Intelligent dust suppression decision-making method and system for dust in storage yard

Abstract

The invention relates to the technical field of intelligent data analysis, in particular to an intelligent dust suppression decision-making method and system for storage yard dust. The method comprises the steps that the dust concentration of a target storage yard monitoring point is obtained, and interpolation is conducted on the dust concentration to obtain a dust concentration distribution diagram; acquiring meteorological data of a target storage yard monitoring point, and calculating a pollution source position and pollution source intensity by using the meteorological data and the dust concentration distribution diagram; a moisture content change trend prediction model is constructed, and a stack surface moisture content prediction value changing along with time is obtained through the moisture content change trend prediction model based on the obtained stacking mechanism characteristics of the pollution source position, the stack surface moisture content and the meteorological data; and when it is judged that the predicted value of the surface moisture content of the stack is smaller than a preset threshold value of the surface moisture content of the stack, water spraying dust suppression is conducted on the stack at the position of the pollution source based on the intensity of the pollution source. According to the invention, a closed loop is formed through monitoring, traceability, prediction and decision making, and the dust suppression accuracy and efficiency are improved.

Description

Technical Field

[0001] This invention relates to the field of data intelligent analysis technology, specifically to an intelligent dust suppression decision-making method and system for stockpile dust. Background Technology

[0002] Industrial storage yards, such as those for storing ores, coal, and building materials, are key storage links in industrial production. However, the stacked materials are prone to generating dust under the influence of wind and loading and unloading operations. This not only causes atmospheric particulate pollution but also harms the health of workers, accelerates equipment wear and tear, and leads to material loss. Therefore, effective dust suppression and control are urgently needed.

[0003] Current dust control methods at stockpile sites largely rely on traditional approaches, which have significant limitations: First, dust suppression methods are crude. For example, manual watering depends on experience and is prone to localized over-watering (wasting water resources and causing material adhesion) or under-watering (dust control failure). Fixed spraying equipment has a fixed coverage area and cannot be dynamically adjusted according to the location of pollution sources, resulting in poor control of mobile or dispersed pollution sources. Second, the ability to locate and trace pollution sources is weak. Traditional monitoring relies on collecting dust concentration data from only a few fixed points, making it difficult to construct a complete concentration distribution from discrete data. This makes it impossible to accurately pinpoint high-pollution areas and strong pollution sources, leading to a lack of targeted dust suppression measures. Third, moisture content control lacks predictability. The moisture content of the stockpile surface is a key indicator for suppressing dust generation, but current technologies mostly only monitor real-time moisture content, often resulting in a passive situation of adding water only after dust has started to rise, missing the optimal time for dust suppression.

[0004] Therefore, a more scientific and effective decision-making method for dust suppression in storage yards is urgently needed. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an intelligent dust suppression decision-making method and system for stockpile dust, which solves the problems of poor dust suppression accuracy and low efficiency in existing technologies.

[0006] One aspect of the present invention provides an intelligent dust suppression decision-making method for stockpile dust, comprising: acquiring dust concentration at a monitoring point of a target stockpile and interpolating the dust concentration to obtain a dust concentration distribution map; acquiring meteorological data at the monitoring point of the target stockpile and calculating the location and intensity of the pollution source using the meteorological data and the dust concentration distribution map; constructing a moisture content change trend prediction model, and obtaining a predicted value of the moisture content of the stack surface changing over time using the moisture content change trend prediction model based on the acquired stacking mechanism characteristics at the location of the pollution source, the moisture content of the stack surface, and the meteorological data; and determining that when the predicted value of the moisture content of the stack surface is less than a preset stack surface moisture content threshold, spraying water to suppress dust at the location of the pollution source based on the pollution source intensity.

[0007] This invention achieves precise source identification of dust concentration in stockpiles and generates a complete concentration distribution map through interpolation. By combining meteorological data to calculate the location and intensity of pollution sources, it avoids the blindness of traditional dust suppression. Based on the characteristics of stockpiling mechanisms, real-time moisture content, and meteorological data, a predictive model is constructed to anticipate changes in moisture content, transforming passive dust suppression into proactive prediction. When the predicted moisture content is below a threshold, spraying parameters are adjusted according to the pollution source intensity, ensuring dust suppression effectiveness in high-pollution areas while avoiding water waste. This forms a closed loop of monitoring, source tracing, prediction, and decision-making, improving the accuracy and efficiency of dust suppression.

[0008] Optionally, acquiring meteorological data from the target storage yard monitoring point includes: determining key meteorological factors; and acquiring meteorological data from the target storage yard monitoring point based on the key meteorological factors.

[0009] This invention improves the quality of meteorological data by first identifying key meteorological factors and then obtaining meteorological data from monitoring points at the target storage yard based on these key factors, thereby accurately identifying meteorological parameters that significantly affect dust diffusion and changes in the moisture content of the stockpiles.

[0010] Optionally, determining the key meteorological factors includes: calculating the historical monthly average meteorological data and historical monthly average dust concentration based on the historical meteorological data and historical dust concentration acquired in advance throughout the year for the target storage yard; performing correlation analysis on the historical monthly average meteorological data and the historical monthly average dust concentration to obtain the Pearson correlation coefficient; and determining the key meteorological factors based on the Pearson correlation coefficient.

[0011] This invention calculates monthly average data using hourly historical data from the target stockpile throughout the year, avoiding the randomness of short-term data and reflecting the long-term correlation between meteorological conditions and dust concentration. Through Pearson correlation coefficient analysis, the correlation strength between meteorological parameters and dust concentration can be quantified, providing objective data support for screening key factors and avoiding subjective judgment bias. Based on the correlation coefficient, key meteorological factors can be determined, accurately identifying parameters that significantly affect dust diffusion and changes in stockpile moisture content, while eliminating irrelevant or weakly correlated parameters, thus improving the scientific rigor and accuracy of key meteorological factor calculation.

[0012] Optionally, the step of calculating the pollution source location and pollution source intensity using the meteorological data and the dust concentration distribution map includes: dividing the target stockpile into a grid to obtain a target grid; determining the initial locations of multiple pollution sources and their corresponding initial pollution source intensities based on the target grid and the dust concentration distribution map; and iteratively obtaining the pollution source location and pollution source intensity using a particle swarm optimization algorithm based on the initial pollution source locations and the initial pollution source intensities.

[0013] This invention, by dividing the target stockpile into grids, can discretize the complex stockpile space into regular units, providing a clear spatial carrier for locating pollution sources and avoiding ambiguity in location determination. By combining the grid with the dust concentration distribution map to determine the initial location and source strength, it can focus on high-pollution areas, providing a reasonable starting point for subsequent optimization and reducing blind search by the algorithm. Furthermore, by iteratively optimizing with the particle swarm optimization algorithm, it can continuously approximate the actual pollution source parameters, improving the accuracy of location and source strength calculation.

[0014] Optionally, the step of iteratively obtaining the pollution source location and pollution source strength using a particle swarm optimization algorithm based on the initial pollution source location and the initial pollution source strength includes: introducing a Gaussian diffusion model and using the Gaussian diffusion model to perform diffusion simulations based on the initial pollution source location, the initial pollution source strength, and the meteorological data to obtain a simulated concentration distribution map; obtaining the location of monitoring points and extracting the simulated dust concentration at the monitoring points based on the simulated concentration distribution map; constructing a fitness function with the goal of minimizing the mean square error between the simulated dust concentration at the monitoring points and the dust concentration; and iteratively obtaining the pollution source location and pollution source strength using a particle swarm optimization algorithm combined with the fitness function based on the initial pollution source location and the initial pollution source strength.

[0015] This invention introduces a Gaussian diffusion model to simulate concentration distribution, which closely matches the physical laws of dust diffusion in stockpiles. This provides a realistic scenario basis for subsequent parameter optimization. The simulated concentration at monitoring points is extracted, and a fitness function is constructed based on the minimum mean square error between the simulated and measured concentrations. This quantifies the deviation between the simulation and reality, providing a clear optimization target for algorithm iteration. Combined with particle swarm optimization algorithm iteration, the pollution source parameters can be continuously adjusted to reduce the deviation, thereby improving the accuracy of locating the pollution source location and intensity.

[0016] Optionally, the step of iteratively obtaining the pollution source location and pollution source strength based on the initial pollution source location and the initial pollution source strength using a particle swarm optimization algorithm combined with the fitness function includes: iteratively obtaining an optimal source strength matrix based on the initial pollution source location and the initial pollution source strength using a particle swarm optimization algorithm combined with the fitness function; setting a source strength threshold, and extracting the pollution source location and pollution source strength from the optimal source strength matrix based on the source strength threshold.

[0017] This invention obtains the optimal source strength matrix by combining particle swarm optimization algorithm with fitness function iteration. It can systematically integrate the global and local optimal solutions in the process of optimizing pollution source parameters, ensuring that the matrix covers the source strength distribution of the entire storage yard area. This provides a complete data foundation for pollution source extraction. By setting a source strength threshold and extracting the location and strength of pollution sources based on it, it can accurately screen out real and effective pollution sources, avoid misjudging weak source strength grids caused by simulation errors as pollution sources, and eliminate areas with no actual dust generation risk, further improving the accuracy of locating pollution source locations and strengths.

[0018] Optionally, the stacking mechanism features include the Biot number and the effective diffusion mass transfer coefficient.

[0019] This invention, through the stacking mechanism characteristics, can characterize the moisture transport characteristics of stacks from the level of the underlying mass transfer law. The Biot number can quantify the ratio of the internal moisture diffusion resistance to the surface mass transfer resistance of the stack, clarifying the dominant resistance link. The effective diffusion mass transfer coefficient can reflect the actual diffusion efficiency of moisture in the stack pores and adapt to dynamic changes in environmental conditions. Both provide mechanistic support for the moisture content change trend prediction model, avoiding the insufficient generalization ability caused by the moisture content change trend prediction model relying solely on surface data, and improving the performance of the moisture content change trend prediction model.

[0020] Optionally, the mass transfer Biot number satisfies the following formula: in, For mass transfer, The convective heat transfer coefficient is... For characteristic length, Porosity For an effective diffusion mass transfer coefficient, The density of reference air, The specific heat capacity of air. For Lewis numbers, This is an empirical constant.

[0021] The Biot number formula of this invention quantifies the ratio of internal moisture diffusion resistance to surface mass transfer resistance by integrating multiple parameters such as flow heat transfer coefficient, characteristic length, porosity, and effective diffusion mass transfer coefficient. It clarifies the dominant link of resistance in the moisture mass transfer process of the stack and improves the accuracy of the moisture content change trend prediction model.

[0022] Optionally, the effective diffusion mass transfer coefficient satisfies the following formula: in, For an effective diffusion mass transfer coefficient, The water diffusion coefficient under standard conditions. Standard atmospheric pressure To measure atmospheric pressure, This is the actual thermodynamic temperature. The reference thermodynamic temperature is used.

[0023] The effective diffusion mass transfer coefficient formula of this invention combines the actual changes in standard moisture diffusion rate, atmospheric pressure and thermodynamic temperature to accurately quantify the actual diffusion efficiency of moisture in the stack pores, and can dynamically adapt to the influence of environmental conditions on moisture transport, thereby improving the accuracy of the moisture content change trend prediction model.

[0024] Another aspect of the present invention provides an intelligent dust suppression decision system for stockpile dust, comprising: a processor, an input device, an output device, and a memory, wherein the processor, the input device, the output device, and the memory are interconnected, wherein the memory is used to store a computer program, the computer program including program instructions, and the processor is configured to invoke the program instructions to execute the intelligent dust suppression decision method for stockpile dust according to any one of the preceding aspects of the present invention.

[0025] The present invention provides an intelligent dust suppression decision system for stockpile dust, which is compact in structure, stable in performance, highly integrated and simple in composition. It can stably execute the intelligent dust suppression decision method for stockpile dust provided in the preceding aspect of the present invention, further improving the overall applicability and practical application capability of the present invention. Attached Figure Description

[0026] Figure 1 This is a flowchart of an intelligent dust suppression decision-making method for stockpile dust according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the structure of an intelligent dust suppression decision system for stockpile dust according to an embodiment of the present invention. Detailed Implementation

[0027] Specific embodiments of the present invention will now be described in detail. It should be noted that the embodiments described herein are for illustrative purposes only and are not intended to limit the invention. In the following description, numerous specific details are set forth in order to provide a thorough understanding of the invention. However, it will be apparent to those skilled in the art that these specific details are not necessary to practice the invention. In other instances, well-known circuits, software, or methods have not been specifically described to avoid obscuring the invention.

[0028] Throughout this specification, references to "an embodiment," "an embodiment," "an example," or "an example" mean that a particular feature, structure, or characteristic described in connection with that embodiment or example is included in at least one embodiment of the invention. Therefore, the phrases "in an embodiment," "in an embodiment," "an example," or "an example" appearing in various places throughout the specification do not necessarily refer to the same embodiment or example. Furthermore, specific features, structures, or characteristics can be combined in one or more embodiments or examples in any suitable combination and / or sub-combination. Moreover, those skilled in the art will understand that the illustrations provided herein are for illustrative purposes and are not necessarily drawn to scale.

[0029] Please see Figure 1 In order to solve the problems in the prior art, in an alternative embodiment, such as Figure 1This paper presents an intelligent dust suppression decision-making method for stockpile dust. It includes the following steps: Step S1: Obtain the dust concentration at the monitoring point of the target stockpile and interpolate the dust concentration to obtain a dust concentration distribution map.

[0030] In this embodiment, dust concentration monitoring points are deployed at the target stockpile, with each monitoring point covering no more than 200,000 square meters. This ensures that the monitoring points can accurately capture concentration data of the core pollution area and diffusion path. However, due to limitations in monitoring equipment cost, stockpile terrain complexity, and deployment construction conditions, the actual number of monitoring points often cannot achieve dense coverage of the entire area. Consequently, the monitoring data can only reflect the dust concentration at local points and cannot fully present the spatial distribution characteristics of the entire stockpile's concentration, resulting in a large number of data gaps. Therefore, after obtaining the dust concentration data from the monitoring points, spatial interpolation algorithms (such as Kriging interpolation, inverse distance weighted interpolation, etc.) need to be used. Based on the measured dust concentration data from existing monitoring points, the correlation and continuity of spatial data are utilized to infer the dust concentration values ​​in areas where no monitoring points have been deployed, thereby generating a complete and visualized dust concentration distribution map.

[0031] Step S2: Obtain meteorological data from the monitoring points of the target storage yard, and use the meteorological data and the dust concentration distribution map to calculate the location and intensity of the pollution source.

[0032] The acquisition of meteorological data from the target storage yard monitoring points includes the following sub-steps: Step S201: Identify key meteorological factors.

[0033] The determination of key meteorological factors specifically includes the following sub-steps: Step S20101: Based on the historical meteorological data and historical dust concentration of the target storage yard acquired in advance, calculate the historical monthly average meteorological data and historical monthly average dust concentration.

[0034] In this embodiment, hourly historical data of the target stockpile for the entire year are collected in advance. The historical meteorological data must include hourly temperature, relative humidity, wind speed, wind direction, air pressure, precipitation, cloud cover, and other core parameters related to dust diffusion and changes in the moisture content of the stockpile. The historical dust concentration data must include hourly concentrations of PM2.5, PM10, and TSP. To improve the accuracy of subsequent calculations, the collected hourly raw data can be preprocessed to remove invalid data caused by monitoring equipment failures, data transmission interruptions, etc. After preprocessing, the valid data are grouped by natural month. For the hourly meteorological data in each group, the arithmetic mean of each meteorological parameter is calculated. For the hourly dust concentration data in each group, the arithmetic mean of PM2.5, PM10, and TSP is calculated. Finally, the historical monthly average meteorological data (such as monthly average temperature, monthly average relative humidity, monthly average air pressure, etc.) and historical monthly average dust concentration data for the target stockpile for the 12 months of the year are generated.

[0035] Step S20102: Perform correlation analysis on the historical monthly average meteorological data and the historical monthly average dust concentration to obtain the Pearson correlation coefficient.

[0036] In this embodiment, before performing correlation analysis on the historical monthly average meteorological data and the historical monthly average dust concentration, the historical monthly average meteorological data (including core parameters such as monthly average temperature, monthly average relative humidity, monthly average wind speed, and monthly average air pressure) is matched one-to-one with the historical monthly average dust concentration data (including monthly average PM2.5, monthly average PM10, and monthly average TSP concentration) according to the time dimension of 12 natural months of the year, forming 12 sets of monthly data pairs with consistent time dimensions. This avoids the correlation analysis results being affected by data misalignment. Subsequently, the organized monthly data pairs are imported into the Origin data analysis software, and the Pearson correlation coefficient analysis function is selected. For each monthly average meteorological parameter (such as monthly average wind speed) and each type of dust monthly average concentration (such as monthly average PM10 concentration), pairwise linear correlation calculations are performed. The correlation coefficient value between the two is solved by the software algorithm, with a value range of [-1, 1]. The closer the absolute value is to 1, the higher the correlation coefficient value. The stronger the linear correlation, the better. Positive values ​​indicate positive correlation and negative values ​​indicate negative correlation. To ensure the reliability of the results, a two-level significance test (such as a p-value test) is performed simultaneously in the software to eliminate spurious correlations caused by random factors and select statistically significant correlation coefficient results.

[0037] Step S20103: Determine key meteorological factors based on the Pearson correlation coefficient.

[0038] In this embodiment, the absolute value of the correlation coefficient is used as the core quantitative criterion. Meteorological parameters that are strongly correlated with the concentration of two or more of the dust types, such as PM2.5, PM10, and TSP, are selected first. At the same time, the technical logic of dust diffusion and changes in the moisture content of the stack is combined, such as wind speed directly affecting the range and rate of dust diffusion, relative humidity being related to the evaporation efficiency of moisture on the stack surface, and temperature affecting the diffusion and mass transfer process of moisture inside the stack. This further verifies and identifies key meteorological factors (such as wind speed, relative humidity, and temperature). The key meteorological factors finally determined must be able to accurately cover the core driving factors of dust concentration changes.

[0039] Step S202: Obtain meteorological data of the target storage yard monitoring point based on the key meteorological factors.

[0040] In this embodiment, a meteorological parameter acquisition module is built simultaneously based on the dust monitoring points already deployed in the target storage yard. At the monitoring points at the edge of the stack, the loading and unloading operation area, and the wind-sensitive areas around the storage yard, meteorological data of corresponding key factors are collected in real time through equipment such as temperature and humidity sensors, wind speed and direction instruments, and air pressure sensors. During the acquisition process, invalid data caused by equipment failure or transmission interruption are simultaneously removed to ensure that the acquired meteorological data is accurately matched with the key factors.

[0041] The calculation of the pollution source location and pollution source intensity using the meteorological data and the dust concentration distribution map specifically includes the following sub-steps: Step S211: The target storage yard is divided into grids to obtain the target grid.

[0042] In this embodiment, considering the balance between positioning accuracy and computational efficiency, the grid size (e.g., 5m×5m or 10m×10m) is determined. A regular rectangular grid discretization method is adopted to uniformly divide the defined storage area into several independent grid units. Each grid unit serves as the basic spatial unit for subsequent identification of potential pollution sources.

[0043] Step S212: Determine the initial locations of multiple pollution sources and their corresponding initial pollution source strengths based on the target grid and the dust concentration distribution map.

[0044] In this embodiment, based on the divided target grid and the interpolated dust concentration distribution map, the high-concentration grid cell clusters with significantly higher dust concentrations than the background value are first located through visualization analysis. The top 20%-30% of grid cells by concentration are preferentially selected, with the specific proportion adjusted according to the size of the stockpile to ensure coverage of the core pollution area. The center coordinates of these high-concentration grid cells are used as the basic candidate locations for potential pollution sources. Simultaneously, a small number of second-highest concentration grid cells (ranked 30%-50%) are supplemented as candidates to avoid incomplete initial population coverage due to a single concentration gradient, thus forming the initial location set required by the particle swarm optimization algorithm. For the initial pollution source strength corresponding to each initial location, reference is made to similar stockpile locations. Historical source strength monitoring data (such as the same stockpiled materials and similar stacking heights), or based on the empirical correlation between concentration and source strength (such as setting the initial source strength of high-concentration grid cells to the middle to high value of the empirical range, and setting the second-highest concentration grid cells to the middle to low value of the empirical range), are combined with the area size of the target grid cells for correction. Finally, a reasonable initial source strength value is matched for each initial location. The initial locations of multiple pollution sources and their corresponding initial source strengths determined in this way are essentially the initial population for constructing the particle swarm optimization algorithm. This ensures that the initial population can focus on high-pollution risk areas and conform to the actual pollution distribution pattern, while also providing diverse starting points for subsequent iterative optimization, avoiding the algorithm from getting trapped in local optima, and laying the foundation for accurately calculating the final pollution source location and source strength.

[0045] Step S213: Based on the initial location of the pollution source and the initial pollution source strength, the pollution source location and pollution source strength are obtained iteratively using the particle swarm optimization algorithm.

[0046] The process of iteratively obtaining the pollution source location and pollution source strength using a particle swarm optimization algorithm based on the initial location and initial pollution source strength specifically includes the following sub-steps: Step S21301: Introduce a Gaussian diffusion model and use the Gaussian diffusion model to perform diffusion simulation based on the initial location of the pollution source, the initial source strength of the pollution, and the meteorological data to obtain a simulated concentration distribution map.

[0047] In this embodiment, a Gaussian diffusion model adapted to dust diffusion scenarios in open environments of stockyards (including steady-state diffusion, neutral and unstable diffusion conditions) is introduced. This model closely matches the actual physical laws of dust diffusion in the target stockyard, providing a reliable theoretical basis for subsequent simulation calculations. Subsequently, the initial location of the pollution source, the corresponding initial source strength, and the selected key meteorological data are used as input parameters for the model. These parameters are substituted into the Gaussian diffusion model one by one for calculation. The model calculates the simulated dust concentration value of each grid cell in the target stockyard under the current input conditions. Then, according to the geographical coordinate correspondence of the grid cells, all simulated concentration values ​​are integrated to generate a simulated concentration distribution map with the same spatial dimension as the measured dust concentration distribution map.

[0048] Step S21302: Obtain the location of the monitoring point, and extract the simulated dust concentration at the monitoring point based on the location of the monitoring point using the simulated concentration distribution map.

[0049] In this embodiment, the specific geographical coordinates of the dust monitoring points that have been deployed in the target storage yard are first obtained. Then, the coordinates of these monitoring points are spatially matched with the grid system of the simulated concentration distribution map to locate the grid cell corresponding to each monitoring point in the simulated concentration distribution map. Finally, the simulated dust concentration value in the grid cell is extracted to obtain the simulated dust concentration corresponding to each monitoring point.

[0050] Step S21303: Construct a fitness function with the goal of minimizing the mean square error between the simulated dust concentration at the monitoring point and the dust concentration.

[0051] In this embodiment, the measured dust concentration at each monitoring point is matched one by one with the simulated dust concentration at the corresponding monitoring point. Then, the mean square error of the two concentrations at each monitoring point is calculated. The overall error value is obtained by summing the mean square errors of all monitoring points (or averaging the mean square errors). The calculation expression of the overall error value is defined as the fitness function. The smaller the fitness function value, the higher the degree of agreement between the simulated concentration and the measured concentration corresponding to the initial location and initial source strength of the pollution source.

[0052] Step S21304: Based on the initial location of the pollution source and the initial pollution source strength, the pollution source location and pollution source strength are obtained iteratively using the particle swarm optimization algorithm combined with the fitness function.

[0053] The method of obtaining the pollution source location and pollution source strength by iteratively using a particle swarm optimization algorithm combined with the fitness function based on the initial location and initial pollution source strength includes: Step S2130401: Based on the initial location of the pollution source and the initial source strength, the optimal source strength matrix is ​​obtained by iteratively using the particle swarm optimization algorithm combined with the fitness function.

[0054] In this embodiment, the initial position and corresponding initial source strength of each independent pollution source are defined as a particle in the algorithm. An initial particle set containing multiple sets of candidate parameters is constructed, and then the algorithm iteration process is initiated: In each iteration, each particle combines its own historical best solution (the "position-source strength" parameter corresponding to the minimum fitness function value in the previous iteration of the particle) with the global best solution of the entire population (the "position-source strength" parameter corresponding to the minimum fitness function value among all particles), and adjusts its corresponding pollution source position (fine-tuning the coordinates within the target grid range) and pollution source strength (correcting the value within a reasonable range) according to the velocity-position update formula of the particle swarm optimization algorithm. After each round of particle parameter update, the new "pollution source position" is... The "position-source strength" parameters are substituted into the Gaussian diffusion model to calculate the simulated concentration. Then, the current fitness value of the particle is evaluated by the fitness function (the sum of the mean square error between the simulated and measured concentrations at the monitoring point) constructed earlier. The particle's own historical optimal solution and the global optimal solution of the population are updated. The above iterative process of "parameter update-fitness evaluation-optimal solution update" is repeated until the preset termination condition is met (such as the number of iterations reaching a set threshold, or the global optimal fitness value of the population not decreasing significantly for several consecutive rounds). At this time, the global optimal "position-source strength" parameter set obtained by the algorithm converges is arranged in the geographical coordinate order of the target grid. The optimal source strength value corresponding to each grid cell is filled into the corresponding coordinate position, and finally the optimal source strength matrix covering the entire target field is formed.

[0055] Step S2130402: Set a source strength threshold, and extract the pollution source location and pollution source strength from the optimal source strength matrix based on the source strength threshold.

[0056] In this embodiment, the source strength threshold must first be scientifically set. This threshold should be determined by considering the actual pollution background of the target storage area, such as the average source strength of the grid during periods without significant dust generation, the lowest effective pollution source strength verified in historical pollution events, or by statistically analyzing the mean and standard deviation of the source strength values ​​of all grids within the optimal source strength matrix. This avoids missing real pollution sources due to an excessively high threshold or introducing false pollution sources due to simulation errors caused by an excessively low threshold. After determining the threshold, all grid cells in the optimal source strength matrix are traversed, and the source strength value of each grid cell is compared with the source strength threshold. Grid cells with source strength values ​​greater than the threshold are selected. These cells represent potential pollution source areas with actual dust emissions. Subsequently, the geographical coordinates of each qualified grid cell are extracted (using the grid center coordinates as the precise location of the pollution source to ensure correspondence with the actual spatial location of the storage area), and the source strength value corresponding to that grid cell is recorded as the pollution source strength.

[0057] To further improve accuracy, spatial correlation verification can be performed on the selected grid cells. Adjacent grid cells with source strength values ​​exceeding the threshold can be merged. These cells usually belong to the same continuous pollution source area. To avoid duplicate statistics, isolated abnormal grid cells with source strength values ​​only slightly higher than the threshold can be removed. These cells are likely caused by simulation calculation errors. Finally, a one-to-one correspondence list of pollution source locations and pollution source strengths is formed.

[0058] Step S3: Construct a moisture content change trend prediction model. Based on the stacking mechanism characteristics of the pollution source location, the moisture content of the stack surface, and the meteorological data, use the moisture content change trend prediction model to obtain the predicted value of the stack surface moisture content changing over time.

[0059] The stacking mechanism features include the mass transfer Biot number and the effective diffusion mass transfer coefficient.

[0060] In this embodiment, a model for predicting the trend of water content change is obtained by using a neural network algorithm as the core framework (such as LSTM or BP neural network, adapted to the dynamic change characteristics of time series data) and training data.

[0061] The input feature layer of the training data covers three core parameters: first, the characteristic parameters of the pollution source location and stacking mechanism (Bioworth number and effective diffusion mass transfer coefficient, where the Bioworth number is calculated from parameters such as convective heat transfer coefficient, characteristic length, and porosity, and the effective diffusion mass transfer coefficient is calculated by combining standard moisture diffusion rate, measured atmospheric pressure, and actual thermodynamic temperature, and both types of parameters need to be recorded synchronously at a daily / hourly frequency); second, real-time environmental parameters (screened key meteorological factors, including hourly temperature, relative humidity, wind speed, and air pressure, from monitoring points in the stockpile). The data collection process includes: 1) real-time data acquisition and alignment with the timestamps of the mechanism characteristic parameters; 2) measured data of the surface moisture content of the stockpile (collected at fixed frequencies using an online moisture meter at the pollution source location, serving as the direct source of the model output labels); 3) supplemented with hourly historical data of the target stockpile throughout the year (historical meteorological parameters, historical stockpile moisture content, and historical mechanism characteristic parameters); and 4) data samples divided into initial training datasets based on the correspondence between input features and output labels, with time intervals of 10 minutes and 30 minutes (matching the response speed of moisture content changes).

[0062] The analysis of moisture evaporation process is based on the concept of wet shrinkage, assuming that there is a wet shrinkage zone within the coal particle, mainly composed of external and internal moisture. Moisture inside the particle undergoes transport through the internal dry zone and interfacial moisture convection mass transfer.

[0063] The formula for calculating the mass transfer Biot number is as follows: In the formula: For mass transfer Biot number; The convective mass transfer coefficient; The characteristic length (where the characteristic radius is the stacking radius) ); Porosity; For an effective diffusion mass transfer coefficient, With convective heat transfer coefficient The relationship is as follows: In the formula: Let be the convective heat transfer coefficient, taken as 11. The complete expression for the mass transfer Biot number is derived from the above formula.

[0064] The mass transfer Biot number satisfies the following formula: in, For mass transfer, The convective heat transfer coefficient is... For characteristic length, Porosity For an effective diffusion mass transfer coefficient, The density of the reference air is taken as 1.239. Let be the specific heat capacity of air, taken as 1005. Let be a Lewis number, and take 1.2. This is an empirical constant.

[0065] The empirical constant is between 0.3 and 0.5. This range represents the typical range of empirical constants under laminar and turbulent conditions in atmospheric boundary layer flow, which is the core flow mode in the storage yard environment. It can adapt to the needs of pollution source tracing and moisture content prediction under different wind directions and speeds.

[0066] Reverse verification based on pre-acquired historical data The value of is determined by first extracting historical data samples showing the correlation between wind speed and the rate of change of moisture content on the stack surface (wind speed is used to determine the flow condition, and the rate of change of moisture content reflects the actual mass transfer efficiency, which is directly related to ). Then, for different wind speed ranges (e.g., wind speed less than 2 m / s is determined to be laminar flow, and wind speed greater than or equal to 2 m / s is determined to be turbulent flow), historical samples of the corresponding range are selected, and the corresponding parameters are substituted into . And by inferring the actual rate of change in historical moisture content (Water evaporation rate and) They are positively correlated (the calculated value can be derived from the mass transfer formula), and finally, the result obtained by reverse calculation... Substitute into the formula and solve for different The theoretical calculation values ​​corresponding to the values ​​(such as 0.3, 0.35, 0.4, 0.45) are selected to minimize the deviation between the theoretical and reverse calculations. Values, for example, in the turbulent region (wind speed ≥ 2 m / s), if If the deviation between the two is less than 5%, and this deviation meets the accuracy requirements for subsequent mass transfer Biot number calculations, then the wind speed range is determined. The laminar flow interval is determined similarly. .

[0067] The effective diffusion mass transfer coefficient satisfies the following formula: in, For an effective diffusion mass transfer coefficient, Let be the water diffusion coefficient under standard conditions, and take . , Standard atmospheric pressure To measure atmospheric pressure, This is the actual thermodynamic temperature. As the reference thermodynamic temperature, take .

[0068] The core function of the Biot number is to quantify the ratio of internal moisture diffusion resistance to surface mass transfer resistance in a stack. It can clearly identify the dominant resistance factor in the moisture transfer process; for example, when the value is small, internal diffusion resistance can be ignored. ), indicating that the moisture concentration in the dry zone inside the stack has reached saturation, when the value is large ( The surface mass transfer resistance can be ignored, and the moisture concentration in the internal dry zone can be determined to be consistent with the surrounding moisture concentration. This is a key basis for accurately depicting the inherent laws of moisture transport in the stack. The effective diffusion mass transfer coefficient directly reflects the actual diffusion efficiency of moisture in the pore structure of the stack, and its value is dynamically adjusted with environmental conditions such as temperature and air pressure, which can adapt to the changes in moisture diffusion characteristics under different working conditions in real time.

[0069] Using both as inputs to the moisture content change trend prediction model essentially provides underlying mechanistic support for the model: on the one hand, it avoids the insufficient generalization ability of the model caused by relying solely on surface data such as meteorological data and measured moisture content, allowing the model to understand the logic of water evaporation and transport from the essential level of mass transfer, rather than simply fitting the data; on the other hand, it enables the model to accurately respond to the impact of different storage conditions (such as different porosities and stacking heights) and meteorological conditions (such as temperature fluctuations and air pressure changes) on water mass transfer, ensuring that the output moisture content prediction value not only conforms to physical laws but also adapts to the variable changes in the actual scenario.

[0070] During prediction, the Biot number (calculated by formula in combination with parameters such as convective heat transfer coefficient, characteristic length, and porosity) and the effective diffusion mass transfer coefficient (derived by formula based on standard moisture diffusion rate, measured atmospheric pressure, and actual thermodynamic temperature) of the stack at the pollution source location are taken as stacking mechanism characteristics. Simultaneously, the surface moisture content of the stack at this location is collected (obtained in real time through an online moisture meter) and key meteorological data (such as temperature, relative humidity, wind speed, and air pressure) are screened to ensure that the timestamps of the three types of data are aligned. After removing outliers, the data are input into the moisture content change trend prediction model. Based on the learned mechanism characteristics and the correlation between meteorological data and moisture content changes, the model outputs the predicted value of the surface moisture content of the stack that changes dynamically over time in a specific future period.

[0071] It should be noted that the training and prediction data for the moisture content change trend prediction model have been normalized.

[0072] Step S4: When the predicted value of the moisture content on the stack surface is less than the preset threshold value of the moisture content on the stack surface, water is sprayed on the stack at the location of the pollution source to suppress dust based on the pollution source strength.

[0073] In this embodiment, a threshold for the surface moisture content of the stacked material is first set based on the critical moisture content requirement for dust-free operation and the historical dust suppression effect data of the target stockpile (including dust concentration monitoring values ​​at different moisture contents). This threshold must ensure that the amount of dust generated can be controlled within the compliant range when the surface moisture content of the stack is not lower than this value. Subsequently, the predicted value of the surface moisture content of the pollution source location is obtained in real time from the moisture content change trend prediction model. The predicted value is compared with the preset threshold for each time period. When the predicted value for a certain time period is less than the threshold, it is determined that there is a risk of dust generation at the location of the pollution source, and water spraying dust suppression operation needs to be initiated. During the spraying process, the pollution source strength calculated in the early stage using particle swarm optimization algorithm combined with Gaussian diffusion model is used as the basis. If the pollution source strength is large, the corresponding dust emission is high and the risk of dust generation is high. In this case, the water volume sprayed per unit time of the spraying equipment in the stacking area is increased and the duration of a single spraying is extended. If the pollution source strength is small, the operating parameters of the spraying equipment are appropriately reduced to avoid water waste. At the same time, it is ensured that the spraying equipment is accurately aimed at the stacking area of ​​the identified pollution source location so that the spraying range is completely matched with the coverage range of the pollution source. Ultimately, on-demand, precise and efficient dust suppression control is achieved, which not only effectively ensures the dust suppression effect, but also meets the energy-saving and consumption-reducing requirements of intelligent dust suppression decision-making.

[0074] like Figure 2As shown, in another aspect, the present invention also provides an intelligent dust suppression decision system for stockyards, comprising: a processor, an input device, an output device, and a memory, wherein the processor, the input device, the output device, and the memory are interconnected, wherein the memory is used to store a computer program, the computer program includes program instructions, and the processor is configured to call the program instructions to execute the relevant steps of a relevant embodiment of the intelligent dust suppression decision method for stockyards of the present invention.

[0075] This invention provides an intelligent dust suppression decision-making system for stockpile dust. The functional components can be integrated into a single processing unit, or each component can exist independently, or two or more components can be integrated into one unit. The integrated components can be implemented in hardware or software.

[0076] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered within the scope of the claims and specification of the present invention.

Claims

1. A smart dust suppression decision-making method for stockpile dust, characterized in that, The method includes: Obtain the dust concentration at the monitoring point of the target stockpile, and interpolate the dust concentration to obtain a dust concentration distribution map; Meteorological data from monitoring points at the target storage yard are acquired, and the location and intensity of the pollution source are calculated using the meteorological data and the dust concentration distribution map. A moisture content change trend prediction model is constructed. Based on the stacking mechanism characteristics of the pollution source location, the moisture content of the stack surface, and the meteorological data, the moisture content change trend prediction model is used to obtain the predicted value of the stack surface moisture content changing over time. When the predicted value of the moisture content on the stack surface is less than the preset threshold value for the moisture content on the stack surface, water is sprayed on the stack at the location of the pollution source to suppress dust based on the pollution source strength.

2. The intelligent dust suppression decision-making method for stockpile dust according to claim 1, characterized in that, The acquisition of meteorological data from the target storage yard monitoring points includes: Identify key meteorological factors; Meteorological data from the target storage yard monitoring points are obtained based on the key meteorological factors.

3. The intelligent dust suppression decision-making method for stockpile dust according to claim 2, characterized in that, The determination of key meteorological factors includes: Based on the historical meteorological data and historical dust concentration of the target storage yard acquired in advance, the historical monthly average meteorological data and historical monthly average dust concentration are calculated. Correlation analysis was performed on the historical monthly average meteorological data and the historical monthly average dust concentration to obtain the Pearson correlation coefficient; Key meteorological factors were determined based on the Pearson correlation coefficient.

4. The intelligent dust suppression decision-making method for stockpile dust according to claim 1, characterized in that, The calculation of pollution source location and pollution source intensity using the meteorological data and the dust concentration distribution map includes: The target storage area is divided into grids to obtain the target grid; Based on the target grid and the dust concentration distribution map, the initial locations of multiple pollution sources and their corresponding initial pollution source intensities are determined. The pollution source location and pollution source strength are obtained iteratively using a particle swarm optimization algorithm based on the initial location and initial pollution source strength.

5. The intelligent dust suppression decision-making method for stockpile dust according to claim 4, characterized in that, The step of iteratively obtaining the pollution source location and pollution source strength using a particle swarm optimization algorithm based on the initial location and initial pollution source strength includes: A Gaussian diffusion model is introduced, and diffusion simulations are performed using the Gaussian diffusion model based on the initial location of the pollution source, the initial source strength of the pollution, and the meteorological data to obtain simulated concentration distribution maps; Obtain the location of the monitoring point, and extract the simulated dust concentration at the monitoring point using the simulated concentration distribution map based on the location of the monitoring point; A fitness function is constructed with the goal of minimizing the mean square error between the simulated dust concentration at the monitoring point and the dust concentration. Based on the initial location and initial intensity of the pollution source, the pollution source location and intensity are obtained iteratively using a particle swarm optimization algorithm combined with the fitness function.

6. The intelligent dust suppression decision-making method for stockpile dust according to claim 5, characterized in that, The step of iteratively obtaining the pollution source location and pollution source strength based on the initial location and initial pollution source strength using a particle swarm optimization algorithm combined with the fitness function includes: The optimal source strength matrix is ​​obtained by iteratively combining the particle swarm optimization algorithm with the fitness function based on the initial location and initial source strength of the pollution source. Set a source strength threshold, and extract the pollution source location and pollution source strength from the optimal source strength matrix based on the source strength threshold.

7. The intelligent dust suppression decision-making method for stockpile dust according to claim 1, characterized in that, The stacking mechanism features include the mass transfer Biot number and the effective diffusion mass transfer coefficient.

8. The intelligent dust suppression decision-making method for stockpile dust according to claim 7, characterized in that, The mass transfer Biot number satisfies the following formula: in, For mass transfer, The convective heat transfer coefficient is... For characteristic length, Porosity For an effective diffusion mass transfer coefficient, The density of reference air, The specific heat capacity of air. For Lewis numbers, This is an empirical constant.

9. The intelligent dust suppression decision-making method for stockpile dust according to claim 7, characterized in that, The effective diffusion mass transfer coefficient satisfies the following formula: in, For an effective diffusion mass transfer coefficient, The water diffusion coefficient under standard conditions. Standard atmospheric pressure To measure atmospheric pressure, The actual thermodynamic temperature, The reference thermodynamic temperature is used.

10. A smart dust suppression decision-making system for stockpile dust, characterized in that, include: The system includes a processor, an input device, an output device, and a memory, all interconnected. The memory stores a computer program, which includes program instructions. The processor is configured to invoke the program instructions to execute a smart dust suppression decision-making method for stockpile dust as described in any one of claims 1 to 9.

Images