Noise monitoring point position automatic determination method fused with noise map

Abstract

The invention relates to the field of environmental noise monitoring, and discloses a noise monitoring point position automatic determination method fused with a noise map. The method comprises the following steps: acquiring a base map and sound source data of a target area, layering the base map according to a preset interval in a vertical direction, and carrying out gridding processing to obtain a three-dimensional grid; performing spatial analysis on the sound source data to obtain a sound source intensity matrix, and mapping the sound source intensity matrix to a three-dimensional grid to obtain a three-dimensional noise map; the base map comprises building structure data; performing three-dimensional path exploration on the three-dimensional noise map based on an ant colony algorithm to obtain a noise hotspot path and a pheromone concentration distribution diagram; and carrying out site selection analysis based on a genetic algorithm on the noise hotspot path and the pheromone concentration distribution diagram to obtain a Pareto optimal monitoring site set, the Pareto optimal monitoring site set being used for indicating site selection of the noise automatic monitoring site. By adopting the method, accurate and efficient automatic noise monitoring site selection in a complex urban environment can be realized, and the urban noise monitoring and treatment level is improved.

Description

Technical Field

[0001] The present invention belongs to the field of environmental noise monitoring, and particularly relates to a method for automatically determining noise monitoring points integrating a noise map. Background Art

[0002] In the densely built-up areas of high-rise buildings in cities, the propagation of noise exhibits complex multi-dimensional characteristics, and the traditional two-dimensional noise monitoring method is difficult to comprehensively reflect the true situation of noise distribution. Due to the significant difference in building heights, the noise intensity at different heights under the same plane coordinates may differ by dozens of decibels, and this vertical noise gradient change has an important impact on the living and working environments of residents and architectural acoustic design.

[0003] When establishing a three-dimensional noise map, it is necessary to arrange multiple monitoring points in the vertical direction, but the special structure of high-rise buildings makes the selection of monitoring points face technical problems. On the one hand, the external facade structure of the building is complex, with multiple reflecting surfaces and obstacles, which may lead to the distortion of noise monitoring data; on the other hand, the internal space of high-rise buildings is limited, it is difficult to install monitoring equipment at multiple heights simultaneously, and the installation position of the equipment may affect the building appearance or normal use.

[0004] Therefore, there is a problem that the coverage of noise monitoring points in urban building groups is insufficient, the distribution points lack representativeness, resulting in low noise monitoring accuracy. Summary of the Invention

[0005] Based on this, in view of the above technical problems, it is necessary to provide a method for automatically determining noise monitoring points integrating a noise map that can improve the accuracy of noise monitoring point selection, which can achieve accurate and efficient noise automatic monitoring point selection in complex urban environments and improve the urban noise monitoring and governance level.

[0006] In a first aspect, the present application provides a method for automatically determining noise monitoring points integrating a noise map, including:

[0007] Obtain the base map and sound source data of the target area, layer the base map vertically at a preset interval and perform grid processing to obtain a three-dimensional grid; perform spatial analysis on the sound source data to obtain a sound source intensity matrix and map it to the three-dimensional grid to obtain a three-dimensional noise map; the base map includes building structure data;

[0008] Perform a three-dimensional path exploration on the three-dimensional noise map based on the ant colony algorithm to obtain a noise hot spot path and a pheromone concentration distribution map;

[0009] Perform site selection analysis on the noise hot spot path and the pheromone concentration distribution map based on the genetic algorithm to obtain a Pareto optimal monitoring point set, and the Pareto optimal monitoring point set is used to indicate the site selection of noise automatic monitoring points.

[0010] Second aspect, the present application also provides a device for automatically determining noise monitoring points integrating a noise map, including:

[0011] A data acquisition module, configured to acquire a base map and sound source data of a target area, layer the base map vertically at a preset interval and perform grid processing to obtain a three-dimensional grid; perform spatial analysis on the sound source data to obtain a sound source intensity matrix and map it to the three-dimensional grid to obtain a three-dimensional noise map; the base map includes building structure data;

[0012] A noise path exploration module, configured to perform three-dimensional path exploration on the three-dimensional noise map based on the ant colony algorithm to obtain a noise hot spot path and a pheromone concentration distribution map;

[0013] A site selection analysis module, configured to perform site selection analysis based on the genetic algorithm on the noise hot spot path and the pheromone concentration distribution map to obtain a Pareto optimal monitoring point set, and the Pareto optimal monitoring point set is used to indicate the site selection of noise automatic monitoring points.

[0014] Third aspect, the present application also provides a computer device, including a memory and a processor, the memory stores a computer program, and when the processor executes the computer program, it implements the method for automatically determining noise monitoring points integrating a noise map as described above.

[0015] Fourth aspect, the present application also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the method for automatically determining noise monitoring points integrating a noise map as described above.

[0016] The above method for automatically determining noise monitoring points integrating a noise map, by acquiring a base map of a target area including building structure data, layering and gridifying it vertically at a preset interval to generate a three-dimensional grid, performing spatial analysis on the sound source data to form a sound source intensity matrix and mapping it to the three-dimensional grid to obtain a three-dimensional noise map; using the ant colony algorithm to perform path exploration in the three-dimensional noise map to obtain a noise hot spot path and a pheromone concentration distribution map, and based on the noise hot spot path and the pheromone concentration distribution map, using the genetic algorithm to perform site selection analysis to find a Pareto optimal monitoring point set from among numerous possible combinations of monitoring points, thereby determining the site selection of noise automatic monitoring points. This method breaks through the static limitations of traditional noise monitoring through three-dimensional modeling, swarm intelligence optimization, and multi-objective decision-making technology, realizes accurate and efficient site selection of noise automatic monitoring points in complex urban environments, and improves the level of urban noise monitoring and governance. Description of the Drawings

[0017] To more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following will briefly introduce the drawings required for the description of the embodiments or related technologies. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0018] Figure 1 It is a schematic flowchart of a method for automatically determining noise monitoring points integrating a noise map provided by an embodiment of the present invention;

[0019] Figure 2 It is a schematic structural diagram of a device for automatically determining noise monitoring points integrating a noise map provided by an embodiment of the present invention. Detailed implementation manners

[0020] In order to make the objectives, technical solutions and advantages of the present application more clear and understandable, the following further details the present application in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0021] First, a brief introduction is made to the nouns involved in the embodiments of the present application.

[0022] Noise map (noise mapping), noise map technology is a technology that closely combines noise prediction technology with geographic information system. It visualizes, visualizes and maps the environmental noise distribution in a certain area. These distribution maps are usually represented by noise contour lines, grids and color bands of different colors.

[0023] Ant Colony Optimization (ACO) is an optimization algorithm that simulates the foraging behavior of ants in nature. It is mainly used to solve combinatorial optimization problems. Ants release pheromones on the path, and other ants will choose the path according to the concentration of pheromones. The higher the concentration of pheromones on the path, the greater the probability of being selected. As time goes by, the concentration of pheromones on the optimal path gradually increases, and finally the entire ant colony will concentrate on the best path, thus finding the optimal solution to the optimization problem.

[0024] Pareto Optimality, also known as Pareto efficiency, refers to an ideal state of resource allocation. The Pareto optimal state means that there is no room for more Pareto improvement. Among them, Pareto improvement is the path and method to achieve Pareto optimality.

[0025] According to the above-mentioned noun explanations, the implementation environment of an automatic determination method for noise monitoring points integrating a noise map provided by an embodiment of the present application is described. Schematically, this implementation environment includes: a storage device, a processor, and a terminal. Among them, the processor communicates with the storage device and the terminal through a network. The terminal is deployed with a data collector, which can be an intelligent sensor or a log collector, and collects raw data such as network traffic and operation logs in real time. The storage device can be a distributed storage device or a disk array, and is connected through a high-speed Ethernet. The processor includes, but is not limited to, a central processing unit (CPU), a graphics processing unit (GPU), an artificial intelligence chip, etc., which are not limited here.

[0026] Combined with the above-mentioned noun explanations and implementation environment, the application scenarios of the embodiments of the present application are described. The automatic determination method for noise monitoring points integrating a noise map provided in the embodiments of the present application can be applied to the following scenarios including but not limited to:

[0027] In the scenario of urban area noise monitoring and planning, in the urban environment, different functional areas such as commercial areas, residential areas, and cultural and educational areas have different requirements for noise control. This technical solution can construct a three-dimensional urban noise map and identify noise hot paths such as traffic arteries and construction sites through the ant colony algorithm. For example, in a newly planned area of a certain city, the Pareto optimal monitoring point set is determined by means of the genetic algorithm, and monitoring points are mainly set around residential areas to monitor the noise situation in real time. Based on the monitoring data, the urban planning department can optimize the road layout, reasonably arrange the distance between commercial areas and residential areas, reduce the interference of noise on residents' lives, and provide strong support for creating a quiet and comfortable urban environment.

[0028] In the scenario of industrial park noise management, there are many factories in the industrial park, and a large amount of noise is generated during the operation of equipment. Using this technology, a three-dimensional noise map is generated based on the park base map and the sound source data of each factory. The ant colony algorithm is used to quickly locate the noise concentration areas, such as the noise propagation paths near large mechanical processing workshops. The genetic algorithm is used to screen out the best monitoring points, and monitoring points are accurately arranged in areas where noise is likely to spread and areas where employees are active frequently. Enterprise managers can adjust the equipment operation time in a timely manner and take noise reduction measures according to the monitoring results, ensuring the health of employees, and at the same time meeting the supervision requirements of the environmental protection department for the noise emissions in the park, achieving the balance between industrial production and environmental protection.

[0029] Schematically, the automatic determination method for noise monitoring points integrating a noise map provided by the embodiments of the present application can also be applied to other application scenarios. Only examples are given here, and the specific application scenarios are not limited.

[0030] In an exemplary embodiment, such as Figure 1As shown, an automatic determination method for noise monitoring points integrating a noise map is provided. In this embodiment, this method is exemplified by being applied to a terminal in the foregoing implementation environment. It can be understood that this system can also be applied to a server, and can also be applied to a system including a terminal and a server, and is implemented through the interaction between the terminal and the server. It includes the following steps 101 to 103.

[0031] Step 101: Obtain the base map and sound source data of the target area, divide the base map into layers at a preset interval in the vertical direction and perform grid processing to obtain a three-dimensional grid; perform spatial analysis on the sound source data to obtain a sound source intensity matrix and map it to the three-dimensional grid to obtain a three-dimensional noise map; the base map includes building structure data.

[0032] Specifically, geographic information system technology can be used to obtain a high-precision base map of the target area from channels such as urban planning databases and surveying and mapping departments to ensure the integrity and accuracy of building structure data (such as building height, location, material, etc.); use a noise sensor network, acoustic monitoring equipment, etc. to collect sound source data, covering the location, intensity, frequency, etc. of various sound sources such as traffic noise, industrial noise, and social life noise; according to the actual height range of the target area, set a reasonable preset interval in the vertical direction. For example, for urban areas, it can be set to 3 - 10 meters per layer. Using the regular grid division method, the base map is gridified in both the horizontal and vertical directions to generate a three-dimensional grid structure, and each grid unit has three-dimensional spatial coordinates; for the collected sound source data, perform spatial analysis based on the acoustic propagation model, calculate the noise intensity values at different positions, construct a sound source intensity matrix, and map the noise intensity values in the matrix one by one according to the spatial coordinates of the three-dimensional grid to complete the construction of the three-dimensional noise map.

[0033] Step 102: Perform a three-dimensional path exploration on the three-dimensional noise map based on the ant colony algorithm to obtain a noise hot spot path and a pheromone concentration distribution map.

[0034] Specifically, the ant colony algorithm has parallelism. Many ants explore paths simultaneously, and can find the noise hot spot path in a relatively short time, greatly improving the search efficiency compared with traditional search algorithms and saving time costs. Ants dynamically select paths according to the noise intensity and pheromone concentration, can automatically adapt to the complex changes of the noise map, and accurately find the key paths of noise propagation. Even in the case of changes in the distribution of noise sources or the presence of obstacles, it can work effectively.

[0035] Step 103: Perform a site selection analysis on the noise hot spot path and the pheromone concentration distribution map based on the genetic algorithm to obtain a Pareto optimal monitoring point set, and the Pareto optimal monitoring point set is used to indicate the site selection of automatic noise monitoring points.

[0036] Specifically, by simulating the biological evolution process, the genetic algorithm conducts a global search in the solution space, with a high probability of finding the Pareto optimal solution, avoiding being trapped in local optima, ensuring that the layout of monitoring points achieves the optimal comprehensive performance, realizing the best balance between monitoring effects and resource utilization, and making the finally determined set of monitoring points more practical and reasonable.

[0037] The above method for automatically determining noise monitoring points by fusing noise maps obtains the base map of the target area including building structure data, stratifies it at preset intervals in the vertical direction and grids it to generate a three-dimensional grid, conducts a spatial analysis on the sound source data, forms a sound source intensity matrix and maps it to the three-dimensional grid to obtain a three-dimensional noise map; uses the ant colony algorithm to conduct path exploration in the three-dimensional noise map to obtain the path to the noise hotspots and the distribution map of pheromone concentration, and based on the path to the noise hotspots and the distribution map of pheromone concentration, uses the genetic algorithm to conduct site selection analysis to find the Pareto optimal set of monitoring points from among numerous possible combinations of monitoring points, thereby determining the site selection of automatic noise monitoring points. This method breaks through the static limitations of traditional noise monitoring through three-dimensional modeling, swarm intelligence optimization and multi-objective decision-making technology, realizes accurate and efficient site selection of automatic noise monitoring points in complex urban environments, and improves the level of urban noise monitoring and governance.

[0038] In one embodiment, the three-dimensional path exploration of the three-dimensional noise map based on the ant colony algorithm to obtain the pheromone concentration distribution map and the path to the noise hotspots may include the following steps:

[0039] S11: Extract obstacles based on the building structure data and mark them on the three-dimensional grid to obtain a set of obstacle grids.

[0040] Specifically, the building structure data can be aligned with the three-dimensional grid, and the grid cells containing obstacles are marked as the "inaccessible" state. Optionally, the sound wave reflection coefficient is defined according to the material properties, and the positions of all obstacles are identified to obtain a set of obstacle grids for avoidance during path selection.

[0041] S12: Use the following formula to initialize the pheromone to obtain the initial pheromone concentration distribution grid, and the initial information concentration distribution grid is used to characterize the spatial distribution of the sound source intensity:

[0042]

[0043] where τ ijk is the pheromone concentration of the grid (i,j,k), S(x i , y j , z k ) is the sound source intensity of the grid (i,j,k), and ∑S xyz is the total sound source intensity of the sound source intensity matrix.

[0044] Specifically, the sound source intensity matrix is normalized to ensure that the initial pheromone concentration in high-noise areas (such as traffic arteries and equipment rooms) is higher. The resulting initial pheromone concentration distribution grid is a quantitative expression of the spatial distribution of the sound source intensity, which is used to guide the initial exploration direction of the ants.

[0045] S13: Use the following formula to calculate the probability of ant path selection and obtain the probability matrix of ant movement path; the ant path selection probability is used to quantify the tendency of ants to choose paths:

[0046]

[0047] in, is the probability of moving from grid (i,j,k) to the adjacent grid (l,m,n), τ lmn is the pheromone concentration of the target grid (l,m,n), η lmn is visibility, d obs is the distance to the nearest obstacle, α is the pheromone weight coefficient, and β is the visibility weight coefficient.

[0048] Specifically, when ants select adjacent grids from the current grid, they prefer paths with high pheromone concentrations and away from obstacles. The probability matrix of ant movement paths is used to quantify the movement tendency of ants in three-dimensional space.

[0049] S14: Using the following formula, update the pheromone to obtain a pheromone concentration distribution map; the pheromone concentration distribution map is used to characterize the noise propagation intensity:

[0050]

[0051] Among them, τ ijk (t+1) is the updated pheromone concentration, ρ k is the pheromone volatilization rate of vertical layer k, is the pheromone increment released by the ath ant in the path, M is the total number of ants, Q is the pheromone intensity constant, L a is the total length of the path of the ath ant, P atha A collection of paths.

[0052] Specifically, after each round of ant movement, the pheromone in the three-dimensional grid is updated. For each grid unit, the pheromone volatilization amount is calculated according to the pheromone volatilization rate of its vertical layer. At the same time, the movement path of each ant is traversed, and the pheromone increment released by each ant in the path is calculated according to the formula. The sum of the pheromone increments released by all ants minus the volatilization amount is used to obtain the updated pheromone concentration. After multiple rounds of iterative updates, a pheromone concentration distribution map that can characterize the intensity of noise propagation is formed.

[0053] S15: Repeat step S14 until the preset number of iterations is reached, extract a set of high-concentration grids whose pheromone concentration exceeds a preset threshold from the pheromone concentration distribution map, and connect the topologically adjacent high-concentration grids into a continuous path based on the connectivity of the grids to obtain a noise hotspot path. The noise hotspot path is used to indicate an efficient path for noise propagation.

[0054] Specifically, after the iteration is completed, the grids whose pheromone concentration exceeds the threshold are extracted, a three-dimensional neighborhood is defined, and adjacent high-concentration grids are connected into a continuous path. Optionally, a breadth-first search (BFS) algorithm is used to merge breakpoints to ensure path continuity, such as the vertical propagation path in the ventilation shaft; noise hotspot paths such as the reflection path along the building facade and the ventilation shaft diffusion path are obtained.

[0055] Furthermore, the noise hotspot path and pheromone concentration distribution map were analyzed based on genetic algorithm, and the Pareto optimal monitoring point set was obtained, including:

[0056] S21: extracting the peak area from the pheromone concentration distribution map, and eliminating invalid grids based on the building structure data to obtain a set of candidate points; the peak area is a set of grids whose pheromone concentration exceeds a preset peak value.

[0057] Specifically, the pheromone concentration distribution map is traversed, a suitable preset peak value is set, and the grid cells with pheromone concentration higher than the peak value are screened out. These cells constitute the peak area, and each grid in the peak area is checked in combination with the building structure data. For example, if the grid cell is located inside a building or in an area that is completely blocked, the grid is invalid as a noise monitoring point, and it is removed from the peak area, and finally a set of candidate points is obtained, which includes potential monitoring points in the key area of noise propagation and not hindered by building structures.

[0058] S22: Use a constraint-based random algorithm to select multiple point sets from the candidate point set to form multiple initial populations, and the initial populations serve as parent populations; wherein each initial population represents a potential feasible noise monitoring point site selection scheme covering the noise hotspot path.

[0059] Exemplarily, a series of constraints are defined, such as the minimum distance between monitoring points, the requirement for full coverage of the monitoring range, points that cannot be deployed under policy conditions, etc., and a random algorithm based on the constraints is designed. The algorithm randomly selects points from the candidate point set to form multiple point sets. Each point set is an initial population. Each initial population can ensure a certain coverage rate of the noise hotspot path while satisfying the constraints, representing a potentially feasible noise monitoring point site selection plan.

[0060] S23: Conduct coverage scoring, cost scoring, and constraint penalty scoring for multiple parental populations, and perform weighted averaging on the coverage scoring, cost scoring, and constraint penalty scoring to obtain a comprehensive fitness score; sort based on the comprehensive fitness score, and use the initial populations with the top n% of the fitness scores as high-quality populations; where n is a preset natural number threshold.

[0061] For each parental population (the initial population in the first round), calculate the coverage score, cost score, and constraint penalty score respectively. The coverage score measures the coverage degree of the population for the noise hot spot paths. For example, it is determined by calculating the ratio of the length of the noise hot spot paths covered by the monitored points to the total length of the noise hot spot paths; the cost score considers factors such as the installation and maintenance costs of monitoring devices and can be evaluated according to a preset cost model; the constraint penalty score deducts points for situations where the constraint conditions are not met, such as too small spacing between monitored points. Perform weighted averaging on the above three scores according to the proportional weight coefficient to obtain the comprehensive fitness score. Sort the comprehensive fitness scores of all parental populations, and select the populations with the top n% of the scores as high-quality populations. Here, the proportional weight coefficient and n are preset according to the actual situation and / or experimental results.

[0062] S24: Exchange the positions of two high-quality populations in the same vertical layer to generate an offspring population. Calculate the pheromone concentration gradient in the neighborhood of the positions of the offspring population, and move the positions to the grid with the highest pheromone concentration gradient in the neighborhood to obtain a new generation of population; where the new generation of population serves as the parental population input for the next round of step S23.

[0063] Specifically, select two populations from the high-quality populations, exchange their positions in the same vertical layer to generate an offspring population. For each position in the offspring population, calculate the pheromone concentration gradient in its neighborhood (a certain range of grid area centered on this position). By comparing the pheromone concentration gradient values of each grid in the neighborhood, move the position to the grid with the highest pheromone concentration gradient. After this operation, a new generation of population is obtained, and this new generation of population will serve as the parental population input for the next round of step S23.

[0064] S25: Calculate the change rate of the comprehensive fitness score, and the change rate is used to characterize the convergence of the algorithm.

[0065] Exemplarily, calculate the change rate of the comprehensive fitness scores of the current generation and the previous generation of populations. The change rate formula can be (the sum of the comprehensive fitness scores of the current generation - the sum of the comprehensive fitness scores of the previous generation) / the sum of the comprehensive fitness scores of the previous generation. This change rate is used to measure the convergence of the algorithm in the current iteration process. If the change rate is small, it indicates that the algorithm is gradually tending to be stable.

[0066] S26: Repeat steps S23 to S25 until the change rate is less than the preset stability threshold, and the new generation of population is the Pareto optimal monitoring point set.

[0067] Specifically, repeat S23 to S25, continuously perform population evolution. When the change rate is less than the preset stability threshold, it indicates that the algorithm has converged. At this time, the new generation of population is the Pareto optimal monitoring point set, and this point set represents a noise monitoring point selection scheme that achieves an optimal balance in aspects such as monitoring coverage effect, cost control, and meeting constraint conditions.

[0068] Preferably, the sound source data is subjected to spatial analysis and mapped to a three-dimensional grid, and the obtained three-dimensional noise map may include:

[0069] Step 201, perform data preprocessing and spatial alignment on the sound source data to obtain a standardized sound source data set. The attributes of the standardized sound source data set include sound source coordinates, sound pressure level, and time tags.

[0070] Specifically, use data processing software to clean and simulate interpolation of the collected sound source data, unify the sound source data from different sources and different formats into a standard format under the same geographic coordinate system, and based on geographic information system technology, perform spatial alignment of the sound source data with the base map of the target area. By matching common geographic features, ensure the accurate position mapping of the sound source data in the three-dimensional grid space to obtain a standardized sound source data set.

[0071] Step 202, perform density clustering based on the sound source coordinates and sound pressure level to obtain sound source clusters and isolated noise points; among them, the attributes of each sound source cluster include center coordinates, average sound pressure level, and spatial range.

[0072] Specifically, the density clustering algorithm can automatically identify sound source clusters and isolated noise points without the need to pre-specify the number of clusters, and can better reflect the actual distribution of the sound source data. For a complex urban noise environment, it can accurately divide different noise source aggregation areas, which helps to carry out targeted noise monitoring and management.

[0073] Step 203, map the average sound pressure level of the sound source cluster to adjacent grid cells according to the distance attenuation model. If multiple sound source clusters cover the same grid, take the maximum value of the mapped sound pressure levels as the final value to obtain a three-dimensional sound source intensity matrix.

[0074] Specifically, for each sound source cluster, determine the adjacent three-dimensional grid cells according to its center coordinates. According to the distance attenuation model, calculate and map the average sound pressure level of the sound source cluster to these adjacent grid cells. If there are multiple sound source clusters covering the same grid cell, compare the sound pressure level values mapped to this grid cell by each sound source cluster, and take the maximum value as the final sound pressure level value of this grid cell, thereby constructing a three-dimensional sound source intensity matrix.

[0075] Step 204: Adjust the grid sound pressure level of the three-dimensional sound source intensity matrix according to the acoustic wave characteristics of the building structure data, and map it to a three-dimensional network to obtain a three-dimensional noise map.

[0076] Specifically, with the aid of acoustic simulation software, analyze the acoustic wave characteristics of the building structure data. According to the analysis results of the acoustic wave characteristics, adjust the sound pressure level values in the three-dimensional sound source intensity matrix. For example, for grid cells close to large concrete buildings and where the acoustic wave propagation direction is blocked, appropriately reduce their sound pressure level values; for grid cells located in open spaces and where acoustic wave propagation is unobstructed, maintain or slightly adjust their sound pressure level values, and map the adjusted sound pressure level values to the three-dimensional grid to generate the final three-dimensional noise map.

[0077] In a possible embodiment, density clustering based on coordinates and sound pressure level is performed to obtain sound source clusters and isolated noise points, which may include:

[0078] Step 301: Calculate the number of sound sources and the average sound pressure level in each grid in the three-dimensional grid according to the sound source coordinates to obtain a sound source density distribution map; the sound source density distribution map is used to quantify the concentration degree of regional sound sources.

[0079] Specifically, traverse the standardized sound source data set. For each sound source data point, determine the grid cell to which it belongs according to its sound source coordinates, count the number of sound sources in each grid cell, and calculate the average sound pressure level of these sound sources. Based on the coordinates of the grid cell, visualize the information of the number of sound sources and the average sound pressure level of each grid cell to generate a sound source density distribution map.

[0080] Step 302: Classify the sound source density distribution map based on the quantile method to obtain high-density regions and low-density regions, and generate corresponding clustering parameters according to the parameter mapping rule; the clustering parameters include the neighborhood radius and the minimum number of samples.

[0081] Specifically, analyze the data in the sound source density distribution map using the quantile method, and pre-establish a set of parameter mapping rules to generate corresponding clustering parameters according to the characteristics of high-density regions and low-density regions. For high-density regions, since the sound source distribution is relatively dense, set a smaller neighborhood radius (such as eps = 30 meters) and a larger minimum number of samples (such as minPts = 15 sound source points) to more accurately identify closely clustered sound source clusters. For low-density regions, set a larger neighborhood radius (such as eps = 60 meters) and a smaller minimum number of samples (such as minPts = 5 sound source points) to ensure that relatively sparse sound source clusters or isolated noise points can be captured.

[0082] Step 303: Perform DBSCAN clustering analysis on the standardized sound source dataset based on the clustering parameters to obtain sound source clusters and isolated noise points; among them, the sound source clusters are subject to adjacent cluster merging and small cluster filtering.

[0083] Specifically, use the DBSCAN algorithm to perform clustering analysis on the standardized sound source dataset. For high-density regions and low-density regions, apply the clustering parameters generated according to the above rules respectively. After clustering, perform adjacent cluster merging processing on the obtained sound source clusters. Check the spatial distances between each sound source cluster. If the distance between two sound source clusters is less than the preset merging threshold, merge these two sound source clusters into a larger sound source cluster to avoid over-clustering caused by noise or data fluctuations and make the clustering result more in line with the actual sound source distribution; set a minimum cluster size threshold to filter the sound source clusters obtained by clustering. Consider the sound source clusters smaller than the minimum cluster size threshold as noise or abnormal conditions, remove them from the set of sound source clusters, retain the relatively large and practically significant sound source clusters, and at the same time mark the data points in the removed small clusters as isolated noise points.

[0084] Preferably, the sound source data is obtained through data fusion based on confidence adjustment weighting for multi-source heterogeneous data, including:

[0085] Step 401: Obtain multi-source heterogeneous data and perform data cleaning and spatio-temporal alignment to obtain multi-source noise data; among them, the multi-source noise data includes fixed sensor data, mobile device crowdsourcing data, and social media complaint data; the social media complaint data includes negative sentiment intensity.

[0086] Exemplarily, collect multi-source heterogeneous data through different channels. Obtain fixed sensor data from the urban environmental monitoring department. These data usually have high accuracy and stability and can continuously monitor the noise situation at specific locations; use mobile applications to collect mobile device crowdsourcing data, which can cover a wider area and make up for the spatial limitations of fixed sensors; collect text data related to noise from social media platforms (such as Weibo, forums, etc.) as social media complaint data. This data contains the public's feedback on noise problems and carries negative sentiment intensity information, which can reflect the public's subjective feelings about noise. The multi-source data fusion makes the obtained sound source data more comprehensive and can more accurately reflect the actual noise situation in the area.

[0087] Step 402: Assign a first weight coefficient to the fixed sensor data; assign a second weight coefficient to the mobile device crowdsourcing data; assign a third weight coefficient to the social media complaint data; the sum of the first weight coefficient, the second weight coefficient, and the third weight coefficient is 1.

[0088] Specifically, considering the characteristics and reliability of various types of data comprehensively, a first weight coefficient is assigned to the fixed sensor data. For example, since the fixed sensor data has high precision and stability, a relatively high weight is given; a second weight coefficient is assigned to the mobile device crowdsourcing data, which has the advantage of wide coverage; a third weight coefficient is assigned to the social media complaint data, which reflects the public's subjective feelings but has relatively low accuracy, and it is ensured that the sum of the three weight coefficients is 1.

[0089] Step 403, when the fixed sensor data shows anomalies for 3 consecutive times or more, reduce the first weight coefficient and correspondingly increase the second weight coefficient based on the reduction adjustment value; when the negative sentiment intensity exceeds the preset value, increase the third weight coefficient and correspondingly reduce the second weight coefficient based on the increase adjustment value.

[0090] Specifically, the weight coefficients are dynamically adjusted according to the real-time performance of the data. When the fixed sensor data shows anomalies, its weight is reduced and the weight of the mobile device crowdsourcing data is increased to ensure that reliable information can still be obtained when some data is abnormal. Adjusting the weight according to the negative sentiment intensity of the social media complaint data can pay more attention to the noise hotspots concerned by the public in a more timely manner, making the fused data more in line with the actual needs and improving the reliability and practicality of the data.

[0091] Step 404, perform weighted fusion on the multi-source noise data based on the adjusted first weight coefficient, second weight coefficient and third weight coefficient to obtain the sound source data.

[0092] Specifically, according to the adjusted first weight coefficient, second weight coefficient and third weight coefficient, perform weighted fusion on the multi-source noise data to obtain more comprehensive and accurate sound source data.

[0093] Preferably, the method may further include the following steps:

[0094] Step 501, map the three-dimensional noise map to the RGB color space based on the intensity of the sound source intensity matrix to obtain the visualized sound source intensity.

[0095] Specifically, a function that maps the sound source intensity value to the RGB color space can be designed, and this function can be linear or non-linearly fitted, which is not limited here.

[0096] Step 502, encode the points of the Pareto optimal monitoring point set into dynamic identifiers to obtain the visualized points.

[0097] Specifically, design dynamic identifiers for each Pareto optimal monitoring point. Circular icons can be used to represent the points, and the icon colors are distinguished according to the importance or other attributes of the monitoring points, and the icon sizes can be adjusted according to factors such as the coverage range of the monitoring points.

[0098] Step 503: Generate an interactive three-dimensional visualization scene file based on the visualized sound source intensity and visualization points.

[0099] Specifically, select a professional three-dimensional visualization software or WebGL framework (such as Three.js) to construct a three-dimensional visualization scene. In the scene, import the three-dimensional noise map mapped to the RGB color space as the background, and at the same time add the encoded visualization points to generate an interactive three-dimensional visualization scene file.

[0100] Step 503: Use the particle system to simulate the noise propagation path, and perform continuous frame simulation based on the three-dimensional visualization scene file to obtain a dynamic three-dimensional noise map, which is used to verify the robustness of the Pareto optimal monitoring point set under temporal conditions.

[0101] Specifically, use the particle system to simulate the noise propagation path and perform continuous frame simulation to generate a dynamic three-dimensional noise map. This dynamic display method can simulate the propagation of noise at different times. By observing the monitoring effect of the monitoring points on the dynamic noise, the robustness of the Pareto optimal monitoring point set under temporal conditions can be effectively verified. Compared with static analysis, it can more truly reflect the effectiveness of the monitoring points in the actual noise change, providing a more reliable basis for optimizing the layout of the monitoring points.

[0102] In summary, the automatic noise monitoring point determination method integrating noise maps provided by the embodiments of the present application obtains reliable multi-source heterogeneous sound source data through integrating fixed sensor, mobile device crowdsourcing, and social media complaint data, cleaning, spatio-temporal alignment, and confidence-based weighted fusion; constructs a three-dimensional grid with the base map of the target area, conducts clustering analysis on the sound source data, and combines with the building structure and distance attenuation model to obtain a three-dimensional noise map; uses the ant colony algorithm to explore the noise hot spot path, and uses the genetic algorithm to screen out the Pareto optimal monitoring point set; visualizes the noise map and sound source intensity, and uses the particle system to simulate the noise propagation to generate a dynamic three-dimensional noise map to verify the temporal robustness of the Pareto optimal monitoring point set. The above solution can break through the limitations of traditional two-dimensional static analysis of noise monitoring, realize the dynamic analysis of the noise propagation path and the accurate location selection of monitoring points in complex urban environments, improve the robustness and adaptability of the noise monitoring network, and provide intelligent decision-making support for urban noise control.

[0103] It should be understood that although the steps in the flowcharts involved in the above-described embodiments are sequentially shown according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless clearly stated in this article, the execution of these steps has no strict order limit, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-described embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily executed at the same moment, but can be executed at different moments. The execution order of these steps or stages is not necessarily sequential, but can be executed alternately or in turn with at least a part of other steps or steps or stages in other steps.

[0104] Based on the same inventive concept, an embodiment of the present application further provides an automatic determination device for noise monitoring points of a fused noise map for implementing the above-mentioned automatic determination method of noise monitoring points of a fused noise map. The solution provided by this device to solve the problem is similar to the solution described in the above method. Therefore, the specific limitations in one or more embodiments of the automatic determination device for noise monitoring points of a fused noise map provided below can refer to the limitations on the automatic determination method of noise monitoring points of a fused noise map in the above text, and will not be repeated here.

[0105] In an exemplary embodiment, as Figure 2 shown, an automatic determination device 20 for noise monitoring points of a fused noise map is provided, including:

[0106] A data acquisition module 21, configured to acquire a base map and sound source data of a target area, layer the base map vertically at a preset interval and perform grid processing to obtain a three-dimensional grid; perform spatial analysis on the sound source data to obtain a sound source intensity matrix and map it to the three-dimensional grid to obtain a three-dimensional noise map; the base map includes building structure data.

[0107] A noise path exploration module 22, configured to perform three-dimensional path exploration on the three-dimensional noise map based on the ant colony algorithm to obtain a noise hot spot path and a pheromone concentration distribution map.

[0108] A site selection analysis module 23, configured to perform site selection analysis based on the genetic algorithm on the noise hot spot path and the pheromone concentration distribution map to obtain a Pareto optimal monitoring point set, and the Pareto optimal monitoring point set is used to indicate the site selection of automatic noise monitoring points.

[0109] In one of the embodiments, in the noise path exploration module 22, performing three-dimensional path exploration on the three-dimensional noise map based on the ant colony algorithm to obtain a pheromone concentration distribution map and a noise hot spot path can be achieved through the following steps:

[0110] S11: Extract obstacles based on building structure data and mark them on a three-dimensional grid to obtain a set of obstacle grids.

[0111] S12: Initialize the pheromone using the following formula to obtain an initial pheromone concentration distribution grid, which is used to characterize the spatial distribution of sound source intensity:

[0112]

[0113] where τ ijk is the pheromone concentration of grid (i, j, k), S(x i , y j , z k ) is the sound source intensity of grid (i, j, k), and ∑S xyz is the total sound source intensity of the sound source intensity matrix.

[0114] S13: Calculate the ant path selection probability using the following formula to obtain a probability matrix of the ant movement path; the ant path selection probability is used to quantify the tendency of the ant to select a path:

[0115]

[0116] where is the probability of moving from grid (i, j, k) to the adjacent grid (l, m, n), τ lmn is the pheromone concentration of the target grid (l, m, n), η lmn is the visibility, d obs is the distance to the nearest obstacle, α is the pheromone weight coefficient, and β is the visibility weight coefficient.

[0117] S14: Update the pheromone using the following formula to obtain a pheromone concentration distribution map; the pheromone concentration distribution map is used to characterize the noise propagation intensity:

[0118]

[0119] where τ ijk (t + 1) is the updated pheromone concentration, ρ k is the pheromone evaporation rate of the vertical layer k, is the pheromone increment released by the a-th ant in the path, M is the total number of ants, Q is the pheromone intensity constant, L a is the total path length of the a-th ant, and Path a is the set of paths.

[0120] S15: Repeat step S14 until the preset number of iterations is reached, extract a set of high-concentration grids whose pheromone concentration exceeds a preset threshold from the pheromone concentration distribution map, and connect the topologically adjacent high-concentration grids into a continuous path based on the connectivity of the grids to obtain a noise hotspot path. The noise hotspot path is used to indicate an efficient path for noise propagation.

[0121] Furthermore, in the site selection analysis module 23, the site selection analysis based on the genetic algorithm is performed on the noise hotspot path and the pheromone concentration distribution map, and the Pareto optimal monitoring point set is obtained by the following steps:

[0122] S21: extracting the peak area from the pheromone concentration distribution map, and eliminating invalid grids based on the building structure data to obtain a set of candidate points; the peak area is a set of grids whose pheromone concentration exceeds a preset peak value;

[0123] S22: using a random algorithm based on constraints to select multiple point sets from the candidate point set to form multiple initial populations, the initial populations serve as parent populations; wherein each initial population represents a potential feasible noise monitoring point location selection scheme covering the noise hotspot path;

[0124] S23: performing coverage scoring, cost scoring and constraint penalty on multiple parent populations, and performing weighted average of the coverage score, cost score and constraint penalty to obtain a comprehensive adaptability score; sorting based on the comprehensive adaptability score, and taking the initial populations with the top n% of the adaptability score as high-quality populations; wherein n is a preset natural number threshold;

[0125] S24: Exchange the points of two high-quality populations in the same vertical layer to generate a daughter population, calculate the pheromone concentration gradient of the field for the points of the daughter population, and move the points to the grid with the highest pheromone concentration gradient in the field to obtain a new generation population; wherein the new generation population is used as the parent population input in the next round of step S23;

[0126] S25: Calculate the change rate of the comprehensive applicability score, which is used to characterize the convergence of the algorithm;

[0127] S26: Repeat steps S23 to S25 until the rate of change is less than a preset stability threshold, and the new generation population is obtained as a Pareto optimal monitoring point set.

[0128] Preferably, the data acquisition module 21 may include the following units:

[0129] The sound source data standardization unit 211 is used to perform data preprocessing and spatial alignment on the sound source data to obtain a standardized sound source data set. The attributes of the standardized sound source data set include sound source coordinates, sound pressure level and time label.

[0130] The density clustering unit 212 is used to perform density clustering based on the sound source coordinates and sound pressure levels to obtain sound source clusters and isolated noise points; among them, the attributes of each sound source cluster include the center coordinates, average sound pressure level, and spatial range.

[0131] The sound source mapping unit 213 is used to map the average sound pressure level of the sound source cluster to the adjacent grid cells according to the distance attenuation model. If multiple sound source clusters cover the same grid, the maximum value of the mapped sound pressure levels is taken as the final value to obtain a three-dimensional sound source intensity matrix.

[0132] The three-dimensional noise mapping unit 214 is used to adjust the grid sound pressure level of the three-dimensional sound source intensity matrix according to the acoustic wave characteristics of the building structure data and map it to a three-dimensional network to obtain a three-dimensional noise map.

[0133] In a possible embodiment, the density clustering unit 212 may include:

[0134] The sound source density distribution mapping unit 2121 is used to calculate the number of sound sources and the average sound pressure level in each grid of the three-dimensional grid according to the sound source coordinates to obtain a sound source density distribution map; the sound source density distribution map is used to quantify the concentration degree of regional sound sources.

[0135] The parameter grading unit 2122 is used to grade the sound source density distribution map based on the quantile method to obtain high-density regions and low-density regions, and generate corresponding clustering parameters according to the parameter mapping rules; the clustering parameters include the neighborhood radius and the minimum number of samples.

[0136] The clustering analysis unit 2123 is used to perform DBSCAN clustering analysis on the standardized sound source data set based on the clustering parameters to obtain sound source clusters and isolated noise points; among them, the sound source clusters are merged with neighboring clusters and small clusters are filtered.

[0137] Preferably, the data acquisition module 21 may include:

[0138] The multi-source data acquisition unit 215 is used to acquire multi-source heterogeneous data and perform data cleaning and spatio-temporal alignment to obtain multi-source noise data; among them, the multi-source noise data includes fixed sensor data, mobile device crowdsourcing data, and social media complaint data; the social media complaint data includes negative sentiment intensity.

[0139] The weight assignment unit 216 is used to assign a first weight coefficient to the fixed sensor data; assign a second weight coefficient to the mobile device crowdsourcing data; assign a third weight coefficient to the social media complaint data; the sum of the first weight coefficient, the second weight coefficient, and the third weight coefficient is 1.

[0140] A weight adjustment unit 217 is configured to reduce the first weight coefficient when the fixed sensor data appears abnormal for 3 consecutive times or more, and correspondingly increase it to the second weight coefficient based on the reduced adjustment value; when the negative sentiment intensity exceeds a preset value, increase the third weight coefficient, and correspondingly reduce the second weight coefficient based on the increased adjustment value.

[0141] A weighted fusion unit 218 is configured to perform weighted fusion on multi-source noise data based on the adjusted first weight coefficient, second weight coefficient, and third weight coefficient to obtain sound source data.

[0142] Preferably, the noise monitoring point position automatic determination device 20 of the fusion noise map further includes:

[0143] A sound source intensity visualization module 24 is configured to map the intensity of the three-dimensional noise map to the RGB color space based on the sound source intensity matrix to obtain the visualized sound source intensity.

[0144] A point position visualization module 25 is configured to encode the point positions of the Pareto optimal monitoring point position set as dynamic identifiers to obtain visualized point positions.

[0145] A scene file generation module 26 is configured to generate an interactive three-dimensional visualization scene file based on the visualized sound source intensity and the visualized point positions.

[0146] A simulation rendering module 27 is configured to simulate the noise propagation path using a particle system, perform continuous frame simulation based on the three-dimensional visualization scene file to obtain a dynamic three-dimensional noise map, and the dynamic three-dimensional noise map is used to verify the robustness of the Pareto optimal monitoring point position set under temporal conditions.

[0147] In one embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the steps of a power supply safety management method as described above are implemented.

[0148] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0149] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to the partial descriptions of the method embodiments. The device embodiments described above are merely illustrative. The components described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the present disclosure solution. Those of ordinary skill in the art can understand and implement it without creative efforts.

[0150] The above embodiments only represent several implementation manners of the embodiments of the present application. The description is relatively specific and detailed, but it should not be construed as a limitation on the patent scope of the embodiments of the application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the embodiments of the present application, several modifications and improvements can still be made, and these all belong to the protection scope of the embodiments of the present application.

Claims

1. An automatic determination method for noise monitoring points integrating a noise map, characterized in that, include: Acquire a base map and sound source data of a target area, layer the base map at preset intervals in a vertical direction, and perform grid processing to obtain a three-dimensional grid; Performing spatial analysis on the sound source data to obtain a sound source intensity matrix and mapping it to the three-dimensional grid to obtain a three-dimensional noise map; the base map includes building structure data; Performing three-dimensional path exploration on the three-dimensional noise map based on an ant colony algorithm to obtain a noise hotspot path and a pheromone concentration distribution map; A site selection analysis based on a genetic algorithm is performed on the noise hotspot path and the pheromone concentration distribution map to obtain a Pareto optimal monitoring point set, and the Pareto optimal monitoring point set is used to indicate the site selection of automatic noise monitoring points.

2. The method according to claim 1, wherein The three-dimensional path exploration of the three-dimensional noise map based on the ant colony algorithm to obtain the pheromone concentration distribution map and the noise hotspot path includes: S11: extracting obstacles based on the building structure data and marking them on the three-dimensional grid to obtain an obstacle grid set; S12: Use the following formula to initialize the pheromone to obtain an initial pheromone concentration distribution grid, which is used to characterize the spatial distribution of the sound source intensity: Among them, τ ijk is the pheromone concentration of the grid (i, j, k), and S(x i , y j , z k ) is the sound source intensity of the grid (i, j, k), and ∑S xyz is the total sound source intensity of the sound source intensity matrix; S13: Calculate the ant path selection probability using the following formula to obtain the probability matrix of the ant movement path; the ant path selection probability is used to quantify the ant's tendency to choose a path: Among them, is the probability of moving from grid (i, j, k) to adjacent grid (l, m, n), τ lmn is the pheromone concentration of the target grid (l, m, n), η lmn is the visibility, d obs is the distance to the nearest obstacle, α is the pheromone weight coefficient, and β is the visibility weight coefficient; S14: Using the following formula, update the pheromone to obtain the pheromone concentration distribution map; the pheromone concentration distribution map is used to characterize the noise propagation intensity: Among them, τ ijk (t + 1) is the updated pheromone concentration, ρ k is the pheromone evaporation rate of the vertical layer k, is the pheromone increment released by the a-th ant in the path, M is the total number of ants, Q is the pheromone intensity constant, L a is the total path length of the a-th ant, Path a is the path set; S15: Repeat step S14 until a preset number of iterations is reached, extract a set of high-concentration grids whose pheromone concentration exceeds a preset threshold from the pheromone concentration distribution map, and connect topologically adjacent high-concentration grids into a continuous path based on the connectivity of the grids to obtain the noise hotspot path, which is used to indicate an efficient path for noise propagation.

3. The method according to claim 2, characterized in that, The site selection analysis based on the genetic algorithm is performed on the noise hotspot path and the pheromone concentration distribution map to obtain the Pareto optimal monitoring point set including: S21: extracting a peak area from the pheromone concentration distribution map, and eliminating invalid grids based on the building structure data to obtain a candidate point set; the peak area is a grid set where the pheromone concentration exceeds a preset peak value; S22: using a random algorithm based on constraints to select multiple point sets from the candidate point set to form multiple initial populations, the initial populations serving as parent populations; wherein each of the initial populations represents a potentially feasible noise monitoring point location selection scheme covering a noise hotspot path; S23: performing coverage scoring, cost scoring and constraint penalty on the plurality of parent populations, and performing weighted average of the coverage score, the cost score and the constraint penalty to obtain a comprehensive adaptability score; sorting based on the comprehensive adaptability score, and taking the initial populations in the top n% of the adaptability score as high-quality populations; wherein n is a preset natural number threshold; S24: Exchange the positions of the two high-quality populations in the same vertical layer to generate an offspring population. Calculate the pheromone concentration gradient in the domain of the positions of the offspring population, and move the positions to the grid with the highest pheromone concentration gradient in the domain to obtain a new generation population; wherein, the new generation population serves as the parental population input in the next round of step S23. S25: Calculate the change rate of the comprehensive applicability score, and the change rate is used to characterize the convergence of the algorithm. S26: Repeat steps S23 to S25 until the change rate is less than a preset stability threshold, and obtain the new generation population as the Pareto optimal monitoring position set.

4. The method according to claim 1, wherein Performing spatial analysis on the sound source data and mapping it to the three-dimensional grid, the obtained three-dimensional noise map includes: Perform data preprocessing and spatial alignment on the sound source data to obtain a standardized sound source data set, and the attributes of the standardized sound source data set include sound source coordinates, sound pressure level, and time tags. Perform density clustering based on the sound source coordinates and the sound pressure level to obtain sound source clusters and isolated noise points; wherein, the attributes of each sound source cluster include central coordinates, average sound pressure level, and spatial range. Map the average sound pressure level of the sound source cluster to adjacent grid cells according to the distance attenuation model. If multiple sound source clusters cover the same grid, take the maximum value of the mapped sound pressure level as the final value to obtain a three-dimensional sound source intensity matrix. Adjust the grid sound pressure level of the three-dimensional sound source intensity matrix according to the acoustic wave characteristics of the building structure data and map it to the three-dimensional network to obtain the three-dimensional noise map.

5. The method according to claim 4, characterized in that, The performing density clustering based on the coordinates and the sound pressure level to obtain sound source clusters and isolated noise points includes: Calculate the number of sound sources and the average sound pressure level in each grid in the three-dimensional grid according to the sound source coordinates to obtain a sound source density distribution map; the sound source density distribution map is used to quantify the concentration degree of regional sound sources. Grade the sound source density distribution map based on the quantile method to obtain high-density regions and low-density regions, and generate corresponding clustering parameters according to the parameter mapping rule; the clustering parameters include domain radius and minimum sample number. Perform DBSCAN clustering analysis on the standardized sound source data set based on the clustering parameters to obtain the sound source clusters and the isolated noise points; wherein, the sound source clusters are merged with adjacent clusters and small clusters are filtered.

6. The method according to claim 4, wherein The sound source data is obtained by data fusion of multi-source heterogeneous data based on confidence adjustment and weighting, including: Obtain the multi-source heterogeneous data and perform data cleaning and spatio-temporal alignment to obtain multi-source noise data; wherein, the multi-source noise data includes fixed sensor data, mobile device crowdsourcing data, and social media complaint data; the social media complaint data includes negative sentiment intensity. Assign a first weight coefficient to the fixed sensor data; assign a second weight coefficient to the mobile device crowdsourcing data; assign a third weight coefficient to the social media complaint data; the sum of the first weight coefficient, the second weight coefficient, and the third weight coefficient is 1. When the fixed sensor data appears abnormal for 3 consecutive times or more, reduce the first weight coefficient, and correspondingly increase it to the second weight coefficient based on the reduction adjustment value; when the negative sentiment intensity exceeds the preset value, increase the third weight coefficient, and correspondingly reduce the second weight coefficient based on the increase adjustment value. Perform weighted fusion on the multi-source noise data based on the adjusted first weight coefficient, second weight coefficient, and third weight coefficient to obtain the sound source data.

7. The method according to claim 2, characterized in that The method further includes: Map the three-dimensional noise map to the RGB color space based on the intensity of the sound source intensity matrix to obtain the visualized sound source intensity. Encode the points in the Pareto optimal monitoring point set as dynamic identifiers to obtain visualized points. Generate an interactive three-dimensional visualization scene file based on the visualized sound source intensity and the visualized points. Use a particle system to simulate the noise propagation path, and perform continuous frame simulation based on the three-dimensional visualization scene file to obtain a dynamic three-dimensional noise map, which is used to test the robustness of the Pareto optimal monitoring point set under temporal conditions.

8. An automatic determination device for noise monitoring points integrating a noise map, characterized in that, The device includes: A data acquisition module, configured to acquire the base map and sound source data of the target area, layer the base map vertically at a preset interval and perform grid processing to obtain a three-dimensional grid; perform spatial analysis on the sound source data to obtain a sound source intensity matrix and map it to the three-dimensional grid to obtain a three-dimensional noise map; the base map includes building structure data. A noise path exploration module, configured to perform three-dimensional path exploration on the three-dimensional noise map based on the ant colony algorithm to obtain a noise hot spot path and a pheromone concentration distribution map. A site selection analysis module, configured to perform site selection analysis based on the genetic algorithm on the noise hot spot path and the pheromone concentration distribution map to obtain a Pareto optimal monitoring point set, which is used to indicate the site selection of the noise automatic monitoring points.

9. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, the method according to any one of claims 1 to 7 is implemented.

Images