Automatic monitoring system for tool wear in full-water-covering complex changeable stratum

Abstract

The invention discloses an automatic monitoring system for cutter wear of a full-water-covering complex changeable stratum, and belongs to the technical field of intelligent monitoring of shield construction, the automatic monitoring system comprises a sensing layer, a transmission layer, a platform layer and an application layer which are connected in sequence, and the sensing layer collects multi-source data such as cutter operation, cutter head panel wear and working condition parameters; the transmission layer performs edge preprocessing on the data and then transmits the data to the platform layer through an anti-interference double link; the platform layer eliminates interference such as muddy water corrosion, scouring and temperature drift through coupling abrasion calibration, calculates the real cutting abrasion amount of a cutter, accurately recognizes the fault type of the cutter by fusing multi-source state features, further evaluates the health state of the cutter in combination with cutter panel abrasion data, and generates a multi-scene linkage management and control strategy. And the application layer is used for visual display and early warning. According to the invention, the interference of a full-water-covering severe muddy water environment on monitoring data is effectively overcome, and high-precision monitoring and intelligent management and control of the health state of the shield cutter head and the cutter are realized.

Description

Technical Field

[0001] This invention belongs to the field of intelligent monitoring and intelligent tunnel construction technology for shield tunneling, specifically relating to an automatic monitoring system for cutter wear in complex and variable strata with full water coverage. Background Technology

[0002] Tunnel boring machine (TBM) construction is the core construction method for underground space development. Among them, complex and variable strata with full water coverage are typical high-risk conditions for TBM construction. These strata generally have characteristics such as high water pressure, strong corrosion, uneven hardness, and development of karst / isolated boulders. The cutterhead of the TBM is subjected to the coupled effects of impact from hard rock, corrosion from mud and water, and scouring from mud and sand. The wear rate is 3 to 5 times that of ordinary strata, and failures such as uneven wear, chipping, and breakage are prone to occur. Moreover, manual inspection under high water pressure conditions poses an extremely high risk of water inrush and collapse, resulting in high operating costs and low efficiency. Therefore, there is an urgent need for automatic cutter wear monitoring technology to achieve non-opening cutter status perception.

[0003] Existing shield tunneling cutter wear monitoring technologies mostly involve deploying wear sensors on the cutters to collect thickness data, which is then transmitted wirelessly or wired to the ground for analysis. This reduces the frequency of opening the tunnel for inspection to some extent, but it still reveals the following shortcomings when dealing with complex and variable strata covered by water: First, the transmission stability is poor. The inside of the cutter head is in a dynamic rotating environment with strong metal shielding and high-pressure mud and water. Traditional communication methods are prone to data loss due to signal attenuation.

[0004] Second, severe environmental interference leads to inaccurate calculations of actual wear. In fully aquifer strata, the tunnel boring machine cutterhead is immersed in a high-pressure slurry environment for extended periods. Existing systems typically use a simple calculation method—subtracting the measured remaining thickness from the initial cutter thickness—to obtain the wear amount. This method fails to consider the thickness loss and measurement errors caused by the unique chemical corrosion of slurry, erosion by sediment, and temperature drift in fully aquifer strata. It cannot distinguish between the actual effective wear generated by the cutter's cutting operation and the ineffective thickness loss caused by environmental factors. Consequently, the monitoring data is severely distorted due to wear measurement errors and cannot accurately reflect the cutting life of the cutter.

[0005] Third, the diagnostic dimensions are limited, lacking multi-source collaboration and coordinated control. Most existing monitoring systems only display single data on wear measurement, failing to accurately distinguish between normal uniform wear of the cutter head and sudden failures such as uneven wear, shutdown, or chipping. Furthermore, cutter wear is often disconnected from cutterhead panel wear and shield tunneling conditions, lacking a global perspective health assessment model and unable to automatically generate targeted tunneling parameter adjustments and cutter replacement control strategies based on different scenarios.

[0006] In summary, there is an urgent need for an automatic tool wear monitoring system that can overcome the interference of mud and water corrosion and erosion, achieve stable data transmission, and integrate multi-source features for accurate fault identification and coordinated control. Summary of the Invention

[0007] To address the problems existing in the background technology, this invention proposes an automatic tool wear monitoring system for complex and variable formations with full water coverage. By constructing a dual-link anti-interference transmission mechanism, a dynamic wear calibration mechanism coupled with mud and water corrosion-erosion-temperature drift, and a multi-source feature fusion fault discrimination and tool head-tool collaborative evaluation and control mechanism, this system solves core technical problems such as high signal attenuation and packet loss rate under strong metal shielding and high-pressure mud and water envelopment in full water coverage formations, distortion of wear calculation under multi-source environmental interference, high rate of false and false fault diagnosis due to single diagnostic dimensions, and lack of global collaborative control capabilities. This system achieves joint optimization of data transmission reliability, wear measurement accuracy, fault identification accuracy, and full-scenario operation and maintenance management efficiency in tool wear monitoring.

[0008] To achieve the above objectives, the present invention adopts the following technical solution: An automatic monitoring system for tool wear in complex and variable formations with full water coverage includes a sensing layer, a transmission layer, a platform layer, and an application layer that are sequentially connected in communication, wherein: The sensing layer is used to collect multi-source status and working condition data under fully water-covered strata during shield tunneling construction. The multi-source status and working condition data includes cutter running status data, cutterhead panel wear data, and working condition parameters. The transport layer is used to receive multi-source status and operating condition data collected by the perception layer, perform edge preprocessing on the multi-source status and operating condition data, and transmit the preprocessed data to the platform layer through an anti-interference dual-link transmission architecture. The platform layer is used to store, analyze, make decisions about, and output control commands for the data uploaded from the transport layer, including: The data storage module is used to store preprocessed multi-source status and operating condition data, historical wear data and fault sample data; The coupled wear calibration module is used to eliminate the interference of mud and water corrosion, mud and sand erosion and temperature drift on wear measurement in fully water-covered formations, and to calculate the actual cutting wear of the tool. A multi-source feature fusion fault discrimination module is used to fuse multi-source state features to identify tool fault types; The tool and cutter collaborative evaluation module is used to evaluate the overall health status of the tool and cutter panel by combining the tool wear status with the tool and cutter panel wear data. The linkage control strategy generation module is used to generate corresponding control strategies based on the fault identification results and the health status of the tool turret; The application layer is used to display the analysis and decision-making results of the platform layer, including a visual monitoring interface, which presents the real-time operating status of the tool, the health status of the tool head, fault warning information, and wear statistical analysis data.

[0009] Specifically, the perception layer includes: Multi-parameter tool sensing unit: includes sensing components set for each tool to be monitored, used to collect three types of tool operating status data: remaining thickness, real-time rotation speed, and surface temperature; Cutterhead panel wear monitoring unit: includes several independent continuous detection oil channels arranged in different wear risk areas of the front and rear panels of the cutterhead. Each continuous detection oil channel is equipped with a pressure detection component at the end, which is used to identify the wear location and wear degree of the cutterhead panel through abnormal changes in oil channel pressure, and obtain cutterhead panel wear data. Working condition linkage acquisition unit: It connects to the shield tunneling main control system and the slurry monitoring system to synchronously acquire cutterhead speed, cutterhead torque, total thrust, penetration depth and slurry parameters, including the electrical conductivity, pH value, sand content and slurry flow rate of the slurry.

[0010] Specifically, the transmission layer includes an edge preprocessing module, several cutterhead relay nodes evenly spaced along the circumference of the cutterhead, and a receiving terminal fixed to the inner wall of the shield's front compartment, wherein: The edge preprocessing module is deployed in each tool head relay node and performs two-level processing on the received raw data from the sensing layer: the first step is to remove abnormal jump values ​​that exceed the preset range, and the second step is to perform mean aggregation on the high-frequency acquired data at a fixed sampling interval to reduce the amount of invalid data transmission. The transmission layer employs a two-stage transmission link to transmit data from the cutterhead to the shield: the first stage is a magnetic induction near-field transmission link from the multi-parameter sensing unit of the cutterhead to the relay node of the cutterhead. Each multi-parameter sensing unit of the cutterhead and the relay node of the cutterhead are equipped with a magnetic induction communication module, which uses the characteristic of alternating magnetic fields to penetrate metal and mud to transmit data; the second stage is a LoRa spread spectrum transmission link from the relay node of the cutterhead to the receiving terminal of the shield. The relay node of the cutterhead is equipped with a LoRa communication module, which uses spread spectrum modulation to compensate for signal attenuation caused by metal shielding.

[0011] Specifically, the process by which the coupled wear calibration module in the platform layer calculates the actual cutting wear of the tool is as follows: S11, Obtain Original wear thickness difference at time , The initial thickness of the cutting tool at the factory and The difference in remaining tool thickness collected by the multi-parameter sensing unit at all times; S12, Calculation The actual cutting wear of the tool at any time : ; The definitions and calculation methods for each correction item are as follows: (1) for The cumulative amount of mud and water corrosion and wear over time is calculated using the following formula: ; in, To determine the corrosion coefficient of the cutting tool material to be monitored, a pre-calibration was performed through a laboratory immersion test simulating the muddy water conditions of the target formation. for The mud-water corrosion intensity factor at any given time is calculated based on the mud-water parameters obtained from the working condition linkage acquisition unit: ; in, for The electrical conductivity of mud and water at any given time. for The pH value of the muddy water at all times; (2) for The cumulative amount of erosion and wear from sediment over time is calculated using the following formula: ; in, To determine the erosion coefficient of the cutting tool material to be monitored, a pre-calibration was performed through a laboratory erosion test simulating the mud conditions of the target formation. for The mud-water scour intensity factor at any given time is calculated based on the mud-water parameters obtained from the working condition linkage acquisition unit: ; in, for The slurry flow rate on the surface of the cutterhead at all times. for The sand content of the mud at any given time. The sediment hardness coefficient of the target stratum was pre-calibrated based on the project's geological exploration report; this item passed... The nonlinear enhancement term reflects the rising characteristics of scour wear under high sand content, which is suitable for the large fluctuation of sand content in fully water-covered strata. (3) for The temperature drift error at any given time is calculated using the following formula: ; in, The temperature drift coefficient of the sensor. for The tool surface temperature is collected by the multi-parameter sensing unit at all times. The reference temperature used to calibrate the sensor.

[0012] Specifically, the execution steps of the multi-source feature fusion fault discrimination module in the platform layer include: S21. Define four types of common tooling faults in fully water-covered formations to be identified, and construct an identification framework: ,in This represents normal wear and tear. Represents uneven wear. The signal is to stop. This indicates chipping or breakage; S22. Extract feature values ​​from four types of independent evidence, including: a. Evidence 1: Corresponding wear characteristics, the characteristic value being the ring-average rate of change of actual cutting wear. The calculation formula is: ; in, For continuous The actual increase in cutting wear of each tunneling ring The number of rings continuously excavated; b. Evidence 2: Corresponding rotational speed characteristics, the characteristic value being the matching coefficient between the tool rotational speed and the tool head rotational speed. The calculation formula is: ; in, For the real-time rotational speed of the cutting tool, This refers to the real-time rotational speed of the cutter head. c. Evidence 3: Corresponding temperature characteristics, the characteristic value being the temperature difference between the cutting tool and the tool head panel. The calculation formula is: ; in, The surface temperature of the cutting tool. This refers to the temperature of the corresponding area of ​​the cutter head. d. Evidence 4: Corresponding to tunneling parameter characteristics, the characteristic value is the deviation of the tunneling coupling factor. The calculation formula is: ; in, This is the actual tunneling coupling factor. The torque of the cutter head. For total thrust, For penetration, The standard tunneling coupling factor for the target stratum under the corresponding tunneling conditions is pre-calibrated using early-stage tunneling data or historical samples from the same stratum. S23. Calculate the basic probability distribution of each piece of evidence, using the following formula: ; in, For the first The characteristic values ​​of each piece of evidence These correspond to four types of evidence. , For the first The evidence corresponds to the first... The characteristic mean and characteristic standard deviation of the fault class are obtained through pre-training using historical fault samples from the target stratum. These correspond to four types of tool failures. For the first The dynamic confidence weights of each piece of evidence satisfy the following: The calculation formula is: ,in The initial weights are pre-calibrated and set based on historical fault identification accuracy. It is a dynamic confidence coefficient, which is negatively correlated with the fluctuation coefficient of the most recent 10 sets of sampled data of the corresponding evidence. The more stable the data, the higher the weight. S24. The improved DS evidence combination rule is used to fuse the four types of evidence. The fusion formula is as follows: ; in, The conflict coefficient is calculated using the following formula: ; The value ranges from 0 to 1, with a larger value indicating a higher degree of conflict among multiple pieces of evidence. As a conflict moderating factor, The higher the conflict coefficient, the better. The closer the value is to 1, the less impact conflicting evidence has on the fusion result; S25. Determine the fault; if the probability of a certain type of fault is after fusion... Greater than the judgment threshold If so, it is determined that the tool has encountered a corresponding type of fault.

[0013] Specifically, the execution steps of the tool head and tool collaborative evaluation module in the platform layer include: S31. The cutter head is divided into three general functional areas according to its radial radius: the central area, the main cutting area, and the edge diameter-maintaining area. The radial radius of the central area... Corresponding to the low wear zone, main cutting zone Corresponding to the medium-to-high wear zone, edge diameter preservation zone This corresponds to the high-incidence and high-wear area of ​​uneven wear. , These are preset partition boundary values ​​based on the radius of the cutter head; S32. Calculate the wear deviation of the tools in each functional area. The calculation formula is as follows: ; in, This represents the average actual cutting wear of all normal tools within the current functional area. This represents the average actual cutting wear of all normal tools on the entire cutter head. This represents the wear deviation of the area. A value greater than 0 indicates that the wear in that area is faster than the average level of the cutter head. The larger the value, the higher the degree of abnormality. S33. Calculate the panel wear coefficient for each functional area: Obtain the rated pressure of the continuous detection oil passage. Real-time pressure of oil passage When satisfied When it is determined that the continuous detection oil passage in this area is cracked due to panel wear, the panel wear coefficient is marked. ,otherwise 0; S34. Three characteristics—abnormal tool wear in the linkage area, tool turret panel wear, and local tool failure—are used to quantitatively assess the health status of the area and calculate the health score for each functional area. The calculation formula is as follows: ; in, As the tool wear weight, Assuming wear weight for the panel, satisfying ; This is the area fault penalty coefficient, corresponding to the penalty coefficient for each faulty tool within the functional area. Increase by 0.1, with a maximum increase of 0.3; The health score of the functional area is given, with a higher score indicating a better health status of the area. S35. The minimum health score of all functional areas is used as the overall health of the cutter head. Risk levels are divided according to fixed thresholds: a health score above 0.8 is marked as normal, a health score between 0.5 and 0.8 is marked as warning, and a health score below 0.5 is marked as high-risk.

[0014] Specifically, the linkage control strategy generation module in the platform layer divides the scenarios into three different categories: single tool scattered failure scenarios, multi-tool batch wear failure scenarios, and tool head high-risk scenarios, and generates different control strategies according to different scenarios.

[0015] Specifically, the criteria for determining the single-tool scattered failure scenario are: the number of faulty tools on the entire tool turret is less than the preset number, and the health of all functional areas is higher than 0.8, and there is no continuous detection of oil passage rupture on the tool turret panel; the generation strategy includes outputting a tool change list sorted by priority, with priority sorted by the severity of the failure as follows: chipped or broken > uneven wear > stoppage. The criteria for determining a multi-tool batch wear failure scenario are: the number of faulty tools in the same functional area is higher than the preset number, or the health of any functional area is in the warning range of 0.5~0.8, and there is no oil passage rupture on the cutterhead panel; the generation strategy includes synchronously outputting a tool replacement priority list and tunneling parameter adjustment instructions. The tunneling parameter adjustment instructions include reducing the advance speed to the preset value, reducing the cutterhead rotation speed to the preset value, and simultaneously increasing the mud circulation pressure to a preset multiple of the current value, so as to avoid further wear expansion of faulty tools and uneven force on the cutterhead causing uneven wear; The criteria for determining high-risk scenarios for the cutterhead are: the health of any functional area is below 0.5, or there is continuous detection of cutterhead wear due to oil passage rupture; the generation strategy includes immediately outputting an emergency shutdown command and pushing it to the operation and maintenance personnel.

[0016] In summary, the beneficial technical effects of the present invention are as follows: 1. Improved transmission reliability: The present invention adopts a dual-link transmission architecture of "magnetic induction near-field coupling + LoRa spread spectrum", which reduces the signal attenuation caused by the metal shielding of the cutterhead and the mud and water coverage in the fully water-covered strata, and ensures the stability of data transmission.

[0017] 2. Improved Wear Measurement Accuracy: This invention uses a coupled wear dynamic calibration algorithm to quantify the nonlinear effects of three types of disturbances unique to fully aquifer formations: mud and water corrosion, mud and sand erosion, and temperature drift. It can accurately separate the actual cutting wear generated by tool cutting operations from the ineffective thickness loss caused by environmental factors, avoiding the waste of cost in premature tool replacement due to environmental interference or the risk of tool head damage due to undetected wear exceeding limits. Wear data can be directly used as a reliable basis for tool replacement decisions.

[0018] 3. Improved fault identification accuracy and adaptability to complex geological conditions with high conflict: This invention adopts an improved DS evidence fusion fault discrimination algorithm, introduces dynamic confidence weight and conflict adjustment mechanism, and integrates four independent features: wear, rotation speed, temperature, and tunneling parameters. It can accurately identify four types of common tool faults in fully water-covered strata: normal wear, uneven wear, stoppage, and chipping / fracture. It can maintain stable identification accuracy even in high-conflict scenarios with features such as sudden changes in strata hardness and impact from isolated boulders, effectively reducing the risk of misjudgment and missed judgment.

[0019] 4. Achieving collaborative evaluation and closed-loop management of cutterhead and cutting tools to reduce construction risks and costs: This invention establishes a collaborative evaluation model for cutterhead and cutting tools, linking discrete cutting tool wear data with continuous cutterhead panel oil passage monitoring data. It can accurately identify the risk of chain wear in local areas and set a higher panel wear weight for the high-risk characteristics of cutterhead wear leakage under full water coverage and high water pressure. With the help of multi-scenario linkage management and control strategies, it automatically outputs a cutter replacement list sorted by priority and automatically issues dynamic adjustment instructions for tunneling parameters in the early stage of batch failures. This effectively avoids the risks of water inrush and collapse when opening high-risk sections, reducing construction costs and safety hazards. Attached Figure Description

[0020] Figure 1 This is a system framework diagram of the present invention; Figure 2 This is a flowchart of the coupling wear calibration module execution in this invention; Figure 3 This is the execution flowchart of the multi-source feature fusion fault discrimination module in this invention; Figure 4 This is the execution flowchart of the tool head and tool collaborative evaluation module in this invention. Detailed Implementation

[0021] To make the technical means, creative features, objectives and effects of this invention clearer and easier to understand, the invention will be further described below in conjunction with the accompanying drawings and specific embodiments.

[0022] Example like Figure 1 As shown, the automatic monitoring system for tool wear in complex and variable formations with full water coverage provided by the present invention includes a sensing layer, a transmission layer, a platform layer, and an application layer that are sequentially connected in communication, wherein: The sensing layer is used to collect multi-source status and operating condition data under fully water-covered strata during shield tunneling construction. This multi-source data includes cutterhead operating status data, cutterhead panel wear data, and operating parameters; specifically: Multi-parameter tool sensing unit: includes sensing components set for each tool to be monitored, used to collect three types of tool operating status data: remaining thickness, real-time rotation speed, and surface temperature; Cutterhead panel wear monitoring unit: includes several independent continuous detection oil channels arranged in different wear risk areas of the front and rear panels of the cutterhead. Each continuous detection oil channel is equipped with a pressure detection component at the end, which is used to identify the wear location and wear degree of the cutterhead panel through abnormal changes in oil channel pressure, and obtain cutterhead panel wear data. Working condition linkage acquisition unit: It connects to the shield tunneling main control system and the slurry monitoring system to synchronously acquire cutterhead speed, cutterhead torque, total thrust, penetration depth and slurry parameters, including the electrical conductivity, pH value, sand content and slurry flow rate of the slurry.

[0023] The transmission layer includes an edge preprocessing module, several cutterhead relay nodes evenly spaced along the circumference of the cutterhead, and a receiving terminal fixed to the inner wall of the shield's forward compartment, wherein: The edge preprocessing module is deployed in each tool head relay node and performs two-level processing on the received raw data from the sensing layer: the first step is to remove abnormal jump values ​​that exceed the preset range, and the second step is to perform mean aggregation on the high-frequency acquired data at a fixed sampling interval, thereby reducing the amount of invalid data transmission. The transmission layer employs a two-stage transmission link to transmit data from the cutterhead to the shield: the first stage is a magnetic induction near-field transmission link from the multi-parameter sensing unit of the cutterhead to the relay node of the cutterhead. Each multi-parameter sensing unit of the cutterhead and the relay node of the cutterhead are equipped with a magnetic induction communication module, which uses the characteristic of alternating magnetic fields to penetrate metal and mud to transmit data; the second stage is a LoRa spread spectrum transmission link from the relay node of the cutterhead to the receiving terminal of the shield. The relay node of the cutterhead is equipped with a LoRa communication module, which uses spread spectrum modulation to compensate for signal attenuation caused by metal shielding.

[0024] Each tool multi-parameter sensing unit is pre-assigned a unique transmission time slot, and only wakes up to send data within its assigned time slot, entering sleep mode at other times; when the tool head speed is below 0.5 r / min, the transmission layer automatically adjusts the sampling interval to 5 min to reduce overall power consumption.

[0025] The platform layer is used to store, analyze, make decisions on, and output control commands for the data uploaded from the transport layer. It includes a data storage module, a coupled wear calibration module, a multi-source feature fusion fault discrimination module, a tool head and tool collaborative evaluation module, and a linkage control strategy generation module. The data storage module is used to store preprocessed multi-source status and operating condition data, historical wear data, and fault sample data; The coupled wear calibration module is used to eliminate the interference of mud-water corrosion, sediment erosion, and temperature drift on wear measurements in fully aquifer formations, and calculates the actual cutting wear of the tool, such as... Figure 2 As shown, the process for calculating the actual cutting wear of the tool is as follows: S11, Obtain Original wear thickness difference at time , The initial thickness of the cutting tool at the factory and The difference in remaining tool thickness collected by the multi-parameter sensing unit at all times; S12, Calculation The actual cutting wear of the tool at any time : ; The definitions and calculation methods for each correction item are as follows: (1) for The cumulative amount of mud and water corrosion wear, i.e., the tool thickness loss caused by chemical corrosion of mud and water in fully aquifer formations, is calculated using the following formula: ; in, To determine the corrosion coefficient of the cutting tool material to be monitored, a pre-calibration was performed through a laboratory immersion test simulating the muddy water conditions of the target formation. for The mud-water corrosion intensity factor at any given time is calculated based on the mud-water parameters obtained from the working condition linkage acquisition unit: ; in, for The electrical conductivity of mud and water at any given time. for The pH value of the muddy water at any time, this correction item is through The nonlinear design highlights the corrosion aggravation effect when the acid or base deviates from neutral, and is adapted to the groundwater acidity and alkalinity fluctuation characteristics of complex strata with full water coverage. (2) for The cumulative wear and tear caused by mud and sand erosion, i.e., the thickness loss caused by mud and sand carrying it and eroding the cutting tool, is calculated using the following formula: ; in, To determine the erosion coefficient of the cutting tool material to be monitored, a pre-calibration was performed through a laboratory erosion test simulating the mud conditions of the target formation. for The mud-water scour intensity factor at any given time is calculated based on the mud-water parameters obtained from the working condition linkage acquisition unit: ; in, for The slurry flow rate on the surface of the cutterhead at all times. for The sand content of the mud at any given time. The sediment hardness coefficient of the target stratum was pre-calibrated based on the project's geological exploration report; this item passed... The nonlinear enhancement term reflects the rising characteristics of scour wear under high sand content, which is suitable for the large fluctuation of sand content in fully water-covered strata. (3) for The temperature drift error at any given time, i.e., the measurement offset of the eddy current wear sensor caused by temperature changes, is calculated using the following formula: ; in, The temperature drift coefficient of the sensor. for The tool surface temperature is collected by the multi-parameter sensing unit at all times. The reference temperature used to calibrate the sensor.

[0026] The multi-source feature fusion fault discrimination module is used to fuse multi-source state features to identify tool fault types; such as Figure 3 As shown, its execution steps include: S21. Define four types of common tooling faults in fully water-covered formations to be identified, and construct an identification framework: ,in This represents normal wear and tear. Represents uneven wear. The signal is to stop. This indicates chipping or breakage; S22. Extract feature values ​​from four types of independent evidence bodies. All feature values ​​are adapted to the design of shield tunneling construction and fully water-covered strata characteristics to avoid interference from fluctuations in working conditions. Specifically, this includes: a. Evidence 1: Corresponding wear characteristics, the characteristic value being the ring-average rate of change of actual cutting wear. The calculation uses the tunneling ring as the statistical benchmark, eliminates invalid interference during shield standby and shutdown periods, and adapts to the characteristics of large fluctuations in tunneling rhythm in fully water-covered strata. The calculation formula is as follows: ; in, For continuous The actual increase in cutting wear of each tunneling ring The number of rings continuously excavated; b. Evidence 2: Corresponding rotational speed characteristics, the characteristic value being the matching coefficient between the tool rotational speed and the tool head rotational speed. The formula for calculating the smoothness of tool rotation is as follows: ; in, For the real-time rotational speed of the cutting tool, This refers to the real-time rotational speed of the cutter head. c. Evidence 3: Corresponding temperature characteristics, the characteristic value being the temperature difference between the cutting tool and the tool head panel. The calculation formula is: ; in, The surface temperature of the cutting tool. This refers to the temperature of the corresponding area of ​​the cutter head. d. Evidence 4: Corresponding to tunneling parameter characteristics, the characteristic value is the deviation of the tunneling coupling factor. This reflects the degree of abnormality in the tunneling parameters, and the calculation formula is: ; in, This is the actual tunneling coupling factor. The torque of the cutter head. For total thrust, For penetration, The standard tunneling coupling factor for the target stratum under the corresponding tunneling conditions is pre-calibrated using early-stage tunneling data or historical samples from the same stratum. S23. Calculate the basic probability distribution of each piece of evidence, using the following formula: ; in, For the first The characteristic values ​​of each piece of evidence These correspond to four types of evidence. , For the first The evidence corresponds to the first... The characteristic mean and characteristic standard deviation of the fault class are obtained through pre-training using historical fault samples from the target stratum. These correspond to four types of tool failures. For the first The dynamic confidence weights of each piece of evidence satisfy the following: The calculation formula is: ,in The initial weights are pre-calibrated and set based on historical fault identification accuracy; the initial weight for wear characteristics is no less than 0.3. It is a dynamic confidence coefficient, which is negatively correlated with the fluctuation coefficient of the most recent 10 sets of sampled data of the corresponding evidence. The more stable the data, the higher the weight. S24. The improved DS evidence combination rule is used to fuse the four types of evidence. The fusion formula is as follows: ; in, The conflict coefficient is calculated using the following formula: ; The value ranges from 0 to 1, with a larger value indicating a higher degree of conflict among multiple pieces of evidence. As a conflict moderating factor, The higher the conflict coefficient, the better. The closer the value is to 1, the less impact conflicting evidence has on the fusion result; S25. Determine the fault; if the probability of a certain type of fault is after fusion... Greater than the judgment threshold If so, it is determined that the tool has encountered a corresponding type of fault.

[0027] The cutter head and tool co-evaluation module combines tool wear status with cutter head panel wear data to assess the overall health of the cutter head; for example... Figure 4 As shown, its execution steps include: S31. The cutter head is divided into three general functional areas according to its radial radius: the central area, the main cutting area, and the edge diameter-maintaining area. The radial radius of the central area... Corresponding to the low wear zone, main cutting zone Corresponding to the medium-to-high wear zone, edge diameter preservation zone This corresponds to the high-incidence and high-wear area of ​​uneven wear. , These are preset partition boundary values ​​based on the radius of the cutter head; S32. Calculate the wear deviation of the tools in each functional area. The wear deviation reflects the degree of abnormality of tool wear in that area relative to the average level of the tool disc, excluding the normal situation of uniform wear across the entire tool disc, and locating the local abnormal wear area. The calculation formula is: ; in, This represents the average actual cutting wear of all normal tools within the current functional area. This represents the average actual cutting wear of all normal tools on the entire cutter head. This represents the wear deviation of the area. A value greater than 0 indicates that the wear in that area is faster than the average level of the cutter head. The larger the value, the higher the degree of abnormality. S33. Calculate the panel wear coefficient for each functional area. The panel wear coefficient reflects the wear degree of the cutter head panel. It is determined based on the pressure change of the continuously monitored oil passage in the corresponding area to avoid misjudging the cutter head failure when the tool is wearing normally. The calibration method is as follows: obtain the rated pressure of the continuously monitored oil passage. Real-time pressure of oil passage When satisfied When it is determined that the continuous detection oil passage in this area is cracked due to panel wear, the panel wear coefficient is marked. ,otherwise 0; S34. Three characteristics—abnormal tool wear in the linkage area, tool turret panel wear, and local tool failure—are used to quantitatively assess the health status of the area and calculate the health score for each functional area. The calculation formula is as follows: ; in, As the tool wear weight, Assuming wear weight for the panel, satisfying To address the fact that the risk of cutter head panel wear leakage is far greater than the risk of cutter wear under full water immersion and high water pressure, pre-calibration is performed. , It can be dynamically adjusted according to the project's risk level; This is the area fault penalty coefficient, corresponding to the penalty coefficient for each faulty tool within the functional area. Increase by 0.1, with a maximum of 0.3, to reflect the risk of cascading wear caused by local tool failure; The health score of the functional area is given, with a higher score indicating a better health status of the area. S35. The minimum health score of all functional areas is used as the overall health of the cutter head. Risk levels are divided according to fixed thresholds: a health score above 0.8 is marked as normal, a health score between 0.5 and 0.8 is marked as warning, and a health score below 0.5 is marked as high-risk.

[0028] The linkage control strategy generation module generates corresponding control strategies based on fault identification results and toolhead health status. This module divides three different scenarios: single-tool scattered fault scenarios, multi-tool batch wear fault scenarios, and high-risk toolhead scenarios. Different control strategies are generated according to different scenarios, including: The criteria for determining the single-tool scattered failure scenario are: the number of faulty tools on the entire tool turret is less than the preset number, the health of all functional areas is higher than 0.8, and there is no continuous detection of oil passage rupture on the tool turret panel; the generation strategy includes outputting a tool change list sorted by priority, with priority sorted by the severity of the failure as follows: chipped or broken > uneven wear > stoppage. The criteria for determining a multi-tool batch wear failure scenario are: the number of faulty tools in the same functional area is higher than the preset number, or the health of any functional area is in the warning range of 0.5~0.8, and there is no oil passage rupture on the cutterhead panel; the generation strategy includes synchronously outputting a tool replacement priority list and tunneling parameter adjustment instructions. The tunneling parameter adjustment instructions include reducing the advance speed to the preset value, reducing the cutterhead rotation speed to the preset value, and simultaneously increasing the mud circulation pressure to a preset multiple of the current value, so as to avoid further wear expansion of faulty tools and uneven force on the cutterhead causing uneven wear; The criteria for determining high-risk scenarios for the cutterhead are: the health of any functional area is below 0.5, or there is continuous detection of cutterhead wear due to oil passage rupture; the generation strategy includes immediately outputting an emergency shutdown command and pushing it to the operation and maintenance personnel.

[0029] The application layer is used to display the analysis and decision-making results of the platform layer, including a visual monitoring interface, which presents the real-time operating status of the tool, the health status of the tool head, fault warning information, and wear statistical analysis data.

[0030] Therefore, this invention provides an automatic tool wear monitoring system for complex and variable formations with full water coverage. By employing a magnetic induction near-field-LoRa spread spectrum anti-interference dual-link transmission architecture, it overcomes the signal attenuation problems caused by strong metal shielding and high-pressure mud and water encapsulation in full water coverage, establishing a stable and reliable underlying sensing and transmission foundation. It utilizes a coupled wear calibration mechanism to eliminate measurement interference caused by mud and water chemical corrosion, high sand content mud erosion, and temperature drift, significantly reducing monitoring data distortion caused by ineffective thickness loss and obtaining the true cutting wear of the tool. Finally, it constructs a multi-source feature fusion fault discrimination and tool head-tool collaborative evaluation and control architecture based on improved DS evidence theory. It realizes a closed-loop monitoring system covering the entire process, from full-domain anti-shielding perception to precise calibration against multi-source interference, multi-feature fault identification, and full-scenario collaborative management. This system solves the pain points of existing shield cutter monitoring systems, such as poor transmission stability due to harsh muddy environments, wear measurement distortion due to the lack of consideration for environmental coupling interference unique to full water coverage, and high rates of false positives and false negatives due to a lack of global collaborative management and a single diagnostic dimension. It improves the accuracy of cutter fault identification and the scientific nature of cutterhead health status assessment, providing a high-precision, robust, and intelligent full-lifecycle cutter status monitoring and management solution for intelligent shield construction in complex geological conditions such as urban rail transit and cross-river and cross-sea tunnels.

[0031] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings. The above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. 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 be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. An automatic monitoring system for tool wear in complex and variable formations with full water coverage, characterized in that, It includes the perception layer, transport layer, platform layer, and application layer, which are connected in sequence, wherein: The sensing layer is used to collect multi-source status and working condition data under fully water-covered strata during shield tunneling construction. The multi-source status and working condition data includes cutter running status data, cutterhead panel wear data, and working condition parameters. The transport layer is used to receive multi-source status and operating condition data collected by the perception layer, perform edge preprocessing on the multi-source status and operating condition data, and transmit the preprocessed data to the platform layer through an anti-interference dual-link transmission architecture. The platform layer is used to store, analyze, make decisions about, and output control commands for the data uploaded from the transport layer, including: The data storage module is used to store preprocessed multi-source status and operating condition data, historical wear data and fault sample data; The coupled wear calibration module is used to eliminate the interference of mud and water corrosion, mud and sand erosion and temperature drift on wear measurement in fully water-covered formations, and to calculate the actual cutting wear of the tool. A multi-source feature fusion fault discrimination module is used to fuse multi-source state features to identify tool fault types; The tool and cutter collaborative evaluation module is used to evaluate the overall health status of the tool and cutter panel by combining the tool wear status with the tool and cutter panel wear data. The linkage control strategy generation module is used to generate corresponding control strategies based on the fault identification results and the health status of the tool turret; The application layer is used to display the analysis and decision-making results of the platform layer, including a visual monitoring interface, which presents the real-time operating status of the tool, the health status of the tool head, fault warning information, and wear statistical analysis data.

2. The automatic monitoring system for tool wear in complex and variable formations with full water coverage as described in claim 1, characterized in that, The perception layer specifically includes: Multi-parameter tool sensing unit: includes sensing components set for each tool to be monitored, used to collect three types of tool operating status data: remaining thickness, real-time rotation speed, and surface temperature; Cutterhead panel wear monitoring unit: includes several independent continuous detection oil channels arranged in different wear risk areas of the front and rear panels of the cutterhead. Each continuous detection oil channel is equipped with a pressure detection component at the end, which is used to identify the wear location and wear degree of the cutterhead panel through abnormal changes in oil channel pressure, and obtain cutterhead panel wear data. Working condition linkage acquisition unit: It connects to the shield tunneling main control system and the slurry monitoring system to synchronously acquire cutterhead speed, cutterhead torque, total thrust, penetration depth and slurry parameters, including the electrical conductivity, pH value, sand content and slurry flow rate of the slurry.

3. The automatic monitoring system for tool wear in complex and variable formations with full water coverage as described in claim 2, characterized in that, The transmission layer includes an edge preprocessing module, several cutterhead relay nodes evenly spaced along the circumference of the cutterhead, and a receiving terminal fixed to the inner wall of the shield's forward compartment, wherein: The edge preprocessing module is deployed in each tool head relay node and performs two-level processing on the received raw data from the sensing layer: the first step is to remove abnormal jump values ​​that exceed the preset range, and the second step is to perform mean aggregation on the high-frequency acquired data at a fixed sampling interval to reduce the amount of invalid data transmission. The transmission layer employs a two-stage transmission link to transmit data from the cutterhead to the shield: the first stage is a magnetic induction near-field transmission link from the multi-parameter sensing unit of the cutterhead to the relay node of the cutterhead. Each multi-parameter sensing unit of the cutterhead and the relay node of the cutterhead are equipped with a magnetic induction communication module, which uses the characteristic of alternating magnetic fields to penetrate metal and mud to transmit data; the second stage is a LoRa spread spectrum transmission link from the relay node of the cutterhead to the receiving terminal of the shield. The relay node of the cutterhead is equipped with a LoRa communication module, which uses spread spectrum modulation to compensate for signal attenuation caused by metal shielding.

4. The automatic monitoring system for tool wear in complex and variable formations with full water coverage as described in claim 3, characterized in that, The process by which the coupled wear calibration module in the platform layer calculates the actual cutting wear of the tool is as follows: S11, Obtain Original wear thickness difference at time , The initial thickness of the cutting tool at the factory and The difference in remaining tool thickness collected by the multi-parameter sensing unit at all times; S12, Calculation The actual cutting wear of the tool at any time : ； The definitions and calculation methods for each correction item are as follows: (1) for The cumulative amount of mud and water corrosion and wear over time is calculated using the following formula: ； in, To determine the corrosion coefficient of the cutting tool material to be monitored, a pre-calibration was performed through a laboratory immersion test simulating the muddy water conditions of the target formation. for The mud-water corrosion intensity factor at any given time is calculated based on the mud-water parameters obtained from the working condition linkage acquisition unit: ； in, for The electrical conductivity of mud and water at any given time. for The pH value of the muddy water at all times; (2) for The cumulative amount of erosion and wear from sediment over time is calculated using the following formula: ； in, To determine the erosion coefficient of the cutting tool material to be monitored, a pre-calibration was performed through a laboratory erosion test simulating the mud conditions of the target formation. for The mud-water scour intensity factor at any given time is calculated based on the mud-water parameters obtained from the working condition linkage acquisition unit: ； in, for The slurry flow rate on the surface of the cutterhead at all times. for The sand content of the mud at any given time. The sediment hardness coefficient of the target stratum was pre-calibrated based on the project's geological exploration report; this item passed... The nonlinear enhancement term reflects the rising characteristics of scour wear under high sand content, which is suitable for the large fluctuation of sand content in fully water-covered strata. (3) for The temperature drift error at any given time is calculated using the following formula: ； in, The temperature drift coefficient of the sensor. for The tool surface temperature is collected by the multi-parameter sensing unit at all times. The reference temperature used to calibrate the sensor.

5. The automatic monitoring system for tool wear in complex and variable formations with full water coverage according to claim 4, characterized in that, The execution steps of the multi-source feature fusion fault discrimination module in the platform layer include: S21. Define four types of common tooling faults in fully water-covered formations to be identified, and construct an identification framework: ,in This represents normal wear and tear. Represents uneven wear. The signal is to stop. This indicates chipping or breakage; S22. Extract feature values ​​from four types of independent evidence, including: a. Evidence 1: Corresponding wear characteristics, the characteristic value being the ring-average rate of change of actual cutting wear. The calculation formula is: ； in, For continuous The actual increase in cutting wear of each tunneling ring The number of rings continuously excavated; b. Evidence 2: Corresponding rotational speed characteristics, the characteristic value being the matching coefficient between the tool rotational speed and the tool head rotational speed. The calculation formula is: ； in, For the real-time rotational speed of the cutting tool, This refers to the real-time rotational speed of the cutter head. c. Evidence 3: Corresponding temperature characteristics, the characteristic value being the temperature difference between the cutting tool and the tool head panel. The calculation formula is: ； in, The surface temperature of the cutting tool. This refers to the temperature of the corresponding area of ​​the cutter head. d. Evidence 4: Corresponding to tunneling parameter characteristics, the characteristic value is the deviation of the tunneling coupling factor. The calculation formula is: ； in, This is the actual tunneling coupling factor. The torque of the cutter head. For total thrust, For penetration, The standard tunneling coupling factor for the target stratum under the corresponding tunneling conditions is pre-calibrated using early-stage tunneling data or historical samples from the same stratum. S23. Calculate the basic probability distribution of each piece of evidence, using the following formula: ； in, For the first The characteristic values ​​of each piece of evidence These correspond to four types of evidence. , For the first The evidence corresponds to the first... The characteristic mean and characteristic standard deviation of the fault class are obtained through pre-training using historical fault samples from the target stratum. These correspond to four types of tool failures. For the first The dynamic confidence weights of each piece of evidence satisfy the following: The calculation formula is: ,in The initial weights are pre-calibrated and set based on historical fault identification accuracy. It is a dynamic confidence coefficient, which is negatively correlated with the fluctuation coefficient of the most recent 10 sets of sampled data of the corresponding evidence. The more stable the data, the higher the weight. S24. The improved DS evidence combination rule is used to fuse the four types of evidence. The fusion formula is as follows: ； in, The conflict coefficient is calculated using the following formula: ； The value ranges from 0 to 1, with a larger value indicating a higher degree of conflict among multiple pieces of evidence. As a conflict moderating factor, The higher the conflict coefficient, the better. The closer the value is to 1, the less impact conflicting evidence has on the fusion result; S25. Determine the fault; if the probability of a certain type of fault is after fusion... Greater than the judgment threshold If so, it is determined that the tool has encountered a corresponding type of fault.

6. The automatic monitoring system for tool wear in complex and variable formations with full water coverage according to claim 5, characterized in that, The execution steps of the tool head and tool collaborative evaluation module in the platform layer include: S31. The cutter head is divided into three general functional areas according to its radial radius: the central area, the main cutting area, and the edge diameter-maintaining area. The radial radius of the central area... Corresponding to the low wear zone, main cutting zone Corresponding to the medium-to-high wear zone, edge diameter preservation zone This corresponds to the high-incidence and high-wear area of ​​uneven wear. , These are preset partition boundary values ​​based on the radius of the cutter head; S32. Calculate the wear deviation of the tools in each functional area. The calculation formula is as follows: ； in, This represents the average actual cutting wear of all normal tools within the current functional area. This represents the average actual cutting wear of all normal tools on the entire cutter head. This represents the wear deviation of the area. A value greater than 0 indicates that the wear in that area is faster than the average level of the cutter head. The larger the value, the higher the degree of abnormality. S33. Calculate the panel wear coefficient for each functional area: Obtain the rated pressure of the continuous detection oil passage. Real-time pressure of oil passage When satisfied When it is determined that the continuous detection oil passage in this area is cracked due to panel wear, the panel wear coefficient is marked. ,otherwise ; S34. Three characteristics—abnormal tool wear in the linkage area, tool turret panel wear, and local tool failure—are used to quantitatively assess the health status of the area and calculate the health score for each functional area. The calculation formula is as follows: ； in, As the tool wear weight, Assuming wear weight for the panel, satisfying ; This is the area fault penalty coefficient, corresponding to the penalty coefficient for each faulty tool within the functional area. Increase by 0.1, with a maximum increase of 0.3; The health score of the functional area is given, with a higher score indicating a better health status of the area. S35. The minimum health score of all functional areas is used as the overall health of the cutter head. Risk levels are divided according to fixed thresholds: a health score above 0.8 is marked as normal, a health score between 0.5 and 0.8 is marked as warning, and a health score below 0.5 is marked as high-risk.

7. The automatic monitoring system for tool wear in complex and variable formations with full water coverage as described in claim 6, characterized in that: The platform layer's linkage control strategy generation module divides the scenarios into three different categories: single tool scattered failure scenarios, multi-tool batch wear failure scenarios, and tool head high-risk scenarios. Different control strategies are generated according to different scenarios.

8. The automatic monitoring system for tool wear in complex and variable formations with full water coverage according to claim 7, characterized in that, The criteria for determining the single-tool scattered failure scenario are: the number of faulty tools on the entire tool turret is less than the preset number, the health of all functional areas is higher than 0.8, and there is no continuous detection of oil passage rupture on the tool turret panel; the generation strategy includes outputting a tool change list sorted by priority, with priority sorted by the severity of the failure as follows: chipped or broken > uneven wear > stoppage. The criteria for determining a multi-tool batch wear failure scenario are: the number of faulty tools in the same functional area is higher than the preset number, or the health of any functional area is in the warning range of 0.5~0.8, and there is no oil passage rupture on the cutterhead panel; the generation strategy includes synchronously outputting a tool replacement priority list and tunneling parameter adjustment instructions. The tunneling parameter adjustment instructions include reducing the advance speed to the preset value, reducing the cutterhead rotation speed to the preset value, and simultaneously increasing the mud circulation pressure to a preset multiple of the current value, so as to avoid further wear expansion of faulty tools and uneven force on the cutterhead causing uneven wear; The criteria for determining high-risk scenarios for the cutterhead are: the health of any functional area is below 0.5, or there is continuous detection of cutterhead wear due to oil passage rupture; the generation strategy includes immediately outputting an emergency shutdown command and pushing it to the operation and maintenance personnel.

Images