A green support shield equipment integrated posture control adjustment system and method

Abstract

The present invention belongs to the technical field of shield machines, and specifically relates to an integrated attitude control and adjustment system for green support shield equipment and a method thereof. The system comprises collecting attitude parameters, formation parameters, and equipment status data of the shield machine, and constructing a weighted directed graph using a mutual information algorithm; generating a control signal based on the output parameters of the weighted directed graph, and adjusting the output force of a differential multi-cylinder linkage mechanism through the control signal; controlling a synchronous grouting control unit and a regenerative material injection device mechanically connected to the synchronous grouting control unit based on the control signal, wherein the regenerative material injection device is driven by a cylinder linkage mechanism; utilizing a neural network model integrated in a remote intelligent control center to predict formation changes and modify the control strategy; and forming a closed-loop pressure circuit between the hydraulic output pipeline of the differential multi-cylinder linkage mechanism and the shield machine propulsion cylinder group. The present invention can improve the accuracy, efficiency, and environmental friendliness of shield tunneling, especially in applications under complex formation conditions.

Description

Technical Field

[0001] The present invention belongs to the technical field of shield machines, and in particular relates to an integrated posture control and adjustment system for green support shield equipment and a method thereof. Background Art

[0002] In modern tunnel engineering, shield construction is widely used due to its advantages such as high efficiency and safety. During the tunneling process, precise control of the shield machine's posture (including inclination angle, deflection angle, roll angle, etc.) is crucial to ensuring the accuracy of the tunnel axis, reducing ground disturbance and improving construction quality.

[0003] Problems with existing technologies:

[0004] Traditional shield machine attitude control methods typically rely on experienced operators to make manual adjustments based on measurement data, or employ automated systems based on simple feedback control (such as PID control). Existing shield machine attitude control technologies have several shortcomings. First, underground geological conditions are complex and changeable. Factors such as surrounding rock strength, groundwater pressure, and soil type can affect the tunneling attitude of the shield machine. Traditional control methods often struggle to fully and in real time perceive and analyze the interactions between these complex environmental factors and the shield machine's own conditions (such as tool wear and hydraulic system pressure). This limits the accuracy and response speed of attitude adjustment. For example, when encountering hard rock or sudden changes in formation, traditional control systems may be unable to adjust thrust and torque in a timely and accurate manner, easily causing the attitude to deviate from the designed axis.

[0005] Secondly, existing shield machine attitude control systems typically treat attitude adjustment and support processes (such as synchronous grouting and segment assembly) as relatively independent links. Attitude adjustment mainly relies on the extension and retraction of the propulsion cylinder, while the support process is carried out based on experience or preset parameters. This separate control method makes it difficult to achieve coordinated optimization of attitude adjustment and support processes, which may lead to problems such as stratum settlement and annular seam misalignment. For example, in soft soil layers, if attitude adjustment does not match the grouting volume, it is easy to cause stratum deformation.

[0006] Therefore, in terms of the posture control of shield equipment, the existing technology lacks an overall solution that can fully perceive multi-dimensional information, deeply analyze the complex relationships between dimensions, achieve coordination between posture adjustment and green support, and effectively utilize waste and reduce carbon emissions. Summary of the Invention

[0007] The purpose of the present invention is to provide an integrated posture control and adjustment system for green support shield equipment and a method thereof, which can improve the accuracy, efficiency and environmental protection of shield tunneling, especially in applications under complex stratum conditions.

[0008] The technical solutions adopted by the present invention are as follows:

[0009] In one possible implementation, a method for controlling and adjusting the overall posture of a green support shield equipment is provided, comprising the following steps:

[0010] Collect the shield machine's attitude parameters, formation parameters, and equipment status data, and use the mutual information algorithm to construct a weighted directed graph based on the collected data;

[0011] Based on the output parameters of the weighted directed graph, a control signal is generated, and the output force of the differential multi-cylinder linkage mechanism is adjusted through the control signal;

[0012] Based on the control signal, the synchronous grouting control unit and the regeneration material injection device mechanically connected to the synchronous grouting control unit are controlled, wherein the regeneration material injection device is driven by a cylinder linkage mechanism;

[0013] Optimize segment assembly angles and establish two-way communication between the shield machine and the remote intelligent control center through the communication module;

[0014] Utilize the neural network model integrated in the remote intelligent control center to predict formation changes and modify control strategies;

[0015] Sharing of data, control signals and correction strategies is achieved through the data bus;

[0016] A closed-loop pressure circuit is formed between the hydraulic output pipeline of the differential multi-cylinder linkage mechanism and the shield machine propulsion cylinder group.

[0017] In one possible implementation, the step of collecting the shield machine's attitude parameters, formation parameters, and equipment status data includes:

[0018] Acquire angle, displacement, partition hardness and equipment status data in real time through multiple sensors.

[0019] In one possible implementation, the step of constructing a weighted directed graph using a mutual information algorithm based on the collected data includes:

[0020] A weighted directed graph is constructed based on the mutual information algorithm to quantify the correlation strength between any two dimensions.

[0021] In one possible implementation, the step of generating a control signal includes:

[0022] Dynamically assign weights according to formation types and generate control signals based on the output parameters of the weighted directed graph;

[0023] The steps of controlling the synchronous grouting control unit and the regenerative material injection device mechanically connected to the synchronous grouting control unit based on the control signal include:

[0024] According to the type of stratum, the synchronous grouting volume and the recycled material injection volume are dynamically adjusted. The grouting volume is increased in soft soil layers, and the proportion of recycled materials is increased in hard rock layers.

[0025] In one possible implementation, the step of adjusting the output force of the differential multi-cylinder linkage mechanism includes:

[0026] Through the PID closed-loop feedback algorithm, the output force of the differential multi-cylinder linkage mechanism is adjusted according to the control signal;

[0027] Monitor the actual output force of the differential multi-cylinder linkage mechanism, compare it with the target output force, and adjust the control signal according to the deviation.

[0028] According to another aspect of an embodiment of the present invention, an electronic device is provided. The electronic device includes a memory and a processor; the memory is used to store programs; and the processor executes the programs to implement any one of the aforementioned methods.

[0029] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is provided. The storage medium stores a computer program. When the computer program is executed by a processor, any one of the aforementioned methods is implemented.

[0030] According to yet another aspect of an embodiment of the present invention, a computer program product is provided, including a computer program, which implements any of the aforementioned methods when executed by a processor.

[0031] In one possible embodiment, a green support shield equipment integrated attitude control and adjustment system is provided, comprising:

[0032] The six-degree interactive monitoring module includes a data input interface, a data processing unit, and a storage unit. The data input interface is used to receive shield machine posture parameters, formation parameters, and equipment status data output by multiple sensors. The data processing unit executes the mutual information algorithm and generates a weighted directed graph. The storage unit is used to store data and the weighted directed graph.

[0033] The differential multi-cylinder linkage mechanism includes a main hydraulic cylinder and an auxiliary cylinder array. The hydraulic pipeline of the main hydraulic cylinder is connected to the hydraulic pipeline of the shield machine propulsion cylinder group through a pressure sensor to form a closed-loop pressure circuit, and a differential connection structure is formed through the electromagnetic reversing valve and the proportional relief valve;

[0034] The vector adaptive control unit includes a control signal generation module and a PID feedback control module. The control signal generation module receives the output parameters of the weighted directed graph and generates a control signal. The PID feedback control module receives the output force feedback signal and control signal of the differential multi-cylinder linkage mechanism and outputs a regulation signal to the differential multi-cylinder linkage mechanism according to the PID closed-loop feedback algorithm.

[0035] Green support collaborative module, including synchronous grouting control unit and segment assembly angle optimization unit;

[0036] A recycled material injection device and a cylinder linkage mechanism mechanically connected thereto;

[0037] a mechanical coupling structure for connecting the regenerative material injection device with the auxiliary cylinder array of the differential multi-cylinder linkage mechanism;

[0038] The cylinder linkage mechanism receives the control signal of the vector adaptive control unit through the signal input port and drives the regeneration material injection device through the mechanical connection structure;

[0039] The remote intelligent control center establishes two-way communication with the six-degree interactive monitoring module and the vector adaptive control unit through the communication module. The remote intelligent control center integrates the LSTM neural network model to predict formation changes;

[0040] The data bus connects the six-degree interactive monitoring module, the vector adaptive control unit and the remote intelligent control center to form a real-time sharing channel for multi-source data.

[0041] In a possible implementation, the six-degree interactive monitoring module receives output signals from the inclination sensor, the displacement sensor, the formation hardness sensor, and the tool force feedback unit via a data interface;

[0042] The output signals include six-dimensional angle dimension, displacement vector dimension, partition hardness coefficient dimension, equipment status dimension, hydraulic cylinder linkage dimension and support adaptation dimension;

[0043] The six-degree interactive monitoring module includes: data fusion unit, correlation analysis unit, dynamic weight allocation unit and prediction correction unit.

[0044] In a possible implementation, the data fusion unit includes a processing module that executes a feature co-occurrence relationship modeling process. The data fusion unit also includes a node mapping module, a weight calculation module, and a topology structure optimization module.

[0045] In a possible implementation, the vector adaptive control unit includes: a formation type identification module, a weight adaptive adjustment module, a control signal generation module, and a PID feedback control module.

[0046] In a possible implementation, the formation type identification module executes a formation type identification algorithm, which determines the current formation type based on the formation parameters output by the six-degree interactive monitoring module.

[0047] The technical effects achieved by the present invention are:

[0048] The present invention is an integrated attitude control and adjustment system for shield equipment based on a six-degree interaction model. The six-degree interaction monitoring module collects multi-dimensional data such as the attitude, formation and equipment status of the shield machine, and uses the mutual information algorithm to construct a weighted directed graph to quantify the correlation strength between each dimension. The vector adaptive control unit receives the output parameters of this weighted directed graph and generates a refined control signal. On the one hand, the control signal drives the differential multi-cylinder linkage mechanism to adjust the attitude of the shield machine. Through the closed-loop pressure circuit formed by the main hydraulic cylinder and the propulsion cylinder group, precise control of the propulsion force is achieved. At the same time, the auxiliary cylinder array is linked with the regenerative material injection device to dynamically inject regenerative materials according to the changes in the formation to actively support the formation.

[0049] The present invention coordinates the control signal with the green support coordination module to optimize the grouting strategy and the segment assembly angle. The remote intelligent control center communicates with the shield machine in real time through communication, uses the LSTM neural network model to predict stratum changes, and modifies the control strategy to ensure the predictability and adaptability of the shield machine's posture adjustment. The data bus ensures the real-time sharing of multi-source data and builds an information closed loop. Through in-depth interactive analysis of multi-dimensional data, it realizes the overall, adaptive and coordinated control of the shield machine's posture under complex stratum conditions, solves the problems of traditional control methods being difficult to cope with complex stratum changes, delayed posture adjustment, and untimely support, and significantly improves the accuracy, safety and efficiency of shield excavation. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 is a flow chart of the method of the present invention;

[0051] Figure 2 It is a schematic diagram of the system structure of the present invention. DETAILED DESCRIPTION

[0052] In order to make the purpose and advantages of the present invention more clearly understood, the present invention is described in detail below with reference to the following examples. It should be understood that the following text is only used to describe one or more specific embodiments of the present invention and does not strictly limit the scope of protection of the present invention.

[0053] It should be noted that the terms "first", "second", etc. in the description and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0054] According to an embodiment of the present invention, a method embodiment of a method for overall posture control adjustment of green support shield equipment is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0055] like Figure 1 As shown, a method for controlling and adjusting the overall posture of a green support shield equipment includes the following steps:

[0056] S1. Collect the shield machine's attitude parameters, formation parameters, and equipment status data. Based on the collected data, use the mutual information algorithm to construct a weighted directed graph to reveal the complex relationships between various dimensions.

[0057] S2. Generate a control signal based on the output parameters of the weighted directed graph. Through the control signal, adjust the output force of the differential multi-cylinder linkage mechanism to achieve precise adjustment of the shield machine posture;

[0058] S3. Based on the control signal, control the synchronous grouting control unit and the regeneration material injection device mechanically connected to the synchronous grouting control unit, wherein the regeneration material injection device is driven by a cylinder linkage mechanism;

[0059] S4. Optimize the segment assembly angle and establish two-way communication between the shield machine and the remote intelligent control center through the communication module;

[0060] S5. Use the neural network model integrated by the remote intelligent control center to predict formation changes and modify the control strategy;

[0061] S6. Sharing of data, control signals and correction strategies is achieved through the data bus;

[0062] S7. A closed-loop pressure circuit is formed between the hydraulic output pipeline of the differential multi-cylinder linkage mechanism and the shield machine propulsion cylinder group to maintain the dynamic balance of the propulsion force.

[0063] According to the above steps, the six-degree interaction model can be used to perceive multi-dimensional information such as the shield machine posture, stratum and equipment status, and use the mutual information algorithm to quantify the correlation strength between each dimension, and construct a weighted directed graph as the control basis; and further realize the coordinated linkage of posture adjustment, green support and remote intelligent regulation, forming a closed-loop control process of "perception-analysis-execution-feedback", so that the system running this method can respond to complex stratum challenges more intelligently and efficiently, and significantly improve excavation efficiency and tunnel forming quality.

[0064] According to the above step S1, the output signals from sensors such as the inclination sensor, displacement sensor, formation hardness sensor and tool force feedback unit are received through the data input interface of the six-degree interactive monitoring module to obtain six-dimensional data such as angle dimension, displacement vector dimension, partition hardness coefficient dimension, equipment status dimension, hydraulic cylinder linkage dimension and support adaptation dimension.

[0065] Furthermore, the data processing unit of the six-degree interaction monitoring module uses the mutual information algorithm to perform:

[0066] ;

[0067] Among them, X and Y are any two feature nodes (such as "tool force feedback" and "formation hardness"), which are used to quantify the nonlinear correlation strength between any two dimensions and construct a weighted directed graph with six dimensions as nodes and mutual information values ​​as weights.

[0068] According to the above step S2, the control signal generation module of the vector adaptive control unit receives the output parameters such as the topological structure and edge weight of the weighted directed graph, combines the recognition results of the formation type recognition module and the adjustment weight of the weight adaptive adjustment module, and generates the control signal for controlling the differential multi-cylinder linkage mechanism and the green support collaborative module.

[0069] Furthermore, the PID feedback control module of the vector adaptive control unit receives the control signal and the output force feedback signal of the differential multi-cylinder linkage mechanism (obtained through the pressure sensor), calculates the adjustment instruction according to the PID closed-loop feedback algorithm, and outputs it to the control unit of the differential multi-cylinder linkage mechanism to adjust the oil or gas flow of the main hydraulic cylinder and the auxiliary cylinder array to change its output force.

[0070] According to step S3 above, the synchronous grouting control unit of the green support collaborative module receives control signals and adjusts the flow rate, pressure, and ratio of the synchronous grouting. Simultaneously, it receives control signals and drives the regenerative material injection device (including the calcium carbonate-based cementitious material synthesis unit and the cylinder-linked nozzle) via the cylinder linkage mechanism to inject the regenerative material.

[0071] According to the above step S4, the segment assembly angle optimization unit of the green support collaborative module calculates the optimal segment assembly angle based on the formation parameters and the shield machine posture parameters, and is executed by the 5G communication module to realize the two-way transmission of data and instructions between the six-degree interactive monitoring module on the shield machine side, the vector adaptive control unit and the remote intelligent control center.

[0072] According to the above step S5, the LSTM neural network model of the remote intelligent control center analyzes and predicts the received formation parameters, and sends the prediction results and the revised control strategy back to the shield machine side through the 5G communication module.

[0073] According to the above step S6, the data bus is used to connect the six-degree interactive monitoring module, the vector adaptive control unit and the remote intelligent control center to ensure the real-time and efficient flow of information between the modules.

[0074] According to the above step S7, it is achieved through physical connection, and the pressure difference is monitored by the pressure sensor in the closed-loop pressure circuit, and the signal is fed back to the vector adaptive control unit for adjustment.

[0075] Through the coordinated execution of the above steps, comprehensive perception, intelligent analysis, precise control and coordinated support of the shield tunneling process are achieved. Six-degree data acquisition and mutual information modeling provide an in-depth understanding of the complex environment; vector adaptive control and green support collaborate to achieve the coordinated optimization of posture and support; remote intelligent control provides forward-looking prediction and strategy correction. The entire process forms an efficient and robust closed-loop control system, which significantly improves the efficiency, safety and tunnel forming quality of shield tunneling.

[0076] In the above steps, the collected data may include the output signals of the inclination sensor, displacement sensor, formation hardness sensor and tool force feedback unit; the constructed weighted directed graph can quantify the correlation strength between any two dimensions, for example, the mutual information weight between tool force feedback and formation hardness is ≥0.8; the generated control signal can dynamically assign weights according to the formation type, and based on the output parameters of the weighted directed graph, the fuzzy control algorithm is applied to generate a nonlinear control strategy; the output force of the differential multi-cylinder linkage mechanism can be adjusted by the PID closed-loop feedback algorithm, thereby shortening the response time; the control of synchronous grouting and recycled material injection can be dynamically adjusted according to the formation type Dynamically adjust the injection volume and proportion. For example, after the grouting volume of the soft soil layer is increased, the proportion of recycled materials in the hard rock layer is increased. In particular, by using different grouting materials, the different properties of the hard rock layer can be improved, such as through pouring; the optimization of the segment assembly angle can use a preset mathematical model or optimization algorithm, such as calculating the optimal angle based on a geometric model or finite element analysis; 5G communication can use a low-latency, high-bandwidth industrial-grade 5G module; the LSTM model can predict sudden changes in the hardness of the stratum ahead and realize early prediction and warning; in the closed-loop pressure circuit, the pressure difference between the main hydraulic cylinder and the propulsion cylinder group can be monitored and adjusted to maintain the propulsion force balance.

[0077] As an optional embodiment, the step of collecting the shield machine's attitude parameters, formation parameters, and equipment status data includes:

[0078] Acquire angle, displacement, partition hardness and equipment status data in real time through multiple sensors.

[0079] According to the above steps, by using multiple sensors to obtain angle, displacement, partition hardness and equipment status data in real time, the comprehensiveness and real-time nature of data collection are further enhanced. A six-degree interaction model is constructed based on the collected data, which is also the basic data source for subsequent intelligent analysis and control decisions; the angle data reflects the pitch, roll and heading of the shield machine, and is obtained through the inclination sensor; the displacement data reflects the longitudinal, lateral and elevation deviations of the shield machine relative to the design axis, and is obtained through the displacement sensor; the partition hardness data reflects the softness and hardness of the stratum in different areas in front of the cutterhead, and is obtained through comprehensive analysis of data from the stratum hardness sensor and the tool force feedback unit; the equipment status data reflects the operating status of key components of the shield machine, such as cutterhead torque, thrust, speed, tool wear, hydraulic system pressure and flow, etc., and is obtained through the tool force feedback unit and other equipment-provided sensors; by collecting these data in real time through multiple sensors, the system can fully perceive the excavation environment and equipment status, providing a rich and reliable data foundation for subsequent intelligent analysis and control.

[0080] This embodiment is completed by the data input interface of the six-degree interactive monitoring module and the various sensors connected thereto. The multiple sensors include but are not limited to inclination sensors, displacement sensors, formation hardness sensors and tool force feedback units. The inclination sensor measures the pitch angle and roll angle of the shield machine in real time and outputs angle data. The displacement sensor measures the longitudinal and lateral deviations of the shield machine relative to the design axis in real time and outputs displacement data. The formation hardness sensor and the tool force feedback unit monitor the cutterhead cutting parameters in real time, comprehensively analyze and generate a partitioned hardness map of the area in front of the cutterhead, and output the partitioned hardness data. The tool force feedback unit and other equipment sensors monitor the cutterhead torque, thrust, speed, hydraulic system pressure flow, etc. in real time and output equipment status data. These multi-sensor data are transmitted to the data input interface of the six-degree interactive monitoring module in real time at a high frequency (for example, 10-100 times per second).

[0081] Furthermore, through the real-time collection of angle, displacement, partition hardness and equipment status data by multiple sensors, the comprehensiveness, real-timeness and accuracy of the input information of the six-degree interactive model are ensured. This comprehensive data perception capability is the prerequisite for achieving precise posture control and collaborative support. For example, the partition hardness data obtained in real time can be directly used to adjust the cutting parameters and posture adjustment force of the cutterhead; the equipment status data obtained in real time can be used to monitor the health of the equipment and adjust the excavation parameters when necessary. The real-time data collection method based on multiple sensors can improve the system's perception of complex formation environments and equipment operating status.

[0082] According to the above, the type and number of multiple sensors can be configured according to the model, size and excavation stratum conditions of the shield machine. For example, a gyroscope sensor can be added to obtain angular velocity information, an acoustic sensor or a seismic sensor can be added to detect the stratum structure ahead, and a temperature sensor and a vibration sensor can be added to monitor the equipment operation status. The data acquisition system should have high synchronization to ensure the consistency of the timestamps between different sensor data to facilitate subsequent data fusion and correlation analysis. The data acquisition frequency should be high enough to capture the rapid changes in the stratum and equipment status. The collected raw data can be stored in the storage unit of the six-degree interactive monitoring module for historical data analysis and model training.

[0083] As an optional embodiment, the steps of constructing a weighted directed graph using a mutual information algorithm based on the collected data include:

[0084] A weighted directed graph is constructed based on the mutual information algorithm to quantify the correlation strength between any two dimensions.

[0085] Based on the above, the mutual information algorithm is used to quantify the correlation strength between any two dimensions, and a weighted directed graph is constructed based on this. Mutual information is a measure of the mutual dependence between two random variables, which can capture linear and nonlinear correlations. By calculating the mutual information values ​​between the six-dimensional data and using these values ​​as the weights of the edges in the graph structure, a weighted directed graph reflecting the strength and direction of the correlation between the six dimensions can be constructed. This graph structure intuitively shows the mutual influence relationship between the dimensions, providing an important basis for the subsequent generation of control strategies.

[0086] Furthermore, the data processing unit (especially the correlation analysis unit and the weight calculation module) of the six-dimensional interactive monitoring module first pre-processes the collected six-dimensional data, including data cleaning, standardization and discretization (if necessary). Then, for any two dimensions X and Y (for example, the "tool force feedback dimension" and the "formation hardness dimension"), the mutual information value between them is calculated:

[0087] , where p(x,y) is the joint probability distribution of X and Y, and p(x) and p(y) are the marginal probability distributions of X and Y, respectively. A higher mutual information value indicates a stronger correlation between the two dimensions. Next, a weighted directed graph is constructed, using the six degrees of dimension as nodes and the calculated mutual information values ​​as edge weights. The direction of the edges can be determined based on experience or further causal analysis. For example, if changes in the "tool force feedback dimension" typically precede changes in the "formation hardness dimension," a directed edge from tool force feedback to formation hardness can be established. The output of this step is a weighted directed graph reflecting the correlation between the six degrees of dimension.

[0088] Furthermore, by using the mutual information algorithm to construct a weighted directed graph, the interdependence between the six dimensions is more accurately quantified. For example, the mutual information algorithm can discover the nonlinear relationship between tool force feedback and formation hardness, or the complex correlation between groundwater pressure and formation deformation; the constructed weighted directed graph intuitively displays these correlations, allowing the control system to understand the mutual influence paths and intensities between various factors, such as the correlation path weight of "groundwater pressure → grouting flow → segment assembly angle". This graph-based correlation analysis provides a solid data foundation for subsequent control signal generation and strategy optimization.

[0089] It should be further explained that in the above scheme, the calculation of the mutual information value can be based on histogram estimation, kernel density estimation or k-nearest neighbor estimation. In order to process continuous data, appropriate discretization and binning can be performed first. When constructing a weighted directed graph, a threshold can be set to retain only edges with mutual information values ​​higher than the threshold. The topological structure and edge weights of the graph can be dynamically updated as the excavation process progresses to reflect changes in formation conditions and equipment status. In addition to the mutual information algorithm, Granger causality analysis, transfer entropy and other methods can also be combined to determine the edge direction of the graph and more accurate association metrics. The constructed weighted directed graph is an important input for the vector adaptive control unit to generate control signals.

[0090] As an optional embodiment, the step of generating a control signal includes:

[0091] Weights are dynamically assigned according to formation types, and control signals are generated based on the output parameters of the weighted directed graph.

[0092] Based on the above, during shield tunneling, the key factors affecting attitude and support vary under different strata conditions. By introducing a mechanism for dynamically assigning weights based on stratum type into the control signal generation step and combining it with the output parameters of a weighted directed graph, adaptive adjustment of the control strategy is achieved. The current stratum type identified by the stratum type recognition module is used by the dynamic weight assignment unit to adjust the priority weights of the six dimensions in the control decision-making process. For example, in hard rock strata, dimensions such as tool force feedback and zone hardness should be given higher weights; in soft soil strata, dimensions such as stratum deformation and hydraulic cylinder pressure should be given more attention. The control signal generation module then receives the weighted output parameters of the weighted directed graph (such as the values ​​of key nodes and the weights of associated paths) after weight adjustment and generates a control signal based on these parameters. This control signal represents the control command for the differential multi-cylinder linkage mechanism, the synchronous grouting control unit, and the segment assembly angle optimization unit. This method of dynamically assigning weights based on stratum type and combining the analysis results of the weighted directed graph to generate control signals ensures that the control strategy can be optimized for the current situation and significantly improves the effectiveness of attitude control and coordinated support.

[0093] Furthermore, the control signal generation module, the stratum type identification module and the weight adaptive adjustment module of the vector adaptive control unit are coordinated to complete the task; first, the stratum type identification module determines the stratum type of the current excavation, and then, the weight adaptive adjustment module dynamically adjusts the weights of the six dimensions in the control decision-making from the preset weight mapping table or through the adaptive algorithm according to the stratum type, for example, in the hard rock layer, the tool force feedback weight is increased to 0.4, and in the soft soil layer, the hydraulic cylinder pressure weight is increased to 0.5. Then, the control signal generation module receives the output parameters of the weighted directed graph, such as the key influencing factors The control signal generation module combines the current values ​​of the nodes (identified by the topology optimization module), the current states of the important associated paths (identified by the associated analysis unit), etc., and generates control signals through control algorithms (such as fuzzy control, model predictive control, or control algorithms based on graph neural networks). The control signals include posture adjustment instructions for the differential multi-cylinder linkage mechanism (such as the extension and contraction amount or pressure of each hydraulic cylinder), grouting instructions for the synchronous grouting control unit (such as grouting volume, pressure, and ratio), and assembly angle instructions for the segment assembly angle optimization unit.

[0094] Furthermore, by dynamically assigning weights based on formation type and generating control signals based on the output parameters of a weighted directed graph, the control strategy is made intelligent and adaptive. Dynamic weight allocation ensures that under different formation conditions, the system can prioritize the most critical influencing factors, improving control efficiency and accuracy. Generating control signals based on the output parameters of the weighted directed graph enables control decisions to fully utilize the complex correlation information between the six degrees of dimension, achieving more refined and comprehensive control. For example, in sand layers, the correlation between groundwater pressure and formation deformation increases. The system will increase the weight of groundwater pressure and formation deformation data and generate corresponding grouting and posture adjustment instructions to prevent formation instability. This intelligent control signal generation method significantly improves the system's ability to cope with complex formation challenges.

[0095] In the above scheme, dynamic weight allocation can be implemented using a fuzzy rule base, expert system, or adaptive weight adjustment algorithm based on reinforcement learning. Weight mapping tables can be constructed and optimized based on historical excavation data and expert experience. The control signal generation algorithm can employ a fuzzy control-based model, using the output parameters of a weighted directed graph as input to the fuzzy controller, with the stratum type as the basis for selecting fuzzy rules and outputting control instructions. Alternatively, a graph neural network-based model can be employed, using a weighted directed graph as input and learning the mapping from graph structure to control instructions through the neural network. The output frequency of the control signal should match the response speed of the actuator to ensure real-time control.

[0096] As an optional embodiment, the step of controlling the synchronous grouting control unit and the regenerative material injection device mechanically connected to the synchronous grouting control unit based on the control signal includes:

[0097] According to the type of stratum, the synchronous grouting volume and the recycled material injection volume are dynamically adjusted. The grouting volume is increased in soft soil layers, and the proportion of recycled materials is increased in hard rock layers.

[0098] Based on the above, the specific process of grouting by controlling the synchronous grouting control unit and the recycled material injection device is as follows. Synchronous grouting is an important support method during shield tunneling, and its effect directly affects the tunnel forming quality and stratum stability. By dynamically adjusting the synchronous grouting volume and the recycled material injection volume according to the stratum type, an intelligent and adaptive grouting strategy is achieved. This step is performed by the synchronous grouting control unit of the green support collaboration module, which receives the control signal generated by the vector adaptive control unit and dynamically adjusts the grouting parameters based on the current stratum type identified by the stratum type identification module. In soft soil layers, since the stratum is prone to deformation and collapse, a larger grouting volume is required to quickly fill the voids and stabilize the surrounding rock, so the grouting volume is increased. In hard rock layers, there are relatively few cracks and voids, and recycled materials have good filling and consolidation properties. The proportion of recycled materials in the injected material can be increased to achieve resource recycling and enhance the support effect. The method of dynamically adjusting the grouting volume and the recycled material ratio according to the stratum type ensures that the grouting strategy can optimally match the actual stratum conditions, improving support efficiency and environmental benefits.

[0099] Furthermore, the synchronous grouting control unit of the green support collaborative module is executed and linked to the recycled material injection device. The synchronous grouting control unit receives a control signal from the vector adaptive control unit, which contains instructions on the grouting volume and the recycled material injection volume. At the same time, the synchronous grouting control unit obtains the current stratum type identified by the stratum type identification module. According to the stratum type, the synchronous grouting control unit dynamically adjusts the grouting volume and the recycled material injection volume parameters in the control signal. For example, when a soft soil layer is identified, the grouting control unit will increase the total grouting volume to ensure that the soil voids can be fully filled, prevent stratum settlement, and increase the grouting volume. When a hard rock layer is identified, the grouting control unit will increase the proportion of recycled materials in the injected materials, increase the proportion of recycled materials, and reduce the use of virgin materials such as cement. The synchronous grouting control unit achieves precise control of the grouting volume and the recycled material injection volume by controlling the cylinder linkage nozzle of the grouting pump and the recycled material injection device.

[0100] This embodiment realizes the intelligence and pertinence of the grouting strategy by dynamically adjusting the synchronous grouting volume and the recycled material injection volume according to the stratum type. Increasing the grouting volume in the soft soil layer can effectively control the stratum deformation and prevent collapse; increasing the proportion of recycled materials in the hard rock layer not only realizes the recycling of resources, but also can utilize the good filling and consolidation properties of recycled materials to improve the crack filling effect. The adaptive grouting strategy significantly improves the support effect and material utilization rate, and reduces resource waste and environmental pollution.

[0101] In the above scheme, dynamic adjustment of grouting and recycled material injection volumes can be achieved using a lookup table or fuzzy control algorithm based on formation type and parameters. For example, a grouting parameter rule library can be established to determine the optimal grouting volume and recycled material ratio based on different formation types and parameters (such as formation deformation rate and groundwater pressure). The recycled material injection volume can be expressed as a portion of the total injection volume or mixed with the traditional slurry in a certain proportion. The cylinder-linked nozzle can control the local injection volume by adjusting its movement speed and injection time. The grouting process can be adjusted through feedback based on formation deformation monitoring data, for example, automatically increasing the grouting volume when the formation settlement rate exceeds a threshold.

[0102] As an optional embodiment, the step of adjusting the output force of the differential multi-cylinder linkage mechanism includes:

[0103] Through the PID closed-loop feedback algorithm, the output force of the differential multi-cylinder linkage mechanism is adjusted according to the control signal;

[0104] Monitor the actual output force of the differential multi-cylinder linkage mechanism, compare it with the target output force, and adjust the control signal according to the deviation.

[0105] As mentioned above, the differential multi-cylinder linkage mechanism is the actuator for the shield machine's attitude adjustment. Precise control of its output force is key to achieving precise attitude adjustment. By adopting a PID closed-loop feedback algorithm, the output force of the differential multi-cylinder linkage mechanism is precisely adjusted based on the control signal generated by the vector adaptive control unit. The PID (proportional-integral-differential) algorithm is a classic closed-loop control algorithm that calculates the control variable based on the deviation between the system output and the target value to reduce the deviation and improve the system's response speed and stability. In this step, the PID feedback control module of the vector adaptive control unit receives the control signal (representing the desired output force or attitude target) and the output force feedback signal of the differential multi-cylinder linkage mechanism (obtained via a pressure sensor or force sensor), calculates the deviation between the current output force and the desired output force, and then calculates the adjustment signal based on the PID algorithm. The signal is output to the control unit of the differential multi-cylinder linkage mechanism, which controls the solenoid reversing valve and proportional relief valve to adjust the oil or gas flow in the main hydraulic cylinder and auxiliary cylinder array, thereby changing the output force to approach the desired value.

[0106] Furthermore, the PID feedback control module of the vector adaptive control unit is executed. The PID feedback control module receives a control signal from the control signal generation module, which represents the expected output force of the differential multi-cylinder linkage mechanism or the force required to achieve the target posture. At the same time, the PID feedback control module receives the actual output force feedback signal of the differential multi-cylinder linkage mechanism, which is obtained through a pressure sensor or a force sensor. The PID feedback control module calculates the deviation between the expected output force and the actual output force, and calculates the adjustment amount according to the proportional (P), integral (I), and differential (D) control laws of the PID algorithm. The proportional term produces a control effect based on the current deviation size; the integral term produces a control effect based on the historical deviation accumulation to eliminate steady-state errors; the differential term produces a control effect based on the deviation change rate to suppress overshoot and improve response speed. The calculated adjustment signal is output to the control unit of the differential multi-cylinder linkage mechanism to control the electromagnetic reversing valve and the proportional relief valve, adjust the oil or gas flow of the main hydraulic cylinder and the auxiliary cylinder array, thereby changing its output force to approach the expected value.

[0107] Furthermore, by using a PID closed-loop feedback algorithm to regulate the output force of the differential multi-cylinder linkage mechanism, precise control of the attitude adjustment actuator is achieved. The closed-loop feedback mechanism can monitor the actual output force in real time and dynamically adjust it based on the deviation, overcoming the shortcoming of open-loop control that is sensitive to external disturbances. As a mature and stable control algorithm, the PID algorithm can effectively reduce attitude deviations and improve the response speed and accuracy of attitude adjustment. For example, when a shield machine attitude deviates slightly, the PID controller can accurately calculate the required small adjustment force based on the deviation size and rate of change, and quickly apply it through the differential multi-cylinder linkage mechanism to restore the attitude to the target position. This precise force control is the key to achieving millimeter-level or even submillimeter-level attitude adjustment.

[0108] It should be further clarified that the PID controller parameters (Kp, Ki, Kd) can be tuned based on the dynamic characteristics and control requirements of the differential multi-cylinder linkage mechanism. Parameter tuning can be performed using empirical methods, the Ziegler-Nichols method, or other optimization algorithms. To improve control performance, modified PID algorithms such as fuzzy PID, adaptive PID, or a combination of feedforward and PID control can be employed. Output force feedback signals can be obtained using pressure sensors or force sensors installed on the main hydraulic cylinder and auxiliary cylinder to ensure feedback signal accuracy. The output frequency of the control signal should match the calculation speed of the PID controller and the response speed of the actuator.

[0109] It should be further explained that the actual output force of the differential multi-cylinder linkage mechanism is monitored, compared with the target output force, and the control signal is adjusted according to the deviation. This step is performed by the PID feedback control module of the vector adaptive control unit. First, the actual output force signal of the differential multi-cylinder linkage mechanism is obtained through a pressure sensor or force sensor. This signal reflects the actual thrust or lateral force applied by the current differential multi-cylinder linkage mechanism to the shield machine posture. Then, the actual output force signal is compared with the target output force represented by the control signal output by the control signal generation module of the vector adaptive control unit, and the deviation e(t) = target output force - actual output force is calculated. Then, the PID feedback control module calculates the adjustment signal according to the PID control law based on the deviation e(t) and its integral and differential: ;

[0110] Where, u(t) is the regulating signal output by the controller (such as the opening of the solenoid reversing valve and the pressure setting value of the proportional relief valve);

[0111] e(t): Deviation at the current moment, i.e., the difference between the target output force and the actual output force e(t) = target output force - actual output force;

[0112] 、 、 : Proportional, integral, and differential gains need to be adjusted according to the dynamic characteristics of the system;

[0113] Finally, the calculated control signal u(t) is output to the control unit of the differential multi-cylinder linkage mechanism. This control unit controls the electromagnetic reversing valve and proportional relief valve to adjust the oil or gas flow in the main hydraulic cylinder and auxiliary cylinder array, thereby changing their output force. This process is repeated in a cycle, with real-time monitoring, comparison, and adjustment, until the deviation is reduced to within the allowable range.

[0114] Furthermore, by monitoring the actual output force, comparing it with the target value and performing PID adjustment according to the deviation, precise closed-loop control of the output force of the differential multi-cylinder linkage mechanism is achieved, which can effectively overcome interference factors such as the nonlinearity of the hydraulic / pneumatic system, friction, and changes in formation reaction force, and ensure that the differential multi-cylinder linkage mechanism can accurately and quickly execute the attitude adjustment instructions issued by the vector adaptive control unit; for example, when the formation resistance suddenly increases and causes the actual output force to deviate from the target value, the PID controller can immediately detect the deviation and quickly adjust the control signal to increase the output force so that the actual output force returns to the target value. This precise force control is an important guarantee for achieving high-precision attitude adjustment and maintaining propulsion force balance.

[0115] In the above scheme, the actual output force can be monitored directly by force sensors installed on the piston rods of the main hydraulic cylinder and the auxiliary cylinder, or by measuring the cavity pressure and calculating it in combination with the effective area of ​​the cylinder (such as through a pressure sensor). The target output force is generated by the vector adaptive control unit based on the analysis results and control strategy of the six-degree interaction model. The calculation of the deviation and the implementation of the PID control algorithm can be completed in the processor of the vector adaptive control unit. The adjustment of the control signal can be a control current or voltage signal directly output to the electro-hydraulic proportional valve group. In order to improve the dynamic performance of the system, a combination of feedforward control and PID control can be adopted to adjust the control signal in advance according to the predicted results of the formation change.

[0116] As an optional embodiment, the step of monitoring the actual output force of the differential multi-cylinder linkage mechanism includes: acquiring the actual output force of the differential multi-cylinder linkage mechanism through a pressure sensor or a force sensor.

[0117] As mentioned above, obtaining the actual output force of the differential multi-cylinder linkage mechanism through pressure sensors or force sensors emphasizes the use of direct or high-precision sensing methods to obtain feedback signals. Pressure sensors can be installed in the oil or gas chambers of the main hydraulic cylinder and auxiliary cylinder array. By measuring the pressure within the chamber and combining it with the effective area of ​​the piston, the output force of the cylinder can be calculated. Force sensors can be directly installed at the connection between the cylinder and the shield machine structure or on the piston rod to directly measure the force applied by the cylinder. By using these sensors, the system can accurately and in real time obtain the actual output force of the differential multi-cylinder linkage mechanism, providing reliable feedback signals for PID closed-loop feedback control, ensuring the performance and stability of the control system.

[0118] As an optional embodiment, the remote intelligent control center predicts the formation changes and corrects the steps of generating the control signal.

[0119] According to the above, the remote intelligent control center uses the formation prediction results to make proactive corrections to the control strategy. By using the LSTM neural network model integrated in the remote intelligent control center to predict formation changes, the prediction results are fed back to the shield machine side for correction of the control strategy. This step is performed before or during the generation of the control signal. According to the future formation change trend predicted by the remote intelligent control center (for example, the appearance of a hard rock layer or a soft, water-rich formation ahead), the control signal generated by the vector adaptive control unit is corrected. For example, if it is predicted that a hard rock layer is about to be entered ahead, the system can adjust the control signal in advance, increase the propulsion force, adjust the cutterhead speed and torque, and prepare posture adjustment and grouting strategies suitable for hard rock. This control strategy correction based on predictive information enables the system to respond to sudden formation changes in advance, reduce posture deviations and safety risks, and improve the smoothness and efficiency of tunneling.

[0120] As an optional embodiment, the optimal segment assembly angle is calculated based on stratum parameters and shield machine attitude parameters. Segment assembly is a key step in tunnel formation. The accuracy of the assembly angle directly affects the misalignment and leakage of the tunnel annular joint, which in turn affects the stability and waterproof performance of the tunnel structure. The optimal segment assembly angle is calculated based on real-time acquired stratum parameters and shield machine attitude parameters. Stratum parameters (such as stratum deformation rate and stratum stress) reflect the impact of the stratum on the compression and deformation of the segment ring. Shield machine attitude parameters (such as shield machine pitch angle, roll angle, heading angle, and axis deviation) reflect the installation position and orientation of the segment ring. The purpose of optimizing the segment assembly angle is to minimize annular joint misalignment and ensure tunnel formation quality by adjusting the segment ring installation angle under current stratum conditions and shield machine attitude. This step is performed by the segment assembly angle optimization unit of the green support collaborative module. Using a mathematical model or optimization algorithm, it comprehensively considers factors such as stratum deformation, shield machine attitude, segment ring geometry, and connection method to calculate the optimal assembly angle for each segment.

[0121] Furthermore, a preset mathematical model or optimization algorithm is used to calculate the optimal segment assembly angle, and the preset mathematical model or optimization algorithm is used to complete this calculation. The preset mathematical model can be a model based on geometry, mechanics or finite element theory, which is used to describe the relationship between stratum deformation, shield machine posture and segment ring shape and position, and establish an optimization function with annular seam misalignment as the target. The optimization algorithm is used to solve this optimization function and find the segment assembly angle combination that minimizes the annular seam misalignment. For example, a geometric model based on the shield machine axis deviation, annular surface deviation and stratum convergence deformation can be established to calculate how to adjust the segment angle in the current state to make the annular seam flush, or an optimization algorithm such as genetic algorithm and particle swarm algorithm can be used to search for the optimal solution with the minimum annular seam misalignment as the objective function and the segment angle as the optimization variable. The application of these mathematical models and optimization algorithms makes the determination of the segment assembly angle no longer rely on experience, but is based on scientific calculations, which significantly improves the assembly accuracy.

[0122] As an optional embodiment, the pressure difference between the master hydraulic cylinder and the thrust cylinder group is monitored and regulated in a closed-loop pressure circuit. During shield tunneling, the master hydraulic cylinder, as part of the differential multi-cylinder linkage mechanism, outputs force from the master hydraulic cylinder and the thrust cylinder group together contribute to the total thrust of the shield machine. The force balance between the two directly impacts the shield machine's propulsion and posture adjustment. By using a pressure sensor in the closed-loop pressure circuit to monitor the pressure difference between the master hydraulic cylinder and the thrust cylinder group, and using this pressure difference as a feedback signal for adjustment, monitoring the pressure difference provides real-time insights into the force distribution between the two. Adjusting the pressure difference aims to maintain a force balance between the two, ensuring that the master hydraulic cylinder's output force for posture adjustment does not adversely affect the normal propulsion of the thrust cylinder group, or that the master hydraulic cylinder can coordinate appropriately when the thrust cylinder group encounters changes in formation resistance. This monitoring and adjustment mechanism helps maintain the stability of the shield machine's overall propulsion force and smooth posture adjustment.

[0123] According to another aspect of an embodiment of the present invention, an electronic device is provided. The electronic device includes a memory and a processor; the memory is used to store programs; and the processor executes the programs to implement any one of the aforementioned methods.

[0124] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is provided. The storage medium stores a computer program. When the computer program is executed by a processor, any one of the aforementioned methods is implemented.

[0125] According to yet another aspect of an embodiment of the present invention, a computer program product is provided, including a computer program, which implements any of the aforementioned methods when executed by a processor.

[0126] Please refer to Figure 2 , an integrated attitude control and adjustment system for green support shield equipment, comprising:

[0127] The six-degree interactive monitoring module includes a data input interface, a data processing unit, and a storage unit. The data input interface is used to receive shield machine posture parameters, formation parameters, and equipment status data output by multiple sensors. The data processing unit executes the mutual information algorithm and generates a weighted directed graph. The storage unit is used to store data and the weighted directed graph.

[0128] The differential multi-cylinder linkage mechanism includes a main hydraulic cylinder and an auxiliary cylinder array. The hydraulic pipeline of the main hydraulic cylinder is connected to the hydraulic pipeline of the shield machine propulsion cylinder group through a pressure sensor to form a closed-loop pressure circuit, and a differential connection structure is formed through the electromagnetic reversing valve and the proportional relief valve;

[0129] The vector adaptive control unit includes a control signal generation module and a PID feedback control module. The control signal generation module receives the output parameters of the weighted directed graph and generates a control signal. The PID feedback control module receives the output force feedback signal and control signal of the differential multi-cylinder linkage mechanism and outputs a regulation signal to the differential multi-cylinder linkage mechanism according to the PID closed-loop feedback algorithm.

[0130] Green support collaborative module, including synchronous grouting control unit and segment assembly angle optimization unit;

[0131] A recycled material injection device and a cylinder linkage mechanism mechanically connected thereto;

[0132] a mechanical coupling structure for connecting the regenerative material injection device with the auxiliary cylinder array of the differential multi-cylinder linkage mechanism;

[0133] The cylinder linkage mechanism receives the control signal of the vector adaptive control unit through the signal input port and drives the regeneration material injection device through the mechanical connection structure;

[0134] The remote intelligent control center establishes two-way communication with the six-degree interactive monitoring module and the vector adaptive control unit through the communication module. The remote intelligent control center integrates the LSTM neural network model to predict formation changes;

[0135] The data bus connects the six-degree interactive monitoring module, the vector adaptive control unit and the remote intelligent control center to form a real-time sharing channel for multi-source data.

[0136] Based on the above, a six-degree interaction model comprehensively perceives multi-source heterogeneous data, utilizes a mutual information algorithm to quantify nonlinear correlations between data, and constructs a weighted directed graph, providing more precise input for attitude control. A differential multi-cylinder linkage mechanism combines hydraulics and pneumatics to provide more flexible attitude adjustment capabilities. The integration of a green support collaborative module and a recycled material injection device enables resource recycling. Calcium carbonate-based cementitious materials are synthesized from shield waste, improving material utilization and reducing CO2 emissions. The LSTM model in the remote intelligent control center predicts stratum changes, providing early warning and reducing the frequency of manual intervention, thereby enhancing the system's intelligence and ability to cope with complex stratum mutations. A data bus enables real-time data sharing across multiple modules, ensuring coordinated linkage among all system components. A closed-loop pressure circuit maintains dynamic propulsion balance, improving tunneling stability. A mechanical coupling structure synchronizes attitude adjustment with support material injection, optimizing tunnel formation quality.

[0137] It should be further explained that the main hydraulic cylinder and auxiliary cylinder array of the differential multi-cylinder linkage mechanism can be customized according to the size of the shield machine and the tunneling force requirements. The auxiliary cylinder array can be composed of multiple independently controlled cylinders to provide more precise differential force control; the electromagnetic reversing valve and proportional relief valve can use industrial-grade valves with high response speed and high precision; the control signal generation module of the vector adaptive control unit can be based on a variety of control algorithms, such as fuzzy control, neural network control or model predictive control, combined with the mutual information algorithm to generate control signals; the parameters of the PID feedback control module can be adaptively adjusted online according to the formation type and tunneling status; the synchronous grouting control unit of the green support collaborative module can use flow meters and proportional control valves to achieve precise control of grouting volume and rate; the segment assembly angle optimization unit can calculate the optimal assembly angle based on three-dimensional scanning data and a preset mathematical model; The calcium carbonate-based cementitious material synthesis unit of the recycled material injection device can adjust reaction conditions and catalyst types based on the waste residue composition and desired material properties. The cylinder linkage nozzle can be designed with multiple or adjustable nozzles to adapt to different injection requirements and formation conditions. The cylinder linkage mechanism can be replaced with an electric push rod or hydraulic cylinder to provide different driving forces and control precision. The mechanical coupling structure can adopt various forms such as connecting rod mechanism, gear drive, or belt drive to ensure reliable linkage between the recycled material injection device and the auxiliary cylinder array of the differential multi-cylinder linkage mechanism. The remote intelligent control center can be deployed on a cloud platform or a local server, and the LSTM neural network model can be continuously trained and optimized based on historical tunneling data. The 5G communication module can use an industrial-grade module that supports low latency and high bandwidth. The data bus can use industrial Ethernet or fiber optic network to ensure stable and real-time data transmission. The closed-loop pressure circuit pressure sensor can use a high-precision pressure sensor with a monitoring range that can cover the maximum operating pressure of the shield machine's propulsion cylinder group.

[0138] As an optional embodiment, the six-degree interactive monitoring module receives output signals of the inclination sensor, displacement sensor, formation hardness sensor and tool force feedback unit through a data interface.

[0139] Based on the above, by introducing multiple sensors such as inclination sensors, displacement sensors, formation hardness sensors and tool force feedback units, a comprehensive perception of the shield machine's posture, formation characteristics and equipment operating status is achieved. The sensor data constitutes the basic input of the six-degree interaction model, providing a rich and reliable information source for subsequent data processing and precise control. The inclination sensor is used to monitor the pitch and roll angles of the shield machine in real time, the displacement sensor is used to monitor the longitudinal and lateral displacement of the shield machine, the formation hardness sensor is used to detect the hardness distribution of the front formation, and the tool force feedback unit provides reaction force data during tool cutting, reflecting the force of the formation on the cutterhead and the wear status of the tool. The fusion of these multi-source data enables the system to more accurately understand the current excavation environment and equipment status, laying the foundation for achieving refined and adaptive posture control.

[0140] Furthermore, the output signal includes six-degree angle dimension, displacement vector dimension, partition hardness coefficient dimension, equipment status dimension, hydraulic cylinder linkage dimension and support adaptation dimension.

[0141] Based on the above, six key dimensions are defined: angle dimension, displacement vector dimension, partition hardness coefficient dimension, equipment status dimension, hydraulic cylinder linkage dimension and support adaptation dimension, which comprehensively cover the key factors affecting shield tunneling and posture control. The angle dimension and displacement vector dimension reflect the spatial posture and position deviation of the shield machine; the partition hardness coefficient dimension and equipment status dimension describe the formation conditions and the operating status of the equipment itself; the hydraulic cylinder linkage dimension and support adaptation dimension are related to the posture adjustment actuator and the support process; by abstracting these seemingly independent factors into a unified six-degree dimension and conducting correlation analysis, the present invention can construct an interactive model that is closer to the actual complex process, providing a comprehensive data basis for realizing overall posture control adjustment.

[0142] Furthermore, the six-degree interaction monitoring module includes: a data fusion unit, a correlation analysis unit, a dynamic weight allocation unit and a prediction correction unit.

[0143] According to the above, by constructing a data processing flow, it aims to extract deep information from six-dimensional data, build an accurate interactive model, and realize the prediction of future stratum changes. The data fusion unit is responsible for integrating the raw data from different sensors, eliminating redundancy and noise, and performing preliminary feature extraction; the correlation analysis unit uses tools such as mutual information algorithm to quantify the correlation strength between different dimensions, construct a weighted directed graph, and reveal complex coupling relationships; the dynamic weight allocation unit dynamically adjusts the weight of each dimension in the control decision according to the current stratum type and excavation status to ensure that key information is processed first; the prediction correction unit uses prediction algorithms such as LSTM neural network model to predict the trend of stratum change, and corrects the model parameters and control strategy according to the prediction results; the collaborative work of the above units enables the six-degree interactive monitoring module to extract information that is crucial to posture control and support coordination from massive data, significantly improving the system's intelligence level and ability to cope with complex environments.

[0144] As an optional embodiment, the data fusion unit includes a processing module, which executes a feature co-occurrence relationship modeling process. The data fusion unit also includes a node mapping module, a weight calculation module, and a topology structure optimization module.

[0145] According to the above, this embodiment introduces a feature co-occurrence relationship modeling process and sets a processing module, a node mapping module, a weight calculation module and a topology optimization module in the data fusion unit, aiming to systematically construct and optimize a model reflecting the correlation between the six degrees of dimension. The processing module is responsible for executing the entire feature co-occurrence relationship modeling process. The node mapping module abstracts and maps the collected six-degree dimension data into feature nodes in the graph structure to prepare for subsequent association analysis. The weight calculation module calculates the correlation strength between any two feature nodes based on methods such as the mutual information algorithm, and uses it as the edge weight in the graph structure to construct a weighted directed graph. The topology optimization module optimizes the constructed weighted directed graph, for example, using the PageRank algorithm to identify key influencing factor nodes, dynamically adjusts the data processing priority, and ensures that the system can quickly focus on the most important information. The synergistic effect of the above modules enables the data fusion unit to efficiently and accurately capture the complex correlations in the six-degree dimension data, provide high-quality input for the association analysis unit and the dynamic weight allocation unit, and significantly improve the analysis capability of the six-degree interaction model and the performance of the overall control system.

[0146] As an optional embodiment, the recycled material injection device includes a calcium carbonate-based cementitious material synthesis unit and a cylinder-linked nozzle. The calcium carbonate-based cementitious material synthesis unit is a chemical reaction device that receives fine-grained slag from the shield machine's slag discharge system, cooling water, and externally supplied CO2 gas. By controlling reaction conditions (such as temperature, pressure, and reaction time), the calcium components in the waste slag undergo a carbonization reaction with the CO2, generating calcium carbonate crystals, which, together with other components, form a cementitious material. The cylinder-linked nozzle is connected to the discharge port of the calcium carbonate-based cementitious material synthesis unit and is located at the rear of the shield machine or in the segment assembly area. The cylinder-linked nozzle is driven by a cylinder linkage mechanism, which allows adjustment of the nozzle's position, angle, and injection rate to ensure that the recycled material is accurately injected into the formation area requiring support.

[0147] Furthermore, the calcium carbonate-based cementitious material synthesis unit is used to synthesize calcium carbonate-based cementitious materials. Using waste soil generated during shield tunneling as the primary raw material, a specific chemical reaction is used to synthesize a calcium carbonate-based material with cementitious properties. This material can replace or partially replace traditional grouting materials for ground support, void filling, and surrounding rock improvement. The synthesis process utilizes industrial byproducts or waste (such as CO2 gas and shield soil), transforming low-value waste into high-value engineering materials. This achieves efficient resource recycling and significantly reduces the project's carbon footprint and environmental impact.

[0148] Furthermore, the calcium carbonate-based cementitious material synthesis unit can utilize a stirred tank reactor, pipeline reactor, or fluidized bed reactor, equipped with heating, pressurization, and gas injection control systems. Reaction conditions can be optimized to a ratio of slag particles: cooling water: CO2 gas of 3:2:1 (mass or volume ratio), a temperature of 80-100°C, a pressure gradient of 0.2 MPa / min, and the addition of 0.5% nano-CaO catalyst to improve reaction efficiency and material strength. The resulting calcium carbonate-based cementitious material can achieve a compressive strength of ≥20 MPa (compared to 15 MPa for conventional cement slurry). The cylinder-driven nozzles can be designed as single or multiple nozzles, with nozzle diameter and shape adjustable based on the viscosity of the injected material and formation conditions. The cylinder-driven nozzles can be controlled by a control unit connected to an electro-hydraulic proportional valve assembly to control injection pressure (e.g., 2.5 MPa) and flow rate (e.g., 0.8-1.2 m³ / h), enabling the mixed injection of recycled materials and conventional slurry, or separate injection.

[0149] As an optional embodiment, the vector adaptive control unit includes: a formation type identification module, a weight adaptive adjustment module, a control signal generation module and a PID feedback control module.

[0150] According to the above, through the formation type identification module and the weight adaptive adjustment module, the control strategy can be adaptively adjusted according to the real-time formation information; the formation type identification module is responsible for judging the current excavation formation type (such as soft soil, sand layer, hard rock, etc.) according to the formation parameters provided by the six-degree interaction monitoring module; the weight adaptive adjustment module dynamically adjusts the weights of each input parameter in the control signal generation module or the parameters of the control algorithm according to the identified formation type and the analysis results of the six-degree interaction model to optimize the control effect; the control signal generation module receives the output parameters of the six-degree interaction model after weight adjustment and generates control instructions; the PID feedback control module receives the feedback signal of the actuator (differential multi-cylinder linkage mechanism) and the generated control signal for closed-loop adjustment; the collaborative work of the above modules enables the vector adaptive control unit to adjust the control strategy in real time according to complex formation conditions and equipment status, thereby realizing accurate and robust control of the shield machine posture.

[0151] As an optional embodiment, the formation type identification module executes a formation type identification algorithm, which determines the current formation type based on the formation parameters output by the six-degree interactive monitoring module.

[0152] According to the above, during the shield tunneling process, the formation conditions are the key factors affecting the tunneling parameters and attitude control strategy; different types of formations (such as soft soil, sand layer, hard rock, composite formation, etc.) have significant differences in the stress, deformation and support requirements of the shield machine; the formation type recognition module is configured to execute the formation type recognition algorithm, and automatically determine the formation type of the current shield machine based on the real-time formation parameters output by the six-degree interactive monitoring module; the six-degree interactive monitoring module provides a wealth of formation-related parameters, such as the partition hardness coefficient dimension, groundwater pressure, formation deformation rate, etc.; the formation type recognition algorithm comprehensively analyzes these parameters and uses pattern recognition or classification technology to classify the current formation into a preset formation type; this real-time, multi-parameter-based formation type recognition capability provides accurate formation information input for the subsequent weight adaptive adjustment and control strategy generation of the vector adaptive control unit, ensuring that the control system can adopt the optimal control strategy for different formations, significantly improving tunneling efficiency and safety.

[0153] In addition, the differential multi-cylinder linkage mechanism includes a control unit connected to the electromagnetic reversing valve and the proportional overflow valve. The control unit is configured to output a control signal to adjust the oil or gas flow of the main hydraulic cylinder and the auxiliary cylinder array. The main hydraulic cylinder and the auxiliary cylinder array form a differential connection structure through hydraulic pipelines and pneumatic pipelines.

[0154] As an optional embodiment, the pressure sensor in the closed-loop pressure circuit is connected to the signal input port of the vector adaptive control unit. The pressure sensor is used to monitor the pressure difference between the main hydraulic cylinder and the thrust cylinder group. The vector adaptive control unit is configured to receive the pressure difference signal and make adjustments.

[0155] Furthermore, a pressure sensor is installed in the closed-loop pressure circuit to monitor the pressure difference between the master hydraulic cylinder and the propulsion cylinder group. This pressure difference signal is input to the signal input port of the vector adaptive control unit. The vector adaptive control unit is configured to receive this pressure difference signal and make adjustments accordingly. By monitoring and adjusting the pressure difference between the master hydraulic cylinder and the propulsion cylinder group, the system can achieve a dynamic balance in the output force of both, ensuring the stability of the overall propulsion force during attitude adjustment, thereby improving the smoothness of excavation and the quality of tunnel formation.

[0156] It should be noted that the above modules can be implemented through software or hardware. For the latter, it can be implemented in the following ways, but not limited to: the above modules are all located in the same processor; or the above modules are located in different processors in any combination.

[0157] The foregoing is merely a preferred embodiment of the present invention. It should be noted that those skilled in the art may make various improvements and modifications without departing from the principles of the present invention, and such improvements and modifications are also within the scope of protection of the present invention. Structures, devices, and operating methods not specifically described or explained herein shall, unless otherwise specified or limited, be implemented in accordance with conventional means in the art.

Claims

1. A green support shield equipment integrated attitude control and adjustment system, characterized in that: include: The six-degree interactive monitoring module includes a data input interface, a data processing unit, and a storage unit. The data input interface is used to receive shield machine posture parameters, formation parameters, and equipment status data output by multiple sensors. The data processing unit executes the mutual information algorithm and generates a weighted directed graph. The storage unit is used to store data and the weighted directed graph. A differential multi-cylinder linkage mechanism includes a main hydraulic cylinder and an auxiliary cylinder array. The hydraulic pipeline of the main hydraulic cylinder is connected to the hydraulic pipeline of the shield machine propulsion cylinder group through a pressure sensor to form a closed-loop pressure circuit, and a differential connection structure is formed through an electromagnetic reversing valve and a proportional relief valve; A vector adaptive control unit includes a control signal generation module and a PID feedback control module. The control signal generation module receives the output parameters of the weighted directed graph and generates a control signal. The PID feedback control module receives the output force feedback signal and the control signal of the differential multi-cylinder linkage mechanism and outputs a regulation signal to the differential multi-cylinder linkage mechanism according to a PID closed-loop feedback algorithm. Green support collaborative module, including synchronous grouting control unit and segment assembly angle optimization unit; A recycled material injection device and a cylinder linkage mechanism mechanically connected thereto; a mechanical coupling structure connecting the regenerative material injection device to the auxiliary cylinder array of the differential multi-cylinder linkage mechanism; The cylinder linkage mechanism receives the control signal of the vector adaptive control unit through the signal input port and drives the regeneration material injection device through the mechanical connection structure; A remote intelligent control center establishes two-way communication with the six-degree interactive monitoring module and the vector adaptive control unit through a communication module. The remote intelligent control center integrates an LSTM neural network model for predicting formation changes; The data bus connects the six-degree interactive monitoring module, the vector adaptive control unit and the remote intelligent control center to form a real-time sharing channel for multi-source data.

2. The green support shield equipment integrated posture control and adjustment system according to claim 1 is characterized by: The six-degree interactive monitoring module receives output signals from the inclination sensor, displacement sensor, formation hardness sensor and tool force feedback unit through a data interface; The output signal includes six-degree angle dimension, displacement vector dimension, partition hardness coefficient dimension, equipment status dimension, hydraulic cylinder linkage dimension and support adaptation dimension; The six-degree interaction monitoring module includes: a data fusion unit, a correlation analysis unit, a dynamic weight allocation unit and a prediction correction unit.

3. The green support shield equipment integrated posture control and adjustment system according to claim 2 is characterized in that: The data fusion unit includes a processing module, which executes a feature co-occurrence relationship modeling process. The data fusion unit also includes a node mapping module, a weight calculation module and a topology structure optimization module.

4. The green support shield equipment integrated posture control and adjustment system according to claim 1 is characterized in that: The vector adaptive control unit includes: a formation type identification module, a weight adaptive adjustment module, a control signal generation module and a PID feedback control module.

5. The green support shield equipment integrated posture control and adjustment system according to claim 4 is characterized in that: The formation type identification module executes a formation type identification algorithm, which determines the current formation type based on the formation parameters output by the six-degree interactive monitoring module.

6. A green support shield equipment integral posture control and adjustment method, applied to the system according to any one of claims 1 to 4, characterized in that: The steps include: Collect the shield machine's attitude parameters, formation parameters, and equipment status data, and use the mutual information algorithm to construct a weighted directed graph based on the collected data; Based on the output parameters of the weighted directed graph, a control signal is generated, and the output force of the differential multi-cylinder linkage mechanism is adjusted through the control signal; Based on the control signal, the synchronous grouting control unit and the regeneration material injection device mechanically connected to the synchronous grouting control unit are controlled, wherein the regeneration material injection device is driven by a cylinder linkage mechanism; Optimize segment assembly angles and establish two-way communication between the shield machine and the remote intelligent control center through the communication module; Utilize the neural network model integrated in the remote intelligent control center to predict formation changes and modify control strategies; Sharing of data, control signals and correction strategies is achieved through the data bus; A closed-loop pressure circuit is formed between the hydraulic output pipeline of the differential multi-cylinder linkage mechanism and the shield machine propulsion cylinder group.

7. A green support shield equipment integral posture control and adjustment method according to claim 6, characterized in that: The steps for collecting shield machine attitude parameters, formation parameters and equipment status data include: Acquire angle, displacement, partition hardness and equipment status data in real time through multiple sensors.

8. The green support shield equipment integral posture control and adjustment method according to claim 6 is characterized in that: Based on the collected data, the steps of constructing a weighted directed graph using a mutual information algorithm include: A weighted directed graph is constructed based on the mutual information algorithm to quantify the correlation strength between any two dimensions.

9. The green support shield equipment integral posture control and adjustment method according to claim 6, characterized in that: The step of generating a control signal comprises: Dynamically assign weights according to formation types and generate control signals based on the output parameters of the weighted directed graph; The steps of controlling the synchronous grouting control unit and the regenerative material injection device mechanically connected to the synchronous grouting control unit based on the control signal include: According to the type of stratum, the synchronous grouting volume and the recycled material injection volume are dynamically adjusted. The grouting volume is increased in soft soil layers, and the proportion of recycled materials is increased in hard rock layers.

10. The green support shield equipment integral posture control and adjustment method according to claim 6, characterized in that: The step of adjusting the output force of the differential multi-cylinder linkage mechanism includes: Through the PID closed-loop feedback algorithm, the output force of the differential multi-cylinder linkage mechanism is adjusted according to the control signal; Monitor the actual output force of the differential multi-cylinder linkage mechanism, compare it with the target output force, and adjust the control signal according to the deviation.

Images