Refined muck cutting system for vertical shaft tunneling of water-rich soft soil foundation

Abstract

The invention discloses a fine muck cutting system for vertical shaft tunneling of a water-rich soft soil foundation, and belongs to the technical field of underground engineering machinery. The system comprises a stratum sensing and self-adaptive cutting roller assembly which is provided with a segmented variable-stiffness double-layer roller structure and a zoning differentiated cutting tooth layout based on stratum prediction; the invention discloses a high-water-pressure bidirectional dynamic compensation intelligent sealing system. A five-stage differential pressure type sealing structure and an active pressure tracking compensation strategy based on model predictive control are adopted; the multi-source information fusion and intelligent cooperative control system is used for realizing low-delay cooperative control of each subsystem based on a time-sensitive network; the invention further provides ground muck closed-loop type fine treatment and recycling complete equipment. Through intelligent sensing, self-adaptive tunneling, high-pressure dynamic sealing, multi-system cooperation and muck resource utilization, the safety, efficiency, economical efficiency and environmental friendliness of vertical shaft construction of the complex stratum are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of underground engineering machinery technology, specifically to a refined slag cutting system for vertical shaft excavation in water-rich soft soil foundations. Background Technology

[0002] In major projects such as cross-river and sea tunnel connections, deep mineral mining, large underground storage facilities, and the development of deep urban underground spaces, large-diameter deep shafts serve as crucial vertical passages for personnel, materials, ventilation, and pipeline transportation, representing a critical "choke point" that restricts the overall project schedule, cost, and safety. Currently, shaft construction primarily employs mechanical tunneling methods (such as shaft drilling rigs and raise shaft drilling rigs), caisson methods, and freezing methods. Among these, mechanized tunneling methods have become the mainstream approach for medium-deep shaft construction due to their advantages such as high tunneling efficiency, relatively safe working environment, and minimal disturbance to the surrounding soil.

[0003] However, the widespread distribution of water-rich, soft soil layers and the common "soft on top, hard on the bottom" composite strata in my country's coastal, riverine, and lacustrine sedimentary areas present unprecedented technical challenges to mechanized shaft excavation. Existing technologies and equipment generally suffer from the following prominent problems when dealing with such complex conditions, severely restricting construction safety, efficiency, and depth:

[0004] Poor geological adaptability makes it difficult to guarantee tunneling efficiency and cutter life. Traditional tunneling machines have fixed cutter head designs, which cannot intelligently adapt to the drastic changes from fluid soft soil to high-strength hard rock. In soft soil, adhesion and blockage are prone to occur, resulting in poor muck removal; in hard rock, the cutting teeth wear severely and the load impact is large. The entire tunneling process lacks real-time strata perception and parameter adaptive adjustment capabilities, relying on manual experience, resulting in low efficiency and serious equipment wear.

[0005] The reliability of dynamic seals in deep-water high-pressure systems is insufficient, limiting the depth of construction. As the depth of the shaft increases, the external water pressure and mud pressure rise significantly and fluctuate dynamically. Traditional multi-lip seals rely on static preload, which is prone to problems such as instantaneous leakage and lip damage under alternating loads, leading to water ingress and failure of the main drive system. This severely limits the reliable construction depth of shafts, making it difficult to break through the 100-meter mark.

[0006] The lack of coordination among multiple subsystems leads to low precision in construction process control. Key subsystems such as tunneling, segment sinking, muck removal, and mud circulation are typically controlled independently, relying on manual coordination, resulting in "information silos" and "slow action." This can easily lead to construction risks such as shaft verticality deviation, sudden or difficult sinking accidents, and mud pressure imbalance causing instability at the excavation face.

[0007] Construction waste disposal methods are crude, costly, and environmentally unfriendly. Large quantities of high-moisture-content construction waste generated from water-rich strata require extensive landfill disposal, with traditional methods of transporting and burying the waste accounting for a significant portion of the disposal costs. Furthermore, direct discharge of sludge causes environmental pollution, failing to meet the requirements of green construction and resource recycling.

[0008] The system has a low level of intelligence and lacks predictive maintenance capabilities. Existing equipment has limited intelligence, insufficient status monitoring and health management of key components, a passive maintenance mode, and a high risk of unplanned downtime.

[0009] In summary, existing technologies lack a systematic solution that integrates intelligent sensing, adaptive tunneling, high-pressure dynamic sealing, precise multi-system coordination, and green resource utilization of excavated soil. Therefore, there is an urgent need to develop a novel intelligent shaft tunneling system to fundamentally improve the overall capabilities of shaft construction in complex geological formations. Summary of the Invention

[0010] The purpose of this invention is to provide a refined soil cutting system for vertical shaft excavation in water-rich soft soil foundations. Through intelligent sensing, adaptive excavation, high-pressure dynamic sealing, multi-system collaboration, and soil resource utilization, it significantly improves the safety, efficiency, economy, and environmental friendliness of vertical shaft construction in complex strata, thereby solving the aforementioned technical problems.

[0011] To achieve the above objectives, the first aspect of the present invention provides a refined soil cutting system for vertical shaft excavation in water-rich soft soil foundations, comprising:

[0012] A formation sensing and adaptive cutting drum assembly, which is connected to the main drive of a vertical shaft rock-embedded drilling rig via a buffer drive connector;

[0013] A high-water-pressure bidirectional dynamic compensation intelligent sealing system is fitted onto the main drive shaft of the drilling rig host to achieve active pressure tracking sealing during the tunneling process;

[0014] The multi-source information fusion and intelligent collaborative control system communicates with the cutting drum assembly, sealing system, segment sinking system, mud pressure control system and ground treatment system through a time-sensitive network.

[0015] The complete set of equipment for closed-loop fine treatment and resource utilization of surface waste soil is connected to the underground system through waste discharge pipelines and mud reinjection pipelines, realizing the real-time sorting, dewatering and resource utilization of waste soil.

[0016] As a further option, the formation sensing and adaptive cutting drum assembly includes:

[0017] The segmented variable stiffness double-layer drum body is welded together from an outer cutting cylinder and an inner support cylinder arranged coaxially.

[0018] The wall thickness of the outer cutting cylinder is designed differently along the axial direction according to the predicted formation hardness distribution. Radial reinforcing ribs and circumferential reinforcing rings are welded to the inner wall of the predicted hard rock section to form a grid-like reinforcing structure.

[0019] The inner support cylinder and the outer cutting cylinder form an annular sealed cavity for introducing high-pressure cooling water and accommodating sensor cables.

[0020] The drum body is sealed at both ends by heavy-duty end caps. The drive end cap is provided with a spline or flange connection structure, and the slag collection end cap is provided with a central slag collection port. The slag collection port is smoothly connected to the conical slag collection groove formed by the inner wall of the inner support cylinder.

[0021] As a further option, the cutting roller assembly also includes:

[0022] Based on the formation prediction model, the zoned differentiated cutting tooth layout divides the drum surface into a high-efficiency main cutting zone and a forced slag collection and guiding zone located at both ends.

[0023] The cutting tooth base of the high-efficiency main cutting zone is arranged along four logarithmic spiral lines;

[0024] The cutting teeth of the forced slag collection and guiding zone are arranged in a double-headed reverse spiral line to collect slag and soil towards the center of the drum.

[0025] All cutting teeth mounting bases are angle-adjustable, allowing for preset installation angles for cutting teeth in different sections based on geological formation predictions.

[0026] As a further option, the high water pressure bidirectional dynamic compensation intelligent sealing system includes:

[0027] The five-stage differential pressure sealing rings arranged sequentially along the main drive shaft include two main sealing rings, two isolation sealing rings, and one dustproof sealing ring.

[0028] Multiple sealing cavities are formed between adjacent sealing rings, which are P0 cavity, P1 cavity, P2 cavity, P3 cavity and P4 collection cavity from the wellbore side inward;

[0029] The P1 cavity is injected with ultra-high viscosity grease, and its pressure is maintained at a level higher than the wellbore pressure by an active grease injection system.

[0030] The P4 chamber is connected to a leaking oil tank equipped with a liquid level sensor via a leaking oil pipe.

[0031] Further options include:

[0032] A ring-shaped multi-point independent micro-pressure grease injection system, including a pressure distributor ring disposed on the outside of the sealing seat;

[0033] The pressure distributor has multiple radial oil outlet holes evenly distributed along the circumference, and each oil outlet hole is connected to an independently controlled miniature high-speed piezoelectric ceramic proportional valve.

[0034] Each proportional valve outlet is connected to a miniature Coriolis mass flow meter for real-time monitoring of the flow rate at the corresponding grease injection point;

[0035] The grease injection system communicates with the sealing controller and receives valve opening commands calculated based on a model predictive control algorithm.

[0036] As a further option, the multi-source information fusion and intelligent collaborative control system includes:

[0037] A deterministic control backbone network based on time-sensitive networking is used to transmit control data and status information between subsystems;

[0038] The central intelligent decision-making unit has a built-in lightweight convolutional neural network model for real-time analysis of vibration spectrum and identification of stratum type and hardness index.

[0039] The segment sinking synchronous controller executes the "virtual spindle" algorithm, using the displacement weighted average of multiple sets of electro-hydraulic proportional servo cylinders as a benchmark to achieve high-precision synchronous control;

[0040] The mud pressure dynamic balance controller uses a dynamic matrix control algorithm to coordinate the frequency of the slag discharge pump and the reinjection pump, maintaining stable pressure at the excavation face.

[0041] As a further option, in the "virtual spindle" algorithm:

[0042] Each control cycle calculates the weighted average of all cylinder displacements as the virtual spindle displacement;

[0043] The weight of each cylinder is dynamically adjusted according to its real-time load pressure;

[0044] The composite synchronization error of the hydraulic cylinder includes the tracking error of its virtual spindle displacement and the coupling error with the displacement of adjacent hydraulic cylinders.

[0045] As a further option, the complete set of equipment for closed-loop fine treatment and resource utilization of ground waste includes:

[0046] The multi-stage sorting unit includes a dual-axis intelligent vibrating screen and a parallel adjustable parameter hydrocyclone for separating aggregates of different particle sizes.

[0047] The deep dewatering unit, including a screw press thickener and a high-pressure diaphragm filter press, is used to dewater fine slurry into slurry cake with low moisture content.

[0048] The online automated production unit includes multiple material bins, a high-precision loss-in-weight metering feeder, a continuous mixer, and a central formula control unit, which are used to dynamically adjust the proportions and produce fluidized solidified soil.

[0049] The mud recycling unit, including a conditioning tank and a reinjection pump, is used to return the treated mud to the well, forming a closed loop.

[0050] A cooperative control method for any of the above-mentioned systems includes the following steps:

[0051] System initialization, establishment of proactive pressure barriers, and startup of all subsystems;

[0052] Real-time acquisition of vibration, torque, and thrust data; formation information is identified through a neural network model.

[0053] Based on the formation identification results, the tunneling parameters are dynamically adjusted, and the sealing pressure is simultaneously controlled to track changes in well pressure.

[0054] After the tunneling process is completed, multiple sets of hydraulic cylinders are controlled by the "virtual spindle" algorithm to achieve synchronous sinking of the tunnel segments;

[0055] The dynamic matrix control algorithm coordinates the slag discharge and reinjection pumps to maintain wellbore pressure balance.

[0056] The excavated soil is transported to the surface treatment system for sorting, dewatering, and resource utilization, and the mud is reinjected into the well.

[0057] Further options include:

[0058] Real-time monitoring of flow rate and sealing cavity pressure at each grease injection point; trend analysis enables early warning of seal wear.

[0059] Continuous monitoring of the condition of key components such as cutting gear vibration and bearing temperature enables predictive maintenance.

[0060] Compared with the prior art, the beneficial effects of the present invention are:

[0061] 1. By employing a segmented variable stiffness double-layer drum structure and a zoned differentiated cutter layout based on stratum prediction, combined with real-time stratum identification technology based on vibration spectrum, the system achieves both "pre-structural adaptation" and "real-time parameter adaptation" for complex strata. Before construction, the strata are predicted by pre-setting the cutter layout and matching the drum stiffness; during tunneling, tunneling parameters are dynamically adjusted to match real-time working conditions. This significantly improves tunneling efficiency, reduces abnormal wear and load impact on the cutters, and extends the equipment's service life.

[0062] 2. An innovative five-stage differential pressure sealing structure was designed, and an active pressure tracking compensation strategy based on model predictive control (MPC) algorithm was adopted. This allows the sealing cavity pressure to track and slightly exceed the external dynamic water pressure in real time and accurately. Through active pressure compensation, this system achieves dynamic isolation of high-pressure mud, significantly improving the sealing reliability and service life of the main drive system in deep water and high-pressure environments, providing crucial support for the construction of vertical shafts exceeding 100 meters in depth.

[0063] 3. A deterministic control backbone network was constructed based on Time-Sensitive Networking (TSN), enabling low-latency and highly reliable information exchange between subsystems. Through central intelligent decision-making, a "virtual master axis" synchronization algorithm, and Dynamic Matrix Control (DMC), the system achieved integrated closed-loop collaborative control of adaptive adjustment of tunneling parameters, high-precision synchronous sinking of tunnel segments, and dynamic balancing of mud pressure. This effectively prevented shaft deviation, sudden sinking, and excavation face instability, fundamentally ensuring construction safety and quality.

[0064] 4. Through a closed-loop fine treatment system for construction waste on the ground, cutting waste is efficiently sorted, dewatered, and automatically produced online into materials such as fluidized solidified soil suitable for engineering projects. Simultaneously, the treated mud is reinjected into the well, forming a closed-loop mud circulation system. This significantly reduces the cost of transporting and disposing of construction waste and the environmental impact, achieving the resource utilization of engineering waste.

[0065] 5. Through multi-source information fusion and intelligent analysis, the system can not only optimize construction parameters in real time, but also monitor and analyze the operating status and trends of key components (such as cutting teeth and sealing status), providing a data foundation for predictive maintenance and helping to improve equipment utilization and construction continuity. Attached Figure Description

[0066] Figure 1 This is a system architecture diagram of the present invention.

[0067] Figure 2 Flowchart of the cutting roller teeth arrangement.

[0068] Figure 3 This is a schematic diagram of the flow of excavated soil from the cutting drum.

[0069] Figure 4 This is a schematic diagram showing the circumferential distribution of the cutting teeth on the cutting drum.

[0070] Figure 5 This is a schematic diagram of a drum-type slag discharge system.

[0071] Figure 6 A schematic diagram of the cutting drum for optimized design.

[0072] Figure 7 This is a schematic diagram of adaptive adjustment of the sealing pressure of the rotary unit.

[0073] Figure 8 This is a schematic diagram of pressure monitoring data in one embodiment.

[0074] Figure 9 This is a schematic diagram of automated construction of a sunken vertical shaft rock-embedded drilling rig.

[0075] Figure 10 This is a flowchart of the steps of the collaborative control method of the present invention. Detailed Implementation

[0076] 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.

[0077] like Figure 1 As shown, this invention provides a highly integrated intelligent shaft tunneling system. This system constructs a complete closed-loop system from intelligent underground tunneling and dynamic pressure control to the immediate resource utilization of excavated soil on the surface. Its core components and connections are as follows:

[0078] The system uses the main body (101) of a vertical shaft rock-embedded drilling rig as its load-bearing and driving base. At its lower part, a formation sensing and adaptive cutting roller assembly (102) that directly interacts with the formation is installed. At its upper part, a segment assembly machine and a synchronous sinking system (103) are installed, responsible for the installation and precise sinking of the wellbore structure. A high-water-pressure bidirectional dynamic compensation intelligent sealing system (104) integrated into the main unit's rotating part provides a reliable high-pressure dynamic seal for the core drive shaft.

[0079] The multi-source information fusion and intelligent collaborative control system (105) is connected to all subsystems through a high-speed deterministic network. The mixture of cutting debris and mud generated is transported to the surface through the downhole slag discharge pipeline (106). The closed-loop fine treatment and resource utilization equipment (107) set up on the surface performs multi-stage sorting, dewatering and resource utilization of the cutting debris. The treated mud that meets the standards is pumped back downhole through the mud reinjection pipeline (108) to form a green closed loop. The modules are closely coupled through mechanical, hydraulic, electrical and information interfaces and work together.

[0080] I. Formation Sensing and Adaptive Cutting Drum Assembly

[0081] This assembly combines structural reinforcement, pre-embedded sensing, and differentiated cutting layout, including:

[0082] 1.1 Segmented Variable Stiffness Double-Layer Roller Body

[0083] As attached Figures 2-6As shown, the drum body adopts a coaxial double-layer welded structure. The outer cutting drum (201) is made of high-strength wear-resistant steel plate. Its core design lies in the fact that the wall thickness is not uniform, but is axially differentiated based on the hardness distribution curve of the strata along the depth predicted by geological survey data and construction BIM model. In the predicted hard rock layer and boulder section, the drum wall is locally thickened, and radial reinforcing ribs and circumferential reinforcing rings are welded on the inner wall to form a high-rigidity grid structure. For example, through finite element analysis and engineering analogy, it is designed with 12 circumferentially distributed radial ribs with a rib height of 30mm, which are orthogonally welded to circumferential ring plates with an axial spacing of 300mm to form a "well"-shaped grid-like reinforcing structure. This design makes the structural stiffness precisely match the expected impact load, avoiding drum deformation during hard rock excavation. In the predicted soft soil section, the foundation wall thickness is maintained to reduce weight.

[0084] The inner support cylinder (202) mainly provides torsional stiffness and forms an internal slag collection channel. The inner and outer cylinders are welded together by heavy-duty end caps (203) at both ends and spaced ribs in the middle. The annular sealed cavity (204) formed between the two has multiple functions: first, it serves as a circulation channel for high-pressure cooling water, effectively cooling the cutting teeth and bearings; second, it can serve as a protection zone for the cables of built-in sensors (such as stress and temperature sensors), improving reliability.

[0085] End Connections and Functional Interfaces: At the drive end, the end cap (203) has an internal spline or flange bolt hole in the center for torque transmission connection with the output flange of the buffer drive connector. At the slag collection end, the end cap has a central slag collection port, which is smoothly connected to the conical slag collection groove (205) formed by the inner wall of the inner support cylinder (202). The end of the slag collection groove (205) is connected to the suction port flange of the downhole slag discharge pipeline (106) through a quick-replaceable wear-resistant alloy short connector.

[0086] 1.2 Differentiated Cutting Tooth Layout Based on Prediction Model

[0087] The layout of the cutting teeth was optimized through mechanical and hydrodynamic simulations to pre-adapt to predicted geological variations before construction. The roller surface is divided into two regions along the axial direction:

[0088] The high-efficiency main cutting zone (Zone I, accounting for approximately 90% of the axial length) features a precisely arranged base of four logarithmic helical lines (two left-handed and two right-handed) with a helix angle of 12°. The left-handed helical lines have a helix angle of +12°, while the right-handed helical lines have a helix angle of -12°. This helix angle is the optimal value obtained through simulation, taking into account both the smoothness of slag removal in soft soil and the stability of cutting hard rock. The high density of the cutting teeth ensures that multiple teeth are simultaneously subjected to force in hard rock, resulting in stable load distribution.

[0089] Forced slag collection and guiding zone (Zone II, approximately 5% at each end): The cutting teeth are arranged in a double-headed reverse spiral line, and their spiral direction is specially designed to gather the slag towards the middle of the drum, effectively preventing the increase in torque caused by slag accumulation at the ends.

[0090] More importantly, all the cutting teeth mounting bases are designed with adjustable angles. Before the drum is lowered into the well, different installation inclination angles can be preset for the cutting teeth in different axial sections based on the formation prediction model. For example, in sections corresponding to soft soil formations, a larger positive inclination angle is preset to facilitate cutting and slag removal; in hard rock sections, a smaller negative inclination angle or zero inclination angle is preset to enhance impact resistance and reduce tooth breakage. This prediction-based static layout optimization significantly improves the cutting system's pre-adaptability to complex formations without excessively increasing the complexity of real-time control.

[0091] II. High-pressure bidirectional dynamic compensation intelligent sealing system

[0092] As attached Figure 7 , Figure 8 As shown, this system solves the problem of dynamic sealing under high pressure in deep water. Its core is the concept of "active pressure tracking compensation," which includes:

[0093] 2.1 Five-stage differential pressure sealing structure and chamber

[0094] Arranged sequentially along the main drive shaft (501):

[0095] The first main sealing ring (502) has a U-shaped cross-section, is made of polyurethane, and has a stainless steel spring clamp on the back to provide initial lip clamping force. Its outer side is the P0 cavity, which communicates with the wellbore.

[0096] The second main sealing ring (503) has the same structure as the first one and is installed back-to-back with the first sealing ring, forming a first-level sealing cavity (P1 cavity) between them.

[0097] The third isolation sealing ring (504) is a double-lip oil seal, with the main lip facing the high-pressure side. It forms a second-stage sealing cavity (P2 cavity) with the second sealing ring.

[0098] The fourth isolation sealing ring (505) has the same structure as the third ring, but is installed in the opposite direction. It forms a third-level sealing cavity (P3 cavity) with the third sealing ring.

[0099] The fifth dustproof sealing ring (506) is a single-lip dustproof ring. It forms a collection chamber (P4 chamber) with the fourth sealing ring. The P4 chamber is connected to a transparent leak oil tank with a liquid level sensor through a leak oil pipe.

[0100] Functions of each chamber: Chamber P0 withstands the total pressure of the wellbore; Chamber P1 establishes the first-level active pressure barrier by injecting ultra-high viscosity EP2 grease, with the pressure set slightly higher than that of Chamber P0; Chambers P2 and P3 are injected with thin lubricating oil, with the pressure decreasing sequentially; Chamber P4 is at normal pressure or slightly negative pressure to collect any possible trace leaks.

[0101] 2.2 Circular Multi-Point Independent Micro-Pressure Grease Injection System

[0102] The system includes a pressure distributor (507) surrounding the sealing seat, multiple independently controlled miniature high-speed piezoelectric ceramic proportional valves (509) and corresponding miniature Coriolis mass flow meters (510).

[0103] The pressure distributor (507) is a high-strength bronze ring press-fitted onto the outside of the sealing seat. The ring body has an annular oil collection groove and multiple (e.g., 16) radial oil outlet holes, each corresponding to a sector area (e.g., 22.5°) on the sealing circumference.

[0104] Each miniature high-speed piezoelectric ceramic proportional valve (509) is threaded into the inlet of each radial oil outlet and can be independently addressed and controlled.

[0105] Each miniature Coriolis mass flow meter (510) is connected to the outlet of the corresponding proportional valve (509) by a clamp, and its measuring tube is directly inserted into the miniature oil pipe leading to the corresponding sealing injection point for real-time measurement of the precise grease injection flow at that point.

[0106] 2.3 Adaptive Stress Management Based on Model Predictive Control (MPC)

[0107] The sealing controller (511) is a dedicated industrial controller configured to execute a model predictive control algorithm.

[0108] Sensor network: A pressure transmitter S1 is installed on the well wall to measure the well mud pressure P_out; pressure sensors S2 and S3 are installed on the sealing seat to measure the pressure of chambers P1 and P2 respectively; similarly, sensors are installed to measure the pressure of chambers P3 and P4; the spindle encoder provides the rotational speed n; multiple (e.g., 16) flow meters (510) provide the grease injection flow rate Q1-Q16 at each point.

[0109] Control Model and Cycle: The controller embeds a simplified discrete state-space model of the sealing system, which describes the dynamic relationship between grease injection pressure, valve opening, and pressure in each sealing cavity. Within each control cycle (e.g., 10ms), the latest data from all sensors is read, and based on the current tunneling speed and depth, the trend of P_out's change over a short future period (e.g., 2 seconds) is predicted.

[0110] Optimization solution: Taking the pressure of P1 and P2 stably tracking their setpoints (P1_set = P_out + ΔP, ΔP is about 0.3-0.8 bar; P2_set = P1 - 0.3 bar) and minimizing the total grease injection flow and valve action amplitude as the objective function J, online rolling optimization solution is performed to obtain the optimal control sequence, including the pressure setpoint P_pump* of the main grease pump (515) and the opening increment ΔV1-ΔVn* of each proportional valve.

[0111] Output execution: The first control quantity in the optimal control sequence obtained from the optimization solution is output to the actuator.

[0112] Wear warning: Continuously analyze the grease injection flow rate at each point. If the flow rate Q_i at a certain point is consistently higher than the average flow rate by a certain percentage (e.g., 20%) for more than a set time (e.g., 1 minute), it is determined that the wear of the sealing ring in that area is accelerating, triggering a warning.

[0113] III. Multi-source information fusion and intelligent collaborative control system

[0114] As attached Figure 9 , 10 As shown, this system serves as the central nervous system for breaking down "information silos" and achieving precise collaboration, including:

[0115] 3.1 Deterministic Networks Based on TSN

[0116] All critical controllers and sensors are interconnected via a Time-Sensitive Networking (TSN) switch (601). The TSN protocol provides deterministic latency and extremely low jitter to standard Ethernet. High-priority fixed time slots are allocated to control data streams with extremely high time requirements, such as sealing pressure commands and hydraulic synchronization pulses, ensuring that their transmission delay is in the microsecond range and essentially constant, laying the network foundation for precise coordination.

[0117] 3.2 Centralized Intelligent Decision-Making and Real-Time Stratigraphic Identification

[0118] The central intelligent controller (602) synchronously acquires multi-source data such as vibration, torque, and thrust at high frequencies (e.g., 1 kHz). Its built-in lightweight convolutional neural network model is used for real-time analysis of vibration spectrum characteristics. This model has a well-defined structure, for example: the input layer receives preprocessed multi-channel vibration spectrum images (time-frequency features), which are then processed through two convolutional layers (for extracting local features), a pooling layer (for dimensionality reduction), and a fully connected layer (for classification and regression), ultimately outputting stratum classification results (e.g., clay, sand, rock) and a continuous hardness index (HI). This model is trained based on a large amount of historical tunneling data and can complete identification in a short time (e.g., within 1 second).

[0119] Subsequently, based on information from three dimensions (formation type, HI, and tunneling depth), the controller queries a pre-calibrated three-dimensional parameter mapping table, calibrated using expert experience and simulation, to obtain recommended target setpoints for tunneling parameters (such as rotational speed, thrust, and mud balance pressure) under the current working conditions. A fuzzy PID controller then fine-tunes these setpoints to smooth the response process.

[0120] 3.3 Synchronous Sinking Control of “Virtual Spindle”

[0121] The segment lowering is performed by multiple sets (e.g., 8 sets) of electro-hydraulic proportional servo cylinders (603). The segment lowering synchronization controller (604) is configured to execute a "virtual spindle" algorithm:

[0122] In each control cycle (e.g., 5ms), the weighted average of the displacements of all cylinders is calculated as the real-time ideal displacement S_virtual of the "virtual spindle". The weight w_i can be dynamically adjusted according to the real-time load pressure of each cylinder. The weight is inversely proportional to the load pressure, for example, w_i = f(P_i). The weight of the cylinder with a large load is appropriately reduced to balance the output of each cylinder.

[0123] Calculate the composite synchronization error E_i for each real cylinder. This error includes not only the tracking error of the virtual spindle displacement, but also the coupling error with the displacement of adjacent cylinders, i.e., E_i = α(S_i - S_virtual) + βΣ(S_i -S_j), where α and β are weighting coefficients.

[0124] The control algorithm incorporates the load pressure of each cylinder as a feedforward to compensate for speed differences caused by uneven load. This method achieves high-precision multi-cylinder electro-hydraulic proportional synchronous control, effectively preventing wellbore deviation.

[0125] 3.4 Dynamic Balance Control of Mud Pressure

[0126] The mud pressure balance controller (605) is configured to use a dynamic matrix control (DMC) algorithm. This algorithm treats the frequency control of the slag discharge pump and the reinjection pump as a coupled multivariable system and incorporates its step response model. The controller continuously optimizes the pump frequency control sequence for future steps based on the deviation between the measured values ​​of the bottom hole pressure and the target values ​​of the wellbore fluid level, and executes the optimal control value at the current moment. This enables the system to proactively and coordinately respond to disturbances such as changes in slag discharge volume and pipeline characteristics, keeping wellbore pressure fluctuations within a very small range (e.g., ±0.1 bar), greatly ensuring the stability of the excavation face.

[0127] IV. Ground Waste Treatment System and Resource Utilization Equipment

[0128] The system realizes a closed-loop cycle of "excavation-treatment-utilization" of excavated soil.

[0129] After being efficiently pumped to the surface by the underground low-pressure generator (701), the slag and mud enter a multi-stage treatment process:

[0130] Intelligent screening and hydrocyclone classification: A dual-shaft intelligent vibrating screen (702) separates large particles. The undersize material enters multiple adjustable-parameter hydrocyclones (703) connected in parallel. The central control system can uniformly adjust the feed pressure and underflow nozzle diameter of each hydrocyclone, thereby accurately controlling the separation particle size (d50) online and stably producing aggregates of different specifications such as coarse sand and fine sand.

[0131] Deep dewatering: The swirling overflow (fine mud) is initially concentrated by a screw press (707) and pressed by a high-pressure diaphragm filter press (708) to produce a dry hard mud cake with a moisture content of less than 35%.

[0132] Online automated production of solidified soil: High-precision loss-in-weight metering feeders (710) are equipped at the bottom of coarse aggregate bins, fine sand bins, mud cake bins, cement bins, admixture bins, etc. The central formula control unit (711) is the core of this module. It receives engineering application instructions (such as "produce C25 wall backfill grout"), retrieves the benchmark formula, and dynamically calculates and finely adjusts the feeding ratio of each material by combining the real-time detected quality data such as the moisture content and fineness modulus of each slag component. Finally, all materials are uniformly mixed in a continuous mixer (712) to directly produce fluidized solidified soil that can be used in engineering. After the filtrate and supernatant are mixed with water quality in the regulating tank (713), they are pumped back down into the well by the reinjection pump (714) to realize the recycling of water and mud materials.

[0133] A brief overview of the system's overall workflow is attached. Figure 10 As shown:

[0134] After system startup, the collaborative control system initializes, and the sealing system establishes an active pressure barrier. The tunneling cycle begins: the ground sensing module identifies soil and rock information in real time, and the central controller dynamically adjusts tunneling parameters; the sealing system synchronously tracks well pressure changes; at the end of one tunneling cycle, the "virtual spindle" algorithm controls multiple sets of hydraulic cylinders to synchronously and precisely lower tunnel segments; the mud balance system dynamically coordinates the slag discharge and reinjection pumps to maintain stable pressure at the excavation face; the excavated soil is immediately transported to the surface, converted into engineering materials via an automated production line, and reused after mud treatment. The entire process is monitored in real time, and the intelligent early warning system analyzes the status of key components and provides maintenance suggestions.

[0135] This invention, through the deep integration and intelligent collaboration of the above modules, constructs a vertical shaft excavation system that can pre-adapt to complex geological structures, optimize tunneling parameters in real time, intelligently resist high-pressure dynamic water pressure, precisely coordinate multiple processes, and achieve green recycling of excavated soil, thereby improving the safety, efficiency, economy, and environmental friendliness of vertical shaft construction on water-rich soft soil foundations.

[0136] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A refined soil cutting system for vertical shaft excavation in water-rich soft soil foundations, characterized in that, include: A formation sensing and adaptive cutting drum assembly, which is connected to the main drive of a vertical shaft rock-embedded drilling rig via a buffer drive connector; A high-water-pressure bidirectional dynamic compensation intelligent sealing system is fitted onto the main drive shaft of the drilling rig host to achieve active pressure tracking sealing during the tunneling process; The multi-source information fusion and intelligent collaborative control system communicates with the cutting drum assembly, sealing system, segment sinking system, mud pressure control system and ground treatment system through a time-sensitive network. The complete set of equipment for closed-loop fine treatment and resource utilization of surface waste soil is connected to the underground system through waste discharge pipelines and mud reinjection pipelines, realizing the real-time sorting, dewatering and resource utilization of waste soil.

2. The system according to claim 1, characterized in that, The formation sensing and adaptive cutting drum assembly includes: The segmented variable stiffness double-layer drum body is welded together from an outer cutting cylinder and an inner support cylinder arranged coaxially. The wall thickness of the outer cutting cylinder is designed differently along the axial direction according to the predicted formation hardness distribution. Radial reinforcing ribs and circumferential reinforcing rings are welded to the inner wall of the predicted hard rock section to form a grid-like reinforcing structure. The inner support cylinder and the outer cutting cylinder form an annular sealed cavity for introducing high-pressure cooling water and accommodating sensor cables. The drum body is sealed at both ends by heavy-duty end caps. The drive end cap is provided with a spline or flange connection structure, and the slag collection end cap is provided with a central slag collection port. The slag collection port is smoothly connected to the conical slag collection groove formed by the inner wall of the inner support cylinder.

3. The system according to claim 2, characterized in that, The cutting roller assembly also includes: Based on the formation prediction model, the zoned differentiated cutting tooth layout divides the drum surface into a high-efficiency main cutting zone and a forced slag collection and guiding zone located at both ends. The cutting tooth base of the high-efficiency main cutting zone is arranged along four logarithmic spiral lines; The cutting teeth of the forced slag collection and guiding zone are arranged in a double-headed reverse spiral line to collect slag and soil towards the center of the drum. All cutting teeth mounting bases are angle-adjustable, allowing for preset installation angles for cutting teeth in different sections based on geological formation predictions.

4. The system according to claim 1, characterized in that, The high-water-pressure bidirectional dynamic compensation intelligent sealing system includes: The five-stage differential pressure sealing rings arranged sequentially along the main drive shaft include two main sealing rings, two isolation sealing rings, and one dustproof sealing ring. Multiple sealing cavities are formed between adjacent sealing rings, which are P0 cavity, P1 cavity, P2 cavity, P3 cavity and P4 collection cavity from the wellbore side inward; The P1 cavity is injected with ultra-high viscosity grease, and its pressure is maintained at a level higher than the wellbore pressure by an active grease injection system. The P4 chamber is connected to a leaking oil tank equipped with a liquid level sensor via a leaking oil pipe.

5. The system according to claim 4, characterized in that, Also includes: A ring-shaped multi-point independent micro-pressure grease injection system, including a pressure distributor ring disposed on the outside of the sealing seat; The pressure distributor has multiple radial oil outlet holes evenly distributed along the circumference, and each oil outlet hole is connected to an independently controlled miniature high-speed piezoelectric ceramic proportional valve. Each proportional valve outlet is connected to a miniature Coriolis mass flow meter for real-time monitoring of the flow rate at the corresponding grease injection point; The grease injection system communicates with the sealing controller and receives valve opening commands calculated based on a model predictive control algorithm.

6. The system according to claim 1, characterized in that, The multi-source information fusion and intelligent collaborative control system includes: A deterministic control backbone network based on time-sensitive networking is used to transmit control data and status information between subsystems; The central intelligent decision-making unit has a built-in lightweight convolutional neural network model for real-time analysis of vibration spectrum and identification of stratum type and hardness index. The segment sinking synchronous controller executes the "virtual spindle" algorithm, using the displacement weighted average of multiple sets of electro-hydraulic proportional servo cylinders as a benchmark to achieve high-precision synchronous control; The mud pressure dynamic balance controller uses a dynamic matrix control algorithm to coordinate the frequency of the slag discharge pump and the reinjection pump, maintaining stable pressure at the excavation face.

7. The system according to claim 6, characterized in that, In the "virtual spindle" algorithm: Each control cycle calculates the weighted average of all cylinder displacements as the virtual spindle displacement; The weight of each cylinder is dynamically adjusted according to its real-time load pressure; The composite synchronization error of the hydraulic cylinder includes the tracking error of its virtual spindle displacement and the coupling error with the displacement of adjacent hydraulic cylinders.

8. The system according to claim 1, characterized in that, The complete set of equipment for closed-loop fine treatment and resource utilization of ground waste includes: The multi-stage sorting unit includes a dual-axis intelligent vibrating screen and a parallel adjustable parameter hydrocyclone for separating aggregates of different particle sizes. The deep dewatering unit, including a screw press thickener and a high-pressure diaphragm filter press, is used to dewater fine slurry into slurry cake with low moisture content. The online automated production unit includes multiple material bins, a high-precision loss-in-weight metering feeder, a continuous mixer, and a central formula control unit, which are used to dynamically adjust the proportions and produce fluidized solidified soil. The mud recycling unit, including a conditioning tank and a reinjection pump, is used to return the treated mud to the well, forming a closed loop.

9. A cooperative control method for the system according to any one of claims 1-8, characterized in that, Includes the following steps: System initialization, establishment of proactive pressure barriers, and startup of all subsystems; Real-time acquisition of vibration, torque, and thrust data; formation information is identified through a neural network model. Based on the formation identification results, the tunneling parameters are dynamically adjusted, and the sealing pressure is simultaneously controlled to track changes in well pressure. After the tunneling process is completed, multiple sets of hydraulic cylinders are controlled by the "virtual spindle" algorithm to achieve synchronous sinking of the tunnel segments; The dynamic matrix control algorithm coordinates the slag discharge and reinjection pumps to maintain wellbore pressure balance. The excavated soil is transported to the surface treatment system for sorting, dewatering, and resource utilization, and the mud is reinjected into the well.

10. The method according to claim 9, characterized in that, Also includes: Real-time monitoring of flow rate and sealing cavity pressure at each grease injection point; trend analysis enables early warning of seal wear. Continuous monitoring of the condition of key components such as cutting gear vibration and bearing temperature enables predictive maintenance.

Images