Dust concentration monitoring and treatment method and system for intelligent furniture production workshop

Abstract

The invention provides a dust concentration monitoring and treatment method and system for an intelligent furniture production workshop, and relates to the technical field of intelligent furniture, and the method comprises the steps: collecting workshop dust concentration data through arranging a sensor, building a workshop dust concentration distribution three-dimensional model through employing a convolutional neural network, predicting the dust concentrations of different regions and different time periods, and obtaining the dust concentration of the workshop. And according to a prediction result, dividing dust danger level areas. And when the dust concentration exceeds the standard, generating early warning information and sending the early warning information to a manager terminal. And the control center determines the treatment priority according to the dust concentration distribution three-dimensional model, controls the ventilation system and the movable dust removal device to perform graded filtration, optimizes the three-dimensional model on line according to the real-time dust concentration data, and adjusts the treatment strategy. According to the invention, precise monitoring and efficient treatment of workshop dust concentration are realized, dust harm is reduced, and health of workers is guaranteed.

Description

Technical Field

[0001] The present invention relates to the technology of intelligent furniture, and particularly to a method and system for monitoring and controlling the dust concentration in the production workshop of intelligent furniture. Background Art

[0002] A large amount of dust is generated during the production process of intelligent furniture. For example, processes such as wood processing, grinding, and painting will release dust particles of different particle sizes. These dusts will not only pollute the workshop environment, but also endanger the respiratory health of workers, causing respiratory diseases and even lung diseases. In addition, high-concentration dust may also trigger fire or explosion accidents, posing a threat to production safety. Therefore, it is crucial to effectively monitor and control the dust concentration in the production workshop of intelligent furniture.

[0003] Traditional dust concentration monitoring methods usually rely on sensors at fixed positions for measurement, and cannot comprehensively reflect the spatial distribution of dust concentration in the workshop. Moreover, traditional dust control methods are often post-treatment, lacking predictability and initiative, and it is difficult to effectively control the dust concentration, resulting in frequent exceeding of the dust concentration standard in the workshop. In addition, traditional dust control systems usually lack intelligent control and cannot dynamically adjust the control strategy according to the real-time change of dust concentration, resulting in energy waste and low control efficiency. Summary of the Invention

[0004] Embodiments of the present invention provide a method and system for monitoring and controlling the dust concentration in the production workshop of intelligent furniture, which can solve the problems in the prior art.

[0005] In the first aspect of the embodiments of the present invention,

[0006] A method for monitoring and controlling the dust concentration in the production workshop of intelligent furniture is provided, including:

[0007] Laser scattering particle counters and gravimetric dust samplers are arranged in the production workshop of intelligent furniture to collect the dust concentration data of different areas in the workshop and transmit it to the control center; based on the workshop dust concentration data, the control center uses a convolutional neural network to train the historical workshop dust concentration data, constructs a three-dimensional model of the workshop dust concentration distribution, and dynamically predicts the dust concentration in different areas and different time periods of the workshop based on the three-dimensional model of the dust concentration distribution;

[0008] Based on the prediction results of the three-dimensional model of the dust concentration distribution, the control center divides the workshop into two dust hazard level areas, one with a dust concentration higher than the preset hazard threshold and the other with a dust concentration lower than the preset hazard threshold; when the laser scattering particle counter and the gravimetric dust sampler detect that the dust concentration in a certain area exceeds the preset hazard threshold, the control center generates a warning message, which records the location of the dust-exceeding area, the current dust concentration value and the duration of the exceedance, and sends the warning message to the on-site management personnel terminal; the control center determines the treatment priority of each area according to the dust concentration change trend calculated by the three-dimensional model of the dust concentration distribution;

[0009] According to the treatment priority of each area, the control center issues control instructions to the ventilation systems of the two dust hazard level areas with dust concentrations higher than and lower than the preset hazard threshold respectively, to adjust the opening degree of the ventilation duct, the rotation speed of the fan and the air supply direction; at the same time, according to the optimal dust removal path calculated by the three-dimensional model of the dust concentration distribution, the control center controls the mobile dust removal device to navigate to the area with the highest dust concentration, and starts the primary filtration unit, the electrostatic dust removal unit and the activated carbon adsorption unit of the mobile dust removal device for hierarchical filtration; the laser scattering particle counter and the gravimetric dust sampler collect the particle dust concentration data during the treatment process in real time and feedback it to the control center, and the control center performs online optimization and update of the three-dimensional model of the dust concentration distribution according to the particle dust concentration data, and adjusts the ventilation system parameters and the operation strategy of the mobile dust removal device accordingly.

[0010] Laser scattering particle counters and gravimetric dust samplers are arranged in the intelligent furniture production workshop to collect the dust concentration data of different areas in the workshop and transmit it to the control center; based on the workshop dust concentration data, the control center uses a convolutional neural network to train the historical workshop dust concentration data, constructs a three-dimensional model of the workshop dust concentration distribution, and dynamically predicts the dust concentration in different areas and at different times in the workshop, including:

[0011] The intelligent furniture production workshop is divided into three functional blocks: the processing area, the grinding area and the painting area according to its functions, and a three-layer three-dimensional monitoring network of the ground layer, the human breathing layer and the upper space layer is constructed in each functional block; laser scattering particle counters are arranged in the three-layer three-dimensional monitoring network of each functional block, and the laser scattering particle counter uses a semiconductor laser as the light source to obtain the dust particle count value based on the measurement of the forward scattered light intensity; gravimetric dust samplers are arranged in the human breathing layer of each functional block, and the gravimetric dust sampler obtains the inhalable dust mass concentration value through a PM2.5 cutter and a quartz filter membrane;

[0012] The dust particle count value of the laser scattering particle counter and the inhalable dust mass concentration value of the gravimetric dust sampler are transmitted to a data acquisition module via an RS-485 bus. The data acquisition module reads and encapsulates the dust particle count value and the inhalable dust mass concentration value using the Modbus RTU communication protocol, and uploads the encapsulated data to a control center via an industrial Ethernet. The control center calibrates the dust particle count value based on the inhalable dust mass concentration value to generate standardized workshop dust concentration data;

[0013] The control center organizes the standardized workshop dust concentration data according to temporal and spatial characteristics to construct a training data set containing historical data for the past three months. Each record in the training data set contains a timestamp, spatial coordinates, and a dust concentration value. A three-dimensional model of the workshop dust concentration distribution is constructed using a three-dimensional convolutional neural network based on the training data set, and spatial and temporal characteristics are extracted through four convolutional blocks and an LSTM layer. The control center uses the three-dimensional model of the workshop dust concentration distribution to dynamically predict the dust concentration in different areas of the workshop for the next four hours.

[0014] Based on the prediction results of the dust concentration distribution three-dimensional model, the control center divides the workshop into two dust hazard level areas: above a preset danger threshold and below a preset danger threshold. When the laser scattering particle counter and the gravimetric dust sampler detect that the dust concentration in a certain area exceeds the preset danger threshold, the control center generates a warning message including:

[0015] The control center performs dust concentration prediction on the workshop based on the dust concentration distribution three-dimensional model to obtain prediction result data. The control center divides the workshop space into two dust hazard level areas: above a preset danger threshold area and below a preset danger threshold area according to the prediction result data;

[0016] The control center performs binary processing on the prediction result data using an adaptive threshold segmentation algorithm, marks the area above the preset danger threshold as a first value, marks the area below the preset danger threshold as a second value, and performs smoothing processing on the marked data using a three-dimensional median filter with a 3×3×3 dimension to obtain a marked result;

[0017] The control center performs connected component analysis on the marked result using a region growing algorithm, detects the twenty-six adjacent points of each marked point, and classifies the adjacent points with the same mark into the same connected component to obtain the workshop dust hazard zoning result;

[0018] When the dust concentration in a certain area is detected by the laser scattering particle counter and the gravimetric dust sampler to exceed the preset dangerous threshold for the first time, the control center starts a sixty - second observation window. During the sixty - second observation window, when more than eighty percent of the sampling points in this area continuously exceed the preset dangerous threshold and the average concentration value exceeds the preset dangerous threshold by twenty percent, the control center generates a warning message.

[0019] The warning message records the location of the dust - exceeding area, the current dust concentration value, and the duration of exceeding the standard, and sends the warning message to the on - site management personnel terminal; the control center determines the treatment priorities of each area according to the dust concentration change trend calculated from the three - dimensional model of dust concentration distribution, including:

[0020] The warning message records the location of the dust - exceeding area, the current dust concentration value, and the duration of exceeding the standard. The control center encapsulates the warning message into a spatial information block, a concentration information block, a time information block, and a trend information block in JSON format;

[0021] The control center sends the warning message to the on - site management personnel terminal. After receiving the warning message, the on - site management personnel terminal displays the location of the dust - exceeding area in a three - dimensional stereoscopic visualization manner and uses different colors to mark the exceeding - standard levels according to the current dust concentration value;

[0022] The control center calculates the dust concentration change trend based on the three - dimensional model of dust concentration distribution, uses the least - squares method to fit the concentration change curve in the recent four hours to obtain the change rate and acceleration, calculates the dust diffusion direction and speed through flow - field simulation to obtain the spatial diffusion risk, and determines the exposure risk coefficient based on the production plan and personnel scheduling;

[0023] The control center uses a weighted scoring method to determine the treatment priorities of each area. The weight of the dust concentration change trend is set to 0.4, the weight of the spatial diffusion risk is set to 0.3, and the weight of the exposure risk coefficient is set to 0.3; when the acceleration is positive and the change rate is greater than 0.5 mg / m³·min, the score of the dust concentration change trend is 5 points; when the acceleration is positive and the change rate is less than 0.5 mg / m³·min, the score of the dust concentration change trend is 4 points; when the acceleration is zero, the score of the dust concentration change trend is 3 points; when the acceleration is negative, the score of the dust concentration change trend is 2 points.

[0024] The control center issues control instructions to the ventilation systems of the two dust hazard level areas above and below the preset danger threshold respectively according to the governance priorities of each area, and adjusts the opening degree of the ventilation duct, the fan speed, and the air supply direction; at the same time, the control center controls the mobile dust removal device to navigate to the area with the highest dust concentration according to the optimal dust removal path calculated by the three-dimensional dust concentration distribution model, and starts the primary filtration unit, electrostatic dust removal unit, and activated carbon adsorption unit of the mobile dust removal device for hierarchical filtration, including:

[0025] The control center calculates the governance priority of each area based on the area volume, the degree of dust concentration exceeding the standard, and the personnel density, and issues control instructions to the ventilation systems of the two dust hazard level areas above and below the preset danger threshold respectively according to the governance priority;

[0026] The control center adjusts the opening degree of the ventilation duct in the area above the preset danger threshold to between 30% and 100% through the control instruction, and adjusts the fan speed to between 70% and 90% of the rated speed; adjusts the opening degree of the ventilation duct in the area below the preset danger threshold to between 20% and 40%, and adjusts the fan speed to between 40% and 60% of the rated speed;

[0027] The control center uses the SIMPLE algorithm to solve the velocity field and pressure field in the area to establish a three-dimensional flow field distribution model, uses the particle tracking method to simulate the dust movement trajectory to determine the optimal air supply angle, and rotates the air supply direction adjustment device within 360 degrees in the horizontal direction and adjusts it within 0 to 60 degrees in the vertical direction according to the optimal air supply angle through a stepper motor;

[0028] The control center calculates the optimal dust removal path based on the three-dimensional dust concentration distribution model, and controls the mobile dust removal device to navigate to the area with the highest dust concentration according to the optimal dust removal path. The calculation of the optimal dust removal path includes using the A* algorithm for global path planning to obtain the shortest collision-free path, and using the dynamic window method for local path planning to avoid moving obstacles;

[0029] The control center starts the primary filtration unit, electrostatic dust removal unit, and activated carbon adsorption unit of the mobile dust removal device for hierarchical filtration. The primary filtration unit uses a metal mesh filter element to filter particles larger than 100 microns, and the control center adjusts the suction fan speed according to the pressure difference at both ends of the primary filtration unit; the electrostatic dust removal unit uses bipolar discharge, and the control center adjusts the electric field strength within the range of 5 to 15 kV / cm according to the corona current; the activated carbon adsorption unit uses honeycomb activated carbon with a specific surface area of 1200 m² / g to adsorb fine dust.

[0030] The control center calculates the optimal dust removal path based on the three-dimensional dust concentration distribution model, and controls the mobile dust removal device to navigate to the area with the highest dust concentration according to the optimal dust removal path. The calculation of the optimal dust removal path includes using the A* algorithm for global path planning to obtain the shortest collision-free path, and using the dynamic window method for local path planning to avoid moving obstacles, including:

[0031] The control center divides the space of the three-dimensional dust concentration distribution model into cubic grids with a side length of 0.1 meter, and constructs a composite cost map based on the three-dimensional dust concentration distribution model;

[0032] The control center calculates the dust concentration cost of each grid node according to the three-dimensional dust concentration distribution model. The dust concentration cost includes a static concentration term and a dynamic diffusion term; the static concentration term is calculated according to the concentration value of the grid node in the three-dimensional dust concentration distribution model; the dynamic diffusion term is calculated according to the concentration gradient and the time change rate in the three-dimensional dust concentration distribution model;

[0033] The control center calculates the obstacle risk degree for each grid node according to the three-dimensional dust concentration distribution model. The obstacle risk degree is calculated using an exponential decay function based on distance, and is weighted by combining the passage difficulty value in the three-dimensional dust concentration distribution model; the control center combines the dust concentration cost and the obstacle risk degree to form the composite cost map;

[0034] The control center calculates the dust concentration gradient field according to the three-dimensional dust concentration distribution model, and divides the space into a gradient area higher than the preset dust threshold and a gradient area lower than the preset dust threshold; the control center uses a search step size of 0.05 meter in the gradient area higher than the preset dust threshold, and uses a search step size of 0.2 meter in the gradient area lower than the preset dust threshold to perform adaptive gradient search on the three-dimensional dust concentration distribution model;

[0035] The control center constructs a multi-objective optimization model by combining the spatio-temporal distribution characteristics in the three-dimensional dust concentration distribution model; the optimization objectives of the multi-objective optimization model include minimizing the path length and maximizing the cumulative dust removal amount; the maximum cumulative dust removal amount is calculated based on the concentration spatio-temporal distribution in the three-dimensional dust concentration distribution model;

[0036] The control center uses the NSGA-II algorithm to solve the multi-objective optimization model. The crossover operator of the NSGA-II algorithm adaptively adjusts the crossover parameters according to the local characteristics of the three-dimensional dust concentration distribution model; the control center verifies the kinematic constraints of the optimized path plan, and verifies the obstacle avoidance constraints based on the three-dimensional dust concentration distribution model;

[0037] The control center monitors the changes of the three-dimensional model of the dust concentration distribution in real time. When it detects that the change in the dust concentration distribution in a local area exceeds the preset dust threshold, it updates the composite cost map of this area and recalculates the optimal dust removal path.

[0038] The laser scattering particle counter and the gravimetric dust sampler collect the particle dust concentration data during the treatment process in real time and feedback it to the control center. The control center performs online optimization and update on the three-dimensional model of the dust concentration distribution according to the particle dust concentration data, and adjusts the ventilation system parameters and the operation strategy of the mobile dust removal device accordingly, including:

[0039] The laser scattering particle counter and the gravimetric dust sampler collect the particle dust concentration data during the treatment process in real time. The gravimetric dust sampler adopts the isokinetic sampling principle, is equipped with a PM10 cutting head and a PM2.5 classifier, and uses the thermobalance principle to measure the particulate matter weight concentration in real time

[0040] The laser scattering particle counter and the gravimetric dust sampler feedback the particle dust concentration data to the control center in real time. The control center performs data fusion on the particle dust concentration data using the Kalman filter algorithm, and judges the outliers based on the fused particle dust concentration data using a neural network model;

[0041] The control center performs online optimization and update on the three-dimensional model of the dust concentration distribution using the recursive least squares method according to the fused particle dust concentration data, and obtains the optimized and updated three-dimensional model of the dust concentration distribution; performs parameter identification calculation every five minutes to obtain the latest model parameter values, and uses the sliding time window method to select the data of the most recent thirty minutes to calculate the prediction error; when the prediction error exceeds the set threshold, the control center triggers the model structure optimization;

[0042] The control center constructs an equation for the change in dust concentration within the prediction time domain based on the optimized and updated three-dimensional model of the dust concentration distribution, takes the ventilation system energy consumption and the degree of dust exceeding the standard as the optimization objectives to solve the multi-objective optimization problem, and obtains the optimal adjustment sequence of the ventilation system parameters;

[0043] The control center adjusts the opening degree of the ventilation duct, the fan speed, and the air supply direction of the ventilation system according to the optimal adjustment sequence of the ventilation system parameters;

[0044] The control center re-plans the operation path of the mobile dust removal device using the ant colony algorithm based on the optimized and updated three-dimensional model of the dust concentration distribution, and adaptively adjusts the operation parameters of each filter unit of the mobile dust removal device according to the measured dust particle size distribution characteristics.

[0045] In the second aspect of the embodiments of the present invention,

[0046] a dust concentration monitoring and control system for an intelligent furniture production workshop is provided, including:

[0047] A first unit is configured to deploy a laser scattering particle counter and a gravimetric dust sampler in the intelligent furniture production workshop, collect workshop dust concentration data in different areas of the workshop, and transmit it to the control center; based on the workshop dust concentration data, the control center uses a convolutional neural network to train the historical workshop dust concentration data, construct a three-dimensional model of the workshop dust concentration distribution, and dynamically predict the dust concentration in different areas and at different times in the workshop based on the three-dimensional model of the dust concentration distribution;

[0048] A second unit is configured to, based on the prediction result of the three-dimensional model of the dust concentration distribution, the control center divides the workshop into two dust hazard level areas, one higher than the preset danger threshold and the other lower than the preset danger threshold; when the laser scattering particle counter and the gravimetric dust sampler detect that the dust concentration in a certain area exceeds the preset danger threshold, the control center generates a warning message, which records the location of the dust exceeding the standard area, the current dust concentration value, and the duration of exceeding the standard, and sends the warning message to the on-site management personnel terminal; the control center determines the treatment priority of each area according to the dust concentration change trend calculated by the three-dimensional model of the dust concentration distribution;

[0049] A third unit is configured to, according to the treatment priority of each area, the control center issues control instructions to the ventilation systems of the two dust hazard level areas, one higher than the preset danger threshold and the other lower than the preset danger threshold, to adjust the opening of the ventilation duct, the rotation speed of the fan, and the air supply direction; at the same time, according to the optimal dust removal path calculated by the three-dimensional model of the dust concentration distribution, the control center controls the mobile dust removal device to navigate to the area with the highest dust concentration, and starts the primary filtration unit, the electrostatic dust removal unit, and the activated carbon adsorption unit of the mobile dust removal device for hierarchical filtration; the laser scattering particle counter and the gravimetric dust sampler collect the particle dust concentration data during the treatment process in real time and feedback it to the control center, and the control center performs online optimization and update of the three-dimensional model of the dust concentration distribution according to the particle dust concentration data, and adjusts the ventilation system parameters and the operation strategy of the mobile dust removal device accordingly.

[0050] In the third aspect of the embodiments of the present invention,

[0051] an electronic device is provided, including:

[0052] a processor;

[0053] a memory for storing processor-executable instructions;

[0054] Among them, the processor is configured to call the instructions stored in the memory to execute the method described above.

[0055] In the fourth aspect of the embodiments of the present invention,

[0056] A computer-readable storage medium is provided, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the method described above is implemented.

[0057] The beneficial effects of this application are as follows:

[0058] 1. Precise monitoring and prediction of the dust concentration in the intelligent furniture production workshop are realized. By using a laser scattering particle counter and a gravimetric dust sampler to collect data in real time, combined with a three-dimensional model constructed by a convolutional neural network, the dust concentration in different areas and at different times in the workshop can be dynamically predicted, effectively grasping the changing trend of the dust concentration.

[0059] 2. The efficiency and pertinence of dust control are improved. According to the dust concentration prediction results, regional division and risk level assessment are carried out, and the treatment priority is determined in combination with the changing trend of the dust concentration, which can more specifically control the ventilation system and mobile dust removal devices, avoid waste of resources, and improve the treatment efficiency.

[0060] 3. Automation and intelligence of dust control are realized. The system automatically generates warning information, issues control instructions, and adjusts operation strategies without manual intervention, realizing the automation of dust control. At the same time, the system can optimize the three-dimensional model online according to real-time data, making the treatment strategy more intelligent and adapting to the dynamic changes of the dust concentration in the workshop. Description of the Drawings

[0061] Figure 1 It is a schematic flowchart of the method for monitoring and controlling the dust concentration in the intelligent furniture production workshop according to the embodiments of the present invention;

[0062] Figure 2 It is a schematic structural diagram of the system for monitoring and controlling the dust concentration in the intelligent furniture production workshop according to the embodiments of the present invention. Detailed Embodiments

[0063] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0064] The technical solution of the present invention will be described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0065] Figure 1 It is a schematic flow chart of the method for monitoring and controlling the dust concentration in the intelligent furniture production workshop according to the embodiment of the present invention. As Figure 1 shown, the method includes:

[0066] S11. Install laser scattering particle counters and gravimetric dust samplers in the intelligent furniture production workshop, collect the workshop dust concentration data of different areas in the workshop and transmit it to the control center; based on the workshop dust concentration data, the control center uses a convolutional neural network to train the historical workshop dust concentration data, constructs a three-dimensional model of the workshop dust concentration distribution, and dynamically predicts the dust concentration in different areas and at different times in the workshop based on the three-dimensional model of the dust concentration distribution;

[0067] S12. Based on the prediction results of the three-dimensional model of the dust concentration distribution, the control center divides the workshop into two dust hazard level areas, one higher than the preset hazard threshold and the other lower than the preset hazard threshold; when the laser scattering particle counter and the gravimetric dust sampler detect that the dust concentration in a certain area exceeds the preset hazard threshold, the control center generates a warning message, which records the location of the dust exceeding the standard area, the current dust concentration value and the duration of exceeding the standard, and sends the warning message to the on-site management personnel terminal; the control center determines the treatment priority of each area according to the dust concentration change trend calculated by the three-dimensional model of the dust concentration distribution.

[0068] S13. The control center issues control instructions to the ventilation systems of the two dust hazard level areas, one higher than the preset hazard threshold and the other lower than the preset hazard threshold, to adjust the opening of the ventilation duct, the rotation speed of the fan and the air supply direction; at the same time, the control center controls the mobile dust removal device to navigate to the area with the highest dust concentration according to the optimal dust removal path calculated by the three-dimensional model of the dust concentration distribution, and starts the primary filtration unit, electrostatic dust removal unit and activated carbon adsorption unit of the mobile dust removal device for hierarchical filtration; the laser scattering particle counter and the gravimetric dust sampler collect the particle dust concentration data during the treatment process in real time and feedback it to the control center, and the control center performs online optimization and update of the three-dimensional model of the dust concentration distribution according to the particle dust concentration data, and adjusts the ventilation system parameters and the operation strategy of the mobile dust removal device accordingly.

[0069] In an alternative embodiment, a laser scattering particle counter and a gravimetric dust sampler are arranged in the intelligent furniture production workshop to collect the workshop dust concentration data of different areas in the workshop and transmit it to the control center; based on the workshop dust concentration data, the control center uses a convolutional neural network to train the historical workshop dust concentration data, constructs a three-dimensional model of the workshop dust concentration distribution, and dynamically predicts the dust concentration in different areas and at different times in the workshop, including:

[0070] According to the functions of the intelligent furniture production workshop, it is divided into three functional blocks: a processing area, a grinding area, and a painting area, and a three-layer three-dimensional monitoring network of a ground layer, a human breathing layer, and an upper space layer is constructed in each of the functional blocks; a laser scattering particle counter is arranged in the three-layer three-dimensional monitoring network of each of the functional blocks, and the laser scattering particle counter uses a semiconductor laser as a light source to obtain the dust particle count value based on the measurement of the forward scattered light intensity; a gravimetric dust sampler is arranged in the human breathing layer of each of the functional blocks, and the gravimetric dust sampler obtains the inhalable dust mass concentration value through a PM2.5 cutter and a quartz filter membrane;

[0071] The dust particle count value of the laser scattering particle counter and the inhalable dust mass concentration value of the gravimetric dust sampler are transmitted to the data acquisition module through the RS-485 bus. The data acquisition module reads and encapsulates the dust particle count value and the inhalable dust mass concentration value using the Modbus RTU communication protocol, and uploads the encapsulated data to the control center through the industrial Ethernet. The control center calibrates the dust particle count value based on the inhalable dust mass concentration value to generate standardized workshop dust concentration data;

[0072] The control center organizes the standardized workshop dust concentration data according to the temporal characteristics and spatial characteristics, constructs a training data set containing historical data of the past three months, and each record of the training data set contains a timestamp, a spatial coordinate, and a dust concentration value; based on the training data set, a three-dimensional model of the workshop dust concentration distribution is constructed using a three-dimensional convolutional neural network, and spatial features and temporal features are extracted through four convolutional blocks and an LSTM layer; the control center uses the three-dimensional model of the workshop dust concentration distribution to dynamically predict the dust concentration in different areas of the workshop in the next four hours.

[0073] The intelligent furniture production workshop dust concentration monitoring and prediction system is used to monitor the dust concentration in the workshop in real time, predict the future change trend of the dust concentration, ensure the health of workers, and improve production efficiency.

[0074] First, according to the functional division of the smart furniture production workshop, it is divided into three functional blocks: the processing area, the grinding area, and the painting area. Within each functional block, a three-layer three-dimensional monitoring network is constructed, namely the ground layer, the human breathing layer (about 1.5 meters above the ground), and the upper space layer (about 3 meters above the ground).

[0075] In the three-layer three-dimensional monitoring network of each functional block, multiple laser scattering particle counters are evenly arranged on each layer. The laser scattering particle counter uses a semiconductor laser as the light source and obtains the dust particle count value by measuring the forward scattered light intensity. For example, 4, 6, and 3 laser scattering particle counters are arranged on the ground layer, the human breathing layer, and the upper space layer of the processing area respectively, and similar arrangements are made for the grinding area and the painting area.

[0076] At the same time, gravimetric dust samplers are arranged on the human breathing layer of each functional block. The gravimetric dust sampler obtains the inhalable dust mass concentration value through a PM2.5 cutter and a quartz filter membrane. For example, 3, 4, and 2 gravimetric dust samplers are arranged on the human breathing layer of the processing area, the grinding area, and the painting area respectively.

[0077] All laser scattering particle counters and gravimetric dust samplers are connected to the data acquisition module through the RS-485 bus. The data acquisition module reads and encapsulates the dust particle count value and the inhalable dust mass concentration value using the Modbus RTU communication protocol. For example, the data acquisition module reads the data of all sensors once a minute and packages the data into JSON format.

[0078] The data acquisition module uploads the encapsulated data to the control center through industrial Ethernet. The control center calibrates the dust particle count value of the laser scattering particle counter based on the inhalable dust mass concentration value obtained by the gravimetric dust sampler. For example, the control center establishes a linear regression model, uses the inhalable dust mass concentration value as the input and the dust particle count value as the output, calibrates the data of the laser scattering particle counter, and generates standardized workshop dust concentration data.

[0079] The control center organizes the standardized workshop dust concentration data according to the temporal and spatial characteristics, and constructs a training data set containing historical data for the past three months. Each record in the training data set contains a timestamp, spatial coordinates (functional block, height layer), and dust concentration value. For example, a record can be expressed as: {"timestamp":"2024-07-27 10:00:00","spatial coordinates":"processing area - human breathing layer","dust concentration value":0.5mg / m 3}.

[0080] The control center constructs a three-dimensional model of the workshop dust concentration distribution using a three-dimensional convolutional neural network based on a training dataset. This model extracts spatial features and temporal features through four convolutional blocks and an LSTM layer. Each convolutional block contains a three-dimensional convolutional layer, an activation function, and a pooling layer. The LSTM layer is used to process time series data. The input of the model is historical dust concentration data, and the output is the predicted value of future dust concentration.

[0081] The control center uses the constructed three-dimensional model of the workshop dust concentration distribution to dynamically predict the dust concentration in different areas of the workshop for the next four hours. For example, the control center can update the prediction results every hour and display them on the monitoring interface.

[0082] The solution of this application can:

[0083] Real-time monitoring and accurate prediction: The system can monitor the dust concentration in different areas and at different heights of the workshop in real time, and accurately predict the future dust concentration based on historical data and the three-dimensional model, providing reliable data support for workshop environmental control. Health protection and safe production: By monitoring and predicting the dust concentration in real time, it is possible to detect situations where the dust exceeds the standard in a timely manner and take corresponding control measures to protect the respiratory health of workers, prevent the occurrence of occupational diseases, and ensure safe production. Intelligent management and efficient control: The system realizes the automatic monitoring and intelligent prediction of the workshop dust concentration, reduces the workload of manual monitoring, improves the efficiency of workshop environmental management, and achieves precise control of dust pollution.

[0084] In an optional implementation manner, based on the prediction results of the three-dimensional model of the dust concentration distribution, the control center divides the workshop into two dust hazard level areas, namely, above the preset hazard threshold and below the preset hazard threshold; when the laser scattering particle counter and the gravimetric dust sampler detect that the dust concentration in a certain area exceeds the preset hazard threshold, the control center generates a warning message including:

[0085] The control center performs dust concentration prediction on the workshop based on the three-dimensional model of the dust concentration distribution, obtains prediction result data, and divides the workshop space into two dust hazard level areas, namely, above the preset hazard threshold area and below the preset hazard threshold area, according to the prediction result data;

[0086] The control center performs binary processing on the prediction result data using an adaptive threshold segmentation algorithm, marks the area above the preset hazard threshold as a first value, marks the area below the preset hazard threshold as a second value, and performs smoothing processing on the marked data using a three-dimensional median filter with a 3×3×3 dimension to obtain a marked result;

[0087] The control center uses the region growing algorithm to perform connected component analysis on the marking result, detects the twenty-six adjacent points of each marked point, classifies the adjacent points with the same mark into the same connected component, and obtains the workshop dust hazard zoning result;

[0088] When the laser scattering particle counter and the gravimetric dust sampler detect that the dust concentration in a certain area exceeds the preset hazard threshold for the first time, the control center starts a sixty-second observation window. During the sixty-second observation window, when more than 80% of the sampling points in this area continuously exceed the preset hazard threshold and the average concentration value exceeds the preset hazard threshold by 20%, the control center generates a warning message.

[0089] The method for predicting dust concentration and warning is as follows in specific implementation steps:

[0090] First, establish a three-dimensional model of the workshop dust concentration distribution. This model can be constructed in various ways. For example, historical dust concentration data can be used, combined with factors such as workshop layout and ventilation conditions, and trained using machine learning algorithms; or based on computational fluid dynamics (CFD) simulation technology, the air flow and dust diffusion laws in the workshop can be simulated to establish a dust concentration distribution model. Assuming the workshop size is 10m x 8m x 5m, the workshop space is discretized into a 100 x 80 x 50 grid, and each grid represents a sampling point. The model can predict the dust concentration value of each grid point.

[0091] Next, perform dust concentration prediction and divide the hazard level areas. Use the established three-dimensional model to predict the dust concentration in each area of the workshop in the future. Assuming the preset hazard threshold is 100mg / m 3 , the prediction result shows that the dust concentration of the grid at the coordinate (3, 5, 2) is 120mg / m 3 , and the dust concentration of the grid at the coordinate (7, 2, 1) is 80mg / m 3 . Then the area where (3, 5, 2) is located is marked as an area above the preset hazard threshold, and the area where (7, 2, 1) is located is marked as an area below the preset hazard threshold. And so on, the entire workshop is divided into two dust hazard level areas.

[0092] Then, the prediction results are binarized and smoothed. The adaptive threshold segmentation algorithm is used to binarize the prediction result data. The area above the preset danger threshold is marked as the value 1, and the area below the preset danger threshold is marked as the value 0. For example, the grid at (3, 5, 2) is marked as 1, and the grid at (7, 2, 1) is marked as 0. Then, a 3D median filter with a 3x3x3 dimension is used to smooth the marked data to eliminate the influence of noise and isolated points. For example, if among the 26 adjacent points around (3, 5, 2), 15 points have a value of 1 and 9 points have a value of 0, then the value of (3, 5, 2) after smoothing is still 1.

[0093] After that, connected component analysis is performed to determine the dust hazard zoning. The region growing algorithm is used to perform connected component analysis on the smoothed data. Starting from a marked point, its twenty-six adjacent points are detected, and the adjacent points with the same mark are grouped into the same connected component. For example, if the value of (3, 5, 2) is 1, and the values of its adjacent points (3, 5, 3) and (3, 6, 2) are also 1, then these three points are grouped into the same connected component. Finally, the workshop dust hazard zoning result is obtained. For example, three connected components above the preset danger threshold can be identified, which are located at different positions in the workshop.

[0094] Finally, the dust concentration is monitored in real time and warning information is generated. The laser scattering particle counter and the gravimetric dust sampler monitor the dust concentration in each area of the workshop in real time. Assume that the preset danger threshold is 100 mg / m 3 , when the dust concentration in a certain area exceeds 100 mg / m for the first time 3 , a 60-second observation window is started. Within 60 seconds, if more than 80% of the sampling points in this area continuously exceed 100 mg / m 3 , and the average concentration value exceeds 120 mg / m 3 (that is, 20% higher than the preset danger threshold), the control center generates warning information, such as "The dust concentration in Area A exceeds the standard, please handle it in time".

[0095] The solution of this application can:

[0096] Improve the accuracy of early warning: By combining technologies such as 3D model prediction of dust concentration distribution, adaptive threshold segmentation, 3D median filtering, and connected component analysis, the dust hazard area can be identified more accurately, avoiding false alarms and missed alarms, and improving the accuracy of early warning. Realize the timeliness of early warning: By adopting real-time monitoring and the 60-second observation window mechanism, the situation of dust concentration exceeding the standard can be detected in time, and warning information can be sent out in time so that the staff can take measures quickly to avoid danger. Enhance the reliability of early warning: By combining two detection methods, the laser scattering particle counter and the gravimetric dust sampler, they can confirm each other, improve the reliability of the monitoring data, and thus enhance the reliability of early warning.

[0097] In an alternative embodiment, the warning information records the location of the dust-exceeding area, the current dust concentration value, and the duration of exceeding the standard, and sends the warning information to the terminal of the on-site management personnel; according to the dust concentration change trend calculated by the three-dimensional model of the dust concentration distribution, the control center determines the treatment priority levels of each area, including:

[0098] The warning information records the location of the dust-exceeding area, the current dust concentration value, and the duration of exceeding the standard. The control center encapsulates the warning information into a spatial information block, a concentration information block, a time information block, and a trend information block in JSON format;

[0099] The control center sends the warning information to the terminal of the on-site management personnel. After receiving the warning information, the terminal of the on-site management personnel displays the location of the dust-exceeding area in a three-dimensional stereoscopic visualization manner, and uses different colors to mark the exceeding standard levels according to the current dust concentration value;

[0100] The control center calculates the dust concentration change trend based on the three-dimensional model of the dust concentration distribution, uses the least squares method to fit the concentration change curve of the most recent four hours to obtain the change rate and acceleration, calculates the dust diffusion direction and speed through flow field simulation to obtain the spatial diffusion risk, and determines the exposure risk coefficient based on the production plan and personnel scheduling;

[0101] The control center uses a weighted scoring method to determine the treatment priority levels of each area, sets the weight of the dust concentration change trend to 0.4, the weight of the spatial diffusion risk to 0.3, and the weight of the exposure risk coefficient to 0.3; when the acceleration is positive and the change rate is greater than 0.5 mg / m³ per minute, the score of the dust concentration change trend is 5 points; when the acceleration is positive and the change rate is less than 0.5 mg / m³ per minute, the score of the dust concentration change trend is 4 points; when the acceleration is zero, the score of the dust concentration change trend is 3 points; when the acceleration is negative, the score of the dust concentration change trend is 2 points.

[0102] Intelligent Warning and Governance System and Method for Dust Exceeding the Standard

[0103] This system aims to monitor, warn, and manage the problem of dust exceeding the standard in the industrial environment in real time, and ensure the health of personnel and production safety. The system includes a dust sensor network, a control center, and a terminal of on-site management personnel.

[0104] First, deploy a dense dust sensor network to collect dust concentration data of each area in real time. The sensor data is transmitted to the control center through a wireless network.

[0105] After the control center receives the sensor data, it determines whether there is an excessive dust concentration. If the dust concentration in a certain area exceeds the preset threshold, it records the location of the area, the current dust concentration value, and the duration of the excess, and encapsulates this information into a warning message.

[0106] The warning message is in JSON format and is divided into a spatial information block, a concentration information block, a time information block, and a trend information block. The spatial information block records the specific location of the area with excessive levels, such as longitude and latitude or specific coordinates within the factory building. The concentration information block records the current dust concentration value, such as 1.2mg / m 3 . The time information block records the start time and duration of the excess, such as starting at "2024-10-27 10:00:00" and lasting for 15 minutes. The trend information block records the change trend of the dust concentration, and this part will be described in detail in the subsequent steps.

[0107] The control center sends the encapsulated JSON-format warning message to the on-site management personnel terminal.

[0108] After the on-site management personnel terminal receives the warning message, it displays the location of the area with excessive dust in a three-dimensional visualization manner. For example, on the three-dimensional model of the factory building, the area with excessive levels is marked in red. At the same time, different colors are used to identify the excess level according to the current dust concentration value. For example, areas with slightly excessive concentrations are marked in orange, and areas with severely excessive concentrations are marked in red.

[0109] The control center calculates the change trend of the dust concentration based on the three-dimensional model of the dust concentration distribution. The specific method is to fit the concentration change curve of the last four hours to obtain the change rate and acceleration. For example, the dust concentrations in the past four hours were 0.8, 1.0, 1.2, and 1.5mg / m 3 , and by analyzing these data, the concentration change rate is obtained as 0.175mg / m 3 / hour, and the acceleration is 0.04mg / m 3 / hour².

[0110] At the same time, the control center calculates the dust diffusion direction and speed through flow field simulation to evaluate the spatial diffusion risk. For example, the simulation results show that the dust is spreading southeast at a speed of 2m / s. In addition, the control center will also determine the exposure risk coefficient based on the production plan and personnel scheduling. For example, if there are a large number of personnel working near the area with excessive levels, the exposure risk coefficient is relatively high.

[0111] The control center uses a weighted scoring method to determine the governance priority of each area. The weight of the dust concentration change trend is set to 0.4, the weight of the spatial diffusion risk is set to 0.3, and the weight of the exposure risk coefficient is set to 0.3.

[0112] The scoring rules for the dust concentration change trend are as follows:

[0113] If the acceleration is positive and the rate of change is greater than 0.5mg / m 3 / minute, the score is 5 points; if the acceleration is positive and the rate of change is less than 0.5mg / m 3 / minute, the score is 4 points; if the acceleration is zero, the score is 3 points; if the acceleration is negative, the score is 2 points. For example, if the acceleration in a certain area is 0.04mg / m 3 / h² and the rate of change is 0.175mg / m 3 / h (less than 0.5mg / m 3 / minute), the score for the changing trend of the dust concentration in this area is 4 points. Assuming that the spatial diffusion risk score is 3 points and the exposure risk coefficient score is 2 points, the governance priority score for this area is 4*0.4 + 3*0.3 + 2*0.3 = 3.1 points.

[0114] The solution of the present application can:

[0115] Improve the timeliness and accuracy of dust over - standard early warning, and can issue an early warning before the dust concentration reaches a dangerous level to avoid accidents. Through three - dimensional visualization and hierarchical identification, on - site management personnel can quickly and intuitively understand the dust over - standard situation, improving the emergency response efficiency. By comprehensively considering the changing trend of dust concentration, spatial diffusion risk and exposure risk coefficient, the optimal allocation of governance resources is realized, improving the governance efficiency.

[0116] In an optional implementation manner, the control center issues control instructions to the ventilation systems of the two dust - hazard level areas above and below the preset danger threshold respectively according to the governance priority of each area, adjusting the opening degree of the ventilation duct, the rotation speed of the fan and the air supply direction; at the same time, the control center controls the mobile dust removal device to navigate to the area with the highest dust concentration according to the optimal dust removal path calculated from the three - dimensional model of dust concentration distribution, and starts the primary filtration unit, electrostatic dust removal unit and activated carbon adsorption unit of the mobile dust removal device for hierarchical filtration, including:

[0117] The control center calculates the governance priority of each area based on the regional volume, the degree of dust concentration exceeding the standard and the personnel density, and issues control instructions to the ventilation systems of the two dust - hazard level areas above and below the preset danger threshold respectively according to the governance priority;

[0118] The control center adjusts the opening degree of the ventilation ducts in the area above the preset danger threshold to between 30% and 100%, and adjusts the fan speed to between 70% and 90% of the rated speed; adjusts the opening degree of the ventilation ducts in the area below the preset danger threshold to between 20% and 40%, and adjusts the fan speed to between 40% and 60% of the rated speed;

[0119] The control center uses the SIMPLE algorithm to solve the velocity field and pressure field in the area to establish a three-dimensional flow field distribution model, uses the particle tracking method to simulate the dust movement trajectory to determine the optimal air supply angle, and drives the air supply direction adjustment device to rotate within 360 degrees in the horizontal direction and adjust within 0 to 60 degrees in the vertical direction according to the optimal air supply angle through a stepper motor;

[0120] The control center calculates the optimal dust removal path based on the three-dimensional model of the dust concentration distribution, and controls the mobile dust removal device to navigate to the area with the highest dust concentration according to the optimal dust removal path. The calculation of the optimal dust removal path includes using the A* algorithm for global path planning to obtain the shortest collision-free path, and using the dynamic window method for local path planning to avoid moving obstacles;

[0121] The control center starts the primary filtration unit, electrostatic dust removal unit and activated carbon adsorption unit of the mobile dust removal device for hierarchical filtration. The primary filtration unit uses a metal mesh filter element to filter particles larger than 100 microns, and the control center adjusts the suction fan speed according to the pressure difference at both ends of the primary filtration unit; the electrostatic dust removal unit uses bipolar discharge, and the control center adjusts the electric field strength within the range of 5 to 15 kV / cm according to the corona current; the activated carbon adsorption unit uses honeycomb activated carbon with a specific surface area of 1200 m² / g to adsorb fine dust.

[0122] An intelligent dust control method aims to efficiently and safely control the dust concentration to protect personnel health and production safety. The core of this method is to synergistically control the ventilation system and the mobile dust removal device according to the regional dust hazard level and treatment priority.

[0123] First, the system collects data such as the dust concentration, regional volume, and personnel density of each area in real time. Taking a 100-cubic-meter area as an example, assuming that the measured dust concentration is 0.8 mg / m 3 , and the personnel density is 5 people / 100 m 2 .

[0124] Then, the system calculates the treatment priority of each area according to a preset formula. This formula comprehensively considers factors such as the volume of the area, the degree of dust concentration exceeding the standard, and the personnel density. For example, a simple priority calculation method can be set: Priority = multiple of dust concentration exceeding the standard * personnel density * volume of the area. Assume the preset danger threshold is 0.5mg / m 3 , then the multiple of the dust concentration exceeding the standard in this area is 0.8 / 0.5 = 1.6 times. Then the treatment priority of this area is 1.6 * 5 * 100 = 800.

[0125] Next, the system divides the areas into two levels: above the preset danger threshold and below the preset danger threshold. Assume the preset danger threshold is 0.5mg / m 3 , then the areas with dust concentration higher than 0.5mg / m 3 are classified as high-risk areas, and the areas below 0.5mg / m 3 are classified as low-risk areas.

[0126] For high-risk areas, the system issues control instructions to adjust the opening degree of the ventilation duct to between 30% and 100%, and adjust the fan speed to between 70% and 90% of the rated speed. For example, the opening degree of the ventilation duct can be set to 80%, and the fan speed can be set to 75% of the rated speed. For low-risk areas, the system issues control instructions to adjust the opening degree of the ventilation duct to between 20% and 40%, and adjust the fan speed to between 40% and 60% of the rated speed. For example, the opening degree of the ventilation duct can be set to 30%, and the fan speed can be set to 50% of the rated speed.

[0127] At the same time, the system uses the computational fluid dynamics method to simulate the flow field distribution in the area, and uses the particle tracking method to simulate the dust movement trajectory to determine the optimal air supply angle. For example, the simulation results show that the optimal air supply angle is 30 degrees in the horizontal direction and 15 degrees in the vertical direction. Then, the system drives the air supply direction adjustment device through a stepper motor to adjust the air supply direction to this angle.

[0128] In addition, the system calculates the optimal dust removal path based on the three-dimensional model of the real-time updated dust concentration distribution. This path planning process combines the global path planning algorithm and the local path planning algorithm to ensure that the mobile dust removal device can quickly and safely reach the area with the highest dust concentration. For example, first use the A* algorithm to plan the shortest collision-free path from the current position to the target position, and then use the dynamic window method to locally adjust the path to avoid moving obstacles.

[0129] Finally, when the mobile dust removal device reaches the target area, the system starts its primary filtration unit, electrostatic precipitation unit, and activated carbon adsorption unit for hierarchical filtration. For example, the primary filtration unit uses a metal mesh filter element to filter particulate matter larger than 100 microns, and the system adjusts the speed of the suction fan according to the pressure difference at both ends of the primary filtration unit. The electrostatic precipitation unit uses bipolar discharge, and the system adjusts the electric field strength according to the corona current within the range of 5 to 15 kV / cm. The activated carbon adsorption unit uses honeycomb activated carbon with a specific surface area of 1200 m² / g to adsorb fine dust.

[0130] The solution of this application can:

[0131] Improve the dust control efficiency. By coordinately controlling the ventilation system and the mobile dust removal device according to the regional dust hazard level and treatment priority, the dust concentration can be reduced more quickly and effectively. Save energy consumption. By dynamically adjusting the parameters of the ventilation system according to the dust concentration, unnecessary energy waste is avoided. Enhance safety. By adopting an advanced path planning algorithm, the safe operation of the mobile dust removal device is ensured, and the occurrence of collision accidents is avoided.

[0132] In an optional implementation manner, the control center calculates the optimal dust removal path based on the three-dimensional dust concentration distribution model, and controls the mobile dust removal device to navigate to the area with the highest dust concentration according to the optimal dust removal path. The calculation of the optimal dust removal path includes using the A* algorithm for global path planning to obtain the shortest collision-free path, and using the dynamic window method for local path planning to avoid moving obstacles, including:

[0133] The control center divides the space of the three-dimensional dust concentration distribution model into cubic grids with a side length of 0.1 meter, and constructs a composite cost map based on the three-dimensional dust concentration distribution model;

[0134] The control center calculates the dust concentration cost of each grid node according to the three-dimensional dust concentration distribution model. The dust concentration cost includes a static concentration term and a dynamic diffusion term; the static concentration term is calculated according to the concentration value of the grid node in the three-dimensional dust concentration distribution model; the dynamic diffusion term is calculated according to the concentration gradient and the time change rate in the three-dimensional dust concentration distribution model;

[0135] The control center calculates the obstacle hazard degree for each grid node according to the three-dimensional dust concentration distribution model. The obstacle hazard degree is calculated using an exponential decay function based on distance and weighted by the passage difficulty value in the three-dimensional dust concentration distribution model; the control center combines the dust concentration cost and the obstacle hazard degree to form the composite cost map;

[0136] The control center calculates the dust concentration gradient field based on the three-dimensional dust concentration distribution model, and divides the space into a gradient area with a dust concentration higher than the preset dust threshold and a gradient area with a dust concentration lower than the preset dust threshold; the control center uses a search step of 0.05 meters in the gradient area with a dust concentration higher than the preset dust threshold, and uses a search step of 0.2 meters in the gradient area with a dust concentration lower than the preset dust threshold to perform adaptive gradient search on the three-dimensional dust concentration distribution model;

[0137] The control center constructs a multi-objective optimization model by combining the spatio-temporal distribution characteristics in the three-dimensional dust concentration distribution model; the optimization objectives of the multi-objective optimization model include minimizing the path length and maximizing the cumulative dust removal amount; the maximum cumulative dust removal amount is calculated based on the concentration spatio-temporal distribution in the three-dimensional dust concentration distribution model;

[0138] The control center uses the NSGA-II algorithm to solve the multi-objective optimization model, and the crossover operator of the NSGA-II algorithm adaptively adjusts the crossover parameters according to the local characteristics of the three-dimensional dust concentration distribution model; the control center verifies the kinematic constraints of the optimized path plan and verifies the obstacle avoidance constraints based on the three-dimensional dust concentration distribution model;

[0139] The control center monitors the changes in the three-dimensional dust concentration distribution model in real time. When it detects that the change in the dust concentration distribution in a local area exceeds the preset dust threshold, it updates the composite cost map of this area and recalculates the optimal dust removal path.

[0140] The control center first obtains the dust concentration data in the three-dimensional space collected by the sensor network in real time and constructs a three-dimensional dust concentration distribution model. This model can be represented by voxels. For example, the space is divided into cubic grids with a side length of 0.1 meters, and each grid records the average dust concentration value in this area. To simulate the dynamic changes of dust, the model also records the concentration change rate of each grid over a period of time.

[0141] Next, the control center constructs a composite cost map based on the three-dimensional model of dust concentration distribution. The dust concentration cost of each grid node is calculated, which consists of two parts: a static concentration term and a dynamic diffusion term. The static concentration term is directly taken from the concentration value of the grid in the three-dimensional model of dust concentration distribution. The dynamic diffusion term is calculated based on the concentration difference between adjacent grids and the concentration change rate of the grid over a period of time. For example, the greater the concentration difference and the faster the change rate, the higher the value of the dynamic diffusion term. At the same time, the control center also calculates the obstacle risk of each grid node. The obstacle information is sourced from the environmental map, and the risk is calculated using an exponential decay function based on distance. The closer to the obstacle, the higher the risk. In addition, according to the passage difficulty value in the three-dimensional model of dust concentration distribution (for example, some areas may be difficult to pass due to equipment limitations), the obstacle risk is weighted. Finally, the dust concentration cost and the weighted obstacle risk are linearly combined to form a composite cost map. For example, the dust concentration cost accounts for 70% and the obstacle risk accounts for 30%.

[0142] In the path planning stage, the control center first calculates the dust concentration gradient field based on the three-dimensional model of dust concentration distribution. The space is divided into gradient regions above a preset dust threshold (for example, the concentration exceeds 10 milligrams per cubic meter) and gradient regions below the preset dust threshold. In the high-concentration gradient region, a finer search step size (for example, 0.05 meters) is adopted, while in the low-concentration gradient region, a larger search step size (for example, 0.2 meters) is adopted to improve the search efficiency. The control center uses the A* algorithm for global path planning, using the composite cost map as the heuristic function to search for the shortest collision-free path from the current position of the mobile dust removal device to the target area (for example, the area with the highest dust concentration). At the same time, the dynamic window method is used for local path planning. According to the kinematic constraints of the mobile dust removal device (such as the maximum speed and maximum acceleration) and the information on moving obstacles detected by the sensors in real time, the global path is locally adjusted to ensure that the dust removal device can safely avoid obstacles.

[0143] To achieve the multi-objective optimization of minimizing the path length and maximizing the cumulative dust removal amount, the control center constructs a multi-objective optimization model based on the three-dimensional model of dust concentration distribution. The cumulative dust removal amount is calculated based on the dust concentration of each grid on the path and the time it stays on the path. The control center uses the NSGA-II algorithm to solve the multi-objective optimization model. To adapt to the local characteristics of the dust concentration distribution, the crossover operator of the NSGA-II algorithm adaptively adjusts the crossover parameters according to the dust concentration gradient. For example, in the high-concentration gradient region, a smaller crossover probability is adopted to maintain excellent gene segments; in the low-concentration gradient region, a larger crossover probability is adopted to increase the population diversity. Finally, the control center verifies the kinematic constraints and obstacle avoidance constraints of the optimized path plan to ensure the feasibility of the path.

[0144] The control center will monitor the changes in the three-dimensional model of the dust concentration distribution in real time. When it detects that the change in the dust concentration distribution in a local area exceeds the preset dust threshold (for example, the concentration change exceeds 5 milligrams per cubic meter), it updates the composite cost map of that area and recalculates the optimal dust removal path to adapt to the dynamic changes in the environment. For example, assume that a sudden dust leak occurs in a certain area, causing the dust concentration in that area to rise sharply. The control center will immediately update the composite cost map of that area and increase the dust concentration cost and risk level of that area, thereby guiding the mobile dust removal device to give priority to dust removal operations in that area.

[0145] The solution of this application can:

[0146] Improve the dust removal efficiency: This method can plan the optimal dust removal path in real time according to the three-dimensional model of the dust concentration distribution, guide the mobile dust removal device to give priority to the area with the highest dust concentration, thereby maximizing the dust removal efficiency. Reduce energy consumption: By optimizing the path length, reduce the running time and energy consumption of the mobile dust removal device. Enhance environmental adaptability: This method can monitor the changes in the dust concentration distribution in real time and dynamically adjust the dust removal path, thereby enhancing the adaptability to complex and changeable environments.

[0147] In an alternative embodiment, the laser scattering particle counter and the gravimetric dust sampler collect the particle dust concentration data during the treatment process in real time and feedback it to the control center. The control center online optimizes and updates the three-dimensional model of the dust concentration distribution according to the particle dust concentration data, and adjusts the ventilation system parameters and the operation strategy of the mobile dust removal device accordingly, including:

[0148] The laser scattering particle counter and the gravimetric dust sampler collect the particle dust concentration data during the treatment process in real time. The gravimetric dust sampler adopts the isokinetic sampling principle, is equipped with a PM10 cutting head and a PM2.5 classifier, and the gravimetric dust sampler measures the particulate matter weight concentration in real time using the thermogravimetric principle

[0149] The laser scattering particle counter and the gravimetric dust sampler feedback the particle dust concentration data to the control center in real time. The control center uses the Kalman filter algorithm to fuse the particle dust concentration data, and judges the outliers based on the fused particle dust concentration data using a neural network model;

[0150] The control center uses the recursive least squares method to perform online optimization and update on the three-dimensional dust concentration distribution model based on the fused particle dust concentration data, and obtains an optimized and updated three-dimensional dust concentration distribution model; parameter identification calculation is performed every five minutes to obtain the latest model parameter values, and the sliding time window method is used to select the data of the most recent thirty minutes to calculate the prediction error; when the prediction error exceeds the set threshold, the control center triggers model structure optimization;

[0151] The control center constructs an equation for the change in dust concentration within the prediction time domain based on the optimized and updated three-dimensional dust concentration distribution model, takes the energy consumption of the ventilation system and the degree of dust exceeding the standard as the optimization objectives to solve the multi-objective optimization problem, and obtains the optimal adjustment sequence of the ventilation system parameters;

[0152] The control center adjusts the opening degree of the ventilation duct, the fan speed, and the air supply direction of the ventilation system according to the optimal adjustment sequence of the ventilation system parameters;

[0153] The control center re-plans the operation path of the mobile dust removal device based on the optimized and updated three-dimensional dust concentration distribution model using the ant colony algorithm, and adaptively adjusts the operation parameters of each filtration unit of the mobile dust removal device according to the measured dust particle size distribution characteristics.

[0154] An intelligent dust control method based on a three-dimensional dust concentration distribution model, which uses a laser scattering particle counter and a gravimetric dust sampler to collect dust concentration data in real time, and the control center processes and analyzes the data, and finally realizes the optimized control of the ventilation system and the mobile dust removal device, so as to achieve the purpose of efficient dust control.

[0155] First, deploy a laser scattering particle counter and a gravimetric dust sampler. The gravimetric dust sampler is equipped with a PM10 cutting head and a PM2.5 classifier, and based on the principles of isokinetic sampling and thermogravimetric balance, it measures the particulate matter weight concentration in real time. For example, the PM10 concentration is 50 μg / m 3 , and the PM2.5 concentration is 25 μg / m 3 . The laser scattering particle counter can measure the particle number concentration in different particle size ranges. For example, the particle number concentration of particles with a particle size of 0.5 μm is 10000 particles / L, and the particle number concentration of particles with a particle size of 2.5 μm is 5000 particles / L.

[0156] Next, the collected particle dust concentration data is transmitted to the control center in real time. The control center uses the Kalman filtering algorithm to fuse the concentration data from the laser scattering particle counter and the gravimetric dust sampler. For example, the weight concentration of PM2.5 and the particle number concentration of particles with a diameter of 2.5 μm are fused to obtain a more accurate PM2.5 concentration value. At the same time, the control center uses an outlier judgment model based on neural network to analyze the fused data, identify and eliminate abnormal data. For example, if the PM10 concentration suddenly rises to 1000 μg / m 3 , the system will judge it as an outlier and eliminate it.

[0157] Then, using the fused data, the control center uses the recursive least squares method to online optimize and update the three-dimensional model of the dust concentration distribution. For example, every five minutes, the system uses the concentration data of the past thirty minutes to recalculate the model parameters and update the three-dimensional model to make it more accurately reflect the current dust concentration distribution. The system continuously calculates the error between the model prediction value and the actual measurement value. When the prediction error exceeds the pre-set threshold, for example, the error exceeds 15%, the system will automatically trigger the model structure optimization, adjust the model structure, and improve the prediction accuracy.

[0158] Based on the updated three-dimensional model of the dust concentration distribution, the control center predicts the change trend of the dust concentration in the future for a period of time. Taking the energy consumption of the ventilation system and the degree of dust exceeding the standard as the optimization objectives, a multi-objective optimization problem is constructed and solved to obtain the optimal adjustment sequence of the ventilation system parameters. For example, within the next hour, the fan speeds are set to 800 rpm, 850 rpm, 900 rpm…, and the air supply directions are set to 30°, 35°, 40°… respectively. The control center adjusts the opening of the ventilation duct, the fan speed, and the air supply direction of the ventilation system in real time according to the calculated optimal adjustment sequence.

[0159] At the same time, based on the updated three-dimensional model of the dust concentration distribution, the control center uses the ant colony algorithm to re-plan the operation path of the mobile dust removal device so that it can remove dust more effectively. And according to the characteristics of the real-time measured dust particle size distribution, for example, if the proportion of PM10 is high, the power of the corresponding filter unit is increased, and the operation parameters of each stage of the filter unit of the mobile dust removal device are adaptively adjusted to improve the dust removal efficiency.

[0160] The solution of this application can:

[0161] Improve the efficiency of dust control: Through real-time monitoring, data fusion, model optimization, and intelligent control, this method can more accurately grasp the dust concentration distribution, and optimize the ventilation and dust removal strategies accordingly, thus significantly improving the efficiency of dust control. Reduce energy consumption: Through a multi-objective optimization algorithm, this method optimizes the parameters of the ventilation system on the premise of ensuring that the dust concentration meets the standard, effectively reducing the energy consumption of the ventilation system. Improve the automation level: This method realizes the automation and intelligence of the dust control process, reduces manual intervention, improves work efficiency, and reduces labor intensity.

[0162] Figure 2 FIG. is a schematic structural diagram of a dust concentration monitoring and control system for an intelligent furniture production workshop according to an embodiment of the present invention, as Figure 2 shown, the system includes:

[0163] The first unit is used to deploy laser scattering particle counters and gravimetric dust samplers in the intelligent furniture production workshop, collect workshop dust concentration data in different areas of the workshop and transmit it to the control center; based on the workshop dust concentration data, the control center uses a convolutional neural network to train the historical workshop dust concentration data, constructs a three-dimensional model of the workshop dust concentration distribution, and dynamically predicts the dust concentration in different areas and at different times of the workshop based on the three-dimensional model of the dust concentration distribution;

[0164] The second unit is used based on the prediction result of the three-dimensional model of the dust concentration distribution. The control center divides the workshop into two dust hazard level areas above and below the preset hazard threshold; when the laser scattering particle counter and the gravimetric dust sampler detect that the dust concentration in a certain area exceeds the preset hazard threshold, the control center generates a warning message, which records the location of the dust exceeding the standard area, the current dust concentration value and the duration of exceeding the standard, and sends the warning message to the on-site management personnel terminal; the control center determines the treatment priority of each area according to the dust concentration change trend calculated by the three-dimensional model of the dust concentration distribution;

[0165] The third unit is used for the control center to issue control instructions to the ventilation systems of the two dust hazard level areas above and below the preset danger threshold respectively according to the governance priorities of each area, and adjust the opening degree of the ventilation duct, the fan speed and the air supply direction; meanwhile, the control center controls the mobile dust removal device to navigate to the area with the highest dust concentration according to the optimal dust removal path calculated by the three-dimensional dust concentration distribution model, and starts the primary filtration unit, the electrostatic dust removal unit and the activated carbon adsorption unit of the mobile dust removal device for hierarchical filtration; the laser scattering particle counter and the gravimetric dust sampler collect the particle dust concentration data during the governance process in real time and feedback it to the control center, and the control center performs online optimization and update on the three-dimensional dust concentration distribution model according to the particle dust concentration data, and adjusts the ventilation system parameters and the operation strategy of the mobile dust removal device accordingly.

[0166] In the third aspect of the embodiments of the present invention,

[0167] A kind of electronic device is provided, including:

[0168] A processor;

[0169] A memory for storing instructions executable by the processor;

[0170] Wherein, the processor is configured to call the instructions stored in the memory to execute the method described above.

[0171] In the fourth aspect of the embodiments of the present invention,

[0172] A computer-readable storage medium is provided, on which computer program instructions are stored, and when the computer program instructions are executed by a processor, the method described above is implemented.

[0173] The present invention can be a method, a device, a system and / or a computer program product. The computer program product can include a computer-readable storage medium, on which computer-readable program instructions for executing various aspects of the present invention are carried.

[0174] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for monitoring and controlling dust concentration in a smart furniture production workshop, characterized in that: include: A laser scattering particle counter and a gravimetric dust sampler are arranged in the smart furniture production workshop to collect workshop dust concentration data in different areas of the workshop and transmit the data to the control center; based on the workshop dust concentration data, the control center uses a convolutional neural network to train historical workshop dust concentration data, constructs a three-dimensional model of workshop dust concentration distribution, and dynamically predicts the dust concentration in different areas and at different times of the workshop based on the three-dimensional model of dust concentration distribution; Based on the prediction results of the three-dimensional model of dust concentration distribution, the control center divides the workshop into two dust hazard level areas, one above the preset hazard threshold and one below the preset hazard threshold; When the laser scattering particle counter and the gravimetric dust sampler detect that the dust concentration in a certain area exceeds the preset danger threshold, the control center generates an early warning message, which records the location of the area where the dust exceeds the standard, the current dust concentration value and the duration of the exceeding standard, and sends the early warning message to the on-site management personnel terminal; the control center determines the control priority of each area based on the dust concentration change trend calculated by the three-dimensional dust concentration distribution model; The control center issues control instructions to the ventilation systems of the two dust hazard level areas above and below the preset hazard threshold, respectively, according to the control priority of each area, and adjusts the ventilation duct opening, fan speed and air supply direction; at the same time, the control center controls the mobile dust removal device to navigate to the area with the highest dust concentration according to the optimal dust removal path calculated by the three-dimensional model of dust concentration distribution, and starts the primary filtration unit, electrostatic dust removal unit and activated carbon adsorption unit of the mobile dust removal device for graded filtration; the laser scattering particle counter and gravimetric dust sampler collect particle dust concentration data in the control process in real time and feed it back to the control center, and the control center optimizes and updates the three-dimensional model of dust concentration distribution online according to the particle dust concentration data, and adjusts the ventilation system parameters and the operation strategy of the mobile dust removal device accordingly.

2. The method according to claim 1, characterized in that A laser scattering particle counter and a gravimetric dust sampler are arranged in the smart furniture production workshop to collect the workshop dust concentration data in different areas of the workshop and transmit it to the control center; the control center uses a convolutional neural network to train the historical workshop dust concentration data based on the workshop dust concentration data, constructs a three-dimensional model of workshop dust concentration distribution, and dynamically predicts the dust concentration in different areas and different time periods of the workshop based on the three-dimensional model of workshop dust concentration distribution, including: According to the functions of the smart furniture production workshop, the processing area, the polishing area, and the painting area are divided into three functional blocks, and a three-layer stereoscopic monitoring network of the ground layer, the human breathing layer, and the upper space layer is constructed in each of the functional blocks; a laser scattering particle counter is arranged in the three-layer stereoscopic monitoring network of each functional block, and the laser scattering particle counter adopts a semiconductor laser as a light source, and obtains the dust particle count value based on the forward scattered light intensity measurement; a gravimetric dust sampler is arranged in the human breathing layer of each functional block, and the gravimetric dust sampler obtains the mass concentration value of the inhalable dust through a PM2.5 cutter and a quartz filter membrane; The dust particle count value of the laser scattering particle counter and the inhalable dust mass concentration value of the gravimetric dust sampler are transmitted to a data acquisition module via an RS-485 bus. The data acquisition module uses a Modbus RTU communication protocol to read and encapsulate the dust particle count value and the inhalable dust mass concentration value, and uploads the encapsulated data to a control center via industrial Ethernet. The control center calibrates the dust particle count value based on the inhalable dust mass concentration value to generate standardized workshop dust concentration data. The control center organizes the standardized workshop dust concentration data according to time series characteristics and spatial characteristics, and constructs a training data set containing historical data of the past three months, where each record of the training data set contains a timestamp, spatial coordinates and a dust concentration value; a three-dimensional model of workshop dust concentration distribution is constructed based on the training data set using a three-dimensional convolutional neural network, and spatial features and time series features are extracted through four convolution blocks and an LSTM layer; the control center uses the three-dimensional model of workshop dust concentration distribution to dynamically predict the dust concentration in different areas of the workshop in the next four hours.

3. The method according to claim 1, characterized in that Based on the prediction results of the three-dimensional model of dust concentration distribution, the control center divides the workshop into two dust hazard level areas, one above the preset hazard threshold and one below the preset hazard threshold; When the laser scattering particle counter and the gravimetric dust sampler detect that the dust concentration in a certain area exceeds the preset danger threshold, the control center generates warning information including: The control center predicts the dust concentration of the workshop based on the three-dimensional dust concentration distribution model to obtain prediction result data, and the control center divides the workshop space into two dust hazard level areas, an area above a preset hazard threshold and an area below the preset hazard threshold, according to the prediction result data; The control center uses an adaptive threshold segmentation algorithm to perform binarization processing on the prediction result data, marks the area above the preset danger threshold as a first value, marks the area below the preset danger threshold as a second value, and uses a 3×3×3 dimensional three-dimensional median filter to smooth the marked data to obtain a marking result; The control center uses a region growing algorithm to perform a connected domain analysis on the marking results, detects the twenty-six adjacent points of each marking point, and classifies the adjacent points with the same marking into the same connected domain to obtain the dust hazard zoning result of the workshop; When the laser scattering particle counter and the gravimetric dust sampler detect that the dust concentration in a certain area exceeds the preset danger threshold for the first time, the control center activates a sixty-second observation window. During the sixty-second observation window, when more than eighty percent of the sampling points in the area continue to exceed the preset danger threshold and the average concentration value exceeds the preset danger threshold by twenty percent, the control center generates an early warning message.

4. The method according to claim 1, characterized in that: The warning information records the location of the dust exceeding the standard area, the current dust concentration value and the duration of the exceeding standard, and sends the warning information to the on-site management personnel terminal; The control center determines the control priority of each area based on the dust concentration change trend calculated by the three-dimensional dust concentration distribution model, including: The warning information records the location of the dust exceeding the standard area, the current dust concentration value and the duration of the exceeding standard. The control center encapsulates the warning information into a space information block, a concentration information block, a time information block and a trend information block in JSON format; The control center sends the warning information to the on-site management personnel terminal. After receiving the warning information, the on-site management personnel terminal displays the location of the dust exceeding standard area in a three-dimensional visual manner, and uses different colors to mark the exceeding standard level according to the current dust concentration value; The control center calculates the dust concentration change trend based on the three-dimensional dust concentration distribution model, uses the least squares method to fit the concentration change curve of the last four hours to obtain the change rate and acceleration, calculates the dust diffusion direction and speed through flow field simulation to obtain the spatial diffusion risk, and determines the exposure risk coefficient based on the production plan and personnel scheduling; The control center uses a weighted scoring method to determine the governance priority of each area, setting the weight of the dust concentration change trend to 0.4, the weight of the spatial diffusion risk to 0.3, and the weight of the exposure risk coefficient to 0.3; when the acceleration is positive and the change rate is greater than 0.5 mg per cubic meter per minute, the score of the dust concentration change trend is five points; when the acceleration is positive and the change rate is less than 0.5 mg per cubic meter per minute, the score of the dust concentration change trend is four points; when the acceleration is zero, the score of the dust concentration change trend is three points; when the acceleration is negative, the score of the dust concentration change trend is two points.

5. The method according to claim 1, characterized in that The control center issues control instructions to the ventilation systems of the two dust hazard level areas above and below the preset hazard threshold, respectively, according to the governance priority of each area, and adjusts the ventilation duct opening, fan speed and air supply direction; at the same time, the control center controls the mobile dust removal device to navigate to the area with the highest dust concentration according to the optimal dust removal path calculated by the three-dimensional model of dust concentration distribution, and starts the primary filter unit, electrostatic dust removal unit and activated carbon adsorption unit of the mobile dust removal device for graded filtration, including: The control center calculates the management priority of each area based on the area volume, the degree of dust concentration exceeding the standard and the density of personnel, and issues control instructions to the ventilation systems of the two dust hazard level areas above the preset hazard threshold and below the preset hazard threshold according to the management priority; The control center adjusts the ventilation duct opening of the area above the preset danger threshold to between 30% and 100%, and adjusts the fan speed to between 70% and 90% of the rated speed through the control command; adjusts the ventilation duct opening of the area below the preset danger threshold to between 20% and 40%, and adjusts the fan speed to between 40% and 60% of the rated speed; The control center uses SIMPLE algorithm to solve the velocity field and pressure field in the area to establish a three-dimensional flow field distribution model, and uses particle tracking method to simulate the dust movement trajectory to determine the optimal air supply angle. According to the optimal air supply angle, the air supply direction adjustment device is driven by a stepper motor to rotate within a range of 360 degrees in the horizontal direction and adjusted within a range of 0 to 60 degrees in the vertical direction; The control center calculates the optimal dust removal path based on the three-dimensional model of dust concentration distribution, and controls the mobile dust removal device to navigate to the area with the highest dust concentration according to the optimal dust removal path. The calculation of the optimal dust removal path includes using the A* algorithm to perform global path planning to obtain the shortest collision-free path, and using the dynamic window method to perform local path planning to avoid moving obstacles. The control center starts the primary filter unit, electrostatic dust removal unit and activated carbon adsorption unit of the mobile dust removal device for graded filtration. The primary filter unit uses a metal mesh filter element to filter particles larger than one hundred microns, and the control center adjusts the speed of the suction fan according to the pressure difference at both ends of the primary filter unit; the electrostatic dust removal unit uses bipolar discharge, and the control center adjusts the electric field strength within the range of five to fifteen kilovolts per centimeter according to the corona current; the activated carbon adsorption unit uses honeycomb activated carbon with a specific surface area of ​​one thousand two hundred square meters per gram to adsorb fine dust.

6. The method according to claim 5, characterized in that The control center calculates the optimal dust removal path based on the three-dimensional model of dust concentration distribution, and controls the mobile dust removal device to navigate to the area with the highest dust concentration according to the optimal dust removal path. The calculation of the optimal dust removal path includes using the A* algorithm to perform global path planning to obtain the shortest collision-free path, and using the dynamic window method to perform local path planning to avoid moving obstacles, including: The control center divides the space of the three-dimensional model of dust concentration distribution into cubic grids with a side length of 0.1 meter, and constructs a composite cost map based on the three-dimensional model of dust concentration distribution; The control center calculates the dust concentration cost of each grid node according to the three-dimensional model of dust concentration distribution, and the dust concentration cost includes a static concentration term and a dynamic diffusion term; the static concentration term is calculated according to the concentration value of the grid node in the three-dimensional model of dust concentration distribution; the dynamic diffusion term is calculated according to the concentration gradient and time change rate in the three-dimensional model of dust concentration distribution; The control center calculates the obstacle hazard for each grid node according to the three-dimensional model of dust concentration distribution, wherein the obstacle hazard is calculated using an exponential decay function based on distance and weighted in combination with the passage difficulty value in the three-dimensional model of dust concentration distribution; the control center combines the dust concentration cost and the obstacle hazard to form the composite cost map; The control center calculates the dust concentration gradient field according to the three-dimensional model of dust concentration distribution, and divides the space into a gradient area above a preset dust threshold and a gradient area below the preset dust threshold; the control center uses a search step of 0.05 meters in the gradient area above the preset dust threshold, and uses a search step of 0.2 meters in the gradient area below the preset dust threshold, and performs an adaptive gradient search on the three-dimensional model of dust concentration distribution; The control center constructs a multi-objective optimization model in combination with the spatiotemporal distribution characteristics in the three-dimensional model of dust concentration distribution; the optimization objectives of the multi-objective optimization model include minimizing the path length and maximizing the cumulative dust removal amount; the cumulative dust removal amount maximization is calculated based on the spatiotemporal distribution of the concentration in the three-dimensional model of dust concentration distribution; The control center uses the NSGA-II algorithm to solve the multi-objective optimization model, and the crossover operator of the NSGA-II algorithm adaptively adjusts the crossover parameters according to the local characteristics of the three-dimensional model of dust concentration distribution; the control center performs kinematic constraint verification on the optimized path plan, and performs obstacle avoidance constraint verification based on the three-dimensional model of dust concentration distribution; The control center monitors the changes of the three-dimensional model of dust concentration distribution in real time. When it is detected that the dust concentration distribution change in a local area exceeds the preset dust threshold, the composite cost map of the area is updated and the optimal dust removal path is recalculated.

7. The method according to claim 1, characterized in that The laser scattering particle counter and the gravimetric dust sampler collect the particle dust concentration data in the treatment process in real time and feed it back to the control center. The control center optimizes and updates the three-dimensional model of dust concentration distribution online according to the particle dust concentration data, and adjusts the ventilation system parameters and the operation strategy of the mobile dust removal device accordingly, including: The laser scattering particle counter and the gravimetric dust sampler collect the particle dust concentration data in real time during the treatment process. The gravimetric dust sampler adopts the isokinetic sampling principle and is equipped with a PM10 cutting head and a PM2.5 classifier. The gravimetric dust sampler adopts the thermobalance principle to measure the particle weight concentration in real time. The laser scattering particle counter and the gravimetric dust sampler feed back the particle dust concentration data to the control center in real time, and the control center uses a Kalman filter algorithm to fuse the particle dust concentration data, and performs abnormal value judgment on the fused particle dust concentration data based on a neural network model; The control center uses the recursive least squares method to perform online optimization and update of the three-dimensional dust concentration distribution model based on the fused particle dust concentration data to obtain an optimized and updated three-dimensional dust concentration distribution model; performs parameter identification calculation every five minutes to obtain the latest model parameter value, and uses the sliding time window method to select the data of the most recent thirty minutes to calculate the prediction error; when the prediction error exceeds a set threshold, the control center triggers model structure optimization; The control center constructs a dust concentration change equation in the prediction time domain based on the optimized and updated three-dimensional dust concentration distribution model, solves the multi-objective optimization problem by taking the energy consumption of the ventilation system and the degree of dust exceeding the standard as optimization targets, and obtains the optimal adjustment sequence of the ventilation system parameters; The control center adjusts the ventilation duct opening, fan speed and air supply direction of the ventilation system according to the optimal adjustment sequence of the ventilation system parameters; The control center uses an ant colony algorithm to replan the operation path of the mobile dust removal device based on the optimized and updated three-dimensional dust concentration distribution model, and adaptively adjusts the operation parameters of the filter units at each level of the mobile dust removal device according to the measured dust particle size distribution characteristics.

8. A dust concentration monitoring and control system for a smart furniture production workshop, used to implement the method described in any one of claims 1 to 7, characterized in that: include: The first unit is used to deploy laser scattering particle counters and gravimetric dust samplers in the smart furniture production workshop, collect workshop dust concentration data in different areas of the workshop and transmit them to the control center; the control center uses a convolutional neural network to train historical workshop dust concentration data based on the workshop dust concentration data, constructs a three-dimensional model of workshop dust concentration distribution, and dynamically predicts the dust concentration in different areas and at different times of the workshop based on the three-dimensional model of dust concentration distribution; The second unit is used for the control center to divide the workshop into two dust hazard level areas, one above the preset hazard threshold and the other below the preset hazard threshold, based on the prediction results of the three-dimensional model of dust concentration distribution; when the laser scattering particle counter and the gravimetric dust sampler detect that the dust concentration in a certain area exceeds the preset hazard threshold, the control center generates an early warning message, which records the location of the dust exceeding area, the current dust concentration value and the duration of the exceeding, and sends the early warning message to the on-site management personnel terminal; the control center determines the control priority of each area according to the dust concentration change trend calculated by the three-dimensional model of dust concentration distribution; The third unit is used for the control center to issue control instructions to the ventilation systems of the two dust hazard level areas above and below the preset hazard threshold according to the control priority of each area, and adjust the ventilation duct opening, fan speed and air supply direction; at the same time, the control center controls the mobile dust removal device to navigate to the area with the highest dust concentration according to the optimal dust removal path calculated by the three-dimensional model of dust concentration distribution, and starts the primary filtration unit, electrostatic dust removal unit and activated carbon adsorption unit of the mobile dust removal device for graded filtration; the laser scattering particle counter and gravimetric dust sampler collect particle dust concentration data in the control process in real time and feed it back to the control center, and the control center optimizes and updates the three-dimensional model of dust concentration distribution online according to the particle dust concentration data, and adjusts the ventilation system parameters and the operation strategy of the mobile dust removal device accordingly.

9. An electronic device, characterized in that: include: processor; a memory for storing processor-executable instructions; The processor is configured to call the instructions stored in the memory to execute the method described in any one of claims 1 to 7.

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

Images