Remote collaborative control method and system for multiple types of tunneling equipment

Abstract

The present application provides a remote collaborative control method and system for multiple types of tunneling equipment, which relates to the field of equipment control technology. The method includes: backtracking and locating the perception data stream, performing adaptive confidence analysis, and performing data fusion of the perception data stream with an updated fusion participation; reading the collaborative operation tasks of multiple types of tunneling equipment and establishing the collaborative equipment status of the tunneling equipment; using the collaborative operation tasks and collaborative equipment status to introduce the collaborative relationship tension between the collaborative tension map modeling equipment, and calculate the collaborative tension map; performing optimization based on the collaborative tension map and establishing collaborative operation parameters; performing collaborative status verification of multiple types of tunneling equipment at the verification node, updating the node offset to the collaborative equipment status, and performing global collaborative control optimization. This application can solve the technical problem of low collaborative efficiency of multiple types of tunneling equipment in the existing technology, and achieve the technical effect of improving the collaborative operation efficiency of multiple devices.

Description

Technical Field

[0001] The present application relates to the field of equipment control technology, and in particular to a remote collaborative control method and system for multiple types of tunneling equipment. Background Art

[0002] Against the backdrop of the current development of collaborative control technology for multiple types of tunneling equipment, tunneling projects in coal mines, subways, tunnels and other fields are gradually moving towards intelligence and automation, and the traditional model driven by manual scheduling is gradually being replaced by intelligent scheduling systems.

[0003] Currently, existing technologies typically employ scheduling mechanisms based on fixed rules or data collection strategies centered around the status of a single device. This has improved operational efficiency and safety to a certain extent. However, with the rapid growth in the number, types, and complexity of devices, traditional scheduling approaches have gradually exposed a series of difficult-to-overcome problems. Among these, there is a lack of systematic modeling of the collaborative relationships between multiple devices. Most approaches rely solely on coarse-grained binding at the task level, ignoring details such as inter-device temporal dependencies, spatial overlap, and operational rhythm coupling. This leads to low collaborative efficiency and can even cause device interference and task conflicts.

[0004] In summary, the existing technology has technical problems such as the lack of refined equipment collaboration relationship modeling and dynamic control mechanism, which leads to the collaborative scheduling between multiple types of tunneling equipment relying on static rules and manual settings, further affecting the collaborative efficiency, operation safety and intelligent adaptive capabilities of the scheduling system in complex operation scenarios. Summary of the Invention

[0005] The purpose of this application is to provide a remote collaborative control method and system for multiple types of tunneling equipment, so as to solve the technical problem in the existing technology that due to the lack of refined equipment collaborative relationship modeling and dynamic control mechanism, the collaborative scheduling between multiple types of tunneling equipment relies on static rules and manual settings, further affecting the collaborative efficiency, operation safety and intelligent adaptive capabilities of the scheduling system in complex operation scenarios.

[0006] In view of the above problems, the present application provides a remote collaborative control method and system for multiple types of tunneling equipment.

[0007] On the first aspect, the present application provides a remote collaborative control method for multiple types of tunneling equipment, which is implemented through a remote collaborative control system for multiple types of tunneling equipment, including: configuring a redundant sensor perception network in the tunneling equipment to establish a perception data stream, wherein the tunneling equipment includes a shield machine, a continuous coal mining machine, a hydraulic support conveyor, a concrete spraying machine, a belt conveyor, a drilling equipment, a blasting equipment, a support equipment, and a ventilation equipment; performing backtracking positioning on the perception data stream, performing adaptive confidence analysis using the backtracking positioning result, updating the fusion participation, performing data fusion of the perception data stream with the updated fusion participation, and establishing a spatial perception identifier; reading multiple types of tunneling equipment The collaborative operation tasks of the equipment are carried out, and the collaborative equipment status of the tunneling equipment is established based on the spatial perception identification; the collaborative operation tasks and collaborative equipment status are used to introduce the collaborative relationship tension between the equipment into the collaborative tension map model, and the edge weight coefficient of the collaborative tension map is calculated based on the process dependency intensity, the degree of overlap of the operation area and the equipment status difference; optimization based on the collaborative tension map is performed under multiple optimization objectives, collaborative operation parameters are established, and verification nodes mapped to the collaborative operation parameters are configured; collaborative status verification of multiple types of tunneling equipment is performed at the verification node, node offset is constructed, the node offset is updated to the collaborative equipment status, and global collaborative control optimization is performed.

[0008] Preferably, the remote collaborative control method for multiple types of tunneling equipment also includes: using the collaborative work tasks and collaborative equipment status to construct a task-equipment association map, the task-equipment association map is used to describe the adaptation relationship and dependency relationship between each work task and the tunneling equipment; deducing the equipment-equipment collaborative path based on the task-equipment association map, and using the equipment-equipment collaborative path to introduce a collaborative tension map to model the collaborative relationship tension between devices.

[0009] Preferably, the remote collaborative control method of multiple types of tunneling equipment also includes: establishing an overlapping observation network based on the redundant sensor perception network; using a sliding window to perform retrospective positioning of the perception data stream and construct a retrospective positioning result; performing a time consistency check on the retrospective positioning result to generate a first confidence level; executing real-time environmental data collection of the tunneling scene, and establishing a second confidence level using the real-time environmental data collection results and the sensor adaptability within the redundant sensor perception network; and completing an adaptive confidence analysis using the first confidence level, the second confidence level and the overlapping observation network.

[0010] Preferably, the remote collaborative control method of multiple types of tunneling equipment also includes: obtaining the operating status self-test data of the sensor, and establishing a third confidence level based on the operating status self-test data; using the first confidence level, the second confidence level, and the third confidence level to configure the data weight of the perception data stream; using the overlapping observation network to perform observation authentication of the perception data stream with data weight, and using the observation authentication results to complete adaptive confidence analysis.

[0011] Preferably, the remote collaborative control method for multiple types of tunneling equipment also includes: establishing a fusion participation fallback mechanism, which is configured with a confidence threshold; using the confidence threshold to trigger and authenticate the adaptive confidence analysis results of each sensor; if the confidence threshold is continuously triggered within a preset stage, the fusion participation fallback mechanism is activated, and the fusion participation upper limit value of the corresponding sensor is reduced according to the activation result.

[0012] Preferably, the remote collaborative control method for multiple types of tunneling equipment also includes: using multiple optimization objectives to establish a weighted multi-objective loss function, the multiple optimization objectives include minimizing the collaborative tension objective, minimizing the equipment idle time objective, minimizing the overall operation time objective, and balancing the energy consumption load objective; using the weighted multi-objective loss function to perform collaborative operation parameter optimization iteration based on the collaborative tension map until the optimization iteration converges, and then establishing the collaborative operation parameters.

[0013] Preferably, the remote collaborative control method of multiple types of tunneling equipment also includes: executing startup monitoring of the tunneling equipment and executing startup timing; obtaining actual working data of the tunneling equipment, using the actual working data and startup duration to perform continuous startup load analysis, and generating a first equipment health status; reading the equipment linkage monitoring data of the tunneling equipment, and constructing a second equipment health status based on the equipment linkage monitoring data; and using the first equipment health status and the second equipment health status to perform collaborative status verification and compensation.

[0014] Preferably, the remote collaborative control method for multiple types of tunneling equipment also includes: recording the actual operation plans of the multiple types of tunneling equipment, and establishing a scheduling strategy scenario template based on the collaborative operation tasks, perception data streams, and actual operation plans of the equipment, and storing it in the migration database; when performing collaborative operations, executing similarity matching of the scheduling strategy scenarios in the migration database, and using the similarity matching results for transfer learning to complete self-learning adaptation of the operation plan.

[0015] Preferably, the remote collaborative control method for multiple types of tunneling equipment also includes: performing operation accuracy analysis to establish a first redundant setting constraint; obtaining the environmental adaptability of the sensor to establish a second redundant setting constraint, and configuring a redundant sensor perception network after weighted fusion of the first redundant setting constraint and the second redundant setting constraint.

[0016] In the second aspect, the present application also provides a remote collaborative control system for multiple types of tunneling equipment, which is used to execute the remote collaborative control method for multiple types of tunneling equipment as described in the first aspect, including: a perception data stream establishment module, which is used to configure a redundant sensor perception network in the tunneling equipment and establish a perception data stream, wherein the tunneling equipment includes a shield machine, a continuous coal mining machine, a hydraulic support conveyor, a concrete spraying machine, a belt conveyor, a drilling equipment, a blasting equipment, a support equipment, and a ventilation equipment; a spatial perception identification establishment module, which is used to perform retrospective positioning on the perception data stream, perform adaptive confidence analysis using the retrospective positioning results, update the fusion participation, perform data fusion of the perception data stream with the updated fusion participation, and establish a spatial perception identification; a collaborative equipment status establishment module, which is used to read The collaborative operation tasks of multiple types of tunneling equipment establish the collaborative equipment status of the tunneling equipment based on the spatial perception identifier; the edge weight coefficient acquisition module is used to use the collaborative operation tasks and collaborative equipment status to introduce the collaborative relationship tension between the collaborative tension map modeling equipment, and the edge weight coefficient of the collaborative tension map is calculated based on the process dependency intensity, the degree of overlap of the operation area and the equipment status difference; the collaborative operation parameter establishment module is used to perform optimization based on the collaborative tension map under multiple optimization objectives, establish collaborative operation parameters, and configure verification nodes mapped to the collaborative operation parameters; the collaborative control optimization module is used to perform collaborative status verification of multiple types of tunneling equipment at the verification node, construct node offset, update the node offset to the collaborative equipment status, and perform global collaborative control optimization.

[0017] The technical solution provided in this application has at least the following technical effects or advantages: by realizing the technical goal of intelligent control optimization of tunneling equipment for multi-source data fusion and dynamic collaborative scheduling, the technical effect of improving the collaborative operation efficiency of multiple devices, enhancing the system's environmental adaptability and scheduling strategy migration capabilities is achieved.

[0018] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, which can be implemented in accordance with the contents of the description, and to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are specifically listed below. It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easy to understand through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in this application or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely exemplary, and a person of ordinary skill in the art can obtain other drawings based on the provided drawings without creative work.

[0020] Figure 1 This is a flow chart of the remote collaborative control method for multiple types of tunneling equipment in this application;

[0021] Figure 2 This is a schematic diagram of the structure of the remote collaborative control system for multiple types of tunneling equipment in this application.

[0022] Explanation of the accompanying drawings: perception data flow establishment module 11, spatial perception identification establishment module 12, collaborative device status establishment module 13, edge weight coefficient acquisition module 14, collaborative operation parameter establishment module 15, collaborative control optimization module 16. DETAILED DESCRIPTION

[0023] This application provides a remote collaborative control method and system for multiple types of tunneling equipment. This addresses the existing technical issues, which arise from a lack of refined equipment collaboration modeling and dynamic control mechanisms, resulting in the collaborative scheduling of multiple types of tunneling equipment relying on static rules and manual settings. This further impacts collaborative efficiency, operational safety, and the intelligent adaptive capabilities of the scheduling system in complex operating scenarios. This application achieves the technical goal of optimizing intelligent control of tunneling equipment for multi-source data fusion and dynamic collaborative scheduling, achieving the technical effects of improving the efficiency of multi-equipment collaborative operations, enhancing the system's environmental adaptability, and migrating scheduling strategies.

[0024] Below, the technical solutions in this application will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of this application, rather than all the embodiments of this application. It should be understood that this application is not limited to the example embodiments described herein. Based on the embodiments of this application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application. It should also be noted that, for the convenience of description, only the parts related to this application, rather than all of them, are shown in the accompanying drawings.

[0025] For example, see the attached Figure 1 This application provides a remote collaborative control method for multiple types of tunneling equipment, which is applied to a remote collaborative control system for multiple types of tunneling equipment, and specifically includes the following steps:

[0026] S1: A redundant sensor perception network is configured on the tunneling equipment to establish a perception data stream. The tunneling equipment includes a shield machine, a continuous coal miner, a hydraulic support conveyor, a concrete spraying machine, a belt conveyor, drilling equipment, blasting equipment, support equipment, and ventilation equipment.

[0027] Specifically, a redundant sensor network is deployed on tunneling equipment. This means each piece of equipment is equipped with more sensors than required for normal operation, and data coordination between the redundant sensors is achieved through a network. Redundant sensors enhance system robustness and reliability. Even if the original sensor fails due to harsh environments, equipment wear, or other factors, the entire sensor system can still operate normally. For example, in addition to the standard positioning sensors, a shield machine is equipped with two additional sets of displacement measurement sensors based on different principles to provide backup.

[0028] The types of equipment involved in tunneling operations include shield machines, continuous miners, hydraulic support conveyors, concrete sprayers, belt conveyors, drilling equipment, blasting equipment, support equipment, and ventilation equipment. Shield machines are used for tunneling, cutting the soil ahead through the rotation of the cutterhead; continuous miners are used in coal mining, continuously crushing and transporting coal seams; hydraulic support conveyors are devices that support and move the mining supports, ensuring the stability and safety of the working space; concrete sprayers are used for tunnel support, belt conveyors transport materials or coal to the surface, and ventilation equipment is used for initial ventilation; drilling equipment is used for rock detection or blasting pretreatment, blasting equipment is used in mining processes that require controlled energy release for crushing, and support equipment is used to reinforce the excavated space to ensure operational safety.

[0029] S2: Performing back-tracking positioning on the perception data stream, performing adaptive confidence analysis using the back-tracking positioning result, updating the fusion participation, performing data fusion of the perception data stream with the updated fusion participation, and establishing a spatial perception identifier.

[0030] Specifically, the perception data stream is retrospectively positioned. After the perception data stream is collected and initially processed, a time window mechanism or a historical data index mechanism is used to go back to a certain time period in the past and relocate the specific position of an event or object in space. This is then used to verify perception errors or track anomalies. By replaying the sensor data of historical time, its displacement trajectory and behavior pattern can be accurately restored.

[0031] Subsequently, adaptive confidence analysis is performed using the retroactive positioning results. Based on the accuracy of historical positioning, combined with real-time environmental data and the status of redundant sensors, the confidence level of the data is dynamically adjusted. For example, when the trajectory of a certain segment of historical data closely matches the predictions of multiple redundant sensors, the confidence level is higher. However, if a redundant sensor exhibits drift or packet loss during that period, the confidence level of the corresponding data will decrease.

[0032] Next, updating the fusion participation level refers to dynamically adjusting the "participation weight" of each redundant sensor in the subsequent data fusion process based on the results of the adaptive confidence analysis. The fusion participation level reflects whether a redundant sensor is trustworthy at the current stage and should play a primary role in perception judgment.

[0033] Then, the perception data streams are fused with the updated fusion participation, recombining data from multiple sensors to generate a more accurate and representative unified perception result. Data fusion includes processing mechanisms such as multi-source information superposition, anomaly filtering, and discrepancy compensation. For example, in complex tunneling environments, spatial morphological data from lidar may be combined with image recognition results to improve the spatial positioning accuracy of target identification. Finally, after completing the fusion of multi-source perception data, a spatial perception identifier is established, generating a unique spatial identifier for a specific target or area, which includes the target's three-dimensional position, orientation, boundary characteristics, and dynamic attributes.

[0034] S3: Read the collaborative operation tasks of multiple types of tunneling equipment, and establish the collaborative equipment status of the tunneling equipment based on the spatial perception identifier.

[0035] Specifically, the system retrieves collaborative tasks for multiple types of tunneling equipment. This involves extracting the specific tasks required for each type of tunneling equipment to collaborate at a specific stage from dispatch instructions or work plans. These tasks include multiple sub-processes, such as excavation, support, slag removal, and ventilation. The task description clearly identifies the operating scope, timeframe, and resource dependencies for each type of tunneling equipment. For example, a shield machine might be responsible for main excavation, a belt conveyor for slag removal, and ventilation equipment activated in a specific area. This information can then be used as a basis for subsequent coordination.

[0036] Subsequently, the collaborative device status of the tunneling equipment is established based on the spatial perception identifier. The spatial perception identifier of each tunneling equipment or object, generated from the fusion of multi-source redundant sensors, is further converted into a description of the tunneling equipment's current collaborative status. The collaborative device status includes the equipment's real-time position, orientation, and motion status, as well as its role in the current task, its progress, and its spatial relationship with other equipment. For example, a tunneling arm with a position coordinate of 150 meters in the X direction and 20 meters in the Y direction and a current status of "awaiting support collaboration" is marked as "position reached, awaiting support," while the status of the corresponding support machinery may be "approaching, estimated arrival time 20 seconds."

[0037] S4: Using the collaborative work tasks and collaborative equipment status, a collaborative tension map is introduced to model the collaborative relationship tension between the equipment. The edge weight coefficient of the collaborative tension map is calculated based on the process dependency intensity, the degree of overlap of the work area and the equipment status difference.

[0038] Specifically, a collaborative tension graph is introduced to model the collaborative tension between equipment using collaborative tasks and equipment status. After clarifying the current or upcoming tasks of each tunneling equipment, as well as the equipment's operating status (such as power, temperature, and task progress), a dynamic graph structure is constructed to describe the interaction difficulty or resource coupling tension between different tunneling equipment during the collaborative process. The collaborative tension graph is a time-varying weighted graph model in which each node represents a tunneling equipment, while edges represent the collaborative relationships between that equipment and other equipment. Tension is represented by the edge weights, indicating the potential interference intensity, dependency density, or scheduling complexity that may arise during the collaborative process. The edge weight coefficients in the collaborative tension graph are then dynamically calculated based on multiple factors, including the process dependency strength. This dependency strength measures the degree of sequential logical relationship between two equipment tasks. The degree of overlap in the operating area is a variable that determines edge weights, measuring whether two equipment items intersect spatially. Equipment status differences are a factor in assessing tension, indicating the differences in health status between different equipment items during operation. By constructing a topological graph of the task (i.e., task flow chart), the dependency relationship is converted into the edges of a directed graph to obtain the process dependency strength. . . represents the shortest operation path length from device i to device j (i.e., the number of devices or logical tasks passed through in the middle); if device i is the direct predecessor of device j, then ,but =0.5; if device i has no dependency on device j, then = 0. If the excavation process is: drilling machine → cutting machine → support machine → shotcrete machine, then the shortest operation path length between the drilling machine and the support machine is 2, The degree of overlap of operating areas It represents the overlap ratio of the operating areas of device i and device j in three-dimensional space, reflecting the potential interference and resource competition between devices in space. . is the spatial overlapping volume of the operating range of equipment i and equipment j; The intersection volume is the union of the operating area volumes of equipment i and equipment j. The intersection volume is calculated by combining the BIM model, LiDAR scanning, or 3D sensor to identify the area boundaries and then performing meshing. .

[0039] Device state discrepancy refers to the degree of difference in the current operating state between two devices in a collaborative operation. This discrepancy reflects whether the devices can operate efficiently and synchronously and whether there is any performance mismatch. Device state can encompass multiple dimensions, such as current power output, load factor, operating temperature, vibration intensity, and transmission efficiency. The state discrepancy calculation process involves the following steps: First, multiple state parameters for each device are extracted and normalized to ensure comparability across different dimensions. Next, the numerical differences between the corresponding state parameters of the two devices are squared and summed. Finally, the square root of the result is taken to obtain the comprehensive state distance between the two devices, which serves as the state discrepancy metric. Smaller state discrepancies indicate closer operating states between the two devices, making them more suitable for collaborative operation. Larger state discrepancies indicate significant operational deviations between the two devices, necessitating reduced coupling within the task to avoid performance interference and the risk of failure.

[0040] S5: Perform optimization based on the collaborative tension map under multiple optimization objectives, establish collaborative operation parameters, and configure verification nodes mapped to the collaborative operation parameters.

[0041] Specifically, under the premise of considering multiple optimization directions, the scheduling parameters in the tunneling operation are systematically optimized in combination with the collaborative tension map formed by the collaborative relationship between tunneling equipment. Multiple optimization goals include minimizing collaborative tension, reducing equipment idling time, shortening overall operation time, balancing equipment energy consumption, etc., representing comprehensive requirements in terms of efficiency, stability and energy consumption. Through the analysis and optimization of the collaborative tension map, a more reasonable operation parameter configuration can be derived, such as adjusting the start-up sequence of the equipment, the operation spacing or the way of mutual cooperation to reduce the overall tension of the system. Then, after the optimization is completed, the collaborative operation parameters are established, and the scheduling and control parameters used to guide the collaborative operation of each tunneling equipment are determined, including task allocation, operation time scheduling, spatial path coordination, equipment start and stop strategies, etc., which can provide accurate operating instructions for various types of tunneling equipment.

[0042] Next, logical checkpoints, or verification nodes, were introduced for real-time monitoring and result verification. These checkpoints confirm that equipment is operating strictly according to optimized parameters. Verification nodes compare the current state with preset operating parameters using sensor data, position feedback, and task execution records to determine any deviations. For example, if a tunneling machine is supposed to enter a designated area at the 80th minute but actually enters at the 90th minute, the verification node will trigger an early warning mechanism or adjust subsequent scheduling strategies. Each verification node corresponds to a specific collaborative operation parameter, enabling dynamic monitoring and closed-loop control.

[0043] S6: Perform collaborative status verification of multiple types of tunneling equipment at the verification node, construct a node offset, update the node offset to the collaborative equipment status, and perform global collaborative control optimization.

[0044] Specifically, at specific inspection locations within the task scheduling system, the operating status of various tunneling equipment involved in the collaborative operation is compared, verified, and evaluated. Tunneling equipment may include propulsion equipment, slag removal equipment, ventilation equipment, and other devices. These different types of equipment must maintain a good coordination and functional linkage. Real-time data is used to determine if there are response delays, execution errors, or functional conflicts between devices. This coordination status is then verified to ensure stable operation.

[0045] During the state verification process, the time or state difference between the actual behavior of the equipment at the predetermined verification node and the expected behavior is identified to construct the node offset. Node offset can manifest as delayed start-up of propulsion equipment, early termination of slag discharge equipment, sensor feedback lag, etc. Updating the node offset to the collaborative equipment state means feeding the difference data back into the equipment state model, thereby dynamically adjusting the operation characterization of each device, which in turn helps to reflect the inconsistencies between devices in actual operation and provide real state input for the scheduling system. By continuously updating the collaborative equipment state, the adaptability and robustness of the scheduling optimization algorithm can be improved. Based on the updated equipment state model, the task allocation, collaboration path, execution timing and other parameters between multiple devices are recalculated, and then global collaborative control optimization is performed to optimize the collaborative efficiency of the entire system and ensure that all devices collaborate efficiently under unified scheduling.

[0046] Furthermore, the present application also includes: using the collaborative work tasks and collaborative equipment status to construct a task-equipment association map, the task-equipment association map is used to describe the adaptation relationship and dependency relationship between each work task and the tunneling work equipment; based on the task-equipment association map, the equipment-equipment collaborative path is deduced, and the collaborative tension map is used to model the collaborative relationship tension between the equipment using the equipment-equipment collaborative path.

[0047] Specifically, a task-equipment association graph is constructed using collaborative work tasks and collaborative equipment status. This means that, based on the current status and task division of each type of tunneling equipment, a multi-node graph structure is constructed to intuitively represent the compatibility between various work tasks and tunneling equipment, as well as the order of their dependencies. A node in the task-equipment association graph can represent a specific task, such as "Support work in tunnel section B," while another node represents a type of equipment, such as a "Concrete sprayer." The edges connecting the nodes in the task-equipment association graph carry weights, representing the suitability of the tunneling equipment for the task. Compatibility relationships may be based on technical capabilities, equipment load, or operating environment assessments. Dependency relationships indicate that a task must wait for the completion of another task before it can begin, such as when excavation work must begin after support work.

[0048] Next, based on the task-equipment association map, equipment-equipment collaboration paths were derived. Further analysis revealed intersections in the collaboration between multiple tunneling equipment during task execution, such as sharing the same area, temporal overlap, or dependencies in the logical order of operations. This led to the creation of a path network representing the collaborative relationships between tunneling equipment. Each path represented a possible direct or indirect collaboration between tunneling equipment. For example, tunneling equipment A must complete material transfer before tunneling equipment B can initiate support operations. Thus, a collaborative path existed between tunneling equipment A and B, with the direction of the path indicating the order of collaboration.

[0049] Then, the collaborative tension map is introduced using the equipment-equipment collaborative path to model the collaborative tension between equipment. Specifically, based on the equipment collaborative path map, a "tension" metric is established for each pair of collaborative tunneling equipment, representing the intensity of resource competition, time scheduling urgency, or the tightness of behavioral coupling in the collaboration. The collaborative tension map is a weighted graph whose edge weights are calculated based on parameters such as the strength of process dependencies, the overlap area of ​​the work areas, the execution time difference, and the consistency of equipment states. For example, if two tunneling equipment need to simultaneously complete overlapping operations in a channel only 200 cm wide and the operation time difference is only 5 seconds, the tension between the two tunneling equipment will be higher than that of two equipment pairs operating at a 30-minute interval and not in the same area. Table 1 shows a partial record of the most recent equipment-equipment collaborative path and collaborative tension map.

[0050] Table 1: Partial record of the most recent device-device collaboration path and collaborative tension map

[0051]

[0052] The collaborative tension graph is a weighted undirected graph used to characterize the difficulty of collaboration, coupling degree, and coordination scheduling complexity between devices. It is denoted as: . V represents the device set, each node Corresponding to a tunneling device. E represents the set of edges that may have collaboration between devices. Each edge Indicates that there is a collaborative scheduling relationship between device i and device j. Represents an edge The tension value is the collaborative tension between device i and device j. The tension value calculation model for each pair of device i and device j is: Tension value In the calculation formula Indicates the process dependency intensity; Indicates the degree of overlap of the operating areas; Indicates the degree of difference in device status. is an adjustable weight coefficient, satisfying , weights are configured according to different scenarios; It can be obtained by normalizing the sequential constraint matrix in the operation plan diagram; Calculated by spatial operation grid overlap ratio, such as: = Overlapping area volume / Union area volume. This is the normalized Euclidean distance between the two devices in terms of their status indicators, such as speed, load factor, and remaining life. Table 2 partially records the mathematical modeling of the collaborative tension map.

[0053] Table 2: Partial record of mathematical modeling methods of synergistic tension map

[0054]

[0055] Furthermore, the present application also includes: establishing an overlapping observation network based on the redundant sensor perception network; using a sliding window to perform retrospective positioning of the perception data stream and construct a retrospective positioning result; performing a time consistency check on the retrospective positioning result to generate a first confidence level; performing real-time environmental data collection of the excavation scene, and establishing a second confidence level using the real-time environmental data collection results and the sensor adaptability within the redundant sensor perception network; and completing an adaptive confidence analysis using the first confidence level, the second confidence level and the overlapping observation network.

[0056] Specifically, an overlapping observation network is established based on a redundant sensor perception network. This involves rationally deploying multiple redundant sensors so that they overlap in their observations of the same physical area or target, thereby constructing a data observation structure with spatial redundancy. This overlapping observation network enhances observation stability and fault tolerance. For example, if three redundant sensors in a given area can simultaneously acquire information about the tunneling head's posture, even if one fails, the other two can still continue to function.

[0057] Subsequently, a sliding window is used to perform retrospective positioning of the perception data stream. A time sliding strategy is employed to analyze data collected continuously over a period of time to infer the spatial position changes of the tunneling equipment during that time period. A sliding window is a technique used for stream data processing that dynamically analyzes data trends over short periods of time. For example, a 5-second window can be set to continuously review data from inertial navigation and laser displacement sensors within that period to restore the equipment's motion trajectory within that period and generate retrospective positioning results.

[0058] Next, the back-localization results are subjected to a temporal consistency check. This assesses the accuracy and reliability of the back-localization results by determining the temporal coherence and logical consistency of the redundant sensor data. This temporal consistency check can identify issues such as data anomalies, drift, or sensor clock drift. If the differences in the measurements of multiple redundant sensors within a certain time period fluctuate minimally, this indicates good data consistency and generates a first confidence level. A higher first confidence level indicates more reliable back-localization data, while a lower confidence level indicates less reliable data.

[0059] At the same time, real-time environmental data collection for the tunneling scenario is performed, specifically to obtain real-time physical parameters within the current tunneling area, such as temperature, humidity, dust concentration, and vibration intensity. By comparing this real-time environmental data with the historical performance of each redundant sensor in that environment, the adaptability of each redundant sensor type in the current situation can be inferred. The resulting evaluation result is the sensor fitness. The higher the sensor fitness, the higher the second confidence level, and vice versa. Combining the real-time environmental data collection results with the sensor fitness, another confidence metric, the second confidence level, can be calculated to quantify the expected value of the sensor data reliability in the current environment.

[0060] Finally, adaptive confidence analysis is performed using the first and second confidence levels, along with an overlapping observation network. This involves a fusion calculation based on spatial redundancy and temporal consistency to derive an overall perception confidence level. This allows for dynamic adjustment of the weight of each redundant sensor's data. For example, if a redundant sensor exhibits high temporal consistency but low adaptability, its weight is lowered to prevent erroneous data from influencing the overall judgment. To verify the feasibility and stability of the retroactive positioning mechanism and adaptive confidence analysis mechanism in perception data fusion, a targeted validation experiment was conducted. The experimental platform included: tunneling equipment types: a drum shearer, a hydraulic support, and a transfer machine; a perception network: 12 redundant sensors deployed, collecting data streams such as displacement, pressure, vibration, current, and temperature; and a simulated tunnel environment with a sampling frequency of 10 times per second. The experiment lasted 1800 seconds.

[0061] The retroactive localization mechanism was validated by introducing interference samples (artificially fabricated or delayed data). A sliding window mechanism (with a window length of 300 seconds) was used to locate the source of the anomaly. The system was then compared to determine whether the sensor and time period to which the anomaly data belonged were correctly identified. Adaptive confidence analysis was performed on the three confidence inputs: temporal consistency confidence, environmental acquisition confidence, and sensor self-test confidence. Adaptive confidence analysis was also performed on the hourly calculated confidence outputs, which were compared with expert scoring results (human-determined validity levels). Adaptive confidence analysis was also performed on the error rate of the analyzed and fused data and the fluctuations in system stability. Table 3 summarizes the experimental results.

[0062] Table 3: Experimental results data summary record

[0063]

[0064] Results show that the introduction of a sliding window backtracking mechanism can effectively identify anomalous data sources and time periods, improving the accuracy of anomaly handling by nearly 23%. Furthermore, an adaptive analysis method based on dynamic calculation of multi-source confidence scores achieves lower false alarm rates and greater output stability compared to traditional static weighting methods, reducing confidence fluctuations by nearly 70%. The error in the fusion results is also significantly reduced, ensuring the engineering applicability of the data fusion output.

[0065] For example, in an interference scenario where the vibration sensor of a certain device delayed its response, the control group data was directly incorporated into the fusion, resulting in a vibration value error of more than 1.8 units. However, the experimental group eliminated outliers through backtracking and only retained data with higher confidence levels. After fusion, the error did not exceed 0.3 units.

[0066] Furthermore, the present application also includes: obtaining the operating status self-test data of the sensor, and establishing a third confidence level based on the operating status self-test data; using the first confidence level, the second confidence level, and the third confidence level to configure the data weight of the perception data stream; using the overlapping observation network to perform observation authentication of the perception data stream with data weight, and using the observation authentication results to complete adaptive confidence analysis.

[0067] Specifically, the system acquires sensor operational self-test data. This involves periodically or in real time collecting health information for each redundant sensor during operation. This includes parameter data on indicators such as voltage stability, temperature drift, data transmission integrity, and signal strength changes. This operational self-test data is generated by the redundant sensor's built-in diagnostic mechanism and is used to determine whether the sensor is currently in a normal, sub-healthy, or faulty state.

[0068] A third confidence level is established based on the self-test data from the operating status. This numerical indicator measures the reliability of the redundant sensor, indicating whether the redundant sensor's internal operation is stable, independent of external environmental factors. The third confidence level ranges from 0 to 1, with higher values ​​indicating more stable sensor status and lower values ​​indicating less stable sensor status.

[0069] Furthermore, the first, second, and third confidence levels are used to configure the data weights of the sensor data streams, assigning a fused credibility weight to each piece of sensor data. The data weight represents the influence of a particular sensor data on the overall decision. For example, if the three confidence levels of a sensor within a certain time period are 0.85, 0.9, and 0.8, respectively, the final data weight might be 0.83, which determines the reference value of the data in spatial modeling or path planning.

[0070] Next, an overlapping observation network is used to authenticate observations of the perception data stream with data weights. This involves cross-validating overlapping observations of the same physical target from multiple redundant sensors, taking into account their respective data weights. The consistency of the observations is evaluated, thereby determining their reliability. The authentication process considers the coordination of redundant observations. For example, if two of the three redundant sensors measure similar values, while one shows a significant deviation, the sensor with the larger deviation may be marked as untrustworthy. The authentication results are then used to adjust the dynamic trust level of each sensor, completing an adaptive confidence analysis.

[0071] Furthermore, the present application also includes: establishing a fusion participation fallback mechanism, which is configured with a confidence threshold; using the confidence threshold to trigger and authenticate the adaptive confidence analysis results of each sensor; if the confidence threshold is continuously triggered within a preset stage, the fusion participation fallback mechanism is activated, and the fusion participation upper limit value of the corresponding sensor is reduced according to the activation result.

[0072] Specifically, a fusion participation fallback mechanism is established. This involves introducing a dynamic control mechanism into the tunneling equipment's multi-sensor perception network, flexibly adjusting the proportion of redundant sensors involved in data fusion based on their performance. Fusion participation measures the weight of redundant sensors in the final data fusion result, reflecting the level of confidence in the reliability of their data. If certain redundant sensors exhibit persistent anomalies or their credibility decreases, the fallback mechanism reduces their weight to limit their interference with the overall perception results, thereby improving the overall stability and robustness of the redundant sensor system.

[0073] To make this fallback mechanism controllable, configure a confidence threshold. The confidence threshold is a criterion that defines the minimum value for determining whether a sensor's confidence level (i.e., the reliability of its sensor data) meets the required standards. For example, if the threshold is set to 0.6, all sensor results with a confidence level below 0.6 will be considered unreliable, triggering the next step. This threshold can be flexibly set based on the complexity of the construction environment. For example, the threshold can be appropriately increased in high-interference underground scenarios and relaxed in open tunnel environments.

[0074] This confidence threshold is used to trigger authentication of each sensor's adaptive confidence analysis results. This means that the confidence results of each sensor in the perception network are individually evaluated to determine if they fall below the confidence threshold, triggering an alert mechanism accordingly. A dynamic score, the adaptive confidence analysis result, is automatically calculated based on multiple factors such as device status, sensor compatibility, and observation consistency. For example, if the confidence score of a redundant sensor falls below 0.4 for three consecutive cycles, its data quality is considered unstable.

[0075] If the confidence threshold is repeatedly triggered within a preset period, the fusion participation fallback mechanism is activated. Specifically, if a redundant sensor repeatedly experiences insufficient confidence within a specified time window (e.g., 10 minutes), the fallback strategy for that redundant sensor is initiated. After activation, the upper limit of the redundant sensor's fusion participation—the maximum contribution it can make to the final fusion data—is dynamically adjusted downward based on the persistence and severity of the fault. For example, a radar sensor with an original upper limit of 0.3 could have its contribution reduced to 0.1 due to frequent anomalies, effectively reducing its influence.

[0076] Furthermore, the present application also includes: using multiple optimization objectives to establish a weighted multi-objective loss function, the multiple optimization objectives include minimizing the collaborative tension objective, minimizing the equipment idle time objective, minimizing the overall operation time objective, and balancing the energy consumption load objective; using the weighted multi-objective loss function to perform collaborative operation parameter optimization iteration based on the collaborative tension map until the optimization iteration converges, and then establishing the collaborative operation parameters.

[0077] Specifically, in the scheduling of complex tunneling operations, by introducing multiple interrelated optimization objectives, a mathematical function is constructed to comprehensively measure the operating efficiency and coordination of the system, thereby quantifying the degree of deviation from the optimal state. Multiple optimization objectives refer to indicators that affect operational efficiency and synergy, including minimizing collaborative tension, minimizing equipment idle time, minimizing overall operating time, and balancing energy load. Minimizing collaborative tension means reducing the collaborative resistance between tunneling equipment caused by inconsistent status, overlapping processes, or resource competition; minimizing equipment idle time means reducing the waiting time of equipment in a non-task state and improving resource utilization; minimizing overall operating time aims to shorten the time required for the entire tunneling process as much as possible; and balancing energy load means avoiding continuous high-load operation of certain equipment while ensuring operational efficiency, thereby extending equipment life and saving energy. A weight is assigned to each optimization objective to represent its importance in the current task scenario, where the weighting coefficient can be set using the coefficient of variation method.

[0078] Next, based on a weighted multi-objective loss function, the collaborative operation parameters are iteratively optimized using the collaborative tension graph. Collaborative operation parameters (such as task allocation, start time, and equipment sequence) are continuously adjusted, and multiple rounds of iterative optimization are conducted to find the parameter combination that minimizes the overall loss function. The optimization process uses the graph information to guide parameter adjustments. For example, if the tension on a particular edge is high, the operating order or interval time of the related equipment may be adjusted to reduce the corresponding loss term.

[0079] When the optimization iterations converge, meaning the loss function has stabilized and approached the optimal solution after multiple rounds of parameter adjustments, the final collaborative operation parameters can be determined, including each device's startup time, specific tasks, and the order of collaboration. This serves as a direct basis for subsequent scheduling. Convergence conditions can be set as either a loss function change of less than 0.001 for three consecutive times or a total number of iterations reaching 1000.

[0080] Furthermore, the present application also includes: executing startup monitoring of the tunneling operation equipment and executing startup timing; obtaining actual working data of the tunneling operation equipment, using the actual working data and startup duration to perform continuous startup load analysis, and generating a first equipment health status; reading the equipment linkage monitoring data of the tunneling operation equipment, and constructing a second equipment health status based on the equipment linkage monitoring data; and using the first equipment health status and the second equipment health status to perform collaborative status verification and compensation.

[0081] Specifically, the tunneling equipment's status is tracked and sampled throughout the entire process of preparation for startup, startup, and completion, including successful startup, current consumption, voltage fluctuations, and mechanical response. The time from when the tunneling equipment receives the start command to when it actually completes the startup is recorded, thereby reflecting the agility and stability of the tunneling equipment's startup. For example, if a tunneling equipment takes 35 seconds from the start signal to when all subsystems are operating normally, the startup duration is 35 seconds.

[0082] Next, actual operating data of the tunneling equipment is obtained. Once the equipment enters normal operation, its operating status, including torque, speed, thrust, temperature, and vibration amplitude, is continuously collected to reflect the equipment's load status and actual task performance. Continuous startup load analysis is conducted by combining actual operating data with startup duration. This examines the equipment's transition stability from standstill to loaded operation, and determines whether there are issues such as startup overload, response lag, or load fluctuation. This generates a primary equipment health status, which comprehensively characterizes the equipment's current startup quality and operational performance.

[0083] Subsequently, as multiple devices operate in coordination, they collect information on their synchronization, command response consistency, and mutual interference, such as whether another device is simultaneously avoiding or supporting one another during propulsion. Based on this coordinated monitoring data, a secondary device health status is constructed, indicating the reliability of the device's behavior when working in coordination with other devices, assessing any issues such as delayed responses, mismatched operations, or command conflicts.

[0084] Finally, the health status of the first device and the health status of the second device are used to verify and compensate the collaborative status, and the individual status of the device is cross-compared with the linkage status to determine whether the overall collaboration has failed due to individual abnormalities, and then a compensation mechanism is adopted, such as lowering the collaborative priority of a certain device, temporarily postponing entry into the shared space, rescheduling the task rhythm, etc., to maintain the stability and efficiency of the overall operating system.

[0085] Furthermore, the present application also includes: recording the actual operation plans of multiple types of tunneling equipment, and establishing a scheduling strategy scenario template based on the collaborative operation tasks, perception data streams, and actual equipment operation plans, and storing it in a migration database; when performing collaborative operations, executing similarity matching of the scheduling strategy scenarios in the migration database, and using the similarity matching results for transfer learning to complete self-learning adaptation of the operation plan.

[0086] Specifically, the actual operation plans for various types of tunneling equipment are recorded and archived. This involves documenting and documenting the parameter configurations, execution paths, operating procedures, and emergency response strategies used by various types of tunneling equipment during actual operations. These types of equipment may include tunnel boring machines, continuous miners, and hydraulic support conveyors. These plans vary significantly depending on their structure, purpose, and operational scenarios. These plans not only reflect how the equipment performs its tasks but also include its historical performance and feedback optimization results for specific tasks.

[0087] Scheduling strategy scenario templates are established based on collaborative work tasks, sensory data streams, and the actual equipment operation plans. These templates combine the task objectives of equipment collaborative work, the real-time environment and status information provided by the sensory system, and the historical operation strategies of each device to construct standardized strategy templates for subsequent decision-making. Scheduling strategy scenario templates encompass key control points such as equipment division of labor, timing coordination, path planning, and task switching strategies. Scheduling strategy scenario templates serve as a knowledge graph or rule base for the scheduling system's multi-device collaborative control. Storing scheduling strategy scenario templates in a migration database allows for the rapid extraction of historical best practices for reuse and optimization when encountering similar operating conditions in the future.

[0088] When collaborating, the system performs a similarity match between scheduling strategies and scenarios within the migration database. This involves comparing factors such as the current job site's task type, environmental conditions, and equipment configuration to identify the most similar scenario templates from the past. This similarity match is based on multi-dimensional features, such as task and process similarity, geological environment consistency, and equipment model compatibility. Transfer learning is then performed using the similarity match results, extracting useful empirical parameters from past successful cases. These parameters are then fine-tuned based on the current environment, enabling self-learning and adaptation of the work plan. This eliminates the need for equipment to plan scheduling strategies from scratch, allowing for intelligent adjustments based on historical experience.

[0089] Furthermore, the present application also includes: performing operation accuracy analysis to establish a first redundant setting constraint; obtaining the environmental adaptability of the sensor to establish a second redundant setting constraint, and configuring a redundant sensor perception network after weighted fusion of the first redundant setting constraint and the second redundant setting constraint.

[0090] Specifically, the accuracy requirements for tunneling operations are assessed by combining historical operational data, on-site geological information, and engineering requirements. This allows the initial redundancy constraints for operational accuracy to be determined, guiding the subsequent deployment of redundant sensors. For example, the deviation of the tunneling path must not exceed a certain range, or the accuracy of geological structure identification must meet a certain standard.

[0091] Next, the sensors' environmental adaptability is assessed, evaluating each sensor's performance stability and data reliability in specific tunneling environments. Different sensors may perform differently in environments with high humidity, dust, or vibration. For example, LiDARs are susceptible to signal attenuation in dusty environments, while inertial navigation systems may drift due to vibration. Through experimental testing and environmental modeling, the adaptability of each sensor type in different operating environments is quantified. This creates a second set of redundant constraints, helping to eliminate devices that perform poorly under specific conditions and thus improving the stability of the overall perception system.

[0092] To comprehensively consider operational accuracy requirements and the sensor's environmental adaptability, a weighted fusion of the first and second redundancy constraints is performed. Weighted fusion is a comprehensive assessment process that quantifies the impact of the two constraints by setting weights. For example, the importance of operational accuracy can be set to 0.6, and the importance of environmental adaptability to 0.4. The weights of weighted fusion are customized by technicians based on actual conditions. The fusion results guide the configuration of the redundant sensor perception network, determining the location, density, and types of sensors to build the redundant but valuable perception network to ensure operational stability and reliability.

[0093] To sum up, the remote collaborative control method of multiple types of tunneling equipment provided in this application has the following technical effects: by realizing the technical goal of intelligent control optimization of tunneling equipment for multi-source data fusion and dynamic collaborative scheduling, the technical effect of improving the collaborative operation efficiency of multiple devices, enhancing the system environment adaptability and scheduling strategy migration capabilities is achieved.

[0094] Example 2: Based on the same inventive concept as the remote collaborative control method for multiple types of tunneling equipment in the previous embodiment, this application also provides a remote collaborative control system for multiple types of tunneling equipment. Figure 2, including: a perception data stream establishment module 11, used to configure a redundant sensor perception network in the tunneling equipment and establish a perception data stream; a spatial perception identification establishment module 12, used to perform back-tracking positioning on the perception data stream, use the back-tracking positioning result to perform adaptive confidence analysis, update the fusion participation, perform data fusion of the perception data stream with the updated fusion participation, and establish a spatial perception identification; a collaborative equipment status establishment module 13, used to read the collaborative operation tasks of multiple types of tunneling equipment, and establish the collaborative equipment status of the tunneling equipment based on the spatial perception identification; an edge weight coefficient acquisition module 14, used to use the collaborative operation tasks The collaborative relationship tension between the equipment is modeled by a collaborative tension map, and the edge weight coefficient of the collaborative tension map is calculated based on the process dependency strength, the degree of overlap of the operation area and the equipment status difference; a collaborative operation parameter establishment module 15 is used to perform optimization based on the collaborative tension map under multiple optimization objectives, establish collaborative operation parameters, and configure verification nodes mapped to the collaborative operation parameters; a collaborative control optimization module 16 is used to perform collaborative status verification of multiple types of tunneling equipment at the verification node, construct a node offset, update the node offset to the collaborative equipment status, and perform global collaborative control optimization.

[0095] Furthermore, the remote collaborative control system of multiple types of tunneling equipment is also used to: construct a task-equipment association map using the collaborative work tasks and collaborative equipment status, and the task-equipment association map is used to describe the adaptation relationship and dependency relationship between each work task and the tunneling equipment; deduce the equipment-equipment collaborative path based on the task-equipment association map, and use the equipment-equipment collaborative path to introduce a collaborative tension map to model the collaborative relationship tension between devices.

[0096] Furthermore, the remote collaborative control system of the multiple types of tunneling equipment is also used to: establish an overlapping observation network based on the redundant sensor perception network; use a sliding window to perform retrospective positioning of the perception data stream and construct a retrospective positioning result; perform a time series consistency check on the retrospective positioning result to generate a first confidence level; perform real-time environmental data collection of the tunneling scene, and establish a second confidence level using the real-time environmental data collection results and the sensor adaptability within the redundant sensor perception network; and complete adaptive confidence analysis using the first confidence level, the second confidence level and the overlapping observation network.

[0097] Furthermore, the remote collaborative control system of multiple types of tunneling equipment is also used to: obtain the operating status self-test data of the sensor, and establish a third confidence level based on the operating status self-test data; use the first confidence level, the second confidence level, and the third confidence level to configure the data weight of the perception data stream; use the overlapping observation network to perform observation authentication of the perception data stream with data weight, and use the observation authentication results to complete adaptive confidence analysis.

[0098] Furthermore, the remote collaborative control system of multiple types of tunneling equipment is also used to: establish a fusion participation fallback mechanism, which is configured with a confidence threshold; use the confidence threshold to trigger and authenticate the adaptive confidence analysis results of each sensor; if the confidence threshold is continuously triggered within a preset stage, the fusion participation fallback mechanism is activated, and the fusion participation upper limit value of the corresponding sensor is reduced according to the activation result.

[0099] Furthermore, the remote collaborative control system of the multiple types of tunneling equipment is also used to: establish a weighted multi-objective loss function using multiple optimization objectives, and the multiple optimization objectives include minimizing the collaborative tension objective, minimizing the equipment idle time objective, minimizing the overall operation time objective, and balancing the energy consumption load objective; using the weighted multi-objective loss function to perform collaborative operation parameter optimization iteration based on the collaborative tension map until the optimization iteration converges, and then establish the collaborative operation parameters.

[0100] Furthermore, the remote collaborative control system of multiple types of tunneling equipment is also used to: perform startup monitoring of tunneling equipment and execute startup timing; obtain actual working data of tunneling equipment, use the actual working data and startup duration to perform continuous startup load analysis, and generate a first equipment health status; read the equipment linkage monitoring data of tunneling equipment, and construct a second equipment health status based on the equipment linkage monitoring data; use the first equipment health status and the second equipment health status to perform collaborative status verification and compensation.

[0101] Furthermore, the remote collaborative control system of the multiple types of tunneling equipment is also used to: record the actual operation plans of the multiple types of tunneling equipment, and establish a scheduling strategy scenario template based on the collaborative operation tasks, perception data streams, and actual operation plans of the equipment, and store it in the migration database; when performing collaborative operations, perform similarity matching of the scheduling strategy scenarios in the migration database, and use the similarity matching results for transfer learning to complete self-learning adaptation of the operation plan.

[0102] Furthermore, the remote collaborative control system of the multiple types of tunneling equipment is also used to: perform operation accuracy analysis and establish a first redundant setting constraint; obtain the environmental adaptability of the sensor and establish a second redundant setting constraint; after weighted fusion of the first redundant setting constraint and the second redundant setting constraint, configure a redundant sensor perception network.

[0103] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The remote collaborative control method and specific examples of multiple types of tunneling equipment in the aforementioned embodiment one are also applicable to the remote collaborative control system of multiple types of tunneling equipment in this embodiment. Through the aforementioned detailed description of the remote collaborative control method of multiple types of tunneling equipment, those skilled in the art can clearly understand the remote collaborative control system of multiple types of tunneling equipment in this embodiment, so for the sake of brevity of the specification, it will not be described in detail here.

[0104] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present application. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application is not limited to the embodiments shown herein, but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

[0105] Obviously, those skilled in the art may make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application is intended to include these modifications and variations.

Claims

1. A remote collaborative control method for multiple types of tunneling equipment, characterized in that: The method comprises: Configuring a redundant sensor perception network on tunneling equipment to establish a perception data stream. The tunneling equipment includes a shield machine, a continuous coal miner, a hydraulic support conveyor, a concrete sprayer, a belt conveyor, drilling equipment, blasting equipment, support equipment, and ventilation equipment. Performing back-tracking positioning on the perception data stream, performing adaptive confidence analysis using the back-tracking positioning result, updating the fusion participation, performing data fusion of the perception data stream with the updated fusion participation, and establishing a spatial perception identifier; Reading collaborative operation tasks of multiple types of tunneling equipment, and establishing collaborative equipment status of the tunneling equipment based on the spatial perception identifier; By using the collaborative work tasks and collaborative equipment status, a collaborative tension map is introduced to model the collaborative relationship tension between the equipment. The edge weight coefficient of the collaborative tension map is calculated based on the process dependency intensity, the degree of overlap of the work area and the difference in equipment status. Conduct optimization based on collaborative tension maps under multiple optimization objectives, establish collaborative operation parameters, and configure verification nodes mapped to collaborative operation parameters; Performing collaborative state verification of multiple types of tunneling equipment at the verification node, constructing a node offset, updating the node offset to the collaborative equipment state, and performing global collaborative control optimization; The method of using the collaborative task and collaborative device status to introduce a collaborative tension map to model the collaborative relationship tension between devices includes: A task-equipment association map is constructed using the collaborative operation tasks and collaborative equipment status, wherein the task-equipment association map is used to describe the adaptation relationship and dependency relationship between each operation task and the tunneling operation equipment; The device-device collaboration path is deduced based on the task-device association graph, and the collaboration tension graph is used to model the collaboration relationship tension between devices using the device-device collaboration path.

2. The remote collaborative control method for multiple types of tunneling equipment according to claim 1, characterized in that: The method of using the backtracking positioning result to perform adaptive confidence analysis, updating the fusion participation, performing data fusion of the perception data stream with the updated fusion participation, and establishing a spatial perception identifier includes: Establishing an overlapping observation network based on the redundant sensor perception network; Use sliding windows to perform back-tracking positioning of the perception data stream and construct back-tracking positioning results; Performing a time sequence consistency check on the retrospective positioning result to generate a first confidence level; Perform real-time environmental data collection of the tunneling scenario, and establish a second confidence level using the real-time environmental data collection results and the sensor fitness within the redundant sensor perception network; An adaptive confidence analysis is performed using the first confidence level, the second confidence level, and an overlapping observation network.

3. The remote collaborative control method for multiple types of tunneling equipment according to claim 2, characterized in that: The method of completing the adaptive confidence analysis by using the first confidence level, the second confidence level, and the overlapping observation network includes: Acquire operating state self-test data of the sensor, and establish a third confidence level based on the operating state self-test data; configuring a data weight of a perception data stream using the first confidence level, the second confidence level, and the third confidence level; The overlapping observation network is used to perform observation authentication of the perception data stream with data weights, and the observation authentication results are used to complete the adaptive confidence analysis.

4. The remote collaborative control method for multiple types of tunneling equipment according to claim 3, characterized in that: The update fusion participation includes: Establishing a fusion participation fallback mechanism, wherein the fusion participation fallback mechanism is configured with a confidence threshold; Using the confidence threshold to trigger authentication of the adaptive confidence analysis result of each sensor; If the confidence threshold is continuously triggered within a preset stage, the fusion participation fallback mechanism is activated, and the fusion participation upper limit value of the corresponding sensor is reduced according to the activation result.

5. The remote collaborative control method for multiple types of tunneling equipment according to claim 1, characterized in that: The optimization based on the collaborative tension map is performed under multiple optimization objectives to establish collaborative operation parameters, including: A weighted multi-objective loss function is established using multiple optimization objectives, including minimizing collaborative tension, minimizing equipment idle time, minimizing overall operation time, and balancing energy consumption load. The weighted multi-objective loss function is used to perform iterative optimization of collaborative operation parameters based on the collaborative tension map until the optimization iteration converges, and then the collaborative operation parameters are established.

6. The remote collaborative control method for multiple types of tunneling equipment according to claim 1, characterized in that: The step of performing collaborative status verification of multiple types of tunneling equipment at the verification node and constructing a node offset includes: Perform startup monitoring and startup timing of tunneling equipment; Acquire actual working data of the tunneling operation equipment, perform continuous starting load analysis using the actual working data and the starting duration, and generate a first equipment health status; Reading equipment linkage monitoring data of the tunneling operation equipment, and establishing a second equipment health status based on the equipment linkage monitoring data; The first device health status and the second device health status are used to perform collaborative status verification and compensation.

7. The remote collaborative control method for multiple types of tunneling equipment according to claim 1, characterized in that: After the global collaborative control optimization is performed, the following steps are included: Record the actual operation plans of various types of tunneling equipment, and establish scheduling strategy scenario templates based on the collaborative operation tasks, perception data streams, and actual equipment operation plans, and store them in the migration database; When performing collaborative work, similarity matching of scheduling strategy scenarios in the migration database is performed, and transfer learning is performed using the similarity matching results to complete self-learning adaptation of the work plan.

8. The remote collaborative control method for multiple types of tunneling equipment according to claim 1, characterized in that: The redundant sensor perception network configured on the tunneling equipment includes: Perform job accuracy analysis and establish first redundancy setting constraints; The environmental adaptability of the sensor is obtained, a second redundancy setting constraint is established, and after weighted fusion of the first redundancy setting constraint and the second redundancy setting constraint, a redundant sensor perception network is configured.

9. A remote collaborative control system for multiple types of tunneling equipment, characterized in that: The steps for implementing the remote collaborative control method of multiple types of tunneling equipment according to any one of claims 1 to 8 include: a sensor data stream establishment module configured to configure a redundant sensor perception network on tunneling equipment and establish a sensor data stream, wherein the tunneling equipment includes a shield machine, a continuous coal miner, a hydraulic support conveyor, a concrete sprayer, a belt conveyor, drilling equipment, blasting equipment, support equipment, and ventilation equipment; A spatial perception identification establishment module is used to perform retroactive positioning on the perception data stream, perform adaptive confidence analysis using the retroactive positioning result, update the fusion participation, perform data fusion of the perception data stream with the updated fusion participation, and establish a spatial perception identification; A collaborative equipment status establishment module, configured to read collaborative operation tasks of multiple types of tunneling equipment and establish collaborative equipment status of the tunneling equipment based on the spatial perception identifier; An edge weight coefficient acquisition module is used to use the collaborative work tasks and collaborative equipment status to introduce a collaborative tension map to model the collaborative relationship tension between equipment. The edge weight coefficient of the collaborative tension map is calculated based on the process dependency intensity, the degree of overlap of the work area, and the difference in equipment status; The collaborative operation parameter establishment module is used to perform optimization based on the collaborative tension map under multiple optimization objectives, establish collaborative operation parameters, and configure verification nodes mapped to the collaborative operation parameters; The collaborative control optimization module is used to perform collaborative status verification of multiple types of tunneling equipment at the verification node, construct node offsets, update the node offsets to the collaborative equipment status, and perform global collaborative control optimization.

Images