Construction dust reduction control system based on environmental monitoring

Abstract

The present invention relates to the technical field of dust reduction control optimization, and discloses a construction dust reduction control system based on environmental monitoring, wherein the system comprises an environmental monitoring module, a dust reduction equipment module and an intelligent control module. The environmental monitoring module collects environmental data of the construction site in real time, including particle concentration, wind speed, wind direction, humidity and temperature. The dust reduction equipment module comprises a spray device, a wind shield and a covering material, which are used to reduce dust. The intelligent control module automatically adjusts the working mode of the dust reduction equipment in combination with environmental monitoring data and construction site image information to achieve precise control. By constructing a construction dust reduction decision model, the dust reduction process is modeled as a Markov decision process, and factors such as dust reduction effect, equipment energy consumption, construction safety and avoiding excessive dust reduction are comprehensively considered to achieve intelligent decision-making. The present invention can significantly improve the construction dust reduction effect, reduce equipment energy consumption, improve construction safety, meet the requirements of green construction, and has broad application prospects and promotion value.

Description

Technical Field

[0001] The present invention relates to the technical field of dust reduction control optimization, and in particular to a construction dust reduction control system based on environmental monitoring. Background Art

[0002] At construction sites, dust problems have always been an important factor affecting environmental quality, construction safety and worker health. At present, construction dust reduction methods mainly rely on on-site manual judgment and manual control of dust reduction equipment, such as spray devices, wind barriers and covering materials. This method lacks real-time monitoring and precise control. The environmental conditions at the construction site, such as particle concentration, wind speed, wind direction, humidity and temperature, are changing dynamically. However, manual judgment and manual control often fail to capture these changes in a timely and accurate manner and make corresponding adjustments, resulting in unstable dust reduction effects. Sometimes, dust reduction may be excessive, wasting resources, and sometimes, dust reduction is insufficient and dust cannot be effectively controlled.

[0003] Moreover, current dust reduction methods lack intelligent decision support. Dust reduction control at construction sites requires comprehensive consideration of multiple factors, including dust reduction effect, equipment energy consumption, construction safety, and avoiding excessive dust reduction. However, manual judgment and manual control often make it difficult to fully consider these factors and make the best decision, resulting in dust reduction control strategies that are not scientific, reasonable, and efficient enough. Summary of the invention

[0004] The purpose of the present invention is to provide a construction dust reduction control system based on environmental monitoring to solve the problems raised in the above background technology.

[0005] To achieve the above object, the present invention provides the following technical solution: a construction dust reduction control system based on environmental monitoring, the system comprising:

[0006] Environmental monitoring module, which integrates particle sensor, anemometer, wind direction meter, humidity sensor and temperature sensor, is used to monitor and collect environmental data of the construction site in real time;

[0007] Dust suppression equipment modules, including spray devices, wind barriers and covering materials, are used to reduce dust on construction sites;

[0008] The intelligent control module automatically adjusts the working mode of the dust suppression equipment based on the environmental data collected by the environmental monitoring module, the real-time image information of the construction site and the current status of the dust suppression equipment, so as to efficiently control the construction dust suppression;

[0009] Among them, the implementation method of the intelligent control module is:

[0010] A construction dust reduction decision model is constructed to model the construction dust reduction process as a Markov decision process. The specific definitions include:

[0011] State space: includes the environmental parameters of the construction site, the status of the dust suppression equipment, and the vector representation of the image features of the construction site. The environmental parameters include particle concentration, wind speed, wind direction, humidity, and temperature.

[0012] Action space: includes the possible actions that dust suppression equipment may take during the construction process, including starting or stopping the spray device, adjusting the spray intensity, deploying or retracting the wind shield, covering or uncovering the covering material;

[0013] Reward: A reward function designed based on the dust reduction effect, equipment energy consumption, and construction safety indicators of the construction site is used to evaluate the results of the action;

[0014] Policy: The rule for choosing an action in a given state, optimized through learning to maximize the cumulative reward;

[0015] Design a reward function. Based on the goal of construction dust reduction, design a reward mechanism that includes significant dust reduction effects, reduced equipment energy consumption, improved construction safety, and avoidance of excessive dust reduction. At the same time, set penalties for dust reduction failure or equipment failure.

[0016] To train the construction dust reduction decision model, we selected the deep Q network DQN algorithm, performed iterative training on actual construction data, and optimized the learning rate and discount factor hyperparameters;

[0017] Evaluate the construction dust reduction decision model, and compare the performance of the model recommendation scheme with the traditional control algorithm in terms of dust reduction effect, equipment energy consumption and construction safety indicators through a combination of simulation tests and field tests;

[0018] Based on the trained construction dust reduction decision-making model, the real-time environmental data, image information and status information of the dust reduction equipment at the construction site are analyzed to provide precise control strategies for the dust reduction equipment.

[0019] Preferably, the design of the reward function includes:

[0020] Defining the reward function , where s represents the current state, a represents the action taken in the current state, Represents the reward value obtained after taking action a in state s;

[0021] Reward Function It consists of multiple sub-reward functions, each of which corresponds to a construction dust reduction control target, including:

[0022] Dust reduction effect sub-reward function , calculated based on the particle concentration in the current state s and the change in particle concentration after action a, using a negative exponential function, that is, ,in, and Respectively represent the particle concentration before and after action a, and It is an adjustment parameter used to control the influence of dust reduction effect on the reward value;

[0023] Equipment Energy Consumption Reward Function , according to the energy consumption of the dust reduction equipment in the current state s and the energy consumption change after action a, a linear function is used, that is, ,in, and Respectively represent the energy consumption of the equipment before and after action a, It is an adjustment parameter used to control the degree of influence of equipment energy consumption on the reward value;

[0024] Construction Safety Sub-Reward Function , according to the construction safety evaluation under the current state s and the safety change calculation after action a, a piecewise function form is adopted, that is, when action a improves the construction safety, , when action a reduces construction safety, ,in, and are adjustment parameters, which are used to control the degree of influence of increasing and decreasing construction safety on the reward value;

[0025] Avoid over-degrading the reward function , calculated according to the dust fall intensity under the current state s and the change of dust fall intensity after action a, using the threshold function form, that is, when the dust fall intensity after action a exceeds the preset threshold, ,otherwise ,in, It is a parameter for adjusting the influence of excessive dustfall on the reward value.

[0026] Final reward function is the weighted sum of each sub-reward function, and the calculation formula is: ,in and is the weight coefficient, which is used to balance the proportion of each sub-reward function in the total reward value.

[0027] Preferably, the step of training the construction dust reduction decision model includes:

[0028] S1: Extract key state variables, executable action space and corresponding reward signals from the historical data of construction dust fall, and pre-process the data in advance, including data denoising, missing value processing and standardization, to form a training data set;

[0029] S2: Initialize DQN network parameters, including learning rate and discount factor;

[0030] S3: Use the ε-greedy strategy to train the model. At each decision, randomly select an action for exploration with probability ε, and select the action with the largest current Q value for utilization with probability 1-ε, so as to balance the exploration and utilization of the model.

[0031] S4: Perform iterative training on actual construction data and update the DQN network parameters by minimizing the loss function. The loss function is defined as: L = (r + γ * max_a' Q(s', a') - Q(s, a))^2, where r is the immediate reward of the current action, γ is the discount factor, s' is the next state, a' is the optional action for the next state, and Q(s, a) is the estimated Q value of the current state-action pair; S5: After each training, use the gradient descent method to update the network parameters to minimize the loss value;

[0032] S6: Repeat the above steps S3 to S5 until the performance index of the model meets the requirements.

[0033] Preferably, in the gradient descent algorithm, the parameter update formula is expressed as: ,in, Represents the model parameters, including all weights and biases that the model needs to learn; Represents the learning rate, which is used to control the step size of parameter update; Represents the loss function About parameters The gradient of , which is a vector pointing in the direction of fastest growth of the loss function; Indicates an assignment operation, that is, updating parameters The value of .

[0034] Preferably, in step S1, for the noise data, a sliding average filtering algorithm is used to perform data denoising, and the algorithm formula is: , where y[n] is the denoised data, x[n] is the original data, and N is the size of the sliding window.

[0035] Preferably, in step S1, a linear interpolation algorithm is used to fill in missing values.

[0036] Preferably, in step S1, the Z-score method is used to standardize the data, and the mean of each numerical data is subtracted and divided by its standard deviation, so that the processed data conforms to the standard normal distribution. The formula for standardization is: ,in, is the original data, is the mean of the original data, is the standard deviation of the original data, is the standardized data.

[0037] Preferably, the method for extracting the real-time image features of the construction site is:

[0038] Acquire real-time image data of the construction site, and collect real-time images of the construction site through a camera device;

[0039] The convolutional neural network (CNN) algorithm is used to extract features from the collected images. The CNN network structure includes an input layer, multiple convolutional layers, a pooling layer, a fully connected layer, and an output layer. The preprocessed image is input into the CNN network, the local features of the image are extracted through the convolutional layer, the feature dimension is reduced through the pooling layer, and the local features are combined into a global feature vector through the fully connected layer.

[0040] The extracted global feature vector is normalized to identify and extract the distribution of particles in the image, including the density, aggregation degree and diffusion range of the particles.

[0041] Preferably, the method for evaluating the construction dust reduction decision model is:

[0042] Step 1: Build a simulation test environment to simulate the actual conditions of the construction site based on historical construction data and environmental monitoring data, including environmental factors such as particle concentration, wind speed, wind direction, humidity, and temperature;

[0043] Step 2: In a simulated test environment, run the control scheme recommended by the construction dust reduction decision model and the traditional control algorithm respectively, and record and compare their performance in dust reduction effect, equipment energy consumption and construction safety indicators; the dust reduction effect is measured by the degree of reduction in particle concentration, the equipment energy consumption is calculated by the operating time and power of the dust reduction equipment, and the construction safety index takes into account the clarity of sight at the construction site and the safety of equipment operation;

[0044] Step 3: Conduct field tests, select representative construction sites, apply the control schemes recommended by the construction dust reduction decision model and the traditional control algorithms, and perform actual dust reduction operations; during the field tests, monitor and record the environmental data of the construction site, the operating status of the dust reduction equipment, and the construction safety indicators in real time;

[0045] Step 4: Conduct a comprehensive analysis of the results of the simulation test and the field test, compare the advantages and disadvantages of the recommended solution of the construction dust reduction decision model with the traditional control algorithm in terms of dust reduction effect, equipment energy consumption and construction safety indicators, and evaluate the practicality and effectiveness of the model.

[0046] Preferably, the intelligent control module also includes a human-computer interaction interface for displaying real-time environmental data of the construction site, the status of the dust reduction equipment and the control strategy recommended by the model, while providing an option for manually adjusting the control strategy.

[0047] Compared with the prior art, the present invention has the following beneficial effects:

[0048] The present invention collects environmental data of the construction site in real time through the environmental monitoring module, including particle concentration, wind speed, wind direction, humidity and temperature, etc., and can accurately capture the dynamic changes of environmental conditions at the construction site. The intelligent control module combines real-time monitoring data and real-time image information of the construction site to automatically adjust the working mode of the dust reduction equipment, ensure that the dust reduction measures match the on-site environment, achieve precise control, and avoid excessive or insufficient dust reduction.

[0049] The present invention constructs a construction dust reduction decision model, models the construction dust reduction process as a Markov decision process, and realizes intelligent decision-making on the construction dust reduction process by defining state space, action space, rewards and strategies. The decision model comprehensively considers multiple factors such as dust reduction effect, equipment energy consumption, construction safety and avoiding excessive dust reduction, and can make scientific, reasonable and efficient dust reduction control strategies to improve the overall effect of construction dust reduction. Through the precise control of the intelligent control module and the optimization strategy of the decision model, the present invention can significantly improve the effect of construction dust reduction and effectively reduce dust pollution at the construction site. At the same time, the present invention can avoid the ineffective operation of dust reduction equipment, reduce equipment energy consumption, achieve the goal of energy conservation and emission reduction, and meet the requirements of green construction. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 It is the overall structure diagram of the present invention;

[0051] Figure 2 This is the overall design diagram of the reward function;

[0052] Figure 3 This is the training flow chart of the construction dust reduction decision model. DETAILED DESCRIPTION

[0053] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0054] See also Figure 1-3 The present invention provides a technical solution: a construction dust reduction control system based on environmental monitoring, including an environmental monitoring module, a dust reduction equipment module and an intelligent control module. The environmental monitoring module integrates a particle sensor, an anemometer, an anemometer, a humidity sensor and a temperature sensor, which are used to monitor and collect environmental data of the construction site in real time.

[0055] The dust suppression equipment module is equipped with a spray device, a windshield and a covering material. The spray device can flexibly adjust the spray intensity according to the instructions of the intelligent control module to effectively reduce dust. The windshield and covering materials can be unfolded or covered when necessary to further reduce dust pollution at the construction site.

[0056] The intelligent control module conducts a comprehensive analysis based on the environmental data collected by the environmental monitoring module, the real-time image information of the construction site, and the current status of the dust reduction equipment. In order to achieve intelligent decision-making, a construction dust reduction decision model is constructed, which models the construction dust reduction process as a Markov decision process. The specific definitions are as follows:

[0057] State space: includes environmental parameters of the construction site (such as particle concentration, wind speed, wind direction, humidity and temperature), the status of dust suppression equipment, and vector representation of construction site image features, which provide input data for the decision model.

[0058] Action space: Contains the actions that dust suppression equipment may take during construction, such as starting or stopping the spray device, adjusting the spray intensity, deploying or retracting the wind shield, covering or uncovering the covering material, etc. These actions constitute the control strategies that the decision model can select.

[0059] Reward: Design a reward function based on the dust reduction effect, equipment energy consumption, and construction safety indicators of the construction site. This function is used to evaluate the results of the action and ensure that the decision model can select the optimal control strategy.

[0060] Strategy: The rule for selecting actions in a given state. Through learning optimization, the strategy can maximize the cumulative reward, thereby achieving the best effect of construction dust reduction.

[0061] After building the construction dust reduction decision model, the reward function is designed. According to the construction dust reduction goal, the design includes reward mechanisms such as significant dust reduction effect, reduced equipment energy consumption, improved construction safety, and avoiding excessive dust reduction. In order to deal with unexpected situations such as dust reduction failure or equipment failure, a corresponding penalty mechanism is also set up.

[0062] The deep Q network DQN algorithm is used to train the construction dust reduction decision model. Iterative training is performed on actual construction data, and hyperparameters such as learning rate and discount factor are tuned to ensure the performance and stability of the model.

[0063] In order to evaluate the effect of the construction dust reduction decision model, a combination of simulation test and field test was used. By comparing the performance of the model recommendation scheme with the traditional control algorithm in terms of dust reduction effect, equipment energy consumption and construction safety indicators, the superiority and practicality of the construction dust reduction control system based on environmental monitoring proposed in this invention were verified.

[0064] Finally, based on the trained construction dust reduction decision model, a comprehensive analysis of the real-time environmental data, image information and dust reduction equipment status information of the construction site is realized. The intelligent control module provides precise control strategies for the dust reduction equipment based on the analysis results.

[0065] The intelligent control module also includes a human-computer interaction interface for displaying real-time environmental data of the construction site, the status of the dust reduction equipment and the control strategy recommended by the model, while providing an option for manually adjusting the control strategy.

[0066] The present invention will be further described below in conjunction with Examples 1 to 4: Embodiment 1:

[0067] This embodiment proposes a design method of reward function, which aims to provide accurate reward signals for the intelligent control module by comprehensively considering multiple factors such as dust reduction effect, equipment energy consumption, construction safety and avoiding excessive dust reduction, so as to optimize the dust reduction strategy. Specifically, it includes:

[0068] Defining the reward function , where s represents the current state, covering key indicators such as particle concentration, equipment energy consumption, construction safety and dust reduction intensity at the construction site; a represents the actions taken under the current state, such as adjusting the power of dust reduction equipment, changing the dust reduction method, etc.; It represents the reward value obtained after taking action a in state s, which is used to guide the learning and optimization of the intelligent control module.

[0069] Reward Function It consists of four sub-reward functions, each of which corresponds to a construction dust reduction control target:

[0070] ① Dust reduction effect sub-reward function :This function is based on the particle concentration under the current state s and the particle concentration after action a The negative exponential function is used to calculate the change of. The specific expression is .in, and It is an adjustment parameter used to control the influence of dust reduction effect on the reward value. Lower than When , it means the dust reduction effect is good and the reward value is positive (due to the characteristics of negative exponential function, it is necessary to adjust and So that when dust reduction is effective, the overall expression results in a positive or close-to-zero reward); otherwise, the reward value is negative to punish ineffective dust reduction behavior.

[0071] ② Equipment Energy Consumption Reward Function :This function reduces the energy consumption of the dust reduction device according to the current state s and the energy consumption after action a The linear function form is used, and the specific expression is .in, It is an adjustment parameter used to control the influence of equipment energy consumption on the reward value. Lower than When , it means that the energy consumption is reduced and the reward value is positive; otherwise, the reward value is negative.

[0072] ③ Construction safety sub-reward function : This function is calculated based on the construction safety assessment under the current state s and the safety change after action a. It adopts the piecewise function form, specifically: when action a improves construction safety, ; When action a reduces construction safety, .in, and are adjustment parameters, which are used to control the degree of influence of improving and reducing construction safety on the reward value.

[0073] ④ Avoid excessive dust sub-reward function ,: This function is calculated based on the dust fall intensity in the current state s and the change in dust fall intensity after action a. It adopts the form of a threshold function, specifically: when the dust fall intensity after action a exceeds the preset threshold, ;otherwise, .in, It is an adjustment parameter used to control the impact of excessive dustfall on the reward value.

[0074] Finally, the reward function is the weighted sum of each sub-reward function, and the specific expression is: ,in and is the weight coefficient, which is used to balance the proportion of each sub-reward function in the total reward value to ensure that the reward function can fully reflect the comprehensive control goal of construction dust reduction.

[0075] Embodiment 2:

[0076] In order to achieve effective training of the construction dust reduction decision model, this implementation provides the following steps:

[0077] S1: Extract key state variables, executable action space and corresponding reward signals from historical data of construction dust reduction. State variables include but are not limited to particulate matter concentration (such as PM2.5, PM10), wind speed, wind direction, humidity, temperature and type of construction activities (such as excavation, transportation, pouring, etc.) at the construction site. The action space covers all possible operations of dust reduction equipment, such as opening / closing the spray device, adjusting the spray intensity, and deploying / retracting the windshield. The reward signal is designed comprehensively based on multiple dimensions such as dust reduction effect, equipment energy consumption, and construction safety.

[0078] After extracting the data, preprocessing is performed, including data denoising, missing value processing, and standardization.

[0079] For noisy data, a sliding average filtering algorithm is used to denoise the data to ensure the accuracy of the data. The algorithm formula is: , where y[n] is the denoised data, x[n] is the original data, and N is the size of the sliding window.

[0080] For missing values, linear interpolation algorithm is used to fill them in to ensure data integrity.

[0081] The Z-score method is used to standardize the data. Each numerical data is subtracted from its mean and divided by its standard deviation so that the processed data conforms to the standard normal distribution. The formula for standardization is: ,in, is the original data, is the mean of the original data, is the standard deviation of the original data, is the standardized data. Standardization scales all data according to a unified standard, eliminates the impact of data of different dimensions, and improves the training efficiency of the model.

[0082] After preprocessing, a training data set is formed.

[0083] S2: Before model training, the parameters of the deep Q network (DQN) need to be initialized. Specifically, they include hyperparameters such as the learning rate, discount factor, and the number of layers and nodes of the neural network. The learning rate controls the step size of the model parameter update. If it is too large, the model may be unstable, while if it is too small, the model may converge too slowly. The discount factor is used to balance the importance of immediate rewards and future rewards. The larger its value, the more the model focuses on long-term returns. The number of layers and nodes of the neural network affects the expressiveness and complexity of the model, and needs to be adjusted according to the specific task and data scale.

[0084] S3: During model training, the ε-greedy strategy is used for action selection. Specifically, at each decision, an action is randomly selected for exploration with probability ε, and the action with the largest current Q value is selected for utilization with probability 1-ε. Exploration can help the model discover new state-action pairs and improve the generalization ability of the model; utilization allows the model to make the best decision based on known information. As training progresses, the value of ε can be gradually reduced to allow the model to make more use of known information.

[0085] S4: Perform iterative training on actual construction data and update the DQN network parameters by minimizing the loss function. The loss function is defined as: L = (r + γ * max_a' Q(s', a') - Q(s, a))^2, where r is the immediate reward of the current action, γ is the discount factor, s' is the next state, a' is the optional action for the next state, and Q(s, a) is the Q value estimate of the current state-action pair.

[0086] S5: After each training, the loss value of the current training round is calculated, and the network parameters are updated using the gradient descent method. The gradient descent method calculates the gradient of the loss function with respect to the parameters and updates the parameters in the opposite direction of the gradient to minimize the loss value. Through continuous iterative training, the performance of the model gradually improves and the loss value gradually decreases. In the gradient descent algorithm, the parameter update formula is expressed as: ,in, Represents the model parameters, including all weights and biases that the model needs to learn; Represents the learning rate, which is used to control the step size of parameter update; Represents the loss function About parameters The gradient of , which is a vector pointing in the direction of fastest growth of the loss function; Indicates an assignment operation, that is, updating parameters The value of .

[0087] S6: Repeat the above steps S3 to S5 until the performance indicators of the model meet the requirements. The performance indicators include dust reduction effect, equipment energy consumption, construction safety and other aspects. Through continuous training and optimization, the model can gradually learn the optimal dust reduction strategy in different construction scenarios and improve the environmental management level of the construction site.

[0088] Embodiment 3:

[0089] This embodiment proposes a method for extracting real-time image features of a construction site, specifically:

[0090] First, the real-time image of the construction site is acquired. In this embodiment, a high-definition camera is installed at key locations of the construction site, such as the entrance to the construction area, the main working surface, etc., to capture the image data of the construction site in real time.

[0091] A convolutional neural network (CNN) is used to extract features from the preprocessed image. In this embodiment, the VGG16 network structure is selected as the CNN model. The VGG16 network contains multiple convolutional layers, pooling layers, and fully connected layers, which can extract deep features of the image. The preprocessed image is input into the VGG16 network, and the local features of the image, such as edges, textures, etc., are extracted through the convolutional layer; the feature dimension is reduced through the pooling layer to reduce the amount of calculation; finally, the local features are combined into a global feature vector through the fully connected layer. The feature extraction process is performed on a high-performance computer to ensure processing speed and accuracy.

[0092] The extracted global feature vector is normalized to identify and extract the distribution of particles in the image, including the density, aggregation degree and diffusion range of the particles.

[0093] Embodiment 4:

[0094] In order to evaluate the performance of the construction dust reduction decision model, this embodiment adopts a method combining simulation test and field test to compare the performance of the model recommendation scheme with the traditional control algorithm in terms of dust reduction effect, equipment energy consumption and construction safety indicators. The implementation steps include:

[0095] Step 1: Simulation test environment construction: Based on historical construction data and environmental monitoring data, a simulation test environment is constructed using computer simulation technology. This environment can simulate the actual conditions of the construction site, including environmental factors such as particle concentration, wind speed, wind direction, humidity, temperature, and the dynamic changes of construction activities. The simulation environment also includes the topographical features of the construction site, such as ups and downs, soil exposure, etc.

[0096] Step 2, simulation test operation and data recording: In the simulation test environment, run the control scheme recommended by the construction dust reduction decision model and the traditional control algorithm respectively. For each control scheme, multiple simulation tests are carried out to ensure the stability and reliability of the results. During the test, the performance of the two in dust reduction effect, equipment energy consumption and construction safety indicators is recorded and compared in real time. The dust reduction effect is measured by the degree of reduction in particle concentration, the equipment energy consumption is calculated by the operating time and power of the dust reduction equipment, and the construction safety index takes into account factors such as the clarity of sight at the construction site and the safety of equipment operation.

[0097] Step 3. Field test preparation and execution: In order to verify the results of the simulation test, a representative construction site was selected for field testing. Before the field test, a detailed survey and measurement of the construction site was carried out to ensure the accuracy and consistency of the test environment. Then, the control scheme recommended by the construction dust reduction decision model and the traditional control algorithm were applied to perform actual dust reduction operations. During the field test, professional monitoring equipment was used to monitor and record the environmental data of the construction site, the operating status of the dust reduction equipment, and the construction safety indicators in real time.

[0098] Step 4, data analysis and evaluation: Comprehensively analyze the results of simulation tests and field tests, and compare the advantages and disadvantages of the construction dust reduction decision model recommendation scheme and the traditional control algorithm in terms of dust reduction effect, equipment energy consumption and construction safety indicators. The specific analysis methods include data statistical analysis and difference significance test. Through the evaluation results, the performance advantages and disadvantages of the construction dust reduction decision model can be objectively reflected.

[0099] According to the evaluation results, the construction dust reduction decision model is adjusted and optimized as necessary. For example, if the dust reduction effect is poor, the reward function or strategy rules in the model can be adjusted; if the equipment energy consumption is too high, the operation mode and parameter settings of the dust reduction equipment can be optimized. Through continuous adjustment and optimization, the performance and effect of the construction dust reduction decision model in practical applications can be improved.

[0100] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0101] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. Construction dust reduction control system based on environmental monitoring, characterized in that: The system comprises: Environmental monitoring module, which integrates particle sensor, anemometer, wind direction meter, humidity sensor and temperature sensor, is used to monitor and collect environmental data of the construction site in real time; Dust suppression equipment modules, including spray devices, wind barriers and covering materials, are used to reduce dust on construction sites; The intelligent control module automatically adjusts the working mode of the dust suppression equipment based on the environmental data collected by the environmental monitoring module, the real-time image information of the construction site and the current status of the dust suppression equipment, so as to efficiently control the construction dust suppression; Among them, the implementation method of the intelligent control module is: A construction dust reduction decision model is constructed to model the construction dust reduction process as a Markov decision process. The specific definitions include: State space: includes the environmental parameters of the construction site, the status of the dust suppression equipment, and the vector representation of the image features of the construction site. The environmental parameters include particle concentration, wind speed, wind direction, humidity, and temperature. Action space: includes the possible actions that dust suppression equipment may take during the construction process, including starting or stopping the spray device, adjusting the spray intensity, deploying or retracting the wind shield, covering or uncovering the covering material; Reward: A reward function designed based on the dust reduction effect, equipment energy consumption, and construction safety indicators of the construction site is used to evaluate the results of the action; Policy: The rule for choosing an action in a given state, optimized through learning to maximize the cumulative reward; Design a reward function. Based on the goal of construction dust reduction, design a reward mechanism that includes significant dust reduction effects, reduced equipment energy consumption, improved construction safety, and avoidance of excessive dust reduction. At the same time, set penalties for dust reduction failure or equipment failure. To train the construction dust reduction decision model, we selected the deep Q network DQN algorithm, performed iterative training on actual construction data, and optimized the learning rate and discount factor hyperparameters; Evaluate the construction dust reduction decision model, and compare the performance of the model recommendation scheme with the traditional control algorithm in terms of dust reduction effect, equipment energy consumption and construction safety indicators through a combination of simulation tests and field tests; Based on the trained construction dust reduction decision-making model, the real-time environmental data, image information and status information of the dust reduction equipment at the construction site are analyzed to provide precise control strategies for the dust reduction equipment.

2. The construction dust reduction control system based on environmental monitoring according to claim 1 is characterized in that: The design of the reward function includes: defining a reward function Where s represents the current state, a represents the action taken in the current state, Represents the reward value obtained after taking action a in state s; Reward Function It consists of multiple sub-reward functions, each of which corresponds to a construction dust reduction control target, including: Dust reduction effect sub-reward function , calculated based on the particle concentration in the current state s and the change in particle concentration after action a, using a negative exponential function, that is, ,in, and Respectively represent the particle concentration before and after action a, and It is an adjustment parameter used to control the influence of dust reduction effect on the reward value; Equipment Energy Consumption Reward Function , according to the energy consumption of the dust reduction equipment in the current state s and the energy consumption change after action a, a linear function is used, that is, ,in, and Respectively represent the energy consumption of the equipment before and after action a, It is an adjustment parameter used to control the degree of influence of equipment energy consumption on the reward value; Construction Safety Sub-Reward Function , according to the construction safety evaluation under the current state s and the safety change calculation after action a, a piecewise function form is adopted, that is, when action a improves the construction safety, , when action a reduces construction safety, ,in, and are adjustment parameters, which are used to control the degree of influence of increasing and decreasing construction safety on the reward value; Avoid over-degrading sub-reward functions , calculated according to the dust fall intensity under the current state s and the change of dust fall intensity after action a, using the threshold function form, that is, when the dust fall intensity after action a exceeds the preset threshold, ,otherwise ,in, It is a parameter for adjusting the influence of excessive dustfall on the reward value. Final reward function is the weighted sum of each sub-reward function, and the calculation formula is: ,in and is the weight coefficient, which is used to balance the proportion of each sub-reward function in the total reward value.

3. The construction dust reduction control system based on environmental monitoring according to claim 2 is characterized in that: The steps for training the construction dust reduction decision model include: S1: Extract key state variables, executable action space and corresponding reward signals from the historical data of construction dust fall, and pre-process the data in advance, including data denoising, missing value processing and standardization, to form a training data set; S2: Initialize DQN network parameters, including learning rate and discount factor; S3: Use the ε-greedy strategy to train the model. At each decision, randomly select an action for exploration with probability ε, and select the action with the largest current Q value for utilization with probability 1-ε, so as to balance the exploration and utilization of the model. S4: Perform iterative training on actual construction data and update the DQN network parameters by minimizing the loss function. The loss function is defined as: L = (r + γ * max_a' Q(s', a') - Q(s, a))^2, where r is the immediate reward of the current action, γ is the discount factor, s' is the next state, a' is the optional action for the next state, and Q(s, a) is the Q value estimate of the current state-action pair; S5: After each training, the gradient descent method is used to update the network parameters to minimize the loss value; S6: Repeat the above steps S3 to S5 until the performance index of the model meets the requirements.

4. The construction dust reduction control system based on environmental monitoring according to claim 3 is characterized in that: In the gradient descent algorithm, the parameter update formula is expressed as: ,in, Represents the model parameters, including all weights and biases that the model needs to learn; Represents the learning rate, which is used to control the step size of parameter update; Represents the loss function About parameters The gradient of , which is a vector pointing in the direction of fastest growth of the loss function; Indicates an assignment operation, that is, updating parameters The value of .

5. The construction dust reduction control system based on environmental monitoring according to claim 3 is characterized in that: In step S1, for the noise data, a sliding average filtering algorithm is used to denoise the data, and the algorithm formula is: , where y[n] is the denoised data, x[n] is the original data, and N is the size of the sliding window.

6. The construction dust reduction control system based on environmental monitoring according to claim 3 is characterized in that: In step S1, a linear interpolation algorithm is used to fill in missing values.

7. The construction dust reduction control system based on environmental monitoring according to claim 3 is characterized in that: In step S1, the Z-score method is used to standardize the data, and the mean of each numerical data is subtracted and divided by its standard deviation, so that the processed data conforms to the standard normal distribution. The formula for standardization is: ,in, is the original data, is the mean of the original data, is the standard deviation of the original data, is the standardized data.

8. The construction dust reduction control system based on environmental monitoring according to claim 1 is characterized in that: The method for extracting real-time image features of the construction site is: Acquire real-time image data of the construction site, and collect real-time images of the construction site through a camera device; The convolutional neural network (CNN) algorithm is used to extract features from the collected images. The CNN network structure includes an input layer, multiple convolutional layers, a pooling layer, a fully connected layer, and an output layer. The preprocessed image is input into the CNN network, the local features of the image are extracted through the convolutional layer, the feature dimension is reduced through the pooling layer, and the local features are combined into a global feature vector through the fully connected layer. The extracted global feature vector is normalized to identify and extract the distribution of particles in the image, including the density, aggregation degree and diffusion range of the particles.

9. The construction dust reduction control system based on environmental monitoring according to claim 1 is characterized in that: The method for evaluating the construction dust reduction decision model is: Step 1: Build a simulation test environment to simulate the actual conditions of the construction site based on historical construction data and environmental monitoring data, including environmental factors such as particle concentration, wind speed, wind direction, humidity, and temperature; Step 2: In a simulated test environment, run the control scheme recommended by the construction dust reduction decision model and the traditional control algorithm respectively, and record and compare their performance in dust reduction effect, equipment energy consumption and construction safety indicators; the dust reduction effect is measured by the degree of reduction in particle concentration, the equipment energy consumption is calculated by the operating time and power of the dust reduction equipment, and the construction safety index takes into account the clarity of sight at the construction site and the safety of equipment operation; Step 3: Conduct field tests, select representative construction sites, apply the control schemes recommended by the construction dust reduction decision model and the traditional control algorithms, and perform actual dust reduction operations; during the field tests, monitor and record the environmental data of the construction site, the operating status of the dust reduction equipment, and the construction safety indicators in real time; Step 4: Conduct a comprehensive analysis of the results of the simulation test and the field test, compare the advantages and disadvantages of the recommended solution of the construction dust reduction decision model with the traditional control algorithm in terms of dust reduction effect, equipment energy consumption and construction safety indicators, and evaluate the practicality and effectiveness of the model.

10. The construction dust reduction control system based on environmental monitoring according to claim 1, characterized in that: The intelligent control module also includes a human-computer interaction interface for displaying real-time environmental data of the construction site, the status of the dust reduction equipment and the control strategy recommended by the model, while providing an option for manually adjusting the control strategy.

Images