Shield muck heavy metal pollution intelligent sorting system and method based on multi-parameter online detection

Abstract

The invention discloses a shield muck heavy metal pollution intelligent sorting system and method based on multi-parameter online detection, and relates to the field of shield muck sorting, the shield muck heavy metal pollution intelligent sorting system comprises a muck sample pretreatment module, an online detection module, a pollution evaluation analysis module, a sorting execution module and a linkage control feedback module; the muck sample pretreatment module is used for separating coarse particles with the particle size larger than 50 mm and impurities in shield muck through a vibrating screen. According to the method, the initial value of the fine-grained soil content is detected in real time, data is automatically corrected based on a preset intelligent densimeter and a traditional densimeter difference correction model, rapid measurement is achieved, the detection efficiency is improved, a dynamic correlation model of fine-grained soil component changes is constructed by combining auxiliary parameters such as the muck weight, the water content and coarse grain grading, and the detection accuracy is improved. And the working condition adaptation module is linked to automatically improve the fine-grained soil detection frequency when the stratum mutation is identified by the shield PLC parameters, so that the data is ensured to closely follow the stratum change.

Description

Technical Field

[0001] This invention relates to the field of shield tunneling excavation soil sorting technology, specifically to an intelligent sorting system and method for heavy metal pollution in shield tunneling excavation soil based on multi-parameter online detection. Background Technology

[0002] Tunnel boring machine (TBM) construction is a core technology for underground space development such as urban rail transit and tunnel engineering. However, each kilometer of TBM excavation will generate about 30,000 cubic meters of excavated soil. These excavated soils generally have heavy metal pollution problems, so it is necessary to conduct heavy metal pollution testing and separation treatment on the TBM excavated soils.

[0003] Currently, traditional methods for detecting the fine soil content based on densitometers are time-consuming, require highly skilled operators, and cannot adapt to rapidly changing geological conditions. However, during tunnel boring machine (TBM) excavation, changes in the composition of fine soil in the excavated soil not only affect the construction measures for excavated soil improvement during the excavation process, but also affect the sorting operations in the subsequent excavated soil reuse stage. Summary of the Invention

[0004] The purpose of this invention is to provide an intelligent sorting system and method for heavy metal pollution in shield tunnel excavation soil based on multi-parameter online detection, in order to solve the problems mentioned in the background art. The traditional method for detecting fine soil content based on densitometers is time-consuming, requires highly skilled operators, and cannot adapt to rapidly changing geological conditions. However, during shield tunneling, changes in the composition of fine soil in the excavation soil not only affect the construction measures for excavation soil improvement during tunneling, but also affect the sorting operation in the subsequent excavation soil reuse stage.

[0005] To achieve the above objectives, the present invention provides the following technical solution: an intelligent sorting system for heavy metal pollution in shield tunnel slag based on multi-parameter online detection, comprising a slag sample pretreatment module, an online detection module, a pollution assessment and analysis module, a sorting execution module, and a linkage control feedback module;

[0006] The slag sample pretreatment module is used to separate coarse particles and impurities with a particle size >50mm in the shield tunnel slag through a vibrating screen, and to separate the undersize material to separate the mud-water mixture and solid slag.

[0007] The online detection module is used to detect the mud-water mixture and solid soil samples separated by the slag sample pretreatment module, wherein the mud-water is tested for fine-grained soil and the solid soil is tested for heavy metals.

[0008] The pollution assessment and analysis module is used to acquire online detection data in real time, build a dynamic database by combining historical data, mine the correlation between fine soil content and heavy metal pollution through neural network learning algorithms, establish a heavy metal pollution model, and classify the pollution level of slag and soil.

[0009] The sorting execution module achieves precise sorting through automated equipment based on the pollution level results from the analysis unit.

[0010] The linkage control feedback module dynamically optimizes the heavy metal pollution model by transmitting the sorting results back to the data unit, and dynamically adjusts the heavy metal detection frequency according to the changes in the strata.

[0011] Preferably, the slag sample pretreatment module includes a coarse screening and impurity removal module and a mud-water separation module;

[0012] The coarse screening module separates coarse particles and impurities from the slag and soil through a vibrating screen. The undersize material enters the next step of processing and includes fine soil, small and medium-sized particles and muddy water.

[0013] The mud-water separation module is used to separate the undersize material to obtain a mud-water mixture containing suspended fine soil particles and solid slag.

[0014] Preferably, the online detection module includes a fine-grained soil content detection module, a heavy metal pollution detection module, and an auxiliary parameter detection module;

[0015] The fine-grained soil content detection module is used to introduce the mud-water mixture into the sedimentation tank, use an intelligent densitometer to detect the density of the mixture in real time, invert the initial value of the fine-grained soil content, and automatically correct the detection error based on the difference model between the intelligent densitometer and the traditional densitometer established by the indoor test.

[0016] The heavy metal pollution detection module is used to mix and homogenize solid slag using a screw conveyor, detect the heavy metal content in the homogenized slag using an XRF spectrometer, and simultaneously acquire the distribution of particles of various sizes in the solid particles using a 3D point cloud scanner, and perform particle size coupling analysis by correlating the fine soil content.

[0017] The auxiliary parameter detection module uses an infrared hygrometer to detect the moisture content of the slag and a laser particle size analyzer to detect the distribution of particles of different sizes in the coarse particles, providing auxiliary parameters for subsequent pollution assessment.

[0018] Preferably, the pollution assessment and analysis module includes a data storage module, a pollution assessment module, and a real-time correction module;

[0019] The data storage module constructs a waste soil parameter data repository based on the waste soil testing data and stores the latest data in real time;

[0020] The pollution assessment module uses a random forest algorithm to mine the correlation between fine-grained soil content and heavy metal concentration to establish a pollution early warning threshold.

[0021] The real-time correction module is used to automatically compare laboratory test data every 50m of tunneling, optimize model parameters, and ensure long-term stability.

[0022] Preferably, the sorting execution module includes a graded conveying module, a physical sorting module, and a sealing control module;

[0023] The graded conveying module uses a three-channel belt conveyor to switch conveying paths according to the pollution level signal. The paths include a resource recovery warehouse, a solidification workshop, and a hazardous waste warehouse.

[0024] The physical sorting module is used to separate fine soil particles with a particle size of less than 0.075 mm from coarse particles using a vibrating screen for yellow slag.

[0025] The sealing control module is used to assist in implementing negative pressure ventilation and spray dust suppression operations in the hazardous waste bin to treat heavy metal dust.

[0026] Preferably, the linkage control feedback module includes a dynamic calibration module, a working condition adaptation module, and a visualization platform module;

[0027] The dynamic calibration module is used to compare and verify data from random sample detection. When an anomaly is detected, it automatically triggers the calibration program and calls the standard solution for calibration.

[0028] The working condition adaptation module is used to acquire tunneling parameters through the shield PLC, identify sudden changes in strata, and automatically adjust the detection frequency.

[0029] The visualization platform module is used to display the flow direction of construction waste, pollution heat map and equipment status in real time through BIM and GIS interfaces, and supports mobile terminal early warning.

[0030] A smart sorting method for heavy metal pollution in tunnel boring machine (TBM) slag based on multi-parameter online detection includes the following steps:

[0031] S1. Based on the characteristics of coarse particles and slurry mixture in shield tunnel slag, a continuous pretreatment is carried out. The pretreatment includes preliminary screening and impurity removal and slurry-solid phase separation treatment to obtain slurry mixture samples and solid slag samples.

[0032] S2. Based on the pretreated samples, the adsorption capacity of heavy metals, the concentration of heavy metals and auxiliary parameters associated with the characteristics of fine-grained soil were simultaneously detected.

[0033] S3. Integrate the characteristics of fine-grained soil, heavy metal concentration and auxiliary parameters and perform data correction. A dynamic classification of pollution levels is achieved through a Bayesian optimized long short-term memory network algorithm.

[0034] S4. Based on the pollution level results from the previous step, the waste soil is graded and sorted through linkage control.

[0035] S5. During the tunnel boring machine (TBM) construction, three sets of samples are randomly selected every 50 meters for laboratory testing and comparison. If the deviation between the online test and the laboratory data is greater than 5%, the parameters of the joint testing model are automatically updated and the weights of the machine learning model are adjusted.

[0036] S6. By analyzing the correlation between real-time parameters and historical data of the tunnel boring machine, the trend of geological changes can be identified 5 to 10 minutes in advance, and the detection model and sorting strategy can be automatically switched. By simulating the system operation under different geological strata in a virtual environment, potential faults can be predicted and equipment parameters can be adjusted in advance.

[0037] Preferably, in step S1, the solid slag is stirred and homogenized by a screw conveyor, the heavy metal content is detected by an XRF spectrometer, a secondary verification of high-concentration samples is automatically triggered by laser-induced breakdown spectroscopy, and moisture content and coarse particle size distribution data are collected simultaneously by an infrared hygrometer and a laser particle size analyzer.

[0038] Preferably, in step S3, the heavy metal concentration correction automatically lowers the heavy metal concentration assessment weight for samples with a coarse particle size distribution greater than 60% based on the coarse particle size distribution data.

[0039] Compared with the prior art, the beneficial effects of the present invention are:

[0040] 1. In this invention, the initial value of fine soil content is detected in real time, and the data is automatically corrected based on the preset intelligent density meter and traditional density meter difference correction model, so as to achieve rapid measurement, improve detection efficiency, provide real-time data support for soil improvement measures during shield tunneling, and avoid mismatch of tunneling parameters caused by sudden changes in fine soil content.

[0041] 2. In this invention, by combining auxiliary parameters such as the weight of the slag, moisture content, and coarse particle size distribution, a dynamic correlation model of the changes in fine soil components is constructed, and the working condition adaptation module is linked. When the sudden change in the stratum is identified by the shield tunneling PLC parameters, the detection frequency of fine soil is automatically increased to ensure that the data closely follows the changes in the stratum.

[0042] 3. In this invention, the correlation between fine soil content data and heavy metal detection data is deeply correlated, and the correlation between the two is mined by the random forest algorithm of the pollution assessment module, providing accurate basis for the sorting execution module, avoiding the resource waste caused by traditional direct sorting, and improving the resource utilization rate of slag soil.

[0043] 4. In this invention, the real-time correction module automatically compares the laboratory test data every 50m of tunneling, dynamically optimizes the difference model parameters, and ensures that the long-term error of fine soil content detection remains within a small range. This solves the problem of data stability decaying over time in traditional methods, and provides a continuous and reliable data benchmark for the analysis of fine soil characteristics throughout the entire life cycle of shield tunnel excavation soil, from tunneling to final reuse. Attached Figure Description

[0044] Figure 1 This is a system block diagram of the intelligent sorting system for heavy metal pollution in shield tunnel slag based on multi-parameter online detection according to the present invention.

[0045] Figure 2 This is a flowchart of the intelligent sorting method for heavy metal pollution in shield tunnel slag based on online detection of multiple parameters, according to the present invention. Detailed Implementation

[0046] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0047] Reference Figure 1 and Figure 2 As shown: A smart sorting system for heavy metal pollution in shield tunnel slag based on multi-parameter online detection includes a slag sample pretreatment module, an online detection module, a pollution assessment and analysis module, a sorting execution module, and a linkage control feedback module;

[0048] The slag sample pretreatment module is used to separate coarse particles and impurities with a particle size >50mm from the slag of the tunnel boring machine using a vibrating screen, and to separate the undersize material to separate the mud-water mixture and solid slag; specifically, it includes a coarse screening module for impurity removal and a mud-water separation module;

[0049] The coarse screening and impurity removal module separates coarse particles and impurities from the slag and soil through a vibrating screen, reducing interference from subsequent detection modules and improving detection accuracy. The undersize material enters the next step of processing, which includes fine soil, small and medium-sized particles, and muddy water.

[0050] The mud-water separation module is used to separate the undersize material to obtain a mud-water mixture containing suspended fine soil particles and solid slag, providing a pure sample for subsequent testing;

[0051] The online detection module is used to detect the mud-water mixture and solid waste samples separated by the waste sample pretreatment module. The mud-water mixture is tested for fine-grained soil, and the solid waste sample is tested for heavy metals. Specifically, it includes a fine-grained soil content detection module, a heavy metal pollution detection module, and an auxiliary parameter detection module.

[0052] The fine-grained soil content detection module is used to introduce mud-water mixture into a sedimentation tank, use an intelligent densitometer to detect the density of the mixture in real time, invert the initial value of fine-grained soil content, and automatically correct detection errors based on the difference model between the intelligent densitometer and the traditional densitometer established by indoor tests, thereby improving detection accuracy.

[0053] The heavy metal pollution detection module is used to mix and homogenize solid slag using a screw conveyor, and then detect the heavy metal content in the homogenized slag using an XRF spectrometer. A non-contact scanning of the solid slag is performed using an Rh target X-ray tube (60kV / 150mA), with a single detection time of less than 60 seconds. The distribution of particles of various sizes in the solid particles is simultaneously acquired using a 3D point cloud scanner, and particle size coupling analysis is performed in conjunction with the fine soil content.

[0054] The auxiliary parameter detection module uses an infrared hygrometer to detect the moisture content of the slag and a laser particle size analyzer to detect the distribution of particles of different sizes in the coarse particles, providing auxiliary parameters for subsequent pollution assessment and offering multi-dimensional data support.

[0055] The pollution assessment and analysis module is used to acquire online detection data in real time, build a dynamic database by combining historical data, mine the correlation between fine soil content and heavy metal pollution through neural network learning algorithms and establish a heavy metal pollution model to classify the pollution level of construction waste; specifically, it includes a data storage module, a pollution assessment module and a real-time correction module.

[0056] The data storage module builds a data repository for waste soil parameters based on waste soil testing data and stores the latest data in real time. It also stores historical data through a dynamic database to facilitate long-term trend analysis.

[0057] The pollution assessment module uses the random forest algorithm to mine the correlation between fine-grained soil content and heavy metal concentration to establish pollution early warning thresholds;

[0058] The real-time correction module is used to automatically compare laboratory test data every 50m of tunneling, optimize model parameters, and ensure long-term stability. By correcting the model parameters, it ensures that the long-term test error is <5%.

[0059] The sorting execution module, based on the contamination level results from the analysis unit, achieves precise sorting through automated equipment; specifically, it includes a graded conveying module, a physical sorting module, and a sealing control module.

[0060] The graded conveying module uses a three-channel belt conveyor to switch conveying paths according to the pollution level signal. The paths include the resource recovery bin, the solidification workshop, and the hazardous waste bin.

[0061] The physical sorting module is used to separate fine soil particles with a particle size of less than 0.075mm from coarse particles in yellow slag using a vibrating screen, thereby reducing the amount of hazardous waste generated.

[0062] The sealing control module is used to assist in the negative pressure ventilation and spray dust suppression operation of the hazardous waste bin to treat heavy metal dust;

[0063] The linkage control feedback module dynamically optimizes the heavy metal pollution model by transmitting the sorting results back to the data unit, and dynamically adjusts the heavy metal detection frequency according to the changes in the strata; specifically, it includes a dynamic calibration module, an operating condition adaptation module, and a visualization platform module.

[0064] The dynamic calibration module is used to compare and verify data from random sample tests. When an anomaly is detected, it automatically triggers the calibration program and calls the standard solution for calibration to ensure test accuracy.

[0065] The working condition adaptation module is used to obtain tunneling parameters through the shield PLC, identify sudden changes in strata, and automatically adjust the detection frequency. By adjusting the detection frequency and model parameters in real time, it solves the problem of poor adaptability of traditional systems in complex strata.

[0066] The visualization platform module is used to display the flow direction of construction waste, pollution heat maps and equipment status in real time through BIM and GIS interfaces. It supports mobile terminal early warning and provides data support for similar projects in the future through dynamic database and BIM / GIS visualization platform.

[0067] A smart sorting method for heavy metal pollution in tunnel boring machine (TBM) slag based on multi-parameter online detection includes the following steps:

[0068] Step 1: Based on the characteristics of coarse particles and slurry mixture in shield tunnel excavation soil, continuous pretreatment is carried out. The pretreatment includes preliminary screening and impurity removal and slurry-solid phase separation treatment to obtain slurry mixture samples and solid excavation soil samples.

[0069] Among them, solid slag is mixed and homogenized by a screw conveyor, heavy metal content is detected by XRF spectrometer, high concentration samples are automatically triggered for secondary verification by laser-induced breakdown spectroscopy, and moisture content and coarse particle size distribution data are collected simultaneously by infrared hygrometer and laser particle size analyzer.

[0070] Step 2: Based on the pretreated samples, simultaneously detect the heavy metal adsorption capacity, heavy metal concentration, and auxiliary parameters associated with the characteristics of fine-grained soil.

[0071] Step 3: Integrate the characteristics of fine-grained soil, heavy metal concentration, and auxiliary parameters, and perform data correction. Then, use a Bayesian-optimized long short-term memory network algorithm to dynamically classify the pollution level.

[0072] Among them, the heavy metal concentration correction is based on the coarse particle gradation data, and automatically lowers the heavy metal concentration assessment weight for samples with a coarse particle ratio of >60%.

[0073] Step 4: Based on the pollution level results from the previous step, implement graded sorting of construction waste through linkage control.

[0074] During the grading and sorting process, when solid waste is transported to the sorting node via a belt conveyor, the magnetic levitation diversion valve (response time 0.05 seconds) switches paths according to the pollution level signal (Level I / II / III):

[0075] Grade I waste soil → Resource recovery storage bin.

[0076] Grade II slag → High-frequency vibrating screen (2mm screen hole) separates coarse particles (for direct reuse) from fine soil particles (for separate solidification), reducing the amount of solidification material used by 40%-60%.

[0077] Grade III waste soil → Sealed hazardous waste storage bin, with an embedded laser dust particle counter to monitor dust concentration in real time and automatically adjust the exhaust power (1000~1500m³ / h). 3 / h), ensuring zero spillover;

[0078] Deep treatment of fine-grained soil: For fine-grained soil corresponding to Class II / III slag, iron powder adsorption and superconducting magnetic separation technology are used: iron powder is added (the amount is 0.5% of the dry matter mass), and the iron powder adsorbed with heavy metals is recovered by magnetic separation, so as to achieve the targeted removal of heavy metals such as arsenic (the concentration is reduced to below 0.01 mg / L).

[0079] Step 5: During the tunnel boring machine (TBM) construction, three sets of samples are randomly selected every 50 meters for laboratory testing and comparison. If the deviation between the online test and laboratory data is greater than 5%, the parameters of the joint testing model are automatically updated and the weights of the machine learning model are adjusted.

[0080] Step 6: By analyzing the correlation between real-time parameters and historical data of the tunnel boring machine, identify the trend of geological changes 5 to 10 minutes in advance, automatically switch the detection model and sorting strategy, and predict potential faults by simulating the system operation under different geological strata in a virtual environment, and adjust the equipment parameters in advance.

[0081] This invention solves the problems of missed detection of light elements and misjudgment of fine-grained soil content in traditional single detection methods by dynamically correcting the fine-grained soil content, verifying heavy metals using dual-spectrum analysis, and coupling analysis of auxiliary parameters. It improves the accuracy of pollution level identification, and provides real-time response from waste soil pretreatment to test result output to adapt to the construction schedule. Through a physical sorting strategy of direct reuse of coarse particles and targeted treatment of fine-grained soil, it reduces the amount of solidification materials used for lightly polluted waste soil. Heavily polluted waste soil is directed to a hazardous waste storage area, avoiding resource waste caused by direct treatment and improving the resource utilization rate of waste soil. A three-channel conveyor belt and dynamic diversion design allow for flexible switching of treatment paths according to the pollution level. To reduce subsequent disposal costs, the correlation analysis between fine-grained soil content and heavy metal adsorption identifies high-pollution-risk fine-grained soil components. Targeted sealing control is then implemented to ensure that dust concentration in the hazardous waste storage meets air pollutant emission standards. Combining shield tunneling PLC excavation parameters with 3D point cloud particle size analysis, the system automatically identifies different geological characteristics such as sand and clay layers, and adjusts the detection frequency and model parameters in real time. This addresses the poor adaptability of traditional systems in complex geological formations. Every 50m of excavation, the system automatically compares laboratory data and calibrates the model. The pollution assessment threshold is iteratively optimized through neural network algorithms to ensure long-term operational accuracy and reduce manual intervention costs.

[0082] Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An intelligent sorting system for heavy metal pollution in tunnel boring machine excavated soil based on multi-parameter online detection, characterized in that: It includes a waste soil sample pretreatment module, an online detection module, a pollution assessment and analysis module, a sorting execution module, and a linkage control feedback module; The slag sample pretreatment module is used to separate coarse particles and impurities with a particle size >50mm in the shield tunnel slag through a vibrating screen, and to separate the undersize material to separate the mud-water mixture and solid slag. The online detection module is used to detect the mud-water mixture and solid soil samples separated by the slag sample pretreatment module, wherein the mud-water is tested for fine-grained soil and the solid soil is tested for heavy metals. The pollution assessment and analysis module is used to acquire online detection data in real time, build a dynamic database by combining historical data, mine the correlation between fine soil content and heavy metal pollution through neural network learning algorithms, establish a heavy metal pollution model, and classify the pollution level of slag and soil. The sorting execution module achieves precise sorting through automated equipment based on the pollution level results from the analysis unit. The linkage control feedback module dynamically optimizes the heavy metal pollution model by transmitting the sorting results back to the data unit, and dynamically adjusts the heavy metal detection frequency according to the changes in the strata.

2. The intelligent sorting system for heavy metal pollution in shield tunnel slag based on multi-parameter online detection according to claim 1, characterized in that: The soil sample pretreatment module includes a coarse screening and impurity removal module and a mud-water separation module; The coarse screening module separates coarse particles and impurities from the slag and soil through a vibrating screen. The undersize material enters the next step of processing and includes fine soil, small and medium-sized particles and muddy water. The mud-water separation module is used to separate the undersize material to obtain a mud-water mixture containing suspended fine soil particles and solid slag.

3. The intelligent sorting system for heavy metal pollution in shield tunnel slag based on multi-parameter online detection according to claim 1, characterized in that: The online detection module includes a fine-grained soil content detection module, a heavy metal pollution detection module, and an auxiliary parameter detection module; The fine-grained soil content detection module is used to introduce the mud-water mixture into the sedimentation tank, use an intelligent densitometer to detect the density of the mixture in real time, invert the initial value of the fine-grained soil content, and automatically correct the detection error based on the difference model between the intelligent densitometer and the traditional densitometer established by the indoor test. The heavy metal pollution detection module is used to mix and homogenize solid slag using a screw conveyor, detect the heavy metal content in the homogenized slag using an XRF spectrometer, and simultaneously acquire the distribution of particles of various sizes in the solid particles using a 3D point cloud scanner, and perform particle size coupling analysis by correlating the fine soil content. The auxiliary parameter detection module uses an infrared hygrometer to detect the moisture content of the slag and a laser particle size analyzer to detect the distribution of particles of different sizes in the coarse particles, providing auxiliary parameters for subsequent pollution assessment.

4. The intelligent sorting system for heavy metal pollution in shield tunnel slag based on multi-parameter online detection according to claim 1, characterized in that: The pollution assessment and analysis module includes a data storage module, a pollution assessment module, and a real-time correction module; The data storage module constructs a waste soil parameter data repository based on the waste soil testing data and stores the latest data in real time; The pollution assessment module uses a random forest algorithm to mine the correlation between fine-grained soil content and heavy metal concentration to establish a pollution early warning threshold. The real-time correction module is used to automatically compare laboratory test data and optimize model parameters every 50m of tunneling.

5. The intelligent sorting system for heavy metal pollution in shield tunnel slag based on multi-parameter online detection according to claim 1, characterized in that: The sorting execution module includes a graded conveying module, a physical sorting module, and a sealing control module; The graded conveying module uses a three-channel belt conveyor to switch conveying paths according to the pollution level signal. The paths include a resource recovery warehouse, a solidification workshop, and a hazardous waste warehouse. The physical sorting module is used to separate fine soil particles with a particle size of less than 0.075 mm from coarse particles using a vibrating screen for yellow slag. The sealing control module is used to assist in implementing negative pressure ventilation and spray dust suppression operations in the hazardous waste bin to treat heavy metal dust.

6. The intelligent sorting system for heavy metal pollution in shield tunnel slag based on multi-parameter online detection according to claim 1, characterized in that: The linkage control feedback module includes a dynamic calibration module, an operating condition adaptation module, and a visualization platform module. The dynamic calibration module is used to compare and verify data from random sample detection. When an anomaly is detected, it automatically triggers the calibration program and calls the standard solution for calibration. The working condition adaptation module is used to acquire tunneling parameters through the shield PLC, identify sudden changes in strata, and automatically adjust the detection frequency. The visualization platform module is used to display the flow direction of construction waste, pollution heat map and equipment status in real time through BIM and GIS interfaces, and supports mobile terminal early warning.

7. A smart sorting method for heavy metal pollution in shield tunnel slag based on multi-parameter online detection, characterized in that, The intelligent sorting system for heavy metal pollution in tunnel boring machine slag based on multi-parameter online detection, as described in any one of claims 1-6, includes the following steps: S1. Based on the characteristics of coarse particles and slurry mixture in shield tunnel slag, a continuous pretreatment is carried out. The pretreatment includes preliminary screening and impurity removal and slurry-solid phase separation treatment to obtain slurry mixture samples and solid slag samples. S2. Based on the pretreated samples, the adsorption capacity of heavy metals, the concentration of heavy metals and auxiliary parameters associated with the characteristics of fine-grained soil were simultaneously detected. S3. Integrate the characteristics of fine-grained soil, heavy metal concentration and auxiliary parameters and perform data correction. A dynamic classification of pollution levels is achieved through a Bayesian optimized long short-term memory network algorithm. S4. Based on the pollution level results from the previous step, the waste soil is graded and sorted through linkage control. S5. During the tunnel boring machine (TBM) construction, three sets of samples are randomly selected every 50 meters for laboratory testing and comparison. If the deviation between the online test and the laboratory data is greater than 5%, the parameters of the joint testing model are automatically updated and the weights of the machine learning model are adjusted. S6. By analyzing the correlation between real-time parameters and historical data of the tunnel boring machine, the trend of geological changes can be identified 5 to 10 minutes in advance, and the detection model and sorting strategy can be automatically switched. By simulating the system operation under different geological strata in a virtual environment, potential faults can be predicted and equipment parameters can be adjusted in advance.

8. The intelligent sorting method for heavy metal pollution in shield tunnel slag based on multi-parameter online detection according to claim 7, characterized in that: In step S1, the solid slag is stirred and homogenized by a screw conveyor, and the heavy metal content is detected by an XRF spectrometer. For high-concentration samples, laser-induced breakdown spectroscopy is automatically triggered for secondary verification. Simultaneously, moisture content and coarse particle size distribution data are collected by an infrared hygrometer and a laser particle size analyzer.

9. The intelligent sorting method for heavy metal pollution in shield tunnel slag based on multi-parameter online detection according to claim 7, characterized in that: In step S3, the heavy metal concentration correction automatically lowers the heavy metal concentration assessment weight for samples with a coarse particle size distribution greater than 60% based on the coarse particle size distribution data.

Images