Large-size cutting arm intelligent telescopic control method and system for tunnel water ditch

Abstract

The application discloses a large-size cutting arm intelligent telescopic control method and system for a tunnel water ditch, relates to the technical field of tunnel cutting equipment control, and comprises the following steps: acquiring real-time environmental information of a construction work surface; performing joint solving operation to generate a cutting operation plan; controlling driving components of the cutting arm and controlling a mechanical locking mechanism to perform locking operation; controlling the first cutting arm and the second cutting arm to perform cutting operation; and when the compared deviation exceeds the preset allowable range, triggering local re-planning to update the subsequent target configuration sequence and / or the coordinated cutting control parameter sequence. The application solves the technical problems of the existing multi-arm equipment configuration and control strategy disconnection, the contradiction between the limited tunnel space and the cutting demand, and the instability and asynchronization of the cutting of non-homogeneous rock strata, realizes the adaptive linkage of the cutting arm configuration and the cutting strategy and the dynamic balance of the double-saw load, and improves the technical effects of the operation stability and the environmental adaptability of the equipment under complex tunnel conditions.

Description

Technical Field

[0001] This invention relates to the field of tunnel cutting equipment control technology, specifically to an intelligent telescopic control method and system for large-size cutting arms used in tunnel drainage ditches. Background Technology

[0002] In underground engineering construction such as tunnel and ditch cutting, multi-arm cutting equipment has become the mainstream construction equipment. However, existing equipment generally adopts fixed-configuration robotic arms or simple folding structures, which has a core defect of disconnect between mechanical configuration and control strategy. On the one hand, due to the limited working space in tunnels, equipment transportation and non-working states need to be highly compact, while actual cutting operations require the arm to have ultra-long stroke and high rigidity. Existing structures cannot simultaneously meet the contradictory requirements of spatial constraints and deep cutting. On the other hand, when facing complex working conditions such as abrupt changes in tunnel cross-section, large obstacles, and heterogeneous anisotropy of rock strata, dual-saw or multi-saw cutting arms cannot sense the differences in cutting resistance in real time, and cutting parameters cannot be dynamically and collaboratively adjusted with load changes. This easily leads to problems such as arm vibration, uneven wear of the cutter head, asynchronous advance, and even instability, resulting in low construction accuracy, short cutter head life, and insufficient equipment environmental adaptability, making it difficult to ensure continuous and efficient construction operations under complex geological conditions.

[0003] Existing multi-arm equipment suffers from technical problems such as fragmented configurations and control strategies, contradictions between limited tunnel space and cutting requirements, and instability and asynchronous cutting of heterogeneous rock strata. Summary of the Invention

[0004] This application provides a method and system for intelligent telescopic control of large-size cutting arms for tunnel drainage ditches, which addresses the technical problems of the disconnect between existing multi-arm equipment configuration and control strategy, the contradiction between limited tunnel space and cutting needs, and the instability and asynchronous cutting of heterogeneous rock strata.

[0005] In view of the above problems, this application provides a method and system for intelligent telescopic control of large-size cutting arms for tunnel drainage ditches.

[0006] A first aspect of this application provides an intelligent telescopic control method for a large-size cutting arm used in tunnel drainage ditches, the method comprising: Real-time environmental information of the construction work surface is acquired. This real-time environmental information includes at least three-dimensional geometric information of the work surface collected by a 3D vision sensor and mechanical information of the interaction between the cutting arm and the rock mass collected by a mechanical sensor. Based on the real-time environmental information, preset target trench shape parameters, and the configuration parameter library of the cutting arm, a joint solution calculation is performed to generate a cutting operation plan. The cutting operation plan includes at least the target configuration sequence of the first and second cutting arms during the cutting operation cycle, the corresponding collaborative cutting control parameter sequence, and predicted mechanical information. According to the cutting operation plan, each driving component of the cutting arm is controlled to transform from the current configuration to the initial target configuration of the target configuration sequence, and the mechanical locking mechanism is controlled to perform a locking operation. When the cutting arm is in the locked initial target configuration, the first and second cutting arms are controlled to perform cutting operations according to the collaborative cutting control parameter sequence. During the cutting operation, the actual mechanical information collected is continuously compared with the predicted mechanical information in the cutting operation plan. When the comparison deviation exceeds a preset allowable range, local replanning is triggered to update the subsequent target configuration sequence and / or collaborative cutting control parameter sequence.

[0007] In one possible implementation, a joint solution operation is performed based on the real-time environmental information, preset target trench shape parameters, and the configuration parameter library of the cutting arm to generate a cutting operation plan. This includes: constructing a spatiotemporal joint optimization model with the cutting arm operation process as the object; mapping the three-dimensional geometric information in the real-time environmental information to the spatial and obstacle avoidance constraints in the spatiotemporal joint optimization model; mapping the interaction mechanics information in the real-time environmental information to the rock mass mechanics property constraints in the spatiotemporal joint optimization model; mapping the configuration parameter library to the mechanical limit constraints of the cutting arm in the spatiotemporal joint optimization model; and, under the condition of satisfying the spatial and obstacle avoidance constraints, rock mass mechanics property constraints, and mechanical limit constraints, simultaneously solving the target configuration sequence, the collaborative cutting control parameter sequence, and the predicted mechanics information with at least one of the total operation time, total energy consumption, and dual-arm load balance within the cutting operation cycle as optimization objectives to generate a globally optimized cutting operation plan.

[0008] In one possible implementation, under the conditions of satisfying the spatial and obstacle avoidance constraints, rock mass mechanical property constraints, and mechanical limit constraints, at least one of the following—total operation time, total energy consumption, and dual-arm load balance—within the cutting operation cycle as the optimization objective, the target configuration sequence, the collaborative cutting control parameter sequence, and the predicted mechanical information are solved simultaneously to generate a globally optimized cutting operation plan. This includes: discretizing the entire cutting operation cycle into multiple consecutive planning control periods, and defining a set of decision variables for each planning control period. The decision variables at least include the cutting arm configuration corresponding to the time period in the target configuration sequence, the collaborative cutting... The control parameter sequence includes the cutting parameter combination for the corresponding time period, and the predicted load for the corresponding time period in the predicted mechanical information. Within the framework of the spatiotemporal joint optimization model, based on the spatial and obstacle avoidance constraints, rock mass mechanical property constraints, and mechanical limit constraints, a state transition equation and a dynamic equation connecting the decision variables of adjacent planning control time periods are established as global constraints. A decision objective function is constructed using at least one of the total operating time, total energy consumption, and dual-arm load balance. The decision variable sequence that satisfies the global constraints and optimizes the decision objective function is solved to obtain the target configuration sequence, the collaborative cutting control parameter sequence, and the predicted mechanical information.

[0009] In one possible implementation, the cutting parameter combination for the corresponding time period in the collaborative cutting control parameter sequence is coupled with the cutting arm configuration for the corresponding time period in the target configuration sequence and the predicted load for the corresponding time period in the predicted mechanical information. The cutting parameter combination includes at least: feed rate and cutting depth parameters that match the reachable workspace of the cutting arm under the cutting arm configuration; rotational speed parameters that match the cutter head wear model and rock mass breaking efficiency model under the predicted load; and phase difference parameters for coordinating the dynamic load balance of the first and second cutting arms under the cutting arm configuration.

[0010] In one possible implementation, according to the cutting operation plan, the driving components of the cutting arm are controlled to transform from the current configuration to the initial target configuration of the target configuration sequence, and the mechanical locking mechanism is controlled to perform a locking operation, including: calculating the target displacement of each joint and the telescopic sleeve according to the initial target configuration in the target configuration sequence, and generating corresponding multi-axis linkage motion commands; after confirming that the cutting arm has reached the initial target configuration through pose feedback, identifying the load-bearing key nodes according to the predicted load distribution map for the initial target configuration in the predicted mechanical information, and controlling the mechanical locking mechanism to apply a preload force matching the predicted load distribution map to the selected load-bearing key nodes to complete rigid locking.

[0011] In one possible implementation, with the cutting arms in a locked initial target configuration, controlling the first and second cutting arms to perform a cutting operation according to the coordinated cutting control parameter sequence includes: converting the coordinated cutting control parameter sequence into drive motor torque commands and drive motor speed commands for the first and second cutting arms; and performing cutting execution control based on the drive motor torque commands and drive motor speed commands.

[0012] In one possible implementation, during the cutting operation, the actual mechanical information collected is continuously compared with the predicted mechanical information in the cutting operation plan. When the comparison deviation exceeds a preset allowable range, local replanning is triggered to update the subsequent target configuration sequence and / or collaborative cutting control parameter sequence. This includes: establishing a reference curve for the predicted mechanical information to change over time, the reference curve containing multiple linearly distributed reference mechanical information values; during the cutting operation, the actual mechanical information is continuously collected and compared in real time with the reference mechanical information values ​​at the same time point, and a real-time deviation index is calculated; when the real-time deviation index continuously exceeds a preset deviation threshold or the accumulated deviation energy exceeds a threshold, local replanning is triggered to update the subsequent target configuration sequence and / or collaborative cutting control parameter sequence.

[0013] In one possible implementation, triggering local replanning to update the subsequent target configuration sequence and / or collaborative cutting control parameter sequence includes: taking the current moment as the new planning starting point, using the currently acquired real-time environmental information, the current cutting arm configuration, and the remaining target trench parameters as input, re-executing the joint solution operation to generate an updated cutting operation plan from the current moment; calculating the difference between the initial configuration of the new target configuration sequence in the updated cutting operation plan and the current configuration of the cutting arm; and executing an update strategy in stages based on the difference: a. If the difference is lower than the first level, only the new collaborative cutting control parameter sequence in the updated cutting operation plan is used; b. If the difference is between the first and second levels, the cutting arm is controlled to move to the initial configuration of the new target configuration sequence and relocked, while the new collaborative cutting control parameter sequence is used; c. If the difference is higher than the second level, the current operation process is stopped and an intervention request is sent.

[0014] In one possible implementation, the acquisition of real-time environmental information of the construction work surface further includes: acquiring the body pose information of the cutting arm in real time by using encoders installed at each joint of the cutting arm and displacement sensors at each telescopic sleeve; and aligning and fusing the body pose information with the three-dimensional geometric information and the interaction mechanical information in time and space to form real-time environmental information with a unified time and space reference.

[0015] A second aspect of this application provides an intelligent telescopic control system for a large-size cutting arm used in tunnel drainage ditches, the system comprising: An environmental information acquisition module is used to acquire real-time environmental information of the construction work surface. This real-time environmental information includes at least three-dimensional geometric information of the work surface collected by a 3D vision sensor, and mechanical information of the interaction between the cutting arm and the rock mass collected by a mechanical sensor. A cutting operation planning generation module is used to perform joint calculations based on the real-time environmental information, preset target trench shape parameters, and the configuration parameter library of the cutting arm to generate a cutting operation plan. This cutting operation plan includes at least the target configuration sequence of the first and second cutting arms during the cutting operation cycle, a sequence of collaborative cutting control parameters corresponding to the target configuration sequence, and predicted mechanical information. A locking operation execution module is also included. The system is configured to control each drive component of the cutting arm to transform from the current configuration to the initial target configuration of the target configuration sequence according to the cutting operation plan, and to control the mechanical locking mechanism to perform a locking operation; the cutting operation execution module is configured to control the first cutting arm and the second cutting arm to perform cutting operations according to the cooperative cutting control parameter sequence when the cutting arm is in the locked initial target configuration; the control parameter sequence update module is configured to continuously compare the actual mechanical information collected with the predicted mechanical information in the cutting operation plan during the cutting operation, and trigger local replanning to update the subsequent target configuration sequence and / or cooperative cutting control parameter sequence when the comparison deviation exceeds the preset allowable range.

[0016] One or more technical solutions provided in this application have at least the following technical effects or advantages: The system acquires real-time environmental information of the construction work surface; performs joint solution calculations to generate a cutting operation plan; based on the cutting operation plan, it controls each drive component of the cutting arm to transform from the current configuration to the initial target configuration of the target configuration sequence, and controls the mechanical locking mechanism to perform a locking operation; with the cutting arm in the locked initial target configuration, it controls the first and second cutting arms to perform cutting operations according to the cooperative cutting control parameter sequence; during the cutting operation, it continuously compares the collected actual mechanical information with the predicted mechanical information in the cutting operation plan, and when the comparison deviation exceeds the preset allowable range, it triggers local replanning to update the subsequent target configuration sequence and / or cooperative cutting control parameter sequence. This achieves the technical effect of realizing adaptive linkage between the cutting arm configuration and the cutting strategy and dynamic load balancing of the dual saws, improving the operational stability and environmental adaptability of the equipment under complex tunnel conditions. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic flowchart of an intelligent telescopic control method for a large-size cutting arm used in tunnel drainage ditches, provided in an embodiment of this application. Figure 2 A schematic diagram of the intelligent telescopic control system for a large-size cutting arm used in tunnel drainage ditches, provided in an embodiment of this application.

[0019] Explanation of reference numerals in the attached diagram: 10 for environmental information acquisition module, 20 for cutting operation planning and generation module, 30 for locking operation execution module, 40 for cutting operation execution module, and 50 for control parameter sequence update module. Detailed Implementation

[0020] This application provides a method and system for intelligent telescopic control of large-size cutting arms for tunnel drainage ditches, which addresses the technical problems of existing multi-arm equipment configuration and control strategies being disconnected, the contradiction between limited tunnel space and cutting needs, and the instability and asynchronous cutting of heterogeneous rock strata.

[0021] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0022] Example 1, as Figure 1 As shown, this application provides an intelligent telescopic control method for a large-size cutting arm used in tunnel drainage ditches. The method is applied to tunnel construction equipment comprising at least two cutting arms, wherein the cutting arms are variable-configuration arms with multiple joints and / or multi-stage telescopic sleeves. The method includes: Step S100: Obtain real-time environmental information of the construction work surface. The real-time environmental information includes at least three-dimensional geometric information of the work surface collected by a 3D vision sensor and mechanical information of the interaction between the cutting arm and the rock mass collected by a mechanical sensor.

[0023] Specifically, the method is applied to tunnel construction equipment containing at least two cutting arms. The cutting arms are variable configuration arms with multiple joints and / or multi-stage telescopic sleeves. The method uses 3D vision sensors mounted on the tunnel construction equipment to collect three-dimensional geometric information of the tunnel drainage ditch construction face in real time. At the same time, mechanical sensors arranged on key stress-bearing parts of the cutting arms collect real-time mechanical information of the interaction between the cutting arms and the rock mass. Combined with the body pose information obtained by encoders of each joint of the cutting arms and displacement sensors of each stage of the telescopic sleeves, the above-mentioned three-dimensional geometric information, interaction mechanical information and body pose information are spatiotemporally aligned and fused to form complete and accurate real-time environmental information of the construction face under a unified spatiotemporal reference, providing reliable data input for subsequent cutting operation planning.

[0024] Step S200: Based on the real-time environmental information, the preset target trench shape parameters, and the configuration parameter library of the cutting arm, perform joint solution calculations to generate a cutting operation plan. The cutting operation plan includes at least the target configuration sequence of the first and second cutting arms during the cutting operation cycle, the cooperative cutting control parameter sequence corresponding to the target configuration sequence, and the predicted mechanical information.

[0025] Specifically, based on the acquired real-time environmental information, pre-set target trench shape parameters, and the configuration parameter library of the cutting arm, a spatiotemporal joint optimization model is constructed with the cutting arm operation process as the object. The three-dimensional geometric information is mapped to spatial and obstacle avoidance constraints, the interaction mechanics information is mapped to rock mass mechanics property constraints, and the configuration parameter library is mapped to mechanical limit constraints. Under the condition of satisfying all the above constraints, at least one of the total operation time, total energy consumption, and dual-arm load balance is taken as the optimization objective. The cutting operation cycle is discretized into multiple continuous planning control periods. Global constraints are established through state transition equations and dynamic equations. The target configuration sequence of the first and second cutting arms in the whole cutting operation cycle, the cooperative cutting control parameter sequence coupled with the target configuration sequence, and the predicted mechanical information of the corresponding time period are solved simultaneously to form a globally optimized cutting operation plan.

[0026] Step S300: According to the cutting operation plan, control each driving component of the cutting arm to change from the current configuration to the initial target configuration of the target configuration sequence, and control the mechanical locking mechanism to perform the locking operation.

[0027] Specifically, based on the generated cutting operation plan, the target displacement of each joint and the multi-stage telescopic sleeve is calculated according to the initial target configuration in the target configuration sequence. Multi-axis linkage motion commands are generated and the actions of each drive component of the cutting arm are controlled, so that the cutting arm can smoothly change from the current configuration to the initial target configuration. After confirming that the cutting arm has reached the target position through posture feedback, the key load-bearing nodes are identified according to the predicted load distribution map corresponding to the initial target configuration. The mechanical locking mechanism is controlled to apply matching preload to the key load-bearing nodes to complete the high-rigidity mechanical locking, providing structural protection for subsequent stable cutting operations.

[0028] Step S400: With the cutting arm in the locked initial target configuration, control the first cutting arm and the second cutting arm to perform the cutting operation according to the cooperative cutting control parameter sequence.

[0029] Specifically, with the cutting arm rigidly locked and maintaining its initial target configuration, the feed rate, cutting depth, rotational speed, and phase difference parameters of the two arms coupled with the configuration and predicted load in the collaborative cutting control parameter sequence are converted into torque and speed commands for the corresponding drive motors of the first and second cutting arms. The two cutting arms are synchronously controlled to perform collaborative cutting operations according to the commands, ensuring balanced load on both arms, synchronous operation, and that the cutting efficiency of the cutter head matches the rock mass conditions.

[0030] Step S500: During the cutting operation, the actual mechanical information collected is continuously compared with the predicted mechanical information in the cutting operation plan. When the comparison deviation exceeds the preset allowable range, local replanning is triggered to update the subsequent target configuration sequence and / or collaborative cutting control parameter sequence.

[0031] Specifically, during the continuous execution of the cutting operation, a reference mechanical curve that changes over time is first established based on the predicted mechanical information. The actual mechanical information of the interaction between the cutting arm and the rock mass is collected in real time and compared with the reference value at the same time point to calculate the real-time deviation index. When the deviation index continuously exceeds the preset deviation threshold or the accumulated deviation energy exceeds the set threshold, i.e., the comparison deviation exceeds the preset allowable range, the current moment is taken as the starting point for the new planning. The joint solution calculation is re-executed based on the latest real-time environmental information, the current cutting arm configuration, and the remaining target trench parameters to generate an updated cutting operation plan. The strategy is then implemented in stages according to the degree of difference between the updated target configuration and the current configuration. Only the collaborative cutting control parameter sequence is updated or the configuration is adjusted and re-locked or the operation is stopped and an intervention request is issued to achieve local replanning and updating of the target configuration sequence and / or the collaborative cutting control parameter sequence.

[0032] In one possible implementation, step S200 further includes: Step S210: Construct a spatiotemporal joint optimization model for the cutting arm operation process.

[0033] Step S220: Map the three-dimensional geometric information in the real-time environmental information to the spatial and obstacle avoidance constraints in the spatiotemporal joint optimization model.

[0034] Step S230: Map the interaction mechanics information in the real-time environmental information to the rock mass mechanics property constraints of the spatiotemporal joint optimization model.

[0035] Step S240: Map the configuration parameter library to the mechanical limit constraints of the cutting arm in the spatiotemporal joint optimization model.

[0036] Step S250: Under the conditions of satisfying the spatial and obstacle avoidance constraints, rock mass mechanical property constraints and mechanical limit constraints, at least one of the total operation time, total energy consumption and dual-arm load balance within the cutting operation cycle is used as the optimization objective. The target configuration sequence, the cooperative cutting control parameter sequence and the predicted mechanical information are solved synchronously to generate a globally optimized cutting operation plan.

[0037] Specifically, a spatiotemporal joint optimization model for the cutting arm operation process is constructed using a deep spatiotemporal fusion neural network. The main body of the network consists of four parts: a 3D environment encoding branch, a cutting arm configuration encoding branch, a temporal feature fusion layer, and a joint decision output layer. The 3D environment encoding branch uses a 3D convolutional neural network (3D-CNN) to extract spatial features from the 3D geometric data of the working surface. The cutting arm configuration encoding branch uses a fully connected neural network to encode configuration parameters such as joint displacement, sleeve stroke, and mechanical dimensions. The temporal feature fusion layer uses a long short-term memory network (LSTM) to model the temporal dependencies of states at multiple time points, splicing and fusing spatial and temporal features. Using historical construction data as the training set, supervised training is performed with loss functions of dual-arm coordination error, mechanical deviation, and operation efficiency. The output model can characterize the coupling relationship between spatial state, time series, and mechanical response, achieving unified modeling and optimal decision reasoning for the entire cutting arm operation process.

[0038] The 3D geometric information of the working face collected by the 3D vision sensor in the real-time environmental information is input into the 3D environment encoding branch of the spatiotemporal joint optimization model. The spatial features such as the tunnel cross-section contour, ditch location, obstacle position, and spatial boundary range are extracted and encoded by 3D-CNN to generate a 3D spatial feature vector. This feature vector is then mapped to spatial constraints and obstacle avoidance constraints that the model can recognize, and the safe working range, maximum extension boundary, prohibited area and obstacle avoidance threshold of the cutting arm are defined as hard constraints that must be met in the model solution process to ensure that the cutting arm does not interfere or collide with the tunnel structure, surrounding equipment and rock obstacles during operation and configuration transformation.

[0039] The mechanical information of the interaction between the cutting arm and the rock mass, collected by mechanical sensors in the real-time environment, is input into the temporal feature fusion layer of the spatiotemporal joint optimization model. The mechanical features such as real-time cutting resistance, load fluctuation, and stress distribution are temporally encoded and fused by LSTM, and mapped into rock mass mechanical property constraints that the model can recognize. This is used to characterize the cutting difficulty, load limit, and mechanical stability boundary of heterogeneous and anisotropic rock layers.

[0040] Mechanical structural parameters such as joint rotation range, multi-stage telescopic sleeve stroke, maximum load capacity, upper limit of drive speed, and rigid locking position, which are pre-stored in the cutting arm configuration parameter library, are input into the cutting arm configuration encoding branch of the spatiotemporal joint optimization model. After normalization encoding by a fully connected neural network, they are mapped to the mechanical limit constraints that the model must follow when making decisions and solving problems. This clarifies the physical boundaries and safety limits of the cutting arm in configuration transformation and operation, avoids violations such as overtravel, overload, and overspeed, and ensures the structural safety and stable operation of the equipment.

[0041] Under the conditions of simultaneously satisfying spatial and obstacle avoidance constraints, rock mass mechanical property constraints, and mechanical limit constraints, the entire cutting operation cycle is first discretized into multiple continuous planning control periods. For each planning control period, a set of decision variables is defined, including the cutting arm configuration, cutting parameter combination, and predicted load. Based on this, state transition equations and dynamic equations connecting the decision variables of adjacent planning control periods are established based on the aforementioned three types of constraints, forming a unified global constraint to ensure continuous configuration, smooth parameters, and reasonable mechanical state between periods. Then, a decision objective function is constructed using at least one of the total operation time, total energy consumption, and dual-arm load balance. Within the global constraint range, all decision variable sequences are iterated and optimized to select the decision combination that satisfies both the constraints and optimizes the decision objective function. Finally, the target configuration sequence, collaborative cutting control parameter sequence, and predicted mechanical information of the first and second cutting arms are solved simultaneously to generate a globally optimized cutting operation plan.

[0042] In one possible implementation, step S250 further includes: Step S251: Discretize the entire cutting operation cycle into multiple consecutive planning control periods, and define a set of decision variables for each planning control period. The decision variables include at least the cutting arm configuration in the target configuration sequence for the corresponding period, the cutting parameter combination in the collaborative cutting control parameter sequence for the corresponding period, and the predicted load in the predicted mechanical information for the corresponding period.

[0043] Step S252: Within the framework of the spatiotemporal joint optimization model, based on the spatial and obstacle avoidance constraints, rock mass mechanical property constraints, and mechanical limit constraints, establish state transition equations and dynamic equations connecting decision variables of adjacent planning control periods as global constraints.

[0044] Step S253: Construct a decision objective function using at least one of the total operation time, total energy consumption, and dual-arm load balance.

[0045] Step S254: Solve for the sequence of decision variables that satisfy the global constraints and optimize the decision objective function to obtain the target configuration sequence, the cooperative cutting control parameter sequence, and the predicted mechanical information.

[0046] Specifically, by sampling at equal time intervals, the continuous cutting operation cycle is broken down into multiple sequential and non-overlapping planning and control periods, thereby achieving time-series discretization of the operation process. For each planning and control period, a set of decision variables is defined for model optimization. This set of decision variables includes at least: the cutting arm configurations such as the joint angle and telescopic sleeve stroke corresponding to the two cutting arms in the current period; the cutting parameter combinations such as the feed rate, cutting depth, rotation speed, and phase difference of the two arms that match the configuration; and the predicted load of the interaction between the cutting arm and the rock mass in the current period, providing standardized decision units for subsequent state recursion and global optimization.

[0047] Within the computational framework of the spatiotemporal joint optimization model, based on spatial and obstacle avoidance constraints, rock mass mechanical property constraints, and mechanical limit constraints as fundamental boundary conditions, state transition equations and dynamic equations are established to connect decision variables of adjacent planning control periods. These two types of equations are explicitly added as global constraints to the optimization solution process. Specifically, the state transition equations use the cutting arm configuration, joint angle, and telescopic sleeve stroke of the previous planning control period as inputs, combined with kinematic continuity constraints and spatial obstacle avoidance constraints, to calculate the cutting arm configuration for the next planning control period. This configuration describes the continuous change relationship of the cutting arm configuration between adjacent periods, ensuring that configuration transformation is abrupt and dynamic. The dynamic trajectory is smooth, does not exceed the spatial boundary, and does not interfere with the structure. The dynamic equation takes the control parameters such as the current cutting arm configuration, feed rate, cutting depth, and saw blade rotation speed as input, and combines the constraints of rock mechanics properties and mechanical limits to calculate the mechanical states such as cutting load, driving torque, and arm stress, which are used to describe the mechanical response law and causal relationship under specific configuration and control parameters. The above state transition equation and dynamic equation together define the evolution rules of decision variables in the time dimension, which serve as the core constraints that must be satisfied in the solution process, ensuring that all decision variable sequences are continuous in time, physically feasible, and mechanically stable.

[0048] Based on actual operational requirements, one or more optimization indicators are selected from the total operation time, total energy consumption, and dual-arm load balance within the cutting operation cycle. A unified decision objective function is constructed through weighted summation. The total operation time indicator is used to characterize operation efficiency and is quantified by the cumulative time consumption of each planned control period. The total energy consumption indicator is calculated by integrating the output power and running time of each drive component in the corresponding period. The dual-arm load balance indicator is characterized by the absolute value or variance of the difference between the predicted loads of the first and second cutting arms. When constructing the objective function, normalization coefficients and weight coefficients are set for each optimization indicator. The weight coefficients are adaptively allocated according to the emphasis of the construction scenario, so that the objective function can comprehensively reflect operation efficiency, energy consumption cost, and dual-arm operation stability, providing a unified quantitative evaluation standard for subsequent optimization of the global optimal decision variable sequence.

[0049] In the spatiotemporal joint optimization model, the feasible region is defined by global constraints consisting of state transition equations, dynamic equations, and various boundary constraints. An iterative optimization algorithm based on sequence decision-making is used to solve the sequence of decision variables. During the calculation, the cutting arm configuration, cutting parameter combination, and candidate predicted load values ​​for different planning control periods are substituted one by one. First, feasible solutions that are motion-continuous, mechanically compliant, and free from spatial interference are screened out through global constraints. Then, the feasible solutions are substituted into the decision objective function to calculate the comprehensive evaluation value. By continuously iterating and updating the candidate decision variables, the solution set that optimizes the decision objective function is gradually approached. Finally, the cutting arm configuration, cutting parameter combination, and predicted load corresponding to each time period are extracted from the optimal solution set and arranged in chronological order to form a continuous sequence. This yields the target configuration sequence, collaborative cutting control parameter sequence, and predicted mechanical information that satisfy global constraints and are globally optimal, thus completing the solution for the overall cutting operation planning.

[0050] In one possible implementation, step S250 further includes: The cutting parameter combination in the corresponding time period of the collaborative cutting control parameter sequence is coupled with the cutting arm configuration in the corresponding time period of the target configuration sequence and the predicted load in the corresponding time period of the predicted mechanical information. The cutting parameter combination includes at least: The feed rate and depth of cut parameters are matched to the working space of the cutting arm under the described cutting arm configuration.

[0051] Rotational speed parameters that match the cutter head wear model and rock mass breaking efficiency model under the predicted load.

[0052] The phase difference parameter is used to coordinate the dynamic load balancing of the first and second cutting arms in the cutting arm configuration.

[0053] Specifically, the combination of cutting parameters corresponding to the planned control period in the collaborative cutting control parameter sequence forms a strong coupling relationship with the cutting arm configuration in the same period in the target configuration sequence and the predicted load in the same period in the predicted mechanical information.

[0054] Among them, the feed rate and cutting depth parameters are the core components of the cutting parameter combination, and their values ​​are directly coupled and matched with the cutting arm configuration corresponding to the current planned control period. The cutting arm configuration includes the joint angles of the cutting arm, the extension length of the telescopic sleeve, and the spatial pose parameters. Based on this configuration, the effective reachable working space, boundary range, and interference-free working area of ​​the cutting arm in the current posture can be determined by kinematic calculations. The feed rate is the feed movement rate of the cutting arm along the excavation direction of the ditch, and the cutting depth is the radial depth dimension of the cutter head cutting into the rock mass. In the parameter determination process, the feed rate and cutting depth are adaptively adjusted according to the spatial posture, extension limit, and spatial obstacle avoidance constraints of the cutting arm in the current configuration, so that both are always within the safe working range of the reachable working space. This ensures that the cutting operation can cover the target ditch contour, while avoiding the cutting arm from exceeding the spatial boundary or interfering with the tunnel structure due to excessive feed rate or cutting depth. This achieves coordinated adaptation between the cutting parameters, the arm configuration, and the working space.

[0055] Rotational speed parameters, as a crucial component of the cutting parameter combination, are coupled and matched with the predicted load, cutter head wear model, and rock mass fragmentation efficiency model corresponding to the current planned control period. The predicted load is the cutter head cutting resistance and load amplitude pre-calculated based on the current cutting arm configuration, cutting depth, and rock mass mechanical properties. The cutter head wear model characterizes the wear rate and lifespan decay of the cutter head under different loads, rotational speeds, and cutting conditions. The rock mass fragmentation efficiency model reflects the quantitative relationship between the cutter head's cutting and fragmenting ability, forming quality, and cutting amount per unit time under different rotational speeds. In the parameter determination process, the predicted load is used as input, and the rotational speed is controlled within a low-wear range based on the cutter head wear model. Simultaneously, the rotational speed is adjusted to the optimal range for rock mass cutting efficiency based on the rock mass fragmentation efficiency model. This ensures that the rotational speed parameters guarantee stable and efficient rock mass fragmentation while suppressing excessively rapid cutter head wear under high loads, achieving synergistic optimization of cutting efficiency and tool life.

[0056] The phase difference parameter, a key parameter in the cutting parameter combination for dual-arm coordinated control, coordinates the cutting timing and force distribution of the first and second cutting arms under the current cutting arm configuration, achieving dynamic load balancing. The cutting arm configuration includes the joint angles, extension strokes, and relative spatial poses of each arm, directly determining the cutting range and force transmission path within the working area. Dynamic load balancing refers to adjusting the action timing during continuous cutting to balance the real-time cutting load and driving torque fluctuations of both arms, preventing one arm from being under high load for extended periods, thus avoiding structural fatigue or efficiency imbalance. The phase difference parameter controls the time offset of the cutting actions of the two cutter heads, ensuring that when one arm is in the cutting force stage, the other is in the no-load or low-load transition stage, thereby staggering the peak load in timing. Combined with the spatial pose and predicted load under the current configuration, the phase difference value is adaptively allocated, ensuring uniform load distribution and matched movement rhythm between the two cutting arms during coordinated operation, resulting in smoother and more reliable overall operation.

[0057] In one possible implementation, step S300 further includes: Step S310: Based on the initial target configuration in the target configuration sequence, calculate the target displacement of each joint and telescopic sleeve, and generate the corresponding multi-axis linkage motion command.

[0058] Step S320: After confirming that the cutting arm has reached the initial target configuration through pose feedback, identify the key load-bearing nodes based on the predicted load distribution map for the initial target configuration in the predicted mechanical information, and control the mechanical locking mechanism to apply a preload force matching the predicted load distribution map to the selected key load-bearing nodes to complete rigid locking.

[0059] Specifically, motion control calculations are performed based on the initial target configuration in the target configuration sequence. The initial target configuration is the reference spatial posture of the cutting arm before the start of the cutting operation, which is determined by the target rotation angle of each rotary joint of the cutting arm and the target extension length of the telescopic sleeve. The system performs inverse kinematics calculations based on the kinematic parameters of the cutting arm itself, such as the link length and joint arrangement, combined with the initial target configuration, to obtain the target angular displacement required for each rotary joint and the target linear displacement required for the telescopic sleeve. After obtaining the target displacement of each axis, the displacement of each axis is converted into pulse commands, speed commands, and position commands that can be recognized by the servo driver according to the interpolation rules of multi-axis coordinated motion. Finally, multi-axis linkage motion commands that can drive the joints and the telescopic sleeve to move synchronously and in coordination are generated to ensure that the cutting arm smoothly and accurately approaches and reaches the initial target configuration.

[0060] By installing encoders, tilt sensors, and other posture detection devices at each joint and telescopic sleeve of the cutting arm, the joint angle, telescopic stroke, and spatial posture information are collected in real time. The actual posture is compared and verified with the initial target configuration in a closed loop. After confirming that the cutting arm has accurately reached the set initial target configuration, the predicted load distribution map corresponding to the initial posture is retrieved from the predicted mechanical information. The predicted load distribution map is a cloud map of the stress and load magnitude at each structural position of the cutting arm during the cutting operation. It can intuitively reflect the stress concentration area and high load-bearing parts. Based on this distribution map, threshold judgment and area identification are used to screen out the key load-bearing nodes that bear significant structural stress and the main cutting load. Then, the controller outputs control signals to the mechanical locking mechanism at the corresponding position according to the predicted load amplitude corresponding to each key load-bearing node. The mechanical locking mechanism applies a preload force adapted to the load magnitude to the key nodes such as joints or telescopic sleeves, so that the arm body forms a gapless and highly rigid locking state before the cutting operation, suppressing vibration, deformation, and displacement during the cutting process, and completing rigid locking.

[0061] In one possible implementation, step S400 further includes: Step S410: Convert the collaborative cutting control parameter sequence into drive motor torque commands and drive motor speed commands for the first and second cutting arms.

[0062] Step S420: Based on the drive motor torque command and drive motor speed command, perform cutting execution control.

[0063] Specifically, the coordinated cutting control parameter sequence obtained through global optimization is converted into electrical control commands to drive the cutting arms to perform cutting actions through a control algorithm. Specifically, the feed rate, cutting depth, rotation speed and phase difference between the two arms in the corresponding time period of the parameter sequence are used as inputs. Combined with the reduction ratio, transmission efficiency, cutting resistance of the cutter head and predicted load of the cutting arm transmission mechanism, the output torque and operating speed required by the feed drive motor and rotation drive motor of the first cutting arm and the second cutting arm respectively are calculated. Then, the calculation results are converted into drive motor torque commands and drive motor speed commands that can be directly recognized by the servo control system, so that the motor output of the two cutting arms matches the coordinated cutting parameters, the rigid locking posture of the arm body and the predicted load.

[0064] Using the generated drive motor torque and speed commands as control inputs, a closed-loop servo control method is adopted to adjust the actuators of the first and second cutting arms in real time. The control system compares the command values ​​with the actual output speed of the motor and the real-time feedback torque and makes dynamic corrections. The drive cutting arms operate stably according to the preset feed speed, rotation speed and phase difference parameters. At the same time, the motor output is constrained within the torque range corresponding to the predicted load, so that the two arms can complete the rock cutting operation synchronously under a rigid locking configuration. This ensures that the cutting process is smooth, the load is balanced and does not exceed the mechanical and load safety boundaries, and finally achieves the cutting execution control that conforms to the plan.

[0065] In one possible implementation, step S500 further includes: Step S510: Establish a reference curve for the change of the predicted mechanical information over time, wherein the reference curve contains multiple linearly distributed reference mechanical information values.

[0066] Step S520: During the cutting operation, continuously collect actual mechanical information and compare it with the reference mechanical information value at the same time point in real time, and calculate the real-time deviation index.

[0067] Step S530: When the real-time deviation index continues to exceed the preset deviation threshold or the accumulated deviation energy exceeds the threshold, local replanning is triggered to update the subsequent target configuration sequence and / or cooperative cutting control parameter sequence.

[0068] Specifically, using the predicted mechanical information corresponding to the entire cutting process as the benchmark data, a mechanical reference curve that changes continuously with the operation time is established according to the time axis sequence of the planned control period. This reference curve uses time as the horizontal axis and predicted mechanical information such as predicted load and driving torque as the vertical axis. The reference mechanical information values ​​of adjacent planned control periods are connected sequentially by straight line segments to form a piecewise linear curve composed of multiple linearly distributed and time-continuous reference mechanical information values. This curve is used to provide a standard comparison benchmark for monitoring the mechanical state during the subsequent actual cutting process.

[0069] During the continuous cutting operation, force sensors and torque sensors located on the cutting arm, drive motor, and mechanical locking mechanism collect actual mechanical information such as actual cutting load and motor output torque in real time. According to the same time sampling points as the reference curve, the actual mechanical information collected at each moment is compared with the corresponding reference mechanical information value in real time. Based on the difference, rate of change, and weighting coefficient, the real-time deviation index is calculated to quantitatively characterize the degree of deviation between the actual cutting mechanical state and the planned reference state.

[0070] The real-time deviation index calculated in real time is continuously monitored and integrated. When the index consistently exceeds the preset deviation threshold in multiple consecutive sampling periods, or when the deviation energy accumulated by the deviation exceeds the set safety threshold, it is determined that the actual cutting condition has significantly deviated from the planned state and there is an abnormal risk. At this time, the system automatically triggers the online local replanning mechanism. Based on the current actual pose, mechanical feedback and rock mass conditions, it re-optimizes and calculates the target configuration sequence and / or collaborative cutting control parameter sequence for subsequent unexecuted periods, so as to achieve adaptive correction of cutting trajectory, motion posture and operation parameters, and ensure that the cutting operation returns to a stable and controllable state.

[0071] In one possible implementation, step S530 further includes: Step S531: Using the current moment as the new planning starting point, and taking the currently collected real-time environmental information, the current cutting arm configuration, and the remaining target trench parameters as input, re-execute the joint solution operation to generate an updated cutting operation plan starting from the current moment.

[0072] Step S532: Calculate the difference between the initial configuration of the new target configuration sequence in the updated cutting operation plan and the current configuration of the cutting arm.

[0073] Based on the degree of difference, a tiered update strategy is implemented: a. If the degree of difference is lower than the first level, then only the new collaborative cutting control parameter sequence in the updated cutting operation plan shall be used.

[0074] b. If the degree of difference is between the first level and the second level, control the cutting arm to move to the starting configuration of the new target configuration sequence and relock it, while adopting a new cooperative cutting control parameter sequence.

[0075] c. If the degree of difference is higher than the second level, then the current operation process is terminated and an intervention request is sent.

[0076] Specifically, the current moment when the deviation exceeds the limit is taken as the starting point for the new planning. The current rock mass working condition, load change amplitude, external disturbance and other environmental information are collected in real time by sensors. Combined with the real-time configuration formed by the current actual joint angle of the cutting arm and the extension length of the sleeve, as well as the remaining target groove parameters such as the width, depth and slope of the uncut ditch, these are all input into the kinematic and dynamic joint solution model. Through iterative optimization calculation, the pose constraint, mechanical constraint and dual-arm load balance constraint are satisfied again, and the updated cutting operation plan for the subsequent operation segment starting from the current moment is generated. It includes a new target configuration sequence adapted to the current working condition and a new cooperative cutting control parameter sequence.

[0077] The initial configuration of the new target configuration sequence in the updated cutting operation plan, as well as the joint rotation vector and telescopic sleeve displacement vector corresponding to the current real-time configuration of the cutting arm, are extracted respectively. By normalizing the position differences of the corresponding joints and sleeves and weighted summing, the difference value representing the degree of pose deviation is calculated. This value comprehensively reflects the overall difference between the planned posture and the actual posture in terms of spatial position and extension length, providing a quantitative judgment basis for the subsequent hierarchical update strategy.

[0078] The calculated difference is compared with preset first-level and second-level thresholds, and the corresponding job update strategy is executed according to the comparison results. When the calculated configuration difference is lower than the set first-level threshold, it means that the current actual configuration of the cutting arm deviates little from the updated planned initial configuration, and the spatial posture and extension state are basically matched, so there is no need to adjust the arm posture or relock. At this time, the control system directly and smoothly switches to the new cooperative cutting control parameter sequence in the updated cutting job plan, and continues to execute the cutting operation according to the updated feed rate, rotation speed, phase difference and other parameters, so as to achieve parameter adaptive optimization without interrupting the operation process.

[0079] When the configuration difference is between the first and second level thresholds, it indicates that there is an adjustable deviation between the current cutting arm posture and the newly planned initial configuration. The system first releases the preload of the original mechanical locking mechanism, then performs inverse kinematics calculation based on the updated initial configuration, and outputs multi-axis linkage commands to control the cutting arm to move smoothly to the target posture. After the position is reached and confirmed by posture feedback, the system re-identifies the key load-bearing nodes based on the predicted load corresponding to the new plan, controls the mechanical locking mechanism to apply matching preload to complete the secondary rigid locking, and simultaneously switches to a new collaborative cutting control parameter sequence to continue the cutting operation with the updated operating parameters.

[0080] When the configuration difference exceeds the set second-level threshold, it indicates that the current actual position of the cutting arm deviates too much from the newly planned initial configuration, and a safe transition cannot be achieved through smooth adjustment. At this time, the control system immediately cuts off the drive output, stops all current operation processes such as cutting and feeding, locks the relevant actuators to prevent accidental actions, and sends a manual intervention request in the form of audible and visual alarms and status messages to the host computer and field terminal, waiting for the operator to confirm and handle before resuming the subsequent process.

[0081] In one possible implementation, step S100 further includes: Step S110: The body posture information of the cutting arm is obtained in real time by encoders installed at each joint of the cutting arm and displacement sensors installed at each telescopic sleeve.

[0082] Step S120: The pose information of the body is spatiotemporally aligned and fused with the three-dimensional geometric information and the interaction mechanical information to form real-time environmental information with a unified spatiotemporal reference.

[0083] Specifically, encoders pre-installed at each rotating joint of the cutting arm are used to collect the rotation angle signals of the joints in real time. At the same time, displacement sensors installed on each telescopic sleeve are used to detect the axial extension length and position signals of the sleeve in real time. The controller collects and calculates the above multi-channel sensor data to obtain the position, attitude and extension state of each section of the cutting arm in space, forming complete body posture information, which provides real-time posture feedback for subsequent environmental fusion and motion control.

[0084] The real-time collected cutting arm body pose information, ditch outline and construction scene 3D geometric information, and interaction mechanical information of the cutting contact area are spatiotemporally aligned under the same timestamp and spatial coordinate system. The data fusion algorithm eliminates the time delay and coordinate deviation between different data sources, and finally forms real-time environmental information with a unified spatiotemporal reference and containing pose, geometry and mechanical multi-dimensional features, providing a complete and consistent input basis for subsequent operation planning and deviation judgment.

[0085] Example 2, based on the same inventive concept as the intelligent telescopic control method for large-size cutting arms used in the aforementioned examples for tunnel drainage ditches, such as... Figure 2 As shown, this application provides an intelligent telescopic control system for a large-size cutting arm used in tunnel drainage ditches. The system and method embodiments in this application are based on the same inventive concept. The system includes: The environmental information acquisition module 10 is used to acquire real-time environmental information of the construction work surface. The real-time environmental information includes at least three-dimensional geometric information of the work surface collected by a 3D vision sensor, and mechanical information of the interaction between the cutting arm and the rock mass collected by a mechanical sensor.

[0086] The cutting operation planning generation module 20 is used to perform joint calculations based on the real-time environmental information, the preset target trench shape parameters, and the configuration parameter library of the cutting arm to generate a cutting operation plan. The cutting operation plan includes at least the target configuration sequence of the first and second cutting arms during the cutting operation cycle, the cooperative cutting control parameter sequence corresponding to the target configuration sequence, and the predicted mechanical information.

[0087] The locking operation execution module 30 is used to control each drive component of the cutting arm to change from the current configuration to the initial target configuration of the target configuration sequence according to the cutting operation plan, and to control the mechanical locking mechanism to perform the locking operation.

[0088] The cutting operation execution module 40 is used to control the first cutting arm and the second cutting arm to perform cutting operations according to the cooperative cutting control parameter sequence when the cutting arm is in the locked initial target configuration.

[0089] The control parameter sequence update module 50 is used to continuously compare the actual mechanical information collected with the predicted mechanical information in the cutting operation plan during the cutting operation. When the comparison deviation exceeds the preset allowable range, it triggers local replanning to update the subsequent target configuration sequence and / or collaborative cutting control parameter sequence.

[0090] Furthermore, the system is also used to implement the following functions: A spatiotemporal joint optimization model is constructed with the cutting arm operation process as the object; the three-dimensional geometric information in the real-time environmental information is mapped to the spatial and obstacle avoidance constraints in the spatiotemporal joint optimization model; the interaction mechanics information in the real-time environmental information is mapped to the rock mass mechanics property constraints in the spatiotemporal joint optimization model; the configuration parameter library is mapped to the mechanical limit constraints of the cutting arm in the spatiotemporal joint optimization model; under the condition of satisfying the spatial and obstacle avoidance constraints, rock mass mechanics property constraints and mechanical limit constraints, at least one of the total operation time, total energy consumption and dual-arm load balance degree in the cutting operation cycle is used as the optimization objective, and the target configuration sequence, the cooperative cutting control parameter sequence and the predicted mechanics information are solved synchronously to generate a globally optimized cutting operation plan.

[0091] Furthermore, the system is also used to implement the following functions: The entire cutting operation cycle is discretized into multiple consecutive planning and control periods, and a set of decision variables is defined for each planning and control period. The decision variables include at least the cutting arm configuration of the corresponding period in the target configuration sequence, the cutting parameter combination of the corresponding period in the collaborative cutting control parameter sequence, and the predicted load of the corresponding period in the predicted mechanical information. Within the framework of the spatiotemporal joint optimization model, based on the spatial and obstacle avoidance constraints, rock mass mechanical property constraints, and mechanical limit constraints, a state transition equation and a dynamic equation connecting the decision variables of adjacent planning and control periods are established as global constraints. The decision objective function is constructed using at least one of the total operation time, total energy consumption, and dual-arm load balance. The decision variable sequence that satisfies the global constraints and optimizes the decision objective function is solved to obtain the target configuration sequence, the collaborative cutting control parameter sequence, and the predicted mechanical information.

[0092] Furthermore, the system is also used to implement the following functions: Feed rate and depth of cut parameters that match the reachable workspace of the cutting arm under the cutting arm configuration; rotational speed parameters that match the cutter head wear model and rock mass breaking efficiency model under the predicted load; and phase difference parameters used to coordinate the dynamic load balance of the first and second cutting arms under the cutting arm configuration.

[0093] Furthermore, the system is also used to implement the following functions: Based on the initial target configuration in the target configuration sequence, the target displacement of each joint and telescopic sleeve is calculated, and corresponding multi-axis linkage motion commands are generated. After the cutting arm reaches the initial target configuration through pose feedback, the key load-bearing nodes are identified based on the predicted load distribution map for the initial target configuration in the predicted mechanical information, and the mechanical locking mechanism is controlled to apply a preload force matching the predicted load distribution map to the selected key load-bearing nodes to complete rigid locking.

[0094] Furthermore, the system is also used to implement the following functions: The coordinated cutting control parameter sequence is converted into drive motor torque commands and drive motor speed commands for the first and second cutting arms; cutting execution control is performed based on the drive motor torque commands and drive motor speed commands.

[0095] Furthermore, the system is also used to implement the following functions: A reference curve is established to show how the predicted mechanical information changes over time. The reference curve contains multiple linearly distributed reference mechanical information values. During the cutting operation, the actual mechanical information is continuously collected and compared with the reference mechanical information values ​​at the same time point in real time to calculate the real-time deviation index. When the real-time deviation index continuously exceeds the preset deviation threshold or the accumulated deviation energy exceeds the threshold, local replanning is triggered to update the subsequent target configuration sequence and / or collaborative cutting control parameter sequence.

[0096] Furthermore, the system is also used to implement the following functions: Using the current moment as the new planning starting point, and taking the currently collected real-time environmental information, the current cutting arm configuration, and the remaining target trench parameters as input, the joint solution operation is re-executed to generate an updated cutting operation plan from the current moment. The difference between the initial configuration of the new target configuration sequence in the updated cutting operation plan and the current configuration of the cutting arm is calculated. Based on the difference, an update strategy is executed in stages: a. If the difference is lower than the first level, only the new collaborative cutting control parameter sequence in the updated cutting operation plan is used; b. If the difference is between the first and second levels, the cutting arm is controlled to move to the initial configuration of the new target configuration sequence and relocked, while the new collaborative cutting control parameter sequence is used; c. If the difference is higher than the second level, the current operation process is stopped and an intervention request is sent.

[0097] Furthermore, the system is also used to implement the following functions: The body pose information of the cutting arm is acquired in real time by encoders installed at each joint of the cutting arm and displacement sensors installed at each telescopic sleeve. The body pose information is spatiotemporally aligned and fused with the three-dimensional geometric information and the interaction mechanical information to form real-time environmental information with a unified spatiotemporal reference.

[0098] It should be noted that the order of the embodiments described above is for descriptive purposes only and does not represent the superiority or inferiority of the embodiments. Specific embodiments of this specification have been described above. Furthermore, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0099] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0100] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application intends to include such modifications and variations.

Claims

1. A method for intelligent telescopic control of a large-size cutting arm used in tunnel drainage ditches, characterized in that, The method is applied to tunnel construction equipment comprising at least two cutting arms, wherein the cutting arms are variable configuration arms having multiple joints and / or multi-stage telescopic sleeves, and the method includes: The real-time environmental information of the construction work surface is obtained, including at least the three-dimensional geometric information of the work surface collected by a 3D vision sensor and the mechanical information of the interaction between the cutting arm and the rock mass collected by a mechanical sensor. Based on the real-time environmental information, the preset target trench shape parameters, and the configuration parameter library of the cutting arm, a joint solution calculation is performed to generate a cutting operation plan. The cutting operation plan includes at least the target configuration sequence of the first and second cutting arms during the cutting operation cycle, the cooperative cutting control parameter sequence corresponding to the target configuration sequence, and the predicted mechanical information. According to the cutting operation plan, control each driving component of the cutting arm to change from the current configuration to the initial target configuration of the target configuration sequence, and control the mechanical locking mechanism to perform the locking operation; With the cutting arm in the locked initial target configuration, the first and second cutting arms are controlled to perform cutting operations according to the cooperative cutting control parameter sequence; During the cutting operation, the actual mechanical information collected is continuously compared with the predicted mechanical information in the cutting operation plan. When the comparison deviation exceeds the preset allowable range, local replanning is triggered to update the subsequent target configuration sequence and / or collaborative cutting control parameter sequence.

2. The intelligent telescopic control method for a large-size cutting arm used in tunnel drainage ditches as described in claim 1, characterized in that, Based on the real-time environmental information, the preset target trench shape parameters, and the configuration parameter library of the cutting arm, a joint solution calculation is performed to generate a cutting operation plan, including: Construct a spatiotemporal joint optimization model focusing on the cutting arm operation process; The three-dimensional geometric information in the real-time environmental information is mapped to the spatial and obstacle avoidance constraints in the spatiotemporal joint optimization model; The interaction mechanics information in the real-time environmental information is mapped to the rock mass mechanical property constraints of the spatiotemporal joint optimization model; The configuration parameter library is mapped to the mechanical limit constraints of the cutting arm in the spatiotemporal joint optimization model; Under the conditions of satisfying the spatial and obstacle avoidance constraints, rock mass mechanical property constraints and mechanical limit constraints, at least one of the total operation time, total energy consumption and dual-arm load balance within the cutting operation cycle is used as the optimization objective. The target configuration sequence, the cooperative cutting control parameter sequence and the predicted mechanical information are solved synchronously to generate a globally optimized cutting operation plan.

3. The intelligent telescopic control method for a large-size cutting arm used in tunnel drainage ditches as described in claim 2, characterized in that, Under the conditions of satisfying the spatial and obstacle avoidance constraints, rock mass mechanical property constraints, and mechanical limit constraints, with at least one of the total operation time, total energy consumption, and dual-arm load balance within the cutting operation cycle as the optimization objective, the target configuration sequence, the cooperative cutting control parameter sequence, and the predicted mechanical information are solved simultaneously to generate a globally optimized cutting operation plan, including: The entire cutting operation cycle is discretized into multiple consecutive planning and control periods, and a set of decision variables is defined for each planning and control period. The decision variables include at least the cutting arm configuration in the target configuration sequence for the corresponding period, the cutting parameter combination in the collaborative cutting control parameter sequence for the corresponding period, and the predicted load in the predicted mechanical information for the corresponding period. Within the framework of the spatiotemporal joint optimization model, based on the spatial and obstacle avoidance constraints, rock mass mechanical property constraints, and mechanical limit constraints, a state transition equation and a dynamic equation connecting the decision variables of adjacent planning control periods are established as global constraints. The decision objective function is constructed using at least one of the total operation time, total energy consumption, and dual-arm load balance. Solve the sequence of decision variables that satisfy the global constraints and optimize the decision objective function to obtain the target configuration sequence, the cooperative cutting control parameter sequence, and the predictive mechanical information.

4. The intelligent telescopic control method for a large-size cutting arm used in tunnel drainage ditches as described in claim 3, characterized in that, The cutting parameter combination in the corresponding time period of the collaborative cutting control parameter sequence is coupled with the cutting arm configuration in the corresponding time period of the target configuration sequence and the predicted load in the corresponding time period of the predicted mechanical information. The cutting parameter combination includes at least: Feed rate and depth of cut parameters that match the reachable workspace of the cutting arm under the described cutting arm configuration; Rotational speed parameters that match the cutter head wear model and rock mass fragmentation efficiency model under the predicted load; The phase difference parameter is used to coordinate the dynamic load balancing of the first and second cutting arms in the cutting arm configuration.

5. The intelligent telescopic control method for a large-size cutting arm used in tunnel drainage ditches as described in claim 1, characterized in that, According to the cutting operation plan, the driving components of the cutting arm are controlled to transform from the current configuration to the initial target configuration of the target configuration sequence, and the mechanical locking mechanism is controlled to perform a locking operation, including: Based on the initial target configuration in the target configuration sequence, the target displacement of each joint and telescopic sleeve is calculated, and the corresponding multi-axis linkage motion command is generated. After confirming that the cutting arm has reached the initial target configuration through pose feedback, the key load-bearing nodes are identified based on the predicted load distribution map for the initial target configuration in the predicted mechanical information. The mechanical locking mechanism is then controlled to apply a preload force matching the predicted load distribution map to the selected key load-bearing nodes to complete rigid locking.

6. The intelligent telescopic control method for a large-size cutting arm used in tunnel drainage ditches as described in claim 1, characterized in that, With the cutting arms in the locked initial target configuration, controlling the first and second cutting arms to perform cutting operations according to the cooperative cutting control parameter sequence includes: The coordinated cutting control parameter sequence is converted into drive motor torque commands and drive motor speed commands for the first and second cutting arms. Cutting execution control is performed based on the drive motor torque command and drive motor speed command.

7. The intelligent telescopic control method for a large-size cutting arm used in tunnel drainage ditches as described in claim 1, characterized in that, During the cutting operation, the actual mechanical information collected is continuously compared with the predicted mechanical information in the cutting operation plan. When the comparison deviation exceeds a preset allowable range, local replanning is triggered to update the subsequent target configuration sequence and / or cooperative cutting control parameter sequence, including: A reference curve is established to show how the predicted mechanical information changes over time. The reference curve contains multiple linearly distributed reference mechanical information values. During the cutting operation, the actual mechanical information is continuously collected and compared with the reference mechanical information value at the same time point in real time to calculate the real-time deviation index. When the real-time deviation index continuously exceeds the preset deviation threshold or the accumulated deviation energy exceeds the threshold, local replanning is triggered to update the subsequent target configuration sequence and / or collaborative cutting control parameter sequence.

8. The intelligent telescopic control method for a large-size cutting arm used in tunnel drainage ditches as described in claim 7, characterized in that, Triggering local replanning to update subsequent target configuration sequences and / or cooperative cutting control parameter sequences, including: Taking the current moment as the new planning starting point, and using the currently collected real-time environmental information, the current cutting arm configuration, and the remaining target trench parameters as input, the joint solution operation is re-executed to generate an updated cutting operation plan starting from the current moment. Calculate the difference between the initial configuration of the new target configuration sequence in the updated cutting operation plan and the current configuration of the cutting arm; Based on the degree of difference, a tiered update strategy is implemented: a. If the degree of difference is lower than the first level, then only the new collaborative cutting control parameter sequence in the updated cutting operation plan shall be used; b. If the degree of difference is between the first level and the second level, control the cutting arm to move to the starting configuration of the new target configuration sequence and relock it, while adopting a new cooperative cutting control parameter sequence; c. If the degree of difference is higher than the second level, then the current operation process is terminated and an intervention request is sent.

9. The intelligent telescopic control method for a large-size cutting arm used in tunnel drainage ditches as described in claim 1, characterized in that, The acquisition of real-time environmental information of the construction work surface also includes: The body posture information of the cutting arm is obtained in real time by encoders installed at each joint of the cutting arm and displacement sensors installed at each telescopic sleeve. The body pose information, the three-dimensional geometric information, and the interaction mechanics information are spatiotemporally aligned and fused to form real-time environmental information with a unified spatiotemporal reference.

10. A large-size cutting arm intelligent telescopic control system for tunnel drainage ditches, characterized in that, The system is used to implement the intelligent telescopic control method for a large-size cutting arm for tunnel drainage ditches as described in any one of claims 1-9, the system comprising: The environmental information acquisition module is used to acquire real-time environmental information of the construction work surface. The real-time environmental information includes at least three-dimensional geometric information of the work surface collected by a 3D vision sensor, and mechanical information of the interaction between the cutting arm and the rock mass collected by a mechanical sensor. The cutting operation planning generation module is used to perform joint calculations based on the real-time environmental information, the preset target trench shape parameters, and the configuration parameter library of the cutting arm to generate a cutting operation plan. The cutting operation plan includes at least the target configuration sequence of the first and second cutting arms during the cutting operation cycle, the cooperative cutting control parameter sequence corresponding to the target configuration sequence, and the predicted mechanical information. The locking operation execution module is used to control each driving component of the cutting arm to change from the current configuration to the initial target configuration of the target configuration sequence according to the cutting operation plan, and to control the mechanical locking mechanism to perform the locking operation. The cutting operation execution module is used to control the first cutting arm and the second cutting arm to perform cutting operations according to the cooperative cutting control parameter sequence when the cutting arm is in the locked initial target configuration; The control parameter sequence update module is used to continuously compare the actual mechanical information collected with the predicted mechanical information in the cutting operation plan during the cutting operation. When the comparison deviation exceeds the preset allowable range, local replanning is triggered to update the subsequent target configuration sequence and / or collaborative cutting control parameter sequence.

Images