Double-wheel milling X, Y direction automatic deviation correction control algorithm

Abstract

This application discloses an automatic correction control algorithm for dual-wheel milling in the X and Y directions, belonging to the field of engineering machinery control technology. It is implemented based on a dual-wheel milling top correction mechanical device and an intelligent correction control system. This invention addresses the problems of delayed correction action, X and Y direction correction coupling interference, and parameters failing to adapt to changes in the geological strata in existing dual-wheel milling correction algorithms. Through a closed-loop control logic throughout the entire process, it uses completely independent dual control links in the X and Y directions to solve the correction parameters, combined with a multi-level early warning mechanism to achieve proactive pre-correction, and performs adaptive parameter optimization based on real-time stratum stiffness. This invention effectively solves core pain points in the industry, achieving an optimal balance between correction effect and construction stability, significantly improving the control accuracy of trench verticality, and is applicable to engineering construction scenarios such as hydraulic and hydropower seepage barriers and ultra-deep underground continuous walls.

Description

Technical Field

[0001] This invention relates to the field of engineering machinery control technology, and in particular to an automatic correction control algorithm for dual-wheel milling machines in the X and Y directions. Background Technology

[0002] Twin-wheel trenching machines are core equipment in the field of underground foundation engineering construction. They are widely used in various major foundation projects such as seepage prevention walls in water conservancy and hydropower strata and ultra-deep underground concrete structure construction. The quality of trenching operations directly determines the structural safety, seepage prevention performance, and long-term service life of the main project. Trench verticality, as the most critical quality control indicator in the twin-wheel trenching process, is a key factor in ensuring the smooth implementation of subsequent procedures and meeting engineering design requirements. It is also a key focus and challenge in on-site construction management.

[0003] To address the verticality control requirements during twin-wheel milling trenching, the industry has developed several mainstream correction technologies. These technologies revolve around real-time monitoring and dynamic adjustment of equipment posture, forming a complete trenching correction system. This system has been widely applied in various conventional geological formations, providing fundamental technical support for trenching verticality control. For example, patent CN113250261B discloses a working angle correction control system for a twin-wheel milling machine. This system can perform real-time correction of the pitch and roll angle deviations of the milling device relative to the horizontal plane, enabling the milling device to operate perpendicular to the trench bottom, thus improving the trenching stability, trenching quality, and work efficiency of the twin-wheel milling machine.

[0004] However, although existing correction control systems can correct the deviation of the milling cutter head, they still have inherent defects that are difficult to overcome in practical applications. First, existing correction control algorithms generally lack a dynamic comparison and early warning mechanism based on preset parameters. Their correction decisions rely solely on threshold judgments based on real-time acquired attitude data, and cannot systematically compare and analyze real-time data with preset equipment parameters, geological parameters, and construction condition parameters. This leads to the problem that when the milling cutter head's attitude deviation gradually accumulates and approaches a dangerous threshold, the control system cannot identify the deviation's development trend in advance, nor can it initiate pre-correction actions in a timely manner when the deviation reaches the warning value. This results in the milling cutter head's correction actions always lagging behind the occurrence and development of the deviation. Especially during ultra-deep trench construction, the small attitude deviations generated by the milling cutter head in shallow construction will amplify exponentially with the increase in milling depth. By the time the system detects significant deviation and initiates correction, the optimal adjustment opportunity has often been missed. This not only significantly increases the difficulty of correction operations but also easily leads to the trench verticality exceeding design specifications, directly affecting the construction quality of subsequent main projects.

[0005] Secondly, existing correction control algorithms have significant shortcomings in coordinated control of the X and Y directions. The correction actions of a twin-wheel trenching machine involve two independent directions: top correction in the Y direction and differential correction in the X direction. However, existing algorithms often employ simple parallel control logic, lacking adaptive parameter optimization mechanisms for different geological conditions and construction scenarios. Specifically, when joint correction in the X and Y directions is required, the correction actions in the two directions are prone to mutual coupling and interference, leading to disordered cutter head posture. Simultaneously, correction control parameters (such as correction force, correction speed, and stability coefficient) cannot be dynamically optimized based on key factors such as real-time geological stiffness, construction depth, and the rate of change of posture deviation. This makes it difficult to achieve the optimal balance between correction effect and stability, easily resulting in over- or under-correction, affecting the control accuracy of trench verticality and construction safety. Summary of the Invention

[0006] The main objective of this invention is to overcome the shortcomings of existing technologies and provide a dual-wheel milling automatic correction control algorithm in the X and Y directions that can establish a dynamic comparison and early warning mechanism with preset parameters, realize the early pre-correction of the milling cutter head, and avoid the situation where the correction actions of the milling cutter head in the X and Y directions are prone to mutual coupling interference and cause posture disorder, so as to significantly improve the control accuracy of groove perpendicularity.

[0007] To achieve the above objectives, this invention provides an automatic deviation correction control algorithm for dual-wheel milling in the X and Y directions, implemented based on a dual-wheel milling top deviation correction mechanical device and an intelligent deviation correction control system. The dual-wheel milling top deviation correction mechanical device includes a milling cutter holder, a Y-axis top deviation correction mechanism disposed at the top of the milling cutter holder, and a dual-milling wheel drive unit disposed at the bottom of the milling cutter holder. The intelligent deviation correction control system includes a perception layer module, an analysis layer module, a decision layer module, and an execution layer module connected in sequence, and includes the following steps:

[0008] S1: Completes algorithm control cycle calibration, presets multi-level verticality warning thresholds and correction safety thresholds, loads the formation adaptation parameter library, verifies the communication link between sensors and actuators, and enters real-time operation state after verification.

[0009] S2: The real-time attitude data of the milling cutter head, milling operation parameters, formation characteristic parameters and construction environment parameters within the current control cycle are synchronously collected through the perception layer module. The collected data is then spatiotemporally aligned, noise-reducing filtered and standardized to generate a standardized dataset.

[0010] S3: Based on a standardized dataset, calculate the real-time verticality deviation values ​​of the milling cutter head in the X and Y directions. Combine the milling feed speed and the formation stiffness parameters to predict the development trend of verticality deviation in the next control cycle. Compare the real-time deviation value and the predicted deviation value with the preset multi-level verticality warning thresholds respectively. When any value reaches the warning threshold of the corresponding level, trigger the corresponding pre-correction control command.

[0011] S4: Based on the triggered pre-correction control command, a completely independent dual control link is adopted in the X and Y directions to solve the pre-correction control parameters in the corresponding directions respectively; among them, the Y direction is based on the milling cutter head center of gravity offset and eccentric gravity moment model to solve the target displacement of the Y-direction top correction mechanism; the X direction is based on the dual milling wheel differential cutting correction torque model to solve the target speed difference and torque difference of the dual milling wheels.

[0012] S5: Compare the obtained X-axis and Y-axis pre-correction control parameters with the preset correction safety threshold, and eliminate parameter solutions that exceed the safety boundary; at the same time, based on the real-time formation stiffness parameters, adaptively optimize and adjust the pre-correction control parameters, and output the final optimal correction control parameters.

[0013] S6: The optimal correction control parameters are converted into action commands for the corresponding actuators through the execution layer module, and the Y-axis top correction mechanism and the X-axis dual milling wheel drive unit are driven to perform the corresponding correction actions synchronously. In the next control cycle, the actual attitude data after the correction is executed is collected through the perception layer module, compared with the preset verticality control target, the correction deviation is calculated and fed back to step S3, and the full closed-loop dynamic correction is completed.

[0014] Preferably, in step S1, the multi-level verticality warning threshold includes a first-level warning threshold, a second-level warning threshold, and a third-level action threshold, with each level of warning threshold corresponding to a different pre-correction control strategy.

[0015] Preferably, in step S3, the specific logic for verticality deviation prediction and early warning triggering is as follows: when the real-time deviation value or the predicted deviation value reaches the first-level early warning threshold, a preventive pre-correction strategy for center of gravity deviation is triggered; when the real-time deviation value or the predicted deviation value reaches the second-level early warning threshold, a progressive conventional pre-correction strategy is triggered; and when the real-time deviation value or the predicted deviation value reaches the third-level action threshold, an emergency locking correction strategy is triggered.

[0016] Preferably, in step S3, the formula for calculating the real-time perpendicularity deviation value of the milling cutter holder in the X direction is as follows: The formula for calculating the real-time perpendicularity deviation of the milling cutter holder in the Y direction is as follows: .

[0017] Preferably, in step S4, the solution model for the target displacement of the Y-direction top correction mechanism is: .

[0018] Preferably, in step S4, the solution model for the target speed difference between the two milling wheels is as follows: The solution model for the target torque difference of the two milling wheels is as follows: .

[0019] Preferably, in step S4, the dual control links in the X and Y directions adopt completely independent computing channels and control cycles; when the milling cutter head has only a single-direction deviation, only the control link in the corresponding direction is activated to perform correction; when the milling cutter head has a bidirectional synchronous deviation, the two control links run synchronously and perform bidirectional joint correction.

[0020] Preferably, in step S5, the specific method for adaptively optimizing and adjusting the pre-correction control parameters is as follows: based on the real-time collected milling torque and feed rate, calculate the equivalent formation stiffness of the current formation. ;when When the value exceeds the preset hard ground threshold, the weight of the correction effect is increased, the upper limit of the amplitude of the pre-correction control parameters is increased, and the correction response cycle is shortened; when When the value is less than the preset threshold for weak strata, the weight of the correction stability is increased, the change amplitude of the pre-correction control parameters within a single control cycle is reduced, and the correction transition cycle is extended.

[0021] Preferably, in step S5, the correction safety threshold includes the milling cutter holder center of gravity offset hard limit threshold, the Y-axis top correction mechanism maximum stroke threshold, the dual milling wheel maximum speed difference threshold, and the hanging mechanism load upper limit threshold; when any pre-correction control parameter exceeds the corresponding safety threshold, the parameter is automatically corrected to within the safety boundary, and a safety warning is triggered simultaneously.

[0022] Preferably, in step S6, the specific logic of the closed-loop feedback correction is as follows: collect the actual posture data of the milling cutter head after the correction is executed, and calculate the difference between the actual verticality deviation correction amount and the target correction amount; if the difference is within the preset allowable range, maintain the current correction control parameters until the deviation falls back to the safe range; if the difference exceeds the preset allowable range, input the difference as a compensation amount into the pre-correction parameter solution process of the next control cycle, dynamically correct the control parameters, until the actual deviation value meets the verticality control requirements.

[0023] Beneficial effects:

[0024] 1. This invention enables the pre-alignment of the milling cutter head correction control point and achieves early pre-alignment of the milling cutter head. It effectively intervenes and adjusts the posture of the milling cutter head in the early stage of deviation accumulation. Especially in the construction of ultra-deep grooves, it can not only effectively avoid the exponential amplification of shallow micro-deviations of the milling cutter head with the increase of milling depth, but also prevent the situation where the best adjustment opportunity has been missed when the milling cutter head has significant skew. This greatly reduces the difficulty of correction operations in ultra-deep grooves and ensures that the verticality of the groove is always within the allowable range of the design specifications, significantly improving the control accuracy of the verticality of the groove.

[0025] 2. This invention employs completely independent dual control links in the X and Y directions. Through completely independent computing channels and control logic, it can effectively avoid the situation where the milling cutter head's correction actions in the X and Y directions are prone to mutual coupling interference and cause attitude disorder. At the same time, the dual independent link architecture has both flexibility and adaptability. When the milling cutter head has only a deviation in one direction, only the control link in the corresponding direction needs to be activated to complete the correction. When the milling cutter head has a bidirectional synchronous deviation, the two links can run synchronously to achieve bidirectional joint correction, which can flexibly adapt to different deviation states of the milling cutter head.

[0026] 3. This invention adaptively optimizes and adjusts the pre-correction control parameters based on real-time formation stiffness parameters, allowing core parameters such as correction force and response speed to be dynamically adjusted according to different formation conditions, construction depth, and deviation change rate. This achieves the optimal balance between correction effect and construction stability, thus avoiding the problems of trench wall enlargement and instability caused by excessive correction force in soft formations, and solving the problem of correction failure caused by insufficient correction force in hard rock formations. This further improves the control accuracy of trench verticality and the safety of the construction process.

[0027] 4. This invention breaks through the limitations of existing serial open-loop control by using a closed-loop dynamic correction mechanism throughout the entire process. This allows the algorithm to dynamically adjust the control strategy and parameters according to the actual correction effect, ensuring that the correction action of the milling cutter head is always matched with the verticality control target of the groove, thus guaranteeing the accuracy and stability of the correction control.

[0028] 5. This invention can be directly embedded into the intelligent correction control system without modifying or adjusting the original hardware structure. It can be deeply adapted to the non-contact correction principle of the top correction mechanical device, thereby giving full play to the correction performance advantages of the dual-wheel milling top correction mechanical device. At the same time, the fully automated control logic greatly reduces manual intervention in the construction process, which not only reduces the risk of human error, but also significantly improves the overall operation efficiency of dual-wheel milling trenching construction. Attached Figure Description

[0029] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0030] Figure 1 This is a schematic diagram of the structure of a twin-wheel groove milling machine according to an embodiment of the present invention;

[0031] Figure 2 This is a flowchart of an embodiment of the automatic correction control algorithm for dual-wheel milling in the X and Y directions according to the present invention; Detailed Implementation

[0032] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, 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 some embodiments of this application, and not all embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0033] Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of the application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0034] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0035] In the description of this application, it should be noted that the use of terms such as "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer" to indicate orientation or positional relationships is based on the orientation or positional relationships shown in the accompanying drawings, or the orientation or positional relationships commonly used when the product is in use. These terms are used solely for the convenience of describing this application and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this application. Furthermore, the use of terms such as "first" and "second" in the description of this application is only used to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0036] Furthermore, the use of terms such as "horizontal" and "vertical" in the description of this application does not imply that the component is required to be absolutely horizontal or suspended, but rather that it may be slightly tilted. For example, "horizontal" simply means that its direction is more horizontal relative to "vertical," and does not mean that the structure must be completely horizontal, but rather that it may be slightly tilted.

[0037] In the description of this application, it should also be noted that, unless otherwise expressly specified and limited, the terms "set up," "install," "connect," and "link" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection between two components. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0038] Example 1:

[0039] This invention proposes an automatic correction control algorithm for dual-wheel milling in the X and Y directions.

[0040] In one embodiment of the present invention, the automatic deviation correction control algorithm in the X and Y directions of a dual-wheel milling machine is implemented based on a dual-wheel milling top deviation correction mechanical device and an intelligent deviation correction control system. The dual-wheel milling top deviation correction mechanical device includes a milling cutter holder, a Y-axis top deviation correction mechanism disposed on the top of the milling cutter holder, and a dual-milling wheel drive unit disposed at the bottom of the milling cutter holder. The intelligent deviation correction control system includes a perception layer module, an analysis layer module, a decision layer module, and an execution layer module that are connected in sequence, and includes the following steps:

[0041] S1: Completes algorithm control cycle calibration, presets multi-level verticality warning thresholds and correction safety thresholds, loads the formation adaptation parameter library, verifies the communication link between sensors and actuators, and enters real-time operation state after verification.

[0042] S2: The real-time attitude data of the milling cutter head, milling operation parameters, formation characteristic parameters and construction environment parameters within the current control cycle are synchronously collected through the perception layer module. The collected data is then spatiotemporally aligned, noise-reducing filtered and standardized to generate a standardized dataset.

[0043] S3: Based on a standardized dataset, calculate the real-time verticality deviation values ​​of the milling cutter head in the X and Y directions. Combine the milling feed speed and the formation stiffness parameters to predict the development trend of verticality deviation in the next control cycle. Compare the real-time deviation value and the predicted deviation value with the preset multi-level verticality warning thresholds respectively. When any value reaches the warning threshold of the corresponding level, trigger the corresponding pre-correction control command.

[0044] S4: Based on the triggered pre-correction control command, a completely independent dual control link is adopted in the X and Y directions to solve the pre-correction control parameters in the corresponding directions respectively; among them, the Y direction is based on the milling cutter head center of gravity offset and eccentric gravity moment model to solve the target displacement of the Y-direction top correction mechanism; the X direction is based on the dual milling wheel differential cutting correction torque model to solve the target speed difference and torque difference of the dual milling wheels.

[0045] S5: Compare the obtained X-axis and Y-axis pre-correction control parameters with the preset correction safety threshold, and eliminate parameter solutions that exceed the safety boundary; at the same time, based on the real-time formation stiffness parameters, adaptively optimize and adjust the pre-correction control parameters, and output the final optimal correction control parameters.

[0046] S6: The optimal correction control parameters are converted into action commands for the corresponding actuators through the execution layer module, and the Y-axis top correction mechanism and the X-axis dual milling wheel drive unit are driven to perform the corresponding correction actions synchronously. In the next control cycle, the actual attitude data after the correction is executed is collected through the perception layer module, compared with the preset verticality control target, the correction deviation is calculated and fed back to step S3, and the full closed-loop dynamic correction is completed.

[0047] It should be noted that the Y direction is... Figure 1 The front-back direction shown is the X-direction. Figure 1 The left and right directions are shown. The automatic correction control algorithm for the X and Y directions of the dual-wheel milling machine disclosed in this embodiment is implemented based on the dual-wheel milling top correction mechanical device and intelligent correction control system. Among them, the dual-wheel milling top correction mechanical device, through the Y-direction top correction mechanism set on the top of the milling cutter holder, enables the milling cutter holder to move horizontally in the Y direction and generate an eccentric gravitational torque, thereby completing the non-contact correction of the milling cutter holder in the Y direction (that is, the Y-direction correction of the milling cutter holder does not rely on the support reaction force provided by the groove wall, which can avoid the phenomenon of "wall eating" and hole enlargement of the groove wall during the correction process, breaking through the dependence of the traditional correction device with announcement number CN202767136U on the support reaction force of the groove wall, and perfectly adapting to the construction scenario of low-rigidity strata such as soft soil layers); at the same time, the dual milling wheel drive unit set at the bottom of the milling cutter holder can realize the attitude correction in the X direction by adjusting the speed difference and torque difference of the dual milling wheels. The intelligent correction control system employs a four-layer hierarchical architecture, consisting of a perception layer module, an analysis layer module, a decision-making layer module, and an execution layer module that are sequentially interconnected. This provides a stable hardware platform and system operating environment for the entire process of this invention. This invention can be directly embedded into the intelligent correction control system without requiring modifications to the existing hardware structure. It achieves deep compatibility with the non-contact correction principle of the top correction mechanism, thereby fully leveraging the correction performance advantages of the dual-wheel milling top correction mechanism.

[0048] Specifically, before the twin-wheel trenching machine officially starts trenching operations, this invention first completes the system initialization work, including calibrating the algorithm control cycle, presetting multi-level verticality warning thresholds and correction safety thresholds, loading the stratum adaptation parameter library, and verifying the communication between the full-link sensors and actuators. After all self-check items pass, the system enters real-time operation. It is worth noting that by completing the full-dimensional parameter presetting and verification before trenching operations, a standardized benchmark system can be established for the systematic comparison of real-time data and preset parameters and dynamic early warning during subsequent construction. It also provides complete and reliable parameter support for triggering the advanced pre-correction action of the milling cutter frame.

[0049] Once the twin-wheel grooving machine enters the grooving construction stage, within each calibrated control cycle, this invention synchronously collects four core parameters—real-time attitude of the milling cutter head, milling operation status, geological features, and construction environment—through the sensing layer module. The collected multi-source raw data is then spatiotemporally aligned, noise-reduced, filtered, and standardized, ultimately generating a standardized dataset for subsequent analysis. On one hand, multi-dimensional data acquisition comprehensively captures the dynamic changes in equipment operation status and the external construction environment, accurately pinpointing the underlying causes of milling cutter head attitude deviation. This effectively avoids the problem of existing technologies relying solely on single attitude data for decision-making, where correction strategies are out of sync with actual working conditions. On the other hand, the standardized preprocessing eliminates acquisition delays, coordinate deviations, and measurement noise between different sensors, providing an accurate and reliable data foundation and support for subsequent deviation prediction, parameter solving, and correction decisions.

[0050] Furthermore, based on the standardized dataset generated by the perception layer module, the analysis layer module calculates the real-time verticality deviation values ​​of the milling cutter head in the X and Y directions. Simultaneously, combining the current milling feed rate and formation stiffness parameters, it predicts the development trend of verticality deviation in the next control cycle. Subsequently, the real-time deviation values ​​and predicted deviation values ​​are compared with preset multi-level warning thresholds. If any value reaches the corresponding warning threshold, a corresponding pre-correction control command is immediately triggered, thereby enabling the milling cutter head correction control node to be moved forward and achieving early pre-correction of the milling cutter head. Compared with existing technologies, this invention can effectively intervene and adjust the attitude of the milling cutter head in the early stages of deviation accumulation. Especially in ultra-deep trench construction, it can not only effectively avoid the exponential amplification of shallow, minor deviations of the milling cutter head with increasing milling depth, but also prevent the situation where the optimal adjustment opportunity has been missed when the milling cutter head shows significant skewness. This significantly reduces the difficulty of correction operations in ultra-deep trenches and ensures that the trench verticality is always within the allowable range of the design specifications, significantly improving the control accuracy of trench verticality.

[0051] Understandably, when the decision-making module receives the pre-correction control command output by the analysis-layer module, this invention employs completely independent dual control links in the X and Y directions to solve for the pre-correction control parameters in the corresponding directions. Specifically, in the Y direction, the target displacement of the top correction mechanism is solved based on the milling cutter head's center of gravity offset and eccentric gravity moment model, while in the X direction, the target speed difference and torque difference between the two wheels are solved based on the differential cutting correction torque model of the two milling wheels. Unlike the simple parallel control logic in existing technologies, this invention, through completely independent computing channels and control logic, can effectively avoid the situation where the milling cutter head's correction actions in the X and Y directions are prone to mutual coupling interference and attitude disorder. At the same time, the dual independent link architecture of this invention combines flexibility and adaptability. When the milling cutter head only has a deviation in a single direction, only the control link in the corresponding direction needs to be activated to complete the correction. When the milling cutter head has a bidirectional synchronous deviation, the two links can run synchronously to achieve bidirectional joint correction, which can flexibly adapt to different deviation states of the milling cutter head.

[0052] Furthermore, after the decision-making module completes the preliminary solution of the pre-correction control parameters, the present invention does not directly issue execution instructions. Instead, it first compares the obtained X-axis and Y-axis control parameters with the pre-set correction safety threshold before construction, and at the same time eliminates parameter solutions that exceed the safety boundary. Then, it adaptively optimizes and adjusts the pre-correction control parameters based on the real-time stratum stiffness parameters, and finally outputs the optimal correction control parameters. Among these features, the safety threshold comparison process ensures that all correction actions on the milling cutter frame remain within the safe operating limits of the equipment, effectively preventing safety risks caused by excessive correction actions. Furthermore, the adaptive optimization adjustment of the pre-correction control parameters based on real-time stratum stiffness parameters allows core parameters such as correction force and response speed to be dynamically adjusted according to different stratum conditions, construction depth, and deviation change rates. This achieves an optimal balance between correction effect and construction stability, thus avoiding both the problems of trench wall enlargement and instability caused by excessive correction force in soft strata and the correction failure caused by insufficient correction force in hard rock strata. This further improves the control accuracy of trench verticality and the safety of the construction process.

[0053] Ultimately, this invention transforms the optimal correction control parameters output by the decision layer module into action commands for the actuator through the execution layer module, thereby synchronously driving the Y-axis top correction mechanism and the X-axis dual milling wheel drive unit to perform corresponding correction actions. In the next control cycle, the actual attitude data after the correction is executed is collected by the perception layer module, compared with the preset verticality control target, and the correction deviation is calculated and fed back to the deviation prediction and early warning link corresponding to step S3, thus completing the closed-loop dynamic correction of the entire process. Understandably, this fully closed-loop feedback correction mechanism breaks through the limitations of existing serial open-loop control technologies (for example, the working angle correction control system of a twin-wheel grooving machine disclosed in patent CN113250261B, which can realize real-time correction of the pitch and roll angle deviation of the grooving device relative to the horizontal plane, adopts serial control logic of attitude detection, deviation judgment and correction execution, so that it can only perform the correction action when the attitude of the grooving device deviates or after it has deviated, and cannot predict the deviance trend of the grooving device in advance). This allows the algorithm to dynamically adjust the control strategy and parameters according to the actual correction effect, ensuring that the correction action of the milling cutter head is always matched with the verticality control target height of the groove. At the same time, the fully automated control logic of this invention greatly reduces manual intervention in the construction process, not only reducing the risk of human error, but also significantly improving the overall operation efficiency of twin-wheel grooving construction.

[0054] Example 2:

[0055] This embodiment, based on Embodiment 1, further refines the multi-level verticality warning thresholds preset in step S1. The multi-level verticality warning thresholds disclosed in this embodiment correspond one-to-one with the pre-correction control strategy, providing an accurate execution benchmark for graded intervention and proactive pre-correction of milling cutter head attitude deviations. Specifically, in step S1, the multi-level verticality warning thresholds include a first-level warning threshold, a second-level warning threshold, and a third-level action threshold, with each level of warning threshold corresponding to a different pre-correction control strategy.

[0056] It should be noted that in water conservancy and hydropower strata seepage prevention walls and ultra-deep underground continuous wall projects, current industry design specifications typically require that the verticality deviation limit of the trench be no greater than 3‰. For ultra-deep trench sections exceeding 60m in depth, some key projects will further tighten the verticality deviation limit to 2‰. In this embodiment, all warning thresholds are set based on the verticality deviation limit of the trench specified in the engineering design documents, ensuring that the threshold settings are completely matched with the actual requirements of the project.

[0057] In this embodiment, the first-level warning threshold is set at 30% of the verticality deviation limit allowed by the design specifications. This threshold serves as the trigger point for preventive intervention, corresponding to the pre-deviation prevention and pre-correction strategy for the center of gravity. The reason for selecting 30% as the first warning point is that the milling cutter head attitude deviation within this range is still at an extremely low level and will not have any substantial impact on the grooving quality. Simultaneously, this point can accurately capture the initial development trend of the deviation, initiating intervention before the deviation accumulates, fundamentally preventing the deviation from amplifying with increasing milling depth, and preventing the lag problem of only being able to passively respond after the deviation occurs. When this threshold is triggered, the milling cutter head attitude deviation is still within the safety redundancy range of the design specifications. The pre-deviation prevention and pre-correction strategy initiated at this time only requires a slight adjustment to the top center of gravity position of the milling cutter head, without changing the milling feed speed or milling wheel operating parameters. It will not interfere with normal grooving milling operations, achieving proactive control of grooving verticality without affecting construction efficiency.

[0058] Furthermore, the secondary warning threshold is set at 60% of the verticality deviation limit allowed by the design specifications. This threshold serves as the trigger point for routine correction adjustments, corresponding to a progressive routine pre-correction strategy. The 60% threshold provides ample operational space for correction adjustments, preventing deviations from rapidly exceeding design limits, while also accurately identifying attitude deviations that have already established a clear trend, thus avoiding frequent correction actions that could impact construction stability. When the real-time or predicted deviation value of the milling cutter head reaches this threshold, it indicates that the attitude deviation of the milling cutter head has shown a clear cumulative trend. If not intervened in time, it will quickly approach the design limit. At this point, the progressive routine pre-correction strategy employs a multi-cycle, small-step adjustment method. Through progressive correction actions over multiple control cycles, the attitude of the milling cutter head is smoothly corrected. This not only quickly curbs further deviation development but also avoids problems such as milling cutter head attitude oscillation and slot wall disturbance caused by large correction actions within a single cycle.

[0059] Furthermore, the threshold for the third-level action is set at 90% of the verticality deviation limit allowed by the design specifications. This threshold serves as the trigger point for emergency corrective intervention, corresponding to an emergency lock-up corrective strategy. Setting this threshold at 90% of the design limit allows for mandatory intervention before the deviation exceeds the design specifications, eliminating the quality risk of excessive verticality in the trench from the source. It also avoids the problem of insufficient corrective response window caused by setting the threshold too close to the design limit, thus reserving sufficient execution time for emergency corrective actions. When the real-time or predicted deviation value of the milling cutter head reaches the threshold, it indicates that the milling cutter head posture is close to the critical value of the design specification, and there is an emergency risk of exceeding the verticality standard. At this time, the emergency locking and correction strategy will first lock the safety boundary of the milling cutter head center of gravity offset, synchronously link the milling feed system of the dual-wheel milling machine and reduce the feed speed, and then quickly correct the milling cutter head posture through the joint correction action of the X and Y dual control links. After the deviation falls back to below the secondary warning threshold, normal milling operation will be gradually restored. This can quickly prevent the deviation from further amplification and ensure the safety of equipment and construction in emergency situations.

[0060] Example 3:

[0061] This embodiment, based on Embodiment 2, further elaborates on the specific logic of verticality deviation prediction and early warning triggering in step S3, as well as the execution method of the pre-correction control strategy corresponding to different early warning thresholds. Specifically, in step S3, the specific logic of verticality deviation prediction and early warning triggering is as follows: when the real-time deviation value or the predicted deviation value reaches the first-level early warning threshold, the center of gravity pre-deviation preventive pre-correction strategy is triggered; when the real-time deviation value or the predicted deviation value reaches the second-level early warning threshold, the progressive conventional pre-correction strategy is triggered; and when the real-time deviation value or the predicted deviation value reaches the third-level action threshold, the emergency locking correction strategy is triggered.

[0062] It should be noted that, in the specific execution process, the analysis layer module, based on the standardized dataset output by the perception layer module in Example 1, first calculates the real-time verticality deviation values ​​of the milling cutter head in the X and Y directions. Then, combined with the current milling feed speed and real-time stratum stiffness parameters, it uses a first-order linear prediction model to predict the development trend of the verticality deviation in the next control cycle. Finally, the real-time deviation value and the predicted deviation value are compared one by one with the preset three-level warning thresholds. In this embodiment, as long as either the real-time deviation value or the predicted deviation value reaches the warning threshold of the corresponding level, the corresponding level of pre-correction control command is immediately triggered. Since the predicted deviation value can reflect the future tilt development trend of the milling cutter head posture, the "trigger upon reaching any threshold" logic can ensure the timeliness of the milling cutter head correction action (for example, even if the current real-time deviation is still at a low level, as long as it is predicted that the subsequent deviation will exceed the safe range, intervention will be initiated in advance).

[0063] Specifically, when the real-time or predicted deviation value reaches the first-level warning threshold, a pre-correction strategy for pre-biasing the center of gravity is triggered. The core of this strategy is to achieve reverse offsetting at the nascent stage of deviation accumulation, representing a non-intrusive, proactive intervention. During execution, the analysis layer module first accurately identifies the development direction and rate of change of the deviation, calculates the reverse pre-eccentricity gravity torque required to offset this trend, and then sends a micro-adjustment command to the decision layer module. Upon receiving the command, the decision layer module drives the hydraulic cylinder of the top correction mechanism to make a micro-displacement adjustment in the corresponding direction only through an independent Y-axis control link, causing the milling cutter head's center of gravity to shift in the opposite direction to the deviation's development trend, while maintaining the synchronous and simultaneous normal operation of the two milling wheels in the X-axis direction. Throughout the pre-correction process, the milling feed speed and milling wheel operating parameters remain in normal operating conditions, without interfering with the continuity and efficiency of trenching construction. This strategy is particularly suitable for predictable working conditions such as stable changes in geological characteristics and the imminent crossing of soft-hard interfaces, preventing the generation and accumulation of attitude deviations at the source without affecting normal construction.

[0064] When the real-time deviation or predicted deviation reaches the secondary warning threshold, a progressive conventional pre-correction strategy is triggered. The core of this strategy is to achieve attitude correction in a stable and controllable manner when the deviation shows a clear cumulative trend, balancing the correction effect with construction stability. In specific execution, the analysis layer module first calculates the total correction target required to bring the deviation back below the primary warning threshold by combining the current real-time deviation, deviation development rate, and real-time stratum stiffness parameters. Then, the total target is evenly distributed across 3-5 consecutive control cycles, with only 20%-30% of the total correction amount executed in each control cycle, thus forming a multi-cycle progressive adjustment logic. For Y-axis correction, the decision layer module gradually adjusts the displacement of the Y-axis top correction mechanism within each control cycle, progressively increasing the reverse eccentric gravity moment. For X-axis correction, the speed difference and torque difference of the dual milling wheels are simultaneously and progressively adjusted to avoid large adjustments within a single cycle causing milling cutter head attitude oscillations and groove wall disturbances. In addition, during the execution of this strategy, the milling feed system of the twin-wheel milling machine simultaneously reduces the milling feed speed by 10%-20%, thereby providing stable construction conditions for the correction action. After the milling cutter head deviation is detected to fall below the first-level warning threshold, the normal milling feed speed is gradually restored, thus adapting to and meeting the attitude correction needs in conventional construction scenarios such as soft soil layers and alternating soft and hard strata.

[0065] When the real-time or predicted deviation value reaches the level 3 action threshold, an emergency locking and correction strategy is triggered. The core of this strategy is to quickly stop the amplification trend of deviation in emergency situations where the deviation is about to exceed the design specification limit, eliminate the quality risk of excessive verticality of the groove, and ensure equipment and construction safety. In the specific execution process, the analysis layer module will trigger the emergency locking logic as soon as possible and send an emergency correction command to the decision layer module and the system control unit simultaneously. First, it will link the milling feed system of the twin-wheel grooving machine to reduce the milling feed speed to less than 30% of the normal operating speed (and directly suspend the milling feed if necessary) to avoid the continuous milling operation from amplifying the milling cutter head posture deviation; at the same time, it will immediately lock the maximum displacement boundary of the Y-axis top correction mechanism to avoid the risk of equipment instability caused by the excessive shift of the milling cutter head center of gravity during the correction process; then, it will start the synchronous joint correction of the X-axis and Y-axis dual independent control links to output the maximum safe correction torque that matches the current deviation state at the fastest response speed, and quickly correct the milling cutter head posture. In addition, during emergency correction, the system will collect the attitude data of the milling cutter head in real time during each control cycle. Once the deviation is detected to fall below the secondary warning threshold, it will automatically switch to a gradual conventional pre-correction strategy to avoid subsequent attitude oscillations caused by emergency large-scale correction. This strategy can perfectly adapt to emergency conditions such as sudden changes in stratum conditions and sudden changes in milling resistance caused by obstacles such as boulders encountered during the milling process, so as to maximize the quality and safety of trenching construction.

[0066] Example 4:

[0067] This embodiment, based on Embodiment 1, further refines the calculation formulas, parameter meanings, and calculation execution logic for the real-time perpendicularity deviation values ​​of the milling cutter head in the X and Y directions in step S3. The calculation formulas disclosed in this embodiment are the core calculation basis for quantifying the milling cutter head's attitude deviation, predicting deviation trends, and comparing early warning thresholds. Their calculation accuracy directly determines the accuracy and rationality of the pre-correction action triggering timing. Specifically, in step S3, the calculation formula for the real-time perpendicularity deviation value of the milling cutter head in the X direction is as follows: The formula for calculating the real-time perpendicularity deviation of the milling cutter holder in the Y direction is as follows: .

[0068] In the formula: This represents the real-time perpendicularity deviation of the milling cutter holder in the X direction. The value is the real-time perpendicularity deviation of the milling cutter holder in the Y direction. Both are in meters (m), and the calculation structure can be directly compared with the groove perpendicularity deviation limit specified in the design specifications. This represents the real-time roll angle of the milling cutter holder in the X direction. The values ​​are real-time pitch angles in the Y direction of the milling cutter head, all in ‰. The data are obtained from the dual-axis tilt sensor and IMU inertial measurement unit installed on the milling cutter head body, and are synchronously collected and processed by the sensing layer module for noise reduction and filtering. The current milling depth is expressed in meters (m). The data comes from the depth encoder of the main winch system of the twin-wheel milling machine, which accurately reflects the real-time working depth of the milling cutter head within the slot. Furthermore, it should be noted that the inclusion of absolute value calculations in the formula is to standardize the skew amplitude of the milling cutter head in both directions, avoiding calculation errors due to different skew directions, and ensuring that the system can accurately capture the deviation regardless of the direction of skew.

[0069] Furthermore, during actual execution, the analysis layer module synchronously acquires the milling cutter head tilt angle and milling depth data for each calibrated control cycle, completing the synchronous calculation of real-time verticality deviation values. The calculation results will be directly used for deviation trend prediction and comparison with the three-level early warning threshold. This formula fully considers the characteristic of deviation amplification with depth during ultra-deep trench construction, and the milling depth... The larger the value, the higher the linear deviation value corresponding to the same angle of skew. This accurately reflects the industry pain point that small deviations in the shallow layer of ultra-deep trench sections amplify exponentially with increasing depth, providing a precise quantitative basis for proactive pre-correction in ultra-deep trench sections. Simultaneously, the formula's calculation logic is simple and computationally efficient, fully adapting to the embedded operating environment of dual-wheel milling vehicle-mounted industrial controllers without increasing the system's computational burden. It enables millisecond-level rapid calculations within each control cycle, ensuring timely response to early warning triggers and pre-correction actions.

[0070] Example 5:

[0071] This embodiment, based on Embodiment 1, further refines the solution model, parameter meanings, and engineering execution logic for the target displacement of the Y-axis top correction mechanism in step S4. The solution model disclosed in this embodiment is the core computational basis of the Y-axis independent control link, deeply compatible with the non-contact correction principle of the eccentric gravity torque of the dual-wheel milling top correction mechanism. It can accurately solve for the target displacement matching the current milling cutter head skew state, providing precise quantitative control basis for the Y-axis pre-correction action. Specifically, in step S4, the solution model for the target displacement of the Y-axis top correction mechanism is as follows: .

[0072] In the formula: The target displacement of the Y-axis top correction mechanism is the core output of the model, and its value directly corresponds to the target extension and retraction stroke of the hydraulic cylinder driven by the Y-axis top correction mechanism. The rotational inertia of the top correction mechanism is a fixed value of the equipment calibration relative to the center of the milling wheel, which is the overall value of the milling cutter holder and the Y-axis top correction mechanism. The pre-calibration is completed through equipment 3D model simulation and on-site pendulum test. This is the second derivative of the milling cutter head attitude angle, i.e., the real-time angular acceleration of the milling cutter head pitch angle in the Y direction. This is the first derivative of the milling cutter head attitude angle, i.e., the real-time angular velocity of the milling cutter head pitch angle in the Y direction. and All data originates from the IMU (Inertial Measurement Unit) of the perception layer module; The mud damping coefficient is used to characterize the damping effect of the circulating mud in the trench on the sway of the milling cutter head. Its value is directly related to the current milling depth, mud density, and milling cutter head dimensions. It can be dynamically updated based on the mud density and trench depth data collected in real time by the sensing layer module. The contact damping coefficient is used to characterize the damping effect on the change of the milling cutter head posture during the cutting contact between the milling wheel and the ground. Its value is directly related to the equivalent ground stiffness calculated in real time by the sensing layer module. That is, the greater the ground stiffness, the higher the corresponding contact damping coefficient. The equivalent formation stiffness coefficient can be directly obtained by using the equivalent formation stiffness calculated in real time by the sensing layer module. Equivalent stratum stiffness It can accurately reflect the constraint effect of the current milling formation on the changes in the milling head posture; The real-time value of the pitch angle in the Y direction of the milling cutter head is obtained from the standardized dataset generated and output by the sensing layer module. The milling resistance torque is acquired in real time by the torque sensor of the milling wheel drive system; The total weight of the milling cutter holder is a fixed parameter of the equipment. This is the lever arm length from the center of gravity of the milling cutter holder to the center of the milling wheel, and it is a pre-calibrated fixed value. Furthermore, the angle in the above formula... The values ​​are dimensionless per mille, and the corresponding angular velocities and angular accelerations are obtained by the IMU in radians and then converted by coordinate transformation. The stiffness coefficients in the model... It has been pre-calibrated according to the per mille system.

[0073] In actual construction, the decision-making module inputs the real-time parameters collected by the perception layer module and processed by the analysis layer module into the solution model of the target displacement of the Y-axis top correction mechanism within each control cycle. This enables millisecond-level rapid solution of the target displacement, and the output result is directly converted into control commands for the hydraulic cylinder of the Y-axis top correction mechanism. This model fully couples the core factors affecting the correction effect, such as the rigid body rotation characteristics of the milling cutter head, the damping effect of mud and stratum, and the interference of milling resistance. Compared with the existing fixed displacement correction method, it can accurately match the correction requirements under different construction depths and different stratum conditions, ensuring the accuracy of the Y-axis correction action from the computational level and avoiding problems such as overshoot or undershoot. At the same time, this model is fully compatible with the control logic of the present invention, which features hard decoupling of X and Y axes (i.e., the calculation and execution of X and Y axis correction control are completely independent, with no signal and logic coupling). The solution process is completely independent of the X-axis control link and will not cause coupling interference with the X-axis correction calculation, further ensuring the stability and independence of the Y-axis correction control.

[0074] Example 6:

[0075] This embodiment, based on Embodiment 1, further refines the solution model, parameter meanings, and engineering execution logic for the target speed difference and target torque difference of the dual milling wheels in step S4. The solution model disclosed in this embodiment is the core computational basis of the X-axis independent control link, fully compatible with the mechanical principles of X-axis differential cutting correction in dual-wheel milling. It can accurately solve the control parameters matching the X-axis skew state of the milling cutter head, providing a quantitative control basis for X-axis pre-correction actions, and is also fully compatible with the hard decoupling control architecture of the X and Y axes of this invention. Specifically, in step S4, the solution model for the target speed difference of the dual milling wheels is as follows: The solution model for the target torque difference of the two milling wheels is as follows: .

[0076] In the formula: The target speed difference between the two milling wheels is expressed in r / min. This value corresponds to the speed adjustment command of the two independent drive motors of the milling wheels, and the sign of the value corresponds to the acceleration / deceleration direction of the milling wheels. The moment of inertia of the milling cutter head in the X direction is a fixed value calibrated by three-dimensional simulation and on-site pendulum test before the equipment leaves the factory, which characterizes the magnitude of the inertia of the milling cutter head rotating about the vertical axis; The second derivative of the roll angle in the X direction of the milling cutter head is the real-time angular acceleration of the roll angle in the X direction of the milling cutter head. The data comes from the IMU inertial measurement unit that is paired with the sensing layer module. X-axis mud damping coefficient is used to characterize the damping effect of circulating mud in the trench on the X-axis attitude swing of the milling cutter head. Its value is directly related to the current milling depth, mud density, and milling cutter head dimensions. It can be dynamically updated based on the mud density and trench depth data collected in real time by the sensing layer module. It is the first derivative of the roll angle in the X direction of the milling cutter head, that is, the real-time angular velocity of the roll angle in the X direction of the milling cutter head. The data comes from the IMU inertial measurement unit that is matched with the sensing layer module. X-direction ground contact damping coefficient is used to characterize the damping effect on the X-direction attitude change of the milling cutter head during the milling wheel's contact with the ground during cutting. Its value is directly related to the equivalent ground stiffness calculated in real time by the sensing layer module. That is, the greater the ground stiffness, the higher the corresponding X-direction ground contact damping coefficient. The equivalent formation stiffness coefficient in the X direction can be directly calculated in real time using the equivalent formation stiffness from the sensing layer module. It can accurately reflect the constraint effect of the current milling formation on the X-axis attitude change of the milling head; The real-time value of the roll angle in the X direction of the milling cutter holder is obtained from the standardized dataset generated and output by the sensing layer module. The conversion factor between rotational speed and cutting resistance is calibrated through standard stratum milling tests before the equipment leaves the factory and is directly related to the diameter, number of teeth, and cutting angle of the dual milling wheels. Where is the nominal radius of the milling wheel, and is a fixed structural parameter of the equipment; The target torque difference between the two milling wheels is the core output of the torque difference solution model. The unit is N·m. The value directly corresponds to the pressure control command of the hydraulic drive system of the two milling wheels. It forms a matching linkage with the target speed difference to ensure the dynamic stability of the differential cutting process. The speed-torque matching coefficient is used to match the power characteristics of the dual-milling wheel hydraulic drive system. The pre-calibration is completed through the hydraulic system calibration test before the equipment leaves the factory, which is used to fix the equipment parameters.

[0077] In actual construction, the decision-making module inputs the real-time parameters collected by the perception layer module and processed by the analysis layer module into the solution model for the target speed difference and target torque difference of the dual milling wheels within each control cycle. This completes the millisecond-level synchronous solution of the two core control parameters. Simultaneously, the output results are directly converted into independent control commands for the dual milling wheel drive unit, and the two milling wheels execute their corresponding speed and torque settings. Through differentiated cutting, a correction torque opposite to the X-axis deviation direction is generated, achieving precise correction of the milling head's X-axis attitude. Understandably, the solution model for the target speed difference and target torque difference of the dual milling wheels fully couples the core factors affecting the X-axis correction effect, such as the rigid body rotation characteristics of the milling head in the X-axis, the damping effect of mud and formation, and the mechanical characteristics of milling cutting. This allows for precise matching of X-axis correction requirements under different construction depths and formation conditions, ensuring the accuracy of the X-axis correction action from a computational perspective and avoiding overshoot or undershoot problems. Furthermore, the solution model for the target speed difference and target torque difference of the dual milling wheels can be fully adapted to the hard decoupling control logic of the X and Y directions of this invention. The solution process is completely independent of the Y-direction control link and will not be coupled with the Y-direction correction calculation, further ensuring the stability, accuracy and independence of the X-direction correction control.

[0078] Example 7:

[0079] This embodiment, based on Embodiment 1 and incorporating the technical content disclosed in Embodiment 5 or Embodiment 6, further elaborates on the architecture design, operating rules, and engineering execution logic of the X-axis and Y-axis dual independent control links in step S4. Specifically, in step S4, the X-axis and Y-axis dual control links employ completely independent computation channels and control cycles; when the milling cutter head experiences only a single-direction deviation, only the corresponding direction's control link is activated to perform correction; when the milling cutter head experiences a bidirectional synchronous deviation, the two control links operate synchronously and perform bidirectional joint correction.

[0080] In this embodiment, the X-axis and Y-axis dual control links employ completely independent computing channels and control cycles. The independent computing channels utilize physically isolated dual-core independent computing units within the dual-wheel milling vehicle-mounted industrial controller. The computing processes of the two links do not interfere with each other, and data transmission is independent. This avoids signal crosstalk and logic coupling during parallel computing at the hardware level, providing a hardware foundation for bidirectional hard decoupling control. Simultaneously, the independent control cycles can be independently calibrated according to the dynamic response characteristics of the actuators in the two directions (for example, the control cycle of the Y-axis top correction mechanism can be set to 50ms to match the response characteristics of the hydraulic actuator; the control cycle of the X-axis dual milling wheel drive unit can be set to 20ms to match the response characteristics of the variable frequency hydraulic drive system). This allows for precise adaptation to the dynamic characteristics of the different actuators in the two directions, significantly improving the response accuracy and operational stability of the correction action.

[0081] Furthermore, during actual construction, the dual independent control links can flexibly adjust their operating modes according to the real-time tilt status of the milling cutter head. When the milling cutter head exhibits a single-direction attitude deviation, the system only activates the control link in that direction to perform correction, while the control link in the other direction remains in standby mode, only acquiring real-time data without outputting execution commands. For example, when the milling cutter head exhibits only a Y-axis pitch angle deviation, only the Y-axis control link is activated to solve for the target displacement and drive the top correction mechanism, while the X-axis maintains a normal milling state with both milling wheels operating synchronously at the same speed. This avoids interference from ineffective correction actions on normal milling operations and reduces the system's computational load and unnecessary wear on the actuators. In addition, when the milling cutter head exhibits synchronous deviations in both the X and Y directions, the two control links start operating synchronously and independently complete the pre-correction parameter solution and correction command output for their respective directions. The bidirectional correction actions throughout the entire process do not interfere with each other, achieving bidirectional joint correction of the milling cutter head. It should be noted that the two control links only share the data output from the perception layer module, while the entire process of parameter solving, safety verification, and command output is completely independent. This completely eliminates the coupling interference of bidirectional correction actions and the problem of milling head posture oscillation caused by the mutual influence of correction actions.

[0082] Example 8:

[0083] This embodiment, based on Embodiment 1 and incorporating the technical content disclosed in Embodiment 5 or Embodiment 6, further elaborates on the adaptive optimization adjustment method, core calculation logic, and execution rules of the pre-correction control parameters in step S5. Specifically, in step S5, the specific method for adaptively optimizing the pre-correction control parameters is as follows: based on the real-time acquired milling torque and feed rate, the equivalent formation stiffness of the current formation is calculated. ;when When the value exceeds the preset hard ground threshold, the weight of the correction effect is increased, the upper limit of the amplitude of the pre-correction control parameters is increased, and the correction response cycle is shortened; when When the value is less than the preset threshold for weak strata, the weight of the correction stability is increased, the change amplitude of the pre-correction control parameters within a single control cycle is reduced, and the correction transition cycle is extended.

[0084] It should be noted that the core basis for the adaptive optimization of the pre-correction control parameters is the equivalent formation stiffness of the current formation. This parameter is calculated by the decision-making module based on the real-time data collected by the perception layer module from the dual-milling wheel milling torque and milling feed rate. This parameter accurately quantifies the hardness and softness characteristics and cutting resistance of the current milling stratum. The calculation results are updated synchronously in each control cycle, thus providing real-time and accurate stratum characteristic data for parameter optimization. Furthermore, the hard stratum threshold and the soft stratum threshold are both derived from the stratum adaptation parameter library pre-loaded in step S1. Construction personnel can calibrate these parameters based on the engineering geological survey report and on-site trial milling results to ensure complete adaptation to the actual stratum conditions of the construction area.

[0085] When the equivalent formation stiffness is calculated in real time When the deviation exceeds the preset hard stratum threshold, the system determines that the current construction scenario is in a high-strength stratum such as hard rock or dense gravel. In such strata, milling resistance is high, and once a deviation in the milling cutter head posture occurs, it is difficult to correct, easily leading to insufficient correction force. At this time, the system automatically increases the weight of the correction effect. On the one hand, it increases the upper limit of the amplitude of the pre-correction control parameters, raising the maximum single-cycle displacement of the Y-axis top correction mechanism and the maximum speed difference of the X-axis dual milling wheels by 20%-30%, matching the larger correction torque required for hard strata correction. On the other hand, it shortens the correction response cycle, compressing the calculation and execution cycle of the corresponding control link by 30%-40%, thereby improving the response speed of the correction action, quickly curbing the development of deviation, and effectively avoiding the problem of excessive verticality caused by correction lag in hard strata.

[0086] And when the equivalent formation stiffness is calculated in real time When the ground level is below the preset threshold for weak strata, the system determines that the current construction scenario is in a low-stiffness weak strata such as silty soil or fine sand. In such strata, the trench wall stability is poor, and large-scale correction movements can easily lead to trench wall enlargement, instability, and collapse. At this point, the system automatically increases the weight of correction stability. On the one hand, it reduces the variation amplitude of the pre-correction control parameters within a single control cycle, limiting the adjustment amount of the correction parameters in a single cycle to within 15% of the total target amount, and adopts a multi-cycle gradual adjustment method. On the other hand, it extends the correction transition cycle to avoid abrupt changes in the correction parameters. This minimizes the disturbance and compression of the trench wall caused by the correction action, ensuring the stability of the trench wall structure while smoothly correcting the milling cutter holder posture, thus mitigating the construction safety risks caused by excessive correction in weak strata.

[0087] Understandably, through the adaptive optimization mechanism of this embodiment, the deviation correction control strategy can be dynamically adjusted according to real-time geological conditions, achieving the optimal balance between deviation correction effect and construction stability under all geological conditions, and further improving the accuracy of trench verticality control and the safety during construction.

[0088] Example 9:

[0089] This embodiment, based on embodiment 8, further elaborates on the composition, setting rules, and over-limit handling logic of the correction safety threshold in step S5. Specifically, in step S5, the correction safety threshold includes the hard limit threshold for the milling cutter head center of gravity offset, the maximum stroke threshold for the Y-axis top correction mechanism, the maximum speed difference threshold for the dual milling wheels, and the upper limit threshold for the load of the suspension mechanism; when any pre-correction control parameter exceeds the corresponding safety threshold, the parameter is automatically corrected to within the safety boundary, and a safety warning is triggered simultaneously.

[0090] It should be noted that the correction safety threshold is a hard constraint boundary that the pre-correction control parameters cannot break. Specifically, it includes four categories: the hard limit threshold for the offset of the milling cutter head center of gravity, the maximum stroke threshold of the Y-axis top correction mechanism, the maximum speed difference threshold of the dual milling wheels, and the upper limit threshold of the load of the suspension mechanism. All thresholds are preset and calibrated in step S1, and are fully matched with the structural parameters of the dual-wheel milling machine and the on-site construction safety specifications.

[0091] Among them, the hard limit threshold for the center of gravity offset of the milling cutter head is based on the overall structural dimensions of the milling cutter head and the pre-calibration of the stability boundary of the suspension system. It is used to constrain the center of gravity offset of the milling cutter head caused by the Y-axis top correction action, so as to avoid the problem of milling cutter head instability and uneven suspension load caused by excessive center of gravity offset; the maximum stroke threshold of the Y-axis top correction mechanism is matched with the physical maximum extension stroke setting of the hydraulic cylinder of the top correction mechanism, which is the absolute boundary of the target displacement in the Y-axis, so as to avoid structural damage caused by the over-stroke operation of the hydraulic cylinder; the maximum speed difference threshold of the dual milling wheels is based on the rated speed of the dual milling wheel drive system and the calibration of the safe operating range of the hydraulic system, which is a hard constraint on the target speed difference in the X-axis, so as to avoid the problem of milling wheel drive motor overload and milling instability caused by excessive speed difference; the upper limit threshold of the suspension mechanism load is matched with the rated load setting of the main winch suspension system, so as to constrain the suspension load fluctuation caused by the correction action, and avoid the safety risks of the suspension system caused by excessive load.

[0092] Understandably, during actual construction, after the decision-making module completes the solution of the pre-correction control parameters, it compares all parameters with their corresponding safety thresholds one by one. If any pre-correction control parameter exceeds the corresponding safety threshold, the system automatically corrects the parameter to the maximum allowable value within the safety boundary. Simultaneously, it triggers a safety warning to the control system of the entire twin-wheel milling machine, displays the over-limit item and warning information on the on-board operating terminal, and reminds on-site operators to pay attention to the operating status of the equipment and the attitude changes of the milling cutter head. Obviously, this embodiment defines an absolute safety boundary for the correction actions in the X and Y directions by calibrating the correction safety threshold. At the same time, it can form a complete cooperation with the adaptive optimization mechanism of Embodiment 8. That is, no matter how the correction parameters are dynamically adjusted, they will not exceed the safe operating range of the equipment. This ensures the effectiveness of the correction action and avoids the equipment and construction safety risks caused by the correction operation, further improving the stability of the algorithm operation and the construction safety.

[0093] Example 10:

[0094] This embodiment, based on Embodiment 1, further elaborates on the specific logic, execution flow, and implementation method of the closed-loop feedback correction in step S6. Specifically, the specific logic of the closed-loop feedback correction in step S6 is as follows: Collect the actual posture data of the milling cutter head after the correction is executed, and calculate the difference between the actual verticality deviation correction amount and the target correction amount; if the difference is within the preset allowable range, maintain the current correction control parameters until the deviation falls back to the safe range; if the difference exceeds the preset allowable range, input the difference as a compensation amount into the pre-correction parameter solution process of the next control cycle, dynamically correct the control parameters until the actual deviation value meets the verticality control requirements.

[0095] It should be noted that the core calculation benchmark for closed-loop feedback correction is the difference between the actual verticality deviation correction and the target correction. The target correction is the theoretical deviation correction value corresponding to the pre-correction control parameters in the previous control cycle, which is calculated and generated synchronously by the decision-making module during parameter solving. The actual verticality deviation correction is calculated by the analysis-layer module based on the actual attitude data of the milling cutter head after correction execution collected by the perception-layer module. The calculation method is completely consistent with the calculation formula for the verticality deviation value in Example 4, ensuring the consistency of the data benchmark. Furthermore, the preset allowable range for difference determination is calibrated in step S1 and is typically set to 5% of the verticality deviation limit allowed by engineering design specifications, serving as a pass / fail benchmark for the correction execution effect.

[0096] Understandably, during actual construction, after the analysis layer module completes the difference calculation within each control cycle, it can execute differentiated control logic based on the magnitude of the difference. Specifically, when the difference is within the preset allowable range, it indicates that the correction action of the previous cycle achieved the expected effect. At this time, the system maintains the current correction control parameters and continues to execute them without additional adjustment until the milling cutter head verticality deviation falls back to a safe range within the first-level warning threshold. This avoids milling cutter head attitude oscillations caused by frequent adjustments to control parameters, ensuring the stability of the correction process. However, when the difference exceeds the preset allowable range, it indicates that the correction action of the previous cycle did not achieve the expected result. In this case, there is insufficient correction or overshoot. The system automatically uses this difference as a feedforward compensation amount, inputting it into the X-axis and Y-axis pre-correction parameter solution process of the next control cycle, and dynamically compensates and corrects the control parameters. This compensation and correction logic is executed cyclically within each control cycle until the actual verticality deviation of the milling cutter head meets the verticality control requirements of the engineering design, thereby effectively correcting the correction execution error caused by factors such as sudden changes in strata and fluctuations in milling resistance. It is evident that the closed-loop feedback correction mechanism of this embodiment can work in conjunction with the advanced pre-correction, dual independent control links, and formation adaptive optimization mechanism of the present invention to build a complete control system for the entire process of milling frame construction. It not only realizes the full closed-loop control from deviation prediction, parameter solution, safety verification to execution feedback, but also further improves the control accuracy of trench verticality and the working condition adaptability of the algorithm.

[0097] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. 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.

Claims

1. An automatic deviation correction control algorithm for dual-wheel milling in the X and Y directions, implemented based on a dual-wheel milling top deviation correction mechanical device and an intelligent deviation correction control system, wherein the dual-wheel milling top deviation correction mechanical device includes a milling cutter holder, a Y-axis top deviation correction mechanism disposed at the top of the milling cutter holder, and a dual-milling wheel drive unit disposed at the bottom of the milling cutter holder; the intelligent deviation correction control system includes a perception layer module, an analysis layer module, a decision layer module, and an execution layer module that are sequentially and communicatively connected, characterized in that... Includes the following steps: S1: Completes algorithm control cycle calibration, presets multi-level verticality warning thresholds and correction safety thresholds, loads the formation adaptation parameter library, verifies the communication link between sensors and actuators, and enters real-time operation state after verification. S2: The real-time attitude data of the milling cutter head, milling operation parameters, formation characteristic parameters and construction environment parameters within the current control cycle are synchronously collected through the perception layer module. The collected data is then spatiotemporally aligned, noise-reducing filtered and standardized to generate a standardized dataset. S3: Based on a standardized dataset, calculate the real-time verticality deviation values ​​of the milling cutter head in the X and Y directions. Combine the milling feed speed and the formation stiffness parameters to predict the development trend of verticality deviation in the next control cycle. Compare the real-time deviation value and the predicted deviation value with the preset multi-level verticality warning thresholds respectively. When any value reaches the warning threshold of the corresponding level, trigger the corresponding pre-correction control command. S4: Based on the triggered pre-correction control command, a completely independent dual control link is adopted in the X and Y directions to solve the pre-correction control parameters in the corresponding directions respectively; among them, the Y direction is based on the milling cutter head center of gravity offset and eccentric gravity moment model to solve the target displacement of the Y-direction top correction mechanism; the X direction is based on the dual milling wheel differential cutting correction torque model to solve the target speed difference and torque difference of the dual milling wheels. S5: Compare the obtained X-axis and Y-axis pre-correction control parameters with the preset correction safety threshold, and eliminate parameter solutions that exceed the safety boundary; at the same time, based on the real-time formation stiffness parameters, adaptively optimize and adjust the pre-correction control parameters, and output the final optimal correction control parameters. S6: The optimal correction control parameters are converted into action commands for the corresponding actuators through the execution layer module, and the Y-axis top correction mechanism and the X-axis dual milling wheel drive unit are driven to perform the corresponding correction actions synchronously. In the next control cycle, the actual attitude data after the correction is executed is collected through the perception layer module, compared with the preset verticality control target, the correction deviation is calculated and fed back to step S3, and the full closed-loop dynamic correction is completed.

2. The automatic correction control algorithm for dual-wheel milling in the X and Y directions according to claim 1, characterized in that, In step S1, the multi-level verticality warning threshold includes a first-level warning threshold, a second-level warning threshold, and a third-level action threshold. Each level of warning threshold corresponds to a different pre-correction control strategy.

3. The automatic correction control algorithm for dual-wheel milling in the X and Y directions according to claim 2, characterized in that, In step S3, the specific logic for verticality deviation prediction and early warning triggering is as follows: when the real-time deviation value or the predicted deviation value reaches the first-level early warning threshold, the center of gravity pre-deviation preventive pre-correction strategy is triggered; when the real-time deviation value or the predicted deviation value reaches the second-level early warning threshold, the gradual conventional pre-correction strategy is triggered; when the real-time deviation value or the predicted deviation value reaches the third-level action threshold, the emergency locking correction strategy is triggered.

4. The automatic correction control algorithm for dual-wheel milling in the X and Y directions according to claim 1, characterized in that, In step S3, the formula for calculating the real-time perpendicularity deviation value of the milling cutter holder in the X direction is as follows: The formula for calculating the real-time perpendicularity deviation of the milling cutter holder in the Y direction is as follows: .

5. The automatic correction control algorithm for dual-wheel milling in the X and Y directions according to claim 1, characterized in that, In step S4, the solution model for the target displacement of the Y-axis top correction mechanism is as follows: .

6. The automatic correction control algorithm for dual-wheel milling in the X and Y directions according to claim 1, characterized in that, In step S4, the solution model for the target speed difference between the two milling wheels is as follows: The solution model for the target torque difference of the two milling wheels is as follows: .

7. The automatic correction control algorithm for dual-wheel milling in the X and Y directions according to claim 5 or 6, characterized in that, In step S4, the dual control links in the X and Y directions use completely independent computing channels and control cycles; when the milling cutter head has only a single-direction deviation, only the control link in the corresponding direction is activated to perform correction; when the milling cutter head has a bidirectional synchronous deviation, the two control links run synchronously and perform bidirectional joint correction.

8. The automatic correction control algorithm for dual-wheel milling in the X and Y directions according to claim 5 or 6, characterized in that, In step S5, the specific method for adaptively optimizing and adjusting the pre-correction control parameters is as follows: based on the real-time collected milling torque and feed rate, the equivalent formation stiffness of the current formation is calculated. ;when When the value exceeds the preset hard ground threshold, the weight of the correction effect is increased, the upper limit of the amplitude of the pre-correction control parameters is increased, and the correction response cycle is shortened; when When the value is less than the preset threshold for weak strata, the weight of the correction stability is increased, the change amplitude of the pre-correction control parameters within a single control cycle is reduced, and the correction transition cycle is extended.

9. The automatic correction control algorithm for dual-wheel milling in the X and Y directions according to claim 8, characterized in that, In step S5, the correction safety thresholds include the milling cutter holder center of gravity offset hard limit threshold, the maximum stroke threshold of the Y-axis top correction mechanism, the maximum speed difference threshold of the dual milling wheels, and the upper limit threshold of the load of the suspension mechanism; when any pre-correction control parameter exceeds the corresponding safety threshold, the parameter is automatically corrected to within the safety boundary, and a safety warning is triggered simultaneously.

10. The automatic correction control algorithm for dual-wheel milling in the X and Y directions according to claim 1, characterized in that, In step S6, the specific logic of closed-loop feedback correction is as follows: collect the actual posture data of the milling cutter holder after the correction is executed, and calculate the difference between the actual verticality deviation correction amount and the target correction amount. If the difference is within the preset allowable range, maintain the current correction control parameters until the deviation falls back to the safe range; If the difference exceeds the preset allowable range, the difference is used as a compensation amount and input into the pre-correction parameter solution process of the next control cycle to dynamically correct the control parameters until the actual deviation value meets the verticality control requirements.

Images