Engineering project construction environmental protection supervision and management method and system

Abstract

The invention discloses an engineering project construction environmental protection supervision and management method and system. The method comprises the following steps: firstly, collecting multi-source heterogeneous environmental protection data; carrying out differential preprocessing on the multi-source heterogeneous environmental protection data by adopting a BP neural network optimized by an improved sparrow search algorithm; establishing a multi-dimensional data association model, calculating the association degree between each construction process and each pollutant index, and setting an association degree threshold value to locate a pollution source; a targeted rectification suggestion is generated by building a grading early warning mechanism; and finally, constructing a block chain closed-loop supervision process, tracking rectification scheme submission, rectification progress implementation and rectification result review processes of a construction unit, and storing whole-process data to an alliance chain node. According to the engineering project construction environmental protection supervision and management method and system, all-weather full-coverage monitoring of construction environmental protection is realized, data reliability and traceability efficiency are improved, and environmental protection supervision requirements of complex construction scenes such as buildings and mines can be effectively met.

Description

Technical Field

[0001] This invention relates to the field of environmental protection supervision technology in engineering projects, and in particular to a method and system for supervising and managing environmental protection during the construction of engineering projects. Background Technology

[0002] With increasing efforts in ecological and environmental protection, environmental supervision during the construction phase of engineering projects has become a key focus for the industry. Improper management of pollutants generated during construction, such as dust, noise, wastewater, and solid waste, can severely damage the surrounding ecological environment, affect residents' daily lives, and even trigger environmental pollution disputes.

[0003] Current environmental supervision and management methods for construction still rely primarily on traditional models. Manual inspections are limited by labor costs, inspection time, and coverage, making it difficult to achieve continuous monitoring across all weather conditions and areas, resulting in significant regulatory blind spots and delays. Single-sensor monitoring solutions can only acquire single-dimensional pollutant indicator data, lacking systematic integration and standardized preprocessing of multi-source data. They also lack clear indicator classifications and parameter definition standards, making data accuracy susceptible to interference from complex construction environments and unable to provide precise support for regulatory decisions. Furthermore, traditional early warning mechanisms often employ a uniform threshold comparison model, failing to design differentiated strategies based on the correlation between construction procedures and pollutant emissions. This results in insufficient targeting of early warning information, weak pollution source tracing capabilities, and a lack of clear direction and inefficiency in rectification efforts.

[0004] While some existing environmental monitoring systems attempt to incorporate sensing devices and data processing technologies, significant shortcomings remain: either the sensor acquisition range, core algorithm parameters, and indicator differentiation rules are not clearly defined, resulting in poor multi-source data fusion and insufficient monitoring accuracy to meet the demands of refined supervision; or although blockchain technology is used for data storage, key technical details such as consensus mechanisms and data storage media are not perfected, and appropriate early warning push schemes are not designed for different pollutant characteristics, leading to insufficient early warning response time and difficulty in meeting the current needs of intelligent and refined environmental supervision.

[0005] In summary, existing technologies generally suffer from problems such as incomplete monitoring coverage, insufficient data accuracy, vague definitions of core parameters, weak pollution source tracing capabilities, non-closed-loop regulatory processes, and a lack of targeted early warning pushes, which cannot meet the environmental supervision needs of complex construction scenarios in various engineering projects such as buildings and mines.

[0006] Therefore, there is an urgent need for a construction environmental supervision and management method and system that features efficient data processing, adaptable early warning strategies, and a controllable closed-loop process. Summary of the Invention

[0007] The purpose of this invention is to provide a method and system for supervising and managing environmental protection during construction projects, aiming to provide intelligent and closed-loop support for environmental protection supervision during construction, so as to achieve the goals of all-weather monitoring, highly reliable data, high-efficiency traceability and accurate early warning.

[0008] To achieve the above objectives, the present invention provides a method for environmental protection supervision and management of engineering project construction, comprising the following steps: S1. Collect multi-source heterogeneous environmental protection data; S2. Differentiated preprocessing of multi-source heterogeneous environmental protection data using a BP neural network optimized by an improved sparrow search algorithm; S3. Establish a multi-dimensional data association model, calculate the correlation between each construction process and each pollutant indicator, and set a correlation threshold to locate the source of pollution. S4. Establish a tiered early warning mechanism and generate targeted rectification suggestions; S5. Construct a blockchain closed-loop supervision process to track the submission of rectification plans by construction units, the implementation of rectification progress, and the review of rectification results, and store the entire process data to the consortium blockchain node.

[0009] Preferably, in S1, the multi-source heterogeneous environmental data pollutant indicators and environmental monitoring auxiliary indicators are specifically as follows: Dust pollutant indicators include PM2.5 concentration, PM10 concentration, and total suspended particulate matter (TSP) concentration collected by dust sensors; Noise pollutant indicators include the equivalent continuous A-weighted sound level, noise peak value, and noise frequency distribution collected by noise sensors; Water quality pollutant indicators include chemical oxygen demand (COD), ammonia nitrogen (NH3-N), suspended solids (SS), pH value, total phosphorus (TP), and total nitrogen (TN) concentrations collected by water quality monitoring sensors; Solid waste pollutant indicators include solid waste weight, solid waste type, stacking area and stacking height collected by solid waste monitoring sensors; Environmental monitoring auxiliary indicators include wind speed, humidity, temperature, wind direction, and atmospheric pressure collected by meteorological monitoring instruments; The raw data is transmitted to the cloud data center via the MQTT protocol using a 5G communication module.

[0010] Preferably, in S2, the improved sparrow search algorithm and optimized BP neural network specifically include the following steps: S21. Introducing adaptive weighting coefficients Adjust the search step size for individual sparrows: ; in, The maximum weighting coefficient, The minimum weight coefficient, This represents the current iteration number. This represents the maximum number of iterations. S22. Initialize the sparrow population using Logistic chaotic mapping, with the following formula: ; in, For the first The chaotic sequence values ​​from the next iteration are used to enhance the uniformity of the initial distribution of the population. S23. The fitness function is determined by minimizing the mean squared error of the BP neural network prediction. The fitness function is as follows: ; in, The fitness value is the mean squared error of the prediction. Sample size For the first The true value of each sample For the first Predicted values ​​for each sample; By iteratively optimizing the behaviors of the discoverer, follower, and watcher in the sparrow search algorithm, the mean square error of the prediction is minimized, and the optimal initial weights and thresholds of the BP neural network are obtained. S24. For discrete pollutant indicators, one-hot encoding is used to convert them into numerical features before inputting them into the network; for continuous pollutant indicators, normalization is applied to the [0,1] interval. The normalization formula is as follows: ; in, The original value, This is the minimum value of the indicator. This is the maximum value of the indicator; The optimized BP neural network structure parameters are as follows: The number of input layer nodes equals the total number of pollutant indicators and environmental monitoring auxiliary indicators. The number of hidden layer nodes is determined by an empirical formula. Sure; in, The number of nodes in the input layer. This represents the number of nodes in the output layer. It is a constant; the number of output layer nodes is 1, which is used to output the integrated environmental protection feature value after fusion.

[0011] Preferably, in S3, a multi-dimensional data correlation model between construction procedures and environmental impacts is established based on an improved grey relational analysis algorithm. The multi-source heterogeneous data preprocessed in S2 and the construction procedure parameters are input into the model to calculate the correlation degree between each construction procedure and each pollutant indicator. The formula is as follows: ; in, For the first The pollutant index and the first The degree of correlation between each construction process; For the preprocessed first The values ​​of each pollutant indicator; For the first The characteristic parameters of each construction process include process duration, equipment power, and material consumption. The resolution coefficient is set to 0.5 to mitigate the influence of the maximum absolute difference. It is the minimum absolute difference between the two layers, that is, the minimum value of the difference between all indicators and all process parameters; It is the maximum absolute difference between the two layers, that is, the maximum value of the difference between all indicators and all process parameters; Set a correlation threshold, when When the correlation threshold is ≥, determine the first The construction process is the first... The main pollution sources for each pollutant indicator; for discrete pollutant indicators, they need to be converted into standardized process characteristic parameters before the correlation calculation.

[0012] Preferably, S4 is as follows: Three levels of early warning thresholds are set for each pollutant indicator. The pollutant indicators after S2 preprocessing are compared with the early warning thresholds. Combined with the pollution source tracing results of S3, early warning information including the type of exceeding indicator, value, and related process is generated and pushed to relevant responsible persons through multiple channels. The content and strategy of pushing information are optimized by linking environmental monitoring auxiliary indicators.

[0013] Preferably, in S4, the criteria for classifying the three-level warning thresholds are as follows: A Level 1 warning is issued when the monitored value exceeds the standard value by 10% to 20%, which is considered a slight exceedance. A Level II warning is issued when the monitored value exceeds the standard value by 20% to 50%, which is considered a moderate exceedance. A Level 3 warning is issued when the monitored value exceeds the standard value by more than 50%, which is considered a severe exceedance. The warning information includes the name of the indicator exceeding the standard, the value exceeding the standard, the duration of the exceeding the standard, the related construction procedures, targeted rectification suggestions, and the rectification deadline; The push channels include SMS, system pop-ups, and sound and light alarms, and the target audience is the on-site manager of the construction unit, environmental protection supervisors, and project supervisors.

[0014] Preferably, in S4, the priority for pushing indicators for dust pollutants is as follows: When any level of warning is triggered, the audible and visual alarm at the construction site will immediately activate the graded audible and visual signals: yellow for Level 1, orange for Level 2, and red for Level 3, continuing until rectification is initiated. Simultaneously, pop-up windows will appear on the terminals of the construction unit, environmental protection supervision department, and project supervision, displaying the pollutant index exceeding the standard, real-time values, related processes, and current wind speed. A text message will be sent to the responsible person within 1 minute, and the related processes will immediately take spray dust suppression measures. For noise pollutant indicators, the priority for push notifications is as follows: When an alert is triggered during non-resident rest periods, a notification is sent via pop-up windows, SMS messages, and audible and visual alarms across all terminals. The pop-up window includes noise frequency distribution data and related procedures. When an alert is triggered during resident rest periods, the on-site audible and visual alarms are turned off, and only a highlighted pop-up window and a red-highlighted emergency nighttime alert SMS message are sent, along with a mandatory rectification suggestion to immediately suspend high-noise procedures. For water quality pollutant indicators, the priority for push notifications is as follows: When an alert is triggered, environmental regulators and project supervisors will see a pop-up window displaying comprehensive pollutant index data, historical trend curves, related processes, and potential hazards; SMS messages will include pollutant indices exceeding water quality standards, their values, and a download link for a rectification plan template; the construction technical supervisor will receive on-site notifications and simultaneously obtain real-time monitoring locations of water quality sensors to pinpoint the discharge point. For solid waste pollutant indicators, the priority for push notifications is as follows: When an alert is triggered, a pop-up window on all terminals displays the precise coordinates of the solid waste storage area, the percentage exceeding the standard, the type of solid waste, and the associated processes; an SMS message is sent with the site number of the storage area; and the project supervision unit simultaneously generates a paper alert notification for rectification review and filing. For auxiliary indicators of environmental monitoring, warnings are not triggered independently, but are only pushed out in conjunction with pollutant indicators, with the following priority: When the wind speed is ≥5m / s, the linkage dust pollutant index will be pushed, indicating that strong winds can easily aggravate dust spread. Please reinforce the covering and start the spray equipment. When the humidity is ≤30%, the indicator for solid waste pollutants will be pushed out, indicating that dust is likely to be generated in dry weather, and please increase the frequency of watering.

[0015] Preferably, in S5, the consortium blockchain nodes include construction unit nodes, environmental protection regulatory department nodes, supervision unit nodes, and third-party testing agency nodes; The construction unit node is used to upload the rectification plan and real-time pollutant indicators and environmental monitoring auxiliary indicator data during the rectification process; Environmental regulatory authorities are responsible for reviewing the relevance and feasibility of rectification plans. The supervisory unit's checkpoint is used to verify the compliance status of pollutant indicators after rectification; Third-party testing agency nodes are used to upload independent reports on all pollutant indicators. Each node adopts the PBFT consensus mechanism and has data read / write permissions and consensus verification permissions.

[0016] This invention also provides an environmental protection supervision and management system for construction projects, comprising: The data acquisition and communication module is used to collect multi-source heterogeneous environmental data through multiple sensors and monitoring equipment, and transmit the raw multi-source heterogeneous data to the cloud data center through the 5G communication module and MQTT protocol. The data fusion and preprocessing module has a built-in BP neural network optimized by the improved sparrow search algorithm, which is used to perform differentiated preprocessing and feature fusion on multi-source heterogeneous environmental protection data and output comprehensive environmental protection feature values. The pollution source tracing and analysis module establishes a multi-dimensional data association model based on an improved grey relational analysis algorithm, calculates and locates the correlation between construction procedures and pollutant indicators, as well as the main pollution sources. The intelligent early warning and push module has a built-in hierarchical early warning mechanism, which is used to generate and push early warning information containing targeted rectification suggestions through multiple channels based on the level of exceeding the standard and the results of pollution source tracing. The blockchain closed-loop supervision module is used to build a consortium blockchain network to track and store data throughout the entire process, from the submission of rectification plans and progress implementation to the review of results.

[0017] Preferably, the data fusion and preprocessing module includes: The computational processing unit is used to execute the iterative optimization process of the improved sparrow search algorithm to determine the optimal initial weights and thresholds of the BP neural network; The data standardization unit is used to perform one-hot encoding on discrete pollutant indicators and normalize continuous pollutant indicators to the [0,1] interval. The blockchain closed-loop supervision module includes a verification and audit unit. Based on the consensus mechanism of the consortium blockchain nodes, it cross-verifies the rectification-related data and reports submitted by each node and generates audit traceability records.

[0018] Therefore, the present invention employs the above-mentioned method and system for environmental supervision and management of engineering project construction, and the beneficial effects are as follows: (1) By defining the specific acquisition indicators of various sensors, this invention constructs a multi-source heterogeneous intelligent sensor network to achieve all-weather, full-coverage monitoring of environmental protection indicators at construction sites. Compared with traditional manual inspection, the monitoring efficiency is greatly improved.

[0019] (2) The multi-dimensional data association model constructed in this invention clarifies the core parameters of grey relational analysis and establishes a correlation between pollutant indicators and construction procedures. It can accurately locate the source procedures of different types of pollutants and provide a clear direction for rectification work.

[0020] (3) The full-indicator coverage hierarchical early warning mechanism established by this invention clarifies the benchmark standard value of pollutant indicators, threshold calculation rules and type-differentiated push strategy, links auxiliary indicator optimization content, and adapts to the pollution characteristics of different pollutant indicators.

[0021] (4) The closed-loop regulatory process and blockchain storage technology of the present invention clarify the consensus mechanism and node permissions of the consortium blockchain, and synchronously store pollutant indicators, auxiliary indicators and regulatory data, effectively solving the problem of shirking responsibility in the existing supervision and improving the standardization and fairness of the regulatory process.

[0022] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0023] Figure 1 This is an overall flowchart of an embodiment of an environmental protection supervision and management method for construction projects according to the present invention; Figure 2 This is an overall block diagram of an embodiment of the environmental protection supervision and management method system for construction projects according to the present invention; Figure 3 This is a graph showing the original data curve of PM10 concentration monitoring error in an embodiment of an environmental supervision and management method and system for construction projects according to the present invention. Detailed Implementation

[0024] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0025] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.

[0026] like Figure 1 As shown, a method for environmental protection supervision and management during construction of an engineering project includes the following steps: S1. Collect multi-source heterogeneous environmental protection data.

[0027] Multi-source heterogeneous environmental data pollutant indicators and environmental monitoring auxiliary indicators are as follows: Dust pollutant indicators include PM2.5 concentration, PM10 concentration, and total suspended particulate matter (TSP) concentration collected by dust sensors.

[0028] Noise pollutant indicators include the equivalent continuous A-weighted sound level, noise peak value, and noise frequency distribution collected by noise sensors.

[0029] Water quality pollutant indicators include chemical oxygen demand (COD), ammonia nitrogen (NH3-N), suspended solids (SS), pH value, total phosphorus (TP), and total nitrogen (TN) concentrations collected by water quality monitoring sensors.

[0030] Solid waste pollutant indicators include solid waste weight, solid waste type, stacking area, and stacking height, collected through solid waste monitoring sensors.

[0031] Environmental monitoring auxiliary indicators include wind speed, humidity, temperature, wind direction, and atmospheric pressure collected by meteorological monitoring instruments.

[0032] The raw data is transmitted to the cloud data center via the MQTT protocol (Message Queuing Telemetry Transport Protocol, a lightweight communication protocol based on the publish / subscribe model) using the 5G communication module.

[0033] S2. A BP neural network optimized by an improved sparrow search algorithm is used to preprocess multi-source heterogeneous environmental protection data.

[0034] This invention sets differentiated preprocessing strategies for different types of data. The original data of pollutant indicators and environmental monitoring auxiliary indicators collected by S1 are input into the optimized BP neural network to complete the one-heat encoding of discrete pollutant indicators (solid waste type, noise frequency), the denoising and completion of continuous pollutant indicators (concentration, weight, temperature, etc.), and the feature-level fusion processing of multi-source data.

[0035] The improved sparrow search algorithm and optimized BP neural network of this invention specifically include the following steps: S21. Introducing adaptive weighting coefficients Adjust the search step size for individual sparrows: ; in, =0.9 is the maximum weighting coefficient. =0.4 is the minimum weighting coefficient. This represents the current iteration number. =1000 is the maximum number of iterations.

[0036] S22. Initialize the sparrow population using Logistic chaotic mapping, with the following formula: ; in, For the first The chaotic sequence values ​​from the next iteration are used to enhance the uniformity of the initial distribution of the population, thereby further enhancing population diversity.

[0037] S23. The fitness function is determined by minimizing the mean squared error of the BP neural network prediction. The fitness function is as follows: ; in, The fitness value is the mean squared error of the prediction. Sample size For the first The true value of each sample For the first The predicted value for each sample.

[0038] By iteratively optimizing the behaviors of discoverers, followers, and watchdogs in the sparrow search algorithm, the mean square error of prediction is minimized, and the optimal initial weights and thresholds of the BP neural network are obtained.

[0039] S24. For discrete pollutant indicators, one-hot encoding is used to convert them into numerical features before inputting them into the network; for continuous pollutant indicators, normalization is applied to the [0,1] interval. The normalization formula is as follows: ; in, The original value, This is the minimum value of the indicator. This is the maximum value of the indicator.

[0040] The optimized BP neural network structure parameters of this invention are as follows: The number of input layer nodes equals the total number of pollutant indicators and environmental monitoring auxiliary indicators, which is 18 in this invention, including 13 pollutant indicators and 5 meteorological auxiliary indicators. The number of hidden layer nodes is determined by an empirical formula. Sure.

[0041] in, =18 is the number of nodes in the input layer. =1 represents the number of output layer nodes. If 5 is a constant, then the number of hidden layer nodes is 10; the number of output layer nodes is 1, used to output the fused comprehensive environmental protection feature value. The neural network is trained for 1000 iterations with a learning rate of 0.01 (weight update step size), and the sigmoid function is used as the activation function to enhance nonlinear fitting ability.

[0042] S3. Based on the improved grey relational analysis algorithm, establish a multi-dimensional data correlation model between construction procedures and environmental impacts. Input the multi-source heterogeneous data preprocessed in S2 and the construction procedure parameters into the model, and calculate the correlation degree between each construction procedure and each pollutant indicator. The formula is as follows: ; in, For the first The pollutant index and the first The degree of correlation between each construction process; For the preprocessed first The values ​​of each pollutant indicator; For the first The characteristic parameters of each construction process include process duration, equipment power, and material consumption. The resolution coefficient is set to 0.5, which is used to reduce the influence of the maximum absolute difference.

[0043] It is the minimum absolute difference between the two layers, that is, the minimum value of the difference between all indicators and all process parameters; It represents the maximum absolute difference between two layers, i.e., the maximum value of the difference between all indicators and all process parameters.

[0044] Set the correlation threshold to 0.7, when When the correlation threshold is ≥0.7, the first [item] is determined. The construction process is the first... The main pollution sources for each pollutant indicator; for discrete pollutant indicators, they need to be converted into standardized process characteristic parameters before the correlation calculation.

[0045] S4. Establish a tiered early warning mechanism to generate targeted rectification suggestions, specifically: Three levels of early warning thresholds are set for each pollutant indicator. The pollutant indicators after S2 preprocessing are compared with the early warning thresholds. Combined with the pollution source tracing results of S3, early warning information including the type of exceeding indicator, value, and related process is generated and pushed to relevant responsible persons through multiple channels. The content and strategy of pushing information are optimized by linking environmental monitoring auxiliary indicators.

[0046] The criteria for classifying the three-level warning threshold are as follows: A Level 1 warning is issued when the monitored value exceeds the standard value by 10% to 20%, which is considered a slight exceedance. A Level II warning is issued when the monitored value exceeds the standard value by 20% to 50%, which is considered a moderate exceedance. A Level 3 warning is issued when the monitored value exceeds the standard value by more than 50%, which is considered a severe exceedance.

[0047] The baseline values ​​for each pollutant indicator are as follows: PM2.5 concentration is 35. PM10 concentration was 150 The total suspended particulate matter (TSP) concentration was 300. The equivalent continuous A-weighted sound level is 70 dB (daytime), and the peak noise level is 85 dB; the pH value is 6–9, the chemical oxygen demand (COD) is 100 mg / L, the ammonia nitrogen (NH3-N) is 1.0 mg / L, the suspended solids (SS) are 70 mg / L, the total phosphorus (TP) is 0.5 mg / L, and the total nitrogen (TN) concentration is 15 mg / L; the height of solid waste stockpiles in a single area must be ≤5m, and the stockpiling area must be ≤50m. The weight of each stack in a single area must be ≤1000kg.

[0048] The early warning information includes the name of the exceeded standard index, the exceeded value, the exceeded duration, the associated construction process, the targeted rectification suggestions, and the rectification time limit; the push channels include text messages, system pop-ups, and audible and visual alarms, and the push targets are the on-site person in charge of the construction unit, environmental protection supervision personnel, and project supervisors. The specific push method is set differently according to the type of pollutant index, and the environmental protection monitoring auxiliary index is used to supplement the push content联动.

[0049] For dust pollutant indicators, this invention adopts an immediate and wide-coverage collaborative push strategy, and the push priority is: When any level of early warning is triggered, the audible and visual alarm at the construction site immediately activates graded audible and visual signals. The first-level early warning is yellow, the second-level early warning is orange, and the third-level early warning is red, lasting until the rectification is started; pop-ups will simultaneously appear on the terminals of the construction unit, environmental protection supervision department, and project supervisor, showing the exceeded pollutant index, real-time value, associated process, and the current wind speed through the environmental protection monitoring auxiliary index to assist in judging the diffusion range; the text message will be sent to the responsible person within 1 minute, and the content will be streamlined to "

Engineering Environmental Protection Early Warning

[0050] For noise pollutant indicators, this invention adopts a time period adaptability and targeted reminder push strategy, and the push priority is: When the early warning is triggered during the non-resident rest period (7:00-22:00), the pop-up window, text message, and audible and visual alarm on all terminals will be pushed collaboratively, and the pop-up window will be attached with noise frequency distribution data and associated processes; when the early warning is triggered during the resident rest period (22:00 - 7:00 the next day), the on-site audible and visual alarm will be turned off, and only the highlighted pop-up window and the text message with a red-highlighted night emergency warning will be pushed, with an additional mandatory rectification suggestion to immediately suspend high-noise processes.

[0051] For water quality pollutant indicators, this invention adopts a data integrity and rectification guidance push strategy, and the push priority is: When the early warning is triggered, the pop-up windows on the terminals of environmental protection supervision personnel and project supervisors will display all-dimensional pollutant index data, historical trend curves, associated processes, and potential hazards; the content of the text message push includes the exceeded water quality pollutant index, value, and the download link of the rectification plan template; the construction technical person in charge receives on-site notifications and simultaneously obtains the real-time monitoring location of the water quality sensor to locate the sewage discharge point.

[0052] For solid waste pollutant indicators, this invention adopts a location accuracy and disposal pertinence push strategy, and the push priority is: When the early warning is triggered, the pop-up window on all terminals will display the precise coordinates of the solid waste stacking, the exceeded proportion, the type of solid waste, and the associated process; the text message will mark the on-site number of the stacking area for quick positioning; the project supervision unit will simultaneously generate a paper early warning notice for rectification review and filing.

[0053] For auxiliary indicators of environmental monitoring, early warnings are not triggered independently, but are only pushed out in conjunction with pollutant indicators. A correlation-based and predictive early warning push strategy is adopted, with the following push priority: When the wind speed is ≥5m / s, the system will push the dust pollutant index and remind you that "strong winds can exacerbate dust dispersion. Please reinforce the covering and start the spray equipment."

[0054] When the humidity is ≤30%, the system will push out indicators for solid waste pollutants, prompting "Dry weather is prone to dust, please increase the frequency of watering", with the content focusing on the impact of meteorological conditions on the spread of pollutants and preventive measures.

[0055] S5. Construct a blockchain closed-loop supervision process to track the submission of rectification plans by construction units, the implementation of rectification progress, and the review of rectification results, and store the entire process data to the consortium blockchain node.

[0056] The consortium blockchain nodes include construction unit nodes, environmental regulatory department nodes, supervision unit nodes, and third-party testing agency nodes. The permissions of each node are as follows: The construction unit node is used to upload rectification plans and real-time pollutant indicators and environmental monitoring auxiliary indicator data during the rectification process.

[0057] Environmental regulatory authorities are responsible for reviewing the relevance and feasibility of rectification plans.

[0058] The supervisory unit's checkpoint is used to verify the compliance status of pollutant indicators after rectification.

[0059] Third-party testing agency nodes are used to upload independent, multi-dimensional pollutant indicator reports; each node adopts the PBFT consensus mechanism, with a fault tolerance rate of 25% when the number of nodes is ≥4, and all nodes have data read and write permissions and consensus verification permissions.

[0060] The time limits for the closed-loop supervision process of this invention are as follows: a rectification plan must be submitted within 2 hours for a Level 1 warning, within 1 hour for a Level 2 warning, and within 30 minutes for a Level 3 warning.

[0061] like Figure 2 As shown, the present invention also provides an environmental protection supervision and management system for construction projects, comprising: The data acquisition and communication module is used to collect multi-source heterogeneous environmental data through multiple sensors and monitoring equipment, and transmit the raw multi-source heterogeneous data to the cloud data center through the 5G communication module and MQTT protocol.

[0062] The data fusion and preprocessing module incorporates an improved sparrow search algorithm and an optimized BP neural network for differentiated preprocessing and feature fusion of multi-source heterogeneous environmental protection data, outputting comprehensive environmental protection feature values. The data fusion and preprocessing module includes: The computational processing unit is used to execute the iterative optimization process of the improved sparrow search algorithm to determine the optimal initial weights and thresholds of the BP neural network.

[0063] The data standardization unit is used to perform one-heat encoding on discrete pollutant indicators and normalize continuous pollutant indicators to the [0,1] interval.

[0064] The pollution source tracing and analysis module establishes a multi-dimensional data association model based on an improved grey relational analysis algorithm to calculate and locate the correlation between construction procedures and pollutant indicators, as well as the main pollution sources. The intelligent early warning and push module has a built-in hierarchical early warning mechanism, which is used to generate and push early warning information containing targeted rectification suggestions through multiple channels based on the level of exceedance and pollution source tracing results.

[0065] The blockchain closed-loop supervision module is used to build a consortium blockchain network to track and store data throughout the entire process, from the submission of rectification plans and their implementation to the review of results. This module includes a verification and audit unit that, based on the consensus mechanism of the consortium blockchain nodes, cross-verifies the rectification-related data and reports submitted by each node and generates audit traceability records.

[0066] Example 1: This embodiment uses an open-pit mine project as the test object, with a construction area of ​​80,000 square meters. The core processes include blasting, ore transportation, and tailings stacking, with a testing period of 30 days. The method and system of this invention were used in parallel monitoring with two existing methods: Group B (traditional manual inspection + single threshold early warning) and Group C (monitoring system without optimized algorithm) to verify the effectiveness of this method. Test parameters: 18 nodes in the input layer of the BP neural network (13 pollutant indicators + 5 auxiliary indicators), 10 nodes in the hidden layer, a grey relational analysis resolution coefficient of 0.5, and the blockchain using the PBFT consensus mechanism to ensure the experiment closely matches the invention.

[0067] Experimental steps: Fifteen dust sensors, ten noise sensors, six water quality sensors, eight solid waste sensors, and three meteorological monitoring instruments were deployed to collect all indicators. The sampling frequency for dust and noise data was once per minute, and the sampling frequency for water quality and solid waste data was once per hour. After collection, the data was transmitted to the cloud via the 5G+MQTT protocol.

[0068] Four core indicators were selected: dust monitoring accuracy, pollution source tracing accuracy, early warning response time, and rectification closure rate. The comparison of core performance indicators and the raw data of PM10 monitoring errors are shown in Tables 1 and 2. Table 1 Comparison of core performance indicators of the three methods

[0069] Table 2. Raw data on PM10 concentration monitoring errors

[0070] Combine Table 2 and Figure 3 It can be seen that Group A, relying on the improved sparrow search algorithm and optimized BP neural network of this invention, achieved differentiated preprocessing of discrete / continuous indicators, with a monitoring accuracy of over 97% for dust and water quality indicators. The final error of PM10 was only 3.2%, which is 84.6% lower than Group B's 21.3% and 63.2% lower than Group C's 8.7%, effectively eliminating interference from the construction environment, and the error optimization trend is clearly visible.

[0071] As shown in Table 1, based on the improved grey relational analysis model, when the correlation threshold is 0.7, the source tracing accuracy of Group A reaches 96.8%, which can accurately locate the correlation between blasting and dust, and transportation and noise. The early warning response time is only 2.8 minutes, which is 85% shorter than the traditional method, and can quickly curb the spread of pollution.

[0072] Meanwhile, as can be seen from the data in Table 1, the four nodes of the blockchain consortium chain collaborate with the PBFT consensus mechanism to achieve full-process data traceability. The rectification closure rate of Group A reached 98.5%, which is significantly higher than the existing methods. This effectively solves the problem of shirking traditional regulatory responsibilities and adapts to the environmental supervision needs of complex construction scenarios such as mines.

[0073] Therefore, this invention adopts the aforementioned method and system for environmental supervision and management of construction projects, achieving an upgrade in construction environmental supervision through a full-chain design. Relying on improved algorithms, dust monitoring accuracy reaches 97.2% and source tracing accuracy reaches 96.8%, significantly shortening early warning time, improving closed-loop rate, solving pain points of traditional supervision, adapting to complex scenarios, and providing efficient and reliable technical support for environmental supervision.

[0074] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for environmental supervision and management of engineering project construction, characterized in that, Includes the following steps: S1. Collect multi-source heterogeneous environmental protection data; S2. Differentiated preprocessing of multi-source heterogeneous environmental protection data using a BP neural network optimized by an improved sparrow search algorithm; S3. Establish a multi-dimensional data association model, calculate the correlation between each construction process and each pollutant indicator, and set a correlation threshold to locate the source of pollution. S4. Establish a tiered early warning mechanism and generate targeted rectification suggestions; S5. Construct a blockchain closed-loop supervision process to track the submission of rectification plans by construction units, the implementation of rectification progress, and the review of rectification results, and store the entire process data to the consortium blockchain node.

2. The method for environmental protection supervision and management of engineering project construction according to claim 1, characterized in that, In S1, the pollutant indicators and auxiliary indicators for environmental monitoring from multiple sources and heterogeneous sources are specifically as follows: Dust pollutant indicators include PM2.5 concentration, PM10 concentration, and total suspended particulate matter (TSP) concentration collected by dust sensors; Noise pollutant indicators include the equivalent continuous A-weighted sound level, noise peak value, and noise frequency distribution collected by noise sensors; Water quality pollutant indicators include chemical oxygen demand (COD), ammonia nitrogen (NH3-N), suspended solids (SS), pH value, total phosphorus (TP), and total nitrogen (TN) concentrations collected by water quality monitoring sensors; Solid waste pollutant indicators include solid waste weight, solid waste type, stacking area and stacking height collected by solid waste monitoring sensors; Environmental monitoring auxiliary indicators include wind speed, humidity, temperature, wind direction, and atmospheric pressure collected by meteorological monitoring instruments; The raw data is transmitted to the cloud data center via the MQTT protocol using a 5G communication module.

3. The method for environmental supervision and management of engineering project construction according to claim 2, characterized in that, In S2, the improved sparrow search algorithm and optimized BP neural network specifically include the following steps: S21. Introducing adaptive weighting coefficients Adjust the search step size for individual sparrows: ； in, The maximum weighting coefficient, The minimum weight coefficient, This represents the current iteration number. This represents the maximum number of iterations. S22. Initialize the sparrow population using Logistic chaotic mapping, with the following formula: ； in, For the first The chaotic sequence values ​​from the next iteration are used to enhance the uniformity of the initial distribution of the population. S23. The fitness function is determined by minimizing the mean squared error of the BP neural network prediction. The fitness function is as follows: ； in, The fitness value is the mean squared error of the prediction. Sample size For the first The true value of each sample For the first Predicted values ​​for each sample; By iteratively optimizing the behaviors of the discoverer, follower, and watcher in the sparrow search algorithm, the mean square error of the prediction is minimized, and the optimal initial weights and thresholds of the BP neural network are obtained. S24. For discrete pollutant indicators, one-hot encoding is used to convert them into numerical features before inputting them into the network; for continuous pollutant indicators, normalization is applied to the [0,1] interval. The normalization formula is as follows: ； in, The original value, This is the minimum value of the indicator. This is the maximum value of the indicator; The optimized BP neural network structure parameters are as follows: The number of input layer nodes equals the total number of pollutant indicators and environmental monitoring auxiliary indicators. The number of hidden layer nodes is determined by an empirical formula. Sure; in, The number of nodes in the input layer. This represents the number of nodes in the output layer. It is a constant; the number of output layer nodes is 1, which is used to output the integrated environmental protection feature value after fusion.

4. The method for environmental protection supervision and management of engineering project construction according to claim 3, characterized in that, In S3, a multi-dimensional data correlation model between construction procedures and environmental impacts is established based on an improved grey relational analysis algorithm. The multi-source heterogeneous data preprocessed in S2 and the construction procedure parameters are input into the model to calculate the correlation degree between each construction procedure and each pollutant indicator. The formula is as follows: ； in, For the first The pollutant index and the first The degree of correlation between each construction process; For the preprocessed first The values ​​of each pollutant indicator; For the first The characteristic parameters of each construction process include process duration, equipment power, and material consumption. The resolution coefficient is set to 0.5 to mitigate the influence of the maximum absolute difference. It is the minimum absolute difference between the two layers, that is, the minimum value of the difference between all indicators and all process parameters; It is the maximum absolute difference between the two layers, that is, the maximum value of the difference between all indicators and all process parameters; Set a correlation threshold, when When the correlation threshold is ≥, determine the first The construction process is the first... The main pollution sources for each pollutant indicator; for discrete pollutant indicators, they need to be converted into standardized process characteristic parameters before the correlation calculation.

5. The method for environmental supervision and management of engineering project construction according to claim 4, characterized in that, S4 specifically refers to: Three levels of early warning thresholds are set for each pollutant indicator. The pollutant indicators after S2 preprocessing are compared with the early warning thresholds. Combined with the pollution source tracing results of S3, early warning information including the type of exceeding indicator, value, and related process is generated and pushed to relevant responsible persons through multiple channels. The content and strategies for pushing information are optimized by linking environmental monitoring auxiliary indicators.

6. The method for environmental protection supervision and management of engineering project construction according to claim 5, characterized in that, In S4, the criteria for classifying the three-level warning thresholds are as follows: A Level 1 warning is issued when the monitored value exceeds the standard value by 10% to 20%, which is considered a slight exceedance. A Level II warning is issued when the monitored value exceeds the standard value by 20% to 50%, which is considered a moderate exceedance. A Level 3 warning is issued when the monitored value exceeds the standard value by more than 50%, which is considered a severe exceedance. The warning information includes the name of the indicator exceeding the standard, the value exceeding the standard, the duration of the exceeding the standard, the related construction procedures, targeted rectification suggestions, and the rectification deadline; The push channels include SMS, system pop-ups, and sound and light alarms, and the target audience is the on-site manager of the construction unit, environmental protection supervisors, and project supervisors.

7. The method for environmental supervision and management of engineering project construction according to claim 6, characterized in that, In S4, the priority for pushing indicators related to dust pollutants is as follows: When any level of warning is triggered, the audible and visual alarm at the construction site will immediately activate the graded audible and visual signals: yellow for Level 1, orange for Level 2, and red for Level 3, continuing until rectification is initiated. Simultaneously, pop-up windows will appear on the terminals of the construction unit, environmental protection supervision department, and project supervision, displaying the pollutant index exceeding the standard, real-time values, related processes, and current wind speed. A text message will be sent to the responsible person within 1 minute, and the related processes will immediately take spray dust suppression measures. For noise pollutant indicators, the priority for push notifications is as follows: When an alert is triggered during non-resident rest periods, a notification is sent via pop-up windows, SMS messages, and audible and visual alarms across all terminals. The pop-up window includes noise frequency distribution data and related procedures. When an alert is triggered during resident rest periods, the on-site audible and visual alarms are turned off, and only a highlighted pop-up window and a red-highlighted emergency nighttime alert SMS message are sent, along with a mandatory rectification suggestion to immediately suspend high-noise procedures. For water quality pollutant indicators, the priority for push notifications is as follows: When an alert is triggered, environmental regulators and project supervisors will see a pop-up window displaying comprehensive pollutant index data, historical trend curves, related processes, and potential hazards; SMS messages will include pollutant indices exceeding water quality standards, their values, and a download link for a rectification plan template; the construction technical supervisor will receive on-site notifications and simultaneously obtain real-time monitoring locations of water quality sensors to pinpoint the discharge point. For solid waste pollutant indicators, the priority for push notifications is as follows: When an alert is triggered, a pop-up window on all terminals displays the precise coordinates of the solid waste storage area, the percentage exceeding the standard, the type of solid waste, and the associated processes; an SMS message is sent with the site number of the storage area; and the project supervision unit simultaneously generates a paper alert notification for rectification review and filing. For auxiliary indicators of environmental monitoring, warnings are not triggered independently, but are only pushed out in conjunction with pollutant indicators, with the following priority: When the wind speed is ≥5m / s, the linkage dust pollutant index will be pushed, indicating that strong winds can easily aggravate dust spread. Please reinforce the covering and start the spray equipment. When the humidity is ≤30%, the indicator for solid waste pollutants will be pushed out, indicating that dust is likely to be generated in dry weather, and please increase the frequency of watering.

8. The method for environmental protection supervision and management of engineering project construction according to claim 7, characterized in that, In S5, the consortium blockchain nodes include construction unit nodes, environmental protection regulatory department nodes, supervision unit nodes, and third-party testing agency nodes. The construction unit node is used to upload the rectification plan and real-time pollutant indicators and environmental monitoring auxiliary indicator data during the rectification process; Environmental regulatory authorities are responsible for reviewing the relevance and feasibility of rectification plans. The supervisory unit's checkpoint is used to verify the compliance status of pollutant indicators after rectification; Third-party testing agency nodes are used to upload independent reports on all pollutant indicators. Each node adopts the PBFT consensus mechanism and has data read / write permissions and consensus verification permissions.

9. A construction environmental protection supervision and management system for engineering projects, used to implement the method described in any one of claims 1-8, characterized in that, include: The data acquisition and communication module is used to collect multi-source heterogeneous environmental data through multiple sensors and monitoring equipment, and transmit the raw multi-source heterogeneous data to the cloud data center through the 5G communication module and MQTT protocol. The data fusion and preprocessing module has a built-in BP neural network optimized by the improved sparrow search algorithm, which is used to perform differentiated preprocessing and feature fusion on multi-source heterogeneous environmental protection data and output comprehensive environmental protection feature values. The pollution source tracing and analysis module establishes a multi-dimensional data association model based on an improved grey relational analysis algorithm, calculates and locates the correlation between construction procedures and pollutant indicators, as well as the main pollution sources. The intelligent early warning and push module has a built-in hierarchical early warning mechanism, which is used to generate and push early warning information containing targeted rectification suggestions through multiple channels based on the level of exceeding the standard and the results of pollution source tracing. The blockchain closed-loop supervision module is used to build a consortium blockchain network to track and store data throughout the entire process, from the submission of rectification plans and progress implementation to the review of results.

10. The construction environmental protection supervision and management system for engineering projects according to claim 9, characterized in that, The data fusion and preprocessing module includes: The computational processing unit is used to execute the iterative optimization process of the improved sparrow search algorithm to determine the optimal initial weights and thresholds of the BP neural network; The data standardization unit is used to perform one-hot encoding on discrete pollutant indicators and normalize continuous pollutant indicators to the [0,1] interval. The blockchain closed-loop supervision module includes a verification and audit unit. Based on the consensus mechanism of the consortium blockchain nodes, it cross-verifies the rectification-related data and reports submitted by each node and generates audit traceability records.

Images