution evolution patterns and fluctuation trends in continuous space from the input pollution data array. The constructed air pollution index mean array is then input into the pollution distribution fitter, which performs fitting interpolation of pollution levels in the spatial dimension and outputs a pollution distribution topology map within the continuous space. Simultaneously, the index fluctuation array is input into the index fluctuation fitter to extract the spatial pattern of pollution fluctuations at each location and output an index fluctuation feature map. The obtained pollution distribution topology map and index fluctuation characteristic map can present the spatial aggregation areas of pollution concentration and its fluctuation sensitive areas from a macro perspective, effectively supporting subsequent dynamic environmental factor correction and sensor layout optimization, and improving the overall monitoring response capability and decision-making accuracy to pollution evolution.

[0025] Furthermore, the present application provides a pre-trained pollution distribution fitter and an exponential fluctuation fitter, including:

[0026] According to the historical environmental monitoring records of the target area, a sample air pollution index mean array set, a sample pollution distribution topology atlas, a sample index volatility array set and a sample index fluctuation feature atlas are obtained through processing; the sample air pollution index mean array set and the sample pollution distribution topology atlas are used to train the generator and discriminator of the generative adversarial network until the network converges to obtain a pollution distribution fitter; the sample index volatility array set and the sample index fluctuation feature atlas are used to train the generator and discriminator of the generative adversarial network until the network converges to obtain an index fluctuation fitter.

[0027] Optionally, first, collect environmental monitoring records of the target area in multiple historical time periods, and extract the PM2.5 sensor layout data corresponding to each historical time period. For each time period, calculate the pollution index mean and index volatility according to the pollution index monitoring sequence of each sensor point, and arrange them in order of sensor position coordinates to construct a sample air pollution index mean array and a sample index volatility array. At the same time, based on the spatial pollution distribution and fluctuation characteristics of each historical time period, generate the corresponding sample pollution distribution topology map and sample index fluctuation feature map. Through the above steps, multiple historical sample pairs are formed, which respectively constitute the sample air pollution index mean array set-sample pollution distribution topology map set, and the sample index volatility array set-sample index fluctuation feature map set. Subsequently, the generator and discriminator in the generative adversarial network (GAN) structure are used to train the above two sample sets separately. Taking the training of pollution distribution fitter as an example, we first build a generative adversarial network architecture, in which the generator is a mapping network based on convolutional neural network (CNN), which is used to receive the input sample air pollution index mean array and output the predicted pollution distribution topology map. The discriminator is also a CNN structure, which is used to receive the input topology map and determine whether it comes from the real pollution distribution topology map set. During the training process, the air pollution index mean array of each group of samples will be input into the generator, and the generator will generate a simulated pollution topology map, which will then be compared with the real pollution distribution topology map of the same group of samples. Figure 1 The input image is fed into the discriminator, which attempts to distinguish whether the input image is a real image or a generated image. Its output is a binary classification result (real / fake). The generator adjusts its parameters based on the feedback from the discriminator, making the generated image increasingly realistic and as close to the real contaminated image as possible. During training, the cross-entropy loss function is used as the optimization target for the discriminator and generator. At the same time, the structural similarity index (SSIM) or mean square error (MSE) can be introduced as auxiliary constraints to improve the fidelity of the topological structure. The training process adopts an alternating optimization strategy, that is, in each training round, the generator is first fixed to train the discriminator, and then the discriminator is fixed to train the generator. The entire adversarial training process continues for multiple rounds until the discriminator's discrimination accuracy can no longer be improved and the contaminated topology map output by the generator is highly consistent with the real image, indicating that the network has converged. At this point, the generator weights are frozen and used as a contamination distribution fitter for subsequent actual contaminated image prediction. Similarly, using the sample exponential fluctuation array as input and the sample exponential fluctuation feature map as the output target, another set of GAN networks was trained to optimize its generation capability and discrimination accuracy, ultimately resulting in an exponential fluctuation fitter. In summary, this training process, through deep learning from historical data, achieves modeling and generalization capabilities for the spatial distribution and fluctuation trends of pollution. This enables rapid and accurate output of pollution topology and fluctuation maps for continuous areas when new data is input, providing a data foundation for dynamically optimizing sensor layout.

[0028] The predicted weather change information and predicted crowd density distribution of the target area within a preset future period are obtained, the pollution distribution topology map and the index fluctuation characteristic map are adjusted, and the corrected pollution distribution topology map and the corrected index fluctuation characteristic map are output.

[0029] In one embodiment, weather forecast information for a target area within a preset future time period is first obtained based on public network base station data. This weather forecast information can reflect the air flow state and pollutant diffusion trends over the next period. Simultaneously, a prediction model is constructed, combining historical pedestrian activity trajectory data and weather change characteristics, to infer the pedestrian density distribution within the target area within the future time period, thereby representing the pollution disturbances that may be caused by crowd activity. Subsequently, the predicted weather change information and pedestrian density distribution are input as input features into a pre-trained feature correction branch, which combines the original pollution distribution topology map and the index fluctuation characteristic map to correct the map. During the correction process, the feature correction branch performs spatial offset correction based on the impact of future meteorological factors on pollutant migration paths. It also adjusts the original map to account for potential pollution intensification or disturbance fluctuations caused by high-density crowd areas, thereby outputting a corrected pollution distribution topology map reflecting the actual future pollution diffusion situation and a corrected index fluctuation characteristic map reflecting the uncertainty level of future pollution. In summary, through this correction step, the original pollution distribution topology map is made adaptable to future environmental factors, improving the foresight and accuracy of subsequent sensor deployment decisions.

[0030] Furthermore, the present application provides a method for obtaining predicted weather change information and predicted crowd density distribution in a target area within a preset future period, adjusting the pollution distribution topology map and the index fluctuation characteristic map, and outputting a corrected pollution distribution topology map and a corrected index fluctuation characteristic map, including:

[0031] Based on the public network base station data, the predicted weather change information of the target area within a preset future time period is obtained, wherein the predicted weather change information includes at least the predicted temperature, the predicted wind direction and the predicted wind speed; based on the public network base station data, the historical crowd density distribution sequence of the target area is obtained, and the predicted crowd density distribution is obtained by combining the predicted weather change information with analysis; according to the predicted weather change information and the predicted crowd density distribution, the pollution distribution topology map and the index fluctuation characteristic map are adjusted, and the corrected pollution distribution topology map and the corrected index fluctuation characteristic map are output.

[0032] Preferably, first, based on the public network base station data, the weather forecast information of the target area in the preset future time period is extracted, including the predicted temperature, predicted wind direction and predicted wind speed, etc. This information is usually provided by a third-party meteorological service platform. By associating the public network base station with the terminal geographic location, high-resolution meteorological data matching at the regional level can be achieved, forming a set of weather change information with time series characteristics in the future time period. Subsequently, based on the historical positioning data of the public network base station, the changes in the number of terminal activities in the target area under multiple historical time periods are counted, and a crowd density distribution sequence is constructed. Combined with the weather condition data corresponding to these historical time periods, a training sample set is formed, and modeling is performed using a long short-term memory network (LSTM) or other time series prediction model. Afterwards, by inputting the predicted weather change information, the model can infer the spatiotemporal distribution trend of crowd activities in the future time period, thereby outputting the predicted crowd density distribution, accurately reflecting the future crowd flow and aggregation characteristics in the region. Then, the system uses the predicted weather change information and predicted pedestrian density distribution as input features, combined with the existing pollution distribution topology map and index fluctuation characteristic map, and inputs them into the pre-trained feature correction branch. By simulating the impact of wind speed and direction on pollution migration paths, the pollution distribution topology map is corrected for spatial offset, diffusion, or convergence. By analyzing the trend of pedestrian density fluctuations, areas likely to experience future pollution fluctuations are identified. Based on this, the index fluctuation map is enhanced or smoothed, resulting in a corrected pollution distribution topology map and a corrected index fluctuation characteristic map. The corrected pollution distribution topology map reflects the actual spatial distribution trend of pollutants under the combined influence of future meteorological conditions and pedestrian activity, while the corrected index fluctuation characteristic map reveals the potential fluctuation intensity of pollution data in various regions. This adjustment process provides a more forward-looking and accurate decision-making basis for subsequent sensor deployment.

[0033] Furthermore, the present application provides a method for obtaining a historical crowd density distribution sequence of a target area based on public network base station data, and combining the predicted weather change information with analysis to obtain a predicted crowd density distribution, including:

[0034] Based on public network base station data, a sample crowd density distribution sequence set and a sample weather change information set are collected, and the historical crowd density distribution after a historical period is obtained as the sample predicted crowd density distribution to obtain a sample predicted crowd density distribution set; the sample crowd density distribution sequence set, the sample weather change information set and the sample predicted crowd density distribution set are used as training data to train a long short-term memory network until convergence to obtain a crowd density prediction model; the crowd density prediction model is used to analyze the historical crowd density distribution sequence and the predicted weather change information to obtain a predicted crowd density distribution.

[0035] Optionally, first, based on the public network base station data, select multiple representative historical time periods, and extract the location activity information of the terminal devices in the target area, count the changes in the number of terminals in each geographical grid per unit time, and construct a sample crowd density distribution sequence set. The sequence data has a clear time order and reflects the dynamic distribution characteristics of the crowd in different time periods. At the same time, the corresponding historical weather change information is extracted during these time periods to form a sample weather change information set, which includes multi-dimensional meteorological factors such as temperature, wind speed, and wind direction. Furthermore, the crowd density distribution data continues to be collected in the continuous time periods after the corresponding time period of each group of sample sequences as a sample predicted crowd density distribution set to represent the model prediction target value. Subsequently, the three types of data were combined into a complete training sample (the input is the crowd density sequence and weather information, and the output is the crowd density in the future time period). A long short-term memory network was used as the structure of the crowd density prediction model. The input layer of the crowd density prediction model accepts the historical crowd density sequence and weather information. The temporal dependency is processed by multiple layers of LSTM units. The output layer predicts the crowd density distribution at each moment in the future. The mean squared error (MSE) is used as the loss function during the training process, and the backpropagation algorithm is used to continuously optimize the model weights until the validation set error stabilizes and the model converges. Finally, the trained crowd density prediction model is applied to the actual prediction task. That is, the current historical crowd density distribution sequence and predicted weather change information are input into the crowd density prediction model for forward calculation, and the predicted crowd density distribution for the preset future time period is output. This predicted crowd density distribution can spatially display the high-incidence areas of crowd gathering and temporally reflect the trend of crowd flow changes, providing a key dynamic basis for pollution map correction and sensor deployment.

[0036] Furthermore, the present application provides for adjusting the pollution distribution topology map and the index fluctuation characteristic map according to the predicted weather change information and the predicted crowd density distribution, including:

[0037] Based on public network base station data, a sample weather change information set, a sample crowd density distribution set, a sample pollution distribution topology map set and a sample index fluctuation characteristic map set are collected, and the historical pollution distribution topology map and the historical index fluctuation characteristic map after the historical period are obtained as the sample correction pollution distribution topology map and the sample correction index fluctuation characteristic map, and a sample correction pollution distribution topology map set and a sample correction index fluctuation characteristic map set are obtained; the sample weather change information set, the sample crowd density distribution set and the sample pollution distribution topology map set are used as input, and the sample correction pollution distribution topology map set is used as supervision to train a generative adversarial network until convergence, to obtain a first feature correction branch; the sample weather change information set, the sample crowd density distribution set and the sample index fluctuation characteristic map set are used as input, and the sample correction index fluctuation characteristic map set is used as supervision to train a generative adversarial network until convergence, to obtain a second feature correction branch; the first feature correction branch and the second feature correction branch are used to adjust the pollution distribution topology map and the index fluctuation characteristic map according to the predicted weather change information and the predicted crowd density distribution.

[0038] Optionally, first, based on public network base station data, sample weather change information sets, sample crowd density distribution sets, sample pollution distribution topology maps, and sample index fluctuation feature map sets are collected and sorted within multiple historical time windows, and then the actual pollution change results of each group of sample time periods are obtained for several hours or days, that is, the corresponding historical pollution distribution topology maps and historical index fluctuation feature maps, which are used as correction targets to respectively constitute a sample correction pollution distribution topology map set and a sample correction index fluctuation feature map set for supervised learning. Subsequently, two generative adversarial networks (GANs) are constructed for feature correction, wherein the first group of networks is used to train the first feature correction branch for pollution distribution map correction. The generator in the first group of networks takes sample weather change information, sample crowd density distribution, and sample pollution topology map as input, and outputs a fitted correction pollution distribution topology map. The discriminator is used to judge the difference between the generated correction pollution distribution topology map and the real sample correction pollution distribution topology map. The specific training steps are similar to those mentioned above and will not be repeated here. The second network is used to train the second feature correction branch for exponential fluctuation map correction. Similarly, the generator inputs sample weather change information, sample crowd density distribution, and sample exponential fluctuation feature maps, and outputs a fitted corrected exponential fluctuation feature map. The discriminator then identifies the authenticity of the generated feature map. Finally, the predicted future weather change information and predicted crowd density distribution map are used as input, combined with the current pollution distribution topology map and fluctuation feature map, and input into the first and second feature correction branches for forward reasoning, respectively. The corrected pollution distribution topology map and corrected exponential fluctuation feature map are obtained. Together, they reflect the pollution intensity distribution and its fluctuation characteristics in the future time period, providing a more realistic basis for pollution environment perception for subsequent sensor deployment optimization, improving the foresight and rationality of subsequent sensor deployment, and ensuring high-precision air quality monitoring.

[0039] The sensor layout scheme is optimized according to the corrected pollution distribution topology map and the corrected index fluctuation characteristic map to obtain an adapted layout scheme, and the sensor layout position is adjusted within the preset future time period.

[0040] In one embodiment, after obtaining the corrected pollution distribution topology map and the corrected index fluctuation characteristic map, the corrected index fluctuation characteristic map and the corrected index fluctuation characteristic map are used to perform adaptability analysis on the randomly generated sensor layout scheme, evaluate the coverage and response capability of the layout scheme in the target area, and comprehensively form a layout fitness score. In each round of iteration, the schemes with higher fitness are retained and adjusted and combined, and the sensor position distribution is gradually optimized until the preset number of optimization rounds or fitness convergence is reached. After the iteration is completed, the layout scheme with the highest fitness score will be used as the adaptive layout scheme, and within a preset future time period, the actual installation position of the sensor will be dynamically adjusted according to the adaptive layout scheme to ensure that the sensors can be deployed preferentially in areas with higher pollution risks and greater volatility, thereby improving the coverage and real-time response capability of the entire monitoring, effectively avoiding redundant deployment of sensors in low-value areas, and improving resource utilization efficiency and the accuracy of air quality monitoring.

[0041] Furthermore, the present application provides a method for optimizing a sensor layout scheme based on the corrected pollution distribution topology map and the corrected index fluctuation characteristic map to obtain an adapted layout scheme, including:

[0042] Based on a preset number of sensor deployments, random deployment is performed in the target area to generate a first sensor deployment scheme; based on the corrected pollution distribution topology map, the corrected index fluctuation characteristic map and the first sensor deployment scheme, a deployment fitness analysis is performed to output the first scheme fitness; deployment schemes are iteratively generated and the scheme fitness is iteratively analyzed until a preset number of iterations is reached, and the sensor deployment scheme with the maximum scheme fitness is output as the adaptation deployment scheme.

[0043] Preferably, first, a preset number of sensor deployments N and the spatial boundary conditions of the target area (such as latitude and longitude ranges and installation area constraints) are set. Based on this, a certain number of initial deployment individuals (i.e., initial population) are randomly generated within the target area. Each individual represents a sensor deployment plan, recorded as the first sensor deployment plan, including a spatial coordinate set of N sensors. Subsequently, for each deployment individual, combined with the previously generated corrected pollution distribution topology map and corrected index fluctuation characteristic map, a deployment fitness analysis is performed to calculate the coverage match degree of the deployment plan for high-pollution value areas and strong fluctuation areas, thereby obtaining the fitness score of each deployment individual, i.e., the fitness of the first plan. Next, based on the fitness of the first solution, a roulette wheel or tournament selection method is used to select individuals with high fitness as parents. Coordinate-level swaps (such as crossovers and replacements of local layout segments) are performed on the offspring to generate a new sensor layout, referred to as the second sensor layout. Furthermore, the sensor positions of some individuals are perturbed (e.g., by slightly shifting coordinates) with a certain probability to enhance local exploration. The individuals with the lowest fitness are then replaced with the newly generated solution, maintaining a constant population size. This iterative process is repeated until the maximum number of iterations is reached or the population fitness converges. At the end of the iteration, the layout with the highest fitness is selected from all individuals in the iteration and used as the final adapted layout. This adapted layout effectively avoids local optimality and can globally optimize sensor location distribution within complex pollution environments and deployment constraints, improving coverage in highly polluted areas and response density in areas of high volatility, significantly enhancing the overall effectiveness and accuracy of monitoring.

[0044] Furthermore, the present application provides a method for performing a layout suitability analysis based on the corrected pollution distribution topology map, the corrected index fluctuation characteristic map, and the first sensor layout scheme, and outputting the suitability of the first scheme, including:

[0045] The corrected pollution distribution topology map is divided according to preset area sizes, and a plurality of high-pollution areas are determined by analysis, wherein the pollution index identification value of the high-pollution area is greater than the preset index threshold; the monitoring coverage range of the PM2.5 sensor is obtained, the first sensor layout plan is rendered, and it is determined whether the first sensor layout plan fully covers the plurality of high-pollution areas. If not, the first sensor layout plan is discarded; if so, based on the corrected pollution distribution topology map and the corrected index fluctuation characteristic map, the sensor matching degrees in the plurality of high-pollution areas are respectively calculated, and the weighted adaptation degree of the first plan is obtained.

[0046] Optionally, according to the set area size (for example, each grid is 50 meters × 50 meters), the corrected pollution distribution topology map is spatially gridded to form multiple independent spatial units. Subsequently, the pollution index identification value in each grid unit is counted. This pollution index identification value is the pollution index mean of the area where the corresponding grid is located, representing the pollution level of the area. By comparing the pollution index identification value of each spatial unit with the preset pollution index threshold, if the pollution index identification value of a certain area is greater than the threshold, it is marked as a high-pollution area. These high-pollution areas will be used for subsequent sensor coverage matching analysis. Afterwards, the monitoring coverage radius of the PM2.5 sensor is obtained (for example, 100 meters), and the coordinate positions of all sensors in the first sensor layout scheme are spatially rendered, that is, the circular range that each sensor can cover is calculated, and these coverage areas are compared with the high-pollution areas to determine whether the first sensor layout scheme achieves complete coverage of all high-pollution areas. If any high-pollution area is not covered, it is deemed that the scheme does not meet the basic layout requirements, and the scheme is directly abandoned without entering the fitness calculation link. On the contrary, if the first sensor deployment plan meets the coverage requirements, it enters the adaptation calculation stage, that is, the sensor matching degree of each high-pollution area is calculated based on the corrected pollution distribution topology map, the corrected index fluctuation characteristic map and the number of sensors in the high-pollution area, and then the first plan adaptation of the first sensor deployment plan is calculated by weighted method. This first plan adaptation can effectively improve the configuration efficiency of sensor resources in pollution hotspots and change-sensitive areas, and enhance the scientificity and practicality of the overall monitoring.

[0047] Furthermore, the present application provides a method for calculating the sensor matching degrees in the multiple high-pollution areas based on the corrected pollution distribution topology map and the corrected index fluctuation characteristic map, and weighting them to obtain the first solution fitness, including:

[0048] Based on the corrected pollution distribution topology map and the corrected index fluctuation characteristic map, multiple pollution index identification values ​​and multiple index fluctuation identification values ​​of the multiple high-pollution areas are obtained, and after dimensionless processing of the multiple pollution index identification values ​​and the multiple index fluctuation identification values, multiple pollution characteristic values ​​are weighted; based on the first sensor layout scheme, the number of multiple sensors in the multiple high-pollution areas is obtained; multiple sensor matching degrees are calculated according to the multiple pollution characteristic values ​​and the number of multiple sensors, wherein the sensor matching degree is the ratio of the number of sensors to the pollution characteristic value; pollution scale weights are set for the multiple high-pollution areas according to the multiple pollution characteristic values, the multiple sensor matching degrees are weightedly calculated, and the first scheme adaptability is output, wherein the pollution scale weight is positively correlated with the pollution characteristic value.

[0049] Optionally, based on the constructed correction pollution distribution topology map and correction index fluctuation characteristic map, for each identified high pollution area, a pollution index identification value and an index fluctuation identification value are extracted. These two identification values ​​are the mean air pollution index and the mean index fluctuation of the corresponding area, respectively. Subsequently, the pollution index identification value and the index fluctuation identification value of all high pollution areas are dimensionlessly processed, for example, by minimum-maximum normalization or Z-score standardization, so that their values ​​are on the same scale for subsequent weighted processing. After normalization, the two standardized indicators of each area are linearly weighted according to the set weights to obtain the pollution characteristic value of the area. Afterwards, based on the current first sensor layout plan, the number of sensors falling into each high pollution area is counted, and the ratio of the number of sensors in each high pollution area to the pollution characteristic value is calculated to obtain the sensor matching degree of each area. The higher the matching degree, the relatively sufficient sensor configuration of the area is, and the lower the matching degree, the insufficient monitoring resources in the area. Then, to comprehensively evaluate the suitability of the entire deployment plan, a weighted summary of the matching degrees of each region is required. To this end, pollution scale weights are set for each region based on its pollution characteristic value. Higher weights indicate a greater pollution risk in the region and should be prioritized for coverage. Typically, the pollution weight is the ratio of the pollution characteristic value of the region to the total pollution characteristic value. Finally, a weighted calculation is performed on the matching degrees of multiple sensors based on the set pollution weights to obtain the first-scheme suitability of the first sensor deployment plan. This first-scheme suitability can scientifically measure the degree of monitoring match of the deployment plan for high-risk pollution areas, providing an accurate basis for optimizing sensor resource allocation and significantly improving the monitoring efficiency and scheduling rationality of the sensor network.

[0050] In summary, the embodiments of the present application have at least the following technical effects:

[0051] The embodiment of the present application first obtains the air quality monitoring data of the target area in the historical time zone based on the public network base station data, and calculates several air pollution index means and several index fluctuations; then, according to the several air pollution index means and several index fluctuations, a pollution distribution topology map and an index fluctuation characteristic map are constructed by fitting; thereafter, the predicted weather change information and the predicted crowd density distribution of the target area in the preset future time period are obtained, the pollution distribution topology map and the index fluctuation characteristic map are adjusted, and the corrected pollution distribution topology map and the corrected index fluctuation characteristic map are output; finally, the sensor layout scheme is optimized according to the corrected pollution distribution topology map and the corrected index fluctuation characteristic map, an adapted layout scheme is obtained, and the sensor layout position is adjusted within the preset future time period. These technical effects jointly solve the technical problem that the PM2.5 sensor layout position lacks dynamic and precise layout in the prior art, resulting in insufficient air quality monitoring accuracy, and achieves the technical effect of dynamically parsing the terminal position of the public network base station data through big data analysis technology, accurately optimizing the sensor layout position, and improving the air quality monitoring accuracy.

[0052] It should be noted that the order in which the embodiments of the present application are presented is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. Furthermore, the foregoing descriptions of specific embodiments of this specification are provided. The processes depicted in the accompanying drawings do not necessarily require the specific order or sequential sequence shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0053] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

[0054] This specification and drawings are merely illustrative of the present application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Obviously, those skilled in the art may make various modifications and variations to this application without departing from the scope of this application. Thus, this application is intended to include such modifications and variations as fall within the scope of this application and its equivalents.

Claims

1. A method for analyzing terminal location distribution based on public network base station data, characterized in that: Methods include: Based on public network base station data, obtain air quality monitoring data of the target area within the historical time zone, and calculate several air pollution index means and several index fluctuations; Constructing a pollution distribution topology map and an index fluctuation characteristic map based on the plurality of air pollution index means and the plurality of index fluctuation degrees; Obtaining predicted weather change information and predicted crowd density distribution in a target area within a preset future period, adjusting the pollution distribution topology map and the index fluctuation characteristic map, and outputting a corrected pollution distribution topology map and a corrected index fluctuation characteristic map; The sensor layout scheme is optimized according to the corrected pollution distribution topology map and the corrected index fluctuation characteristic map to obtain an adapted layout scheme, and the sensor layout position is adjusted within the preset future time period.

2. The method for analyzing terminal location distribution based on public network base station data according to claim 1, characterized in that: Based on public network base station data, we obtain air quality monitoring data for the target area within the historical time zone and calculate several air pollution index means and several index fluctuations, including: Acquiring air quality monitoring data for a target area within a historical time zone based on public network base station data, wherein the air quality monitoring data is obtained based on an initial sensor deployment plan and includes several air pollution index sequences, wherein the initial sensor deployment plan includes several PM2.5 sensors at different locations; Calculating the mean of each of the air pollution index sequences to obtain a plurality of air pollution index means; Performing volatility analysis on the plurality of air pollution index sequences respectively, and outputting a plurality of index volatility, wherein the index volatility is a ratio of an index standard deviation to an index mean.

3. The method for analyzing terminal location distribution based on public network base station data according to claim 2, characterized in that: A pollution distribution topology map and an index fluctuation characteristic map are constructed based on the plurality of air pollution index means and the plurality of index fluctuation degrees, including: Arranging the plurality of air pollution index means and the plurality of index fluctuations according to the position coordinates of the plurality of sensors in the initial sensor layout scheme to construct an air pollution index mean array and an index fluctuation array; Pre-trained pollution distribution fitter and exponential fluctuation fitter; The pollution distribution fitter and the index fluctuation fitter are used to perform fitting according to the air pollution index mean array and the index fluctuation array, and output a pollution distribution topology map and an index fluctuation characteristic map.

4. The method for analyzing terminal location distribution based on public network base station data according to claim 3, characterized in that: Pre-trained pollution distribution fitter and exponential fluctuation fitter, including: Based on the historical environmental monitoring records of the target area, a sample air pollution index mean array set, a sample pollution distribution topology atlas, a sample index fluctuation array set, and a sample index fluctuation characteristic atlas are obtained; Using the sample air pollution index mean array set and the sample pollution distribution topology atlas, the generator and the discriminator of the generative adversarial network are trained until the network converges to obtain a pollution distribution fitter; The sample exponential fluctuation array set and the sample exponential fluctuation feature map set are used to train the generator and discriminator of the generative adversarial network until the network converges to obtain an exponential fluctuation fitter.

5. The method for analyzing terminal location distribution based on public network base station data according to claim 1, characterized in that: Obtaining predicted weather change information and predicted crowd density distribution in a target area within a preset future period, adjusting the pollution distribution topology map and the index fluctuation characteristic map, and outputting a corrected pollution distribution topology map and a corrected index fluctuation characteristic map, including: Obtaining predicted weather change information for a target area within a preset future time period based on public network base station data, wherein the predicted weather change information includes at least predicted temperature, predicted wind direction, and predicted wind speed; Obtaining a historical crowd density distribution sequence of the target area based on public network base station data, and analyzing the predicted weather change information to obtain a predicted crowd density distribution; The pollution distribution topology map and the index fluctuation characteristic map are adjusted according to the predicted weather change information and the predicted crowd density distribution, and a corrected pollution distribution topology map and a corrected index fluctuation characteristic map are output.

6. The method for analyzing terminal location distribution based on public network base station data according to claim 5, characterized in that: The historical crowd density distribution sequence of the target area is obtained based on the public network base station data, and the predicted crowd density distribution is obtained by combining the predicted weather change information with the analysis, including: Based on public network base station data, a sample crowd density distribution sequence set and a sample weather change information set are collected, and the historical crowd density distribution after the historical period is obtained as the sample predicted crowd density distribution to obtain the sample predicted crowd density distribution set; Using the sample crowd density distribution sequence set, the sample weather change information set, and the sample predicted crowd density distribution set as training data, training the long short-term memory network until convergence, and obtaining a crowd density prediction model; The crowd density prediction model is used to analyze the historical crowd density distribution sequence and the predicted weather change information to obtain the predicted crowd density distribution.

7. The method for analyzing terminal location distribution based on public network base station data according to claim 5, characterized in that: Adjusting the pollution distribution topology map and the index fluctuation characteristic map according to the predicted weather change information and the predicted crowd density distribution includes: Based on public network base station data, a sample weather change information set, a sample crowd density distribution set, a sample pollution distribution topology map set, and a sample index fluctuation characteristic map set are collected. The historical pollution distribution topology map and the historical index fluctuation characteristic map after the historical period are obtained as the sample correction pollution distribution topology map and the sample correction index fluctuation characteristic map, and the sample correction pollution distribution topology map set and the sample correction index fluctuation characteristic map set are obtained. Taking the sample weather change information set, the sample crowd density distribution set, and the sample pollution distribution topology map set as input, and using the sample correction pollution distribution topology map set as supervision to train a generative adversarial network until convergence, a first feature correction branch is obtained; Taking the sample weather change information set, the sample crowd density distribution set, and the sample index fluctuation feature map set as input, and using the sample correction index fluctuation feature map set as supervision to train a generative adversarial network until convergence, a second feature correction branch is obtained; The first feature correction branch and the second feature correction branch are used to adjust the pollution distribution topology map and the index fluctuation feature map according to the predicted weather change information and the predicted crowd density distribution.

8. The method for analyzing terminal location distribution based on public network base station data according to claim 1, characterized in that: The sensor layout scheme is optimized according to the correction pollution distribution topology map and the correction index fluctuation characteristic map to obtain an adapted layout scheme, including: Generating a first sensor layout scheme by randomly deploying sensors in the target area based on a preset number of sensor deployments; Performing a layout suitability analysis based on the corrected pollution distribution topology map, the corrected index fluctuation characteristic map, and the first sensor layout scheme, and outputting a first scheme suitability; Iterative generation of the layout scheme and iterative analysis of the scheme adaptability are performed until a preset number of iterations is reached, and the sensor layout scheme with the maximum scheme adaptability is output as the adapted layout scheme.

9. The method for analyzing terminal location distribution based on public network base station data according to claim 8, characterized in that: Performing a layout suitability analysis based on the corrected pollution distribution topology map, the corrected index fluctuation characteristic map, and the first sensor layout scheme, and outputting the first scheme suitability, including: Dividing the corrected pollution distribution topology map according to preset area sizes, and analyzing and determining a plurality of high-pollution areas, wherein the pollution index identification value of the high-pollution area is greater than a preset index threshold; Obtaining the monitoring coverage of the PM2.5 sensor, rendering the first sensor deployment plan, and determining whether the first sensor deployment plan fully covers the multiple high-pollution areas. If not, discarding the first sensor deployment plan; If so, based on the corrected pollution distribution topology map and the corrected index fluctuation characteristic map, the sensor matching degrees in the multiple high-pollution areas are calculated respectively, and the first solution adaptation degree is obtained by weighting.

10. The method for analyzing terminal location distribution based on public network base station data according to claim 9, characterized in that: Based on the corrected pollution distribution topology map and the corrected index fluctuation characteristic map, respectively calculating the sensor matching degrees in the multiple high-pollution areas, and weighting them to obtain the first solution fitness, including: Based on the corrected pollution distribution topology map and the corrected index fluctuation characteristic map, a plurality of pollution index identification values ​​and a plurality of index fluctuation identification values ​​are obtained for the plurality of high-pollution areas, and the plurality of pollution index identification values ​​and the plurality of index fluctuation identification values ​​are dimensionlessly processed and weighted to obtain a plurality of pollution characteristic values; Based on the first sensor deployment scheme, obtaining a number of sensors in the plurality of high-pollution areas; Calculating a plurality of sensor matching degrees according to the plurality of pollution characteristic values ​​and the plurality of sensor quantities, wherein the sensor matching degree is a ratio of the number of sensors to the pollution characteristic value; Pollution scale weights are set for the multiple high-pollution areas according to the multiple pollution characteristic values, and weighted calculation is performed on the multiple sensor matching degrees to output a first solution fitness, wherein the pollution scale weights are positively correlated with the pollution characteristic values.

Images