A data-driven hydraulic outrigger dynamic leveling method and working equipment

Abstract

The application discloses a hydraulic outrigger dynamic leveling method and operation equipment based on data driving. The method comprises the following steps: collecting attitude parameters of the operation equipment under all working conditions, and establishing a dynamic reference model; intercepting an outrigger pressure sequence, comparing a real-time pressure curve with a normalized standard pressure curve by using a sliding window correlation coefficient or a dynamic time warping algorithm, and combining an inclination angle trend to determine a virtual leg phenomenon; matching historical safe reference attitude parameters after the determination, calculating target center coordinates and determining an error range, deducing influence weights of each outrigger linkage on the center of gravity based on a multi-body space kinematics model of the operation equipment, screening an optimal action combination through a cost function, and controlling support cylinder actions to realize leveling. The application overcomes the defect that a static threshold judgment is easy to misjudge under a dynamic load, and can prevent vehicle body shaking and mechanism interference caused by single-leg leveling.

Description

Technical Field

[0001] This invention relates to the field of automated control technology for construction machinery, and in particular to a data-driven dynamic leveling method and equipment for hydraulic outriggers. Background Technology

[0002] Multi-point support work equipment (such as wheeled walking vehicles) is widely used in complex scenarios such as power construction, emergency rescue, and mountain construction. Its chassis features a multi-point support mechanism composed of multiple sets of telescopic outriggers, swing legs, and claws, while the superstructure integrates various heavy-duty work modules. In actual construction, this type of equipment often operates in unstructured terrains such as mountainous areas, slopes, and loose deposits, where the bearing stiffness of the foundation exhibits spatial non-uniformity and time-varying uncertainty. The foundation media contacted by the ends of each outrigger vary significantly in lithology, density, and moisture content, leading to inconsistent elastoplastic settlement at each support point under load, thus disrupting the mechanical balance of the support system.

[0003] During continuous construction operations, as the superstructure platform performs actions such as rotation, luffing, lifting, or drilling, the center of gravity of the superstructure dynamically shifts in three-dimensional space. The resulting overturning moment and inertial load are directly transmitted to the chassis support system, causing a drastic redistribution of the load-bearing capacity of each outrigger. The multi-point support system is a statically indeterminate structure; the distribution of reaction forces in each outrigger depends not only on the magnitude and direction of the external load but also on the relative stiffness of each support point and its compatibility with foundation deformation. When the local stiffness of the foundation decreases or the center of gravity of the superstructure shifts, causing a continuous decrease in the vertical reaction force of a particular outrigger, the contact normal stress between the outrigger and the ground decreases accordingly. When this normal stress approaches zero, the outrigger partially or completely detaches from the foundation, resulting in a "virtual outrigger" phenomenon.

[0004] The occurrence of the "empty leg" phenomenon reduces the number of effective support points, increases the load on the remaining outriggers, and may trigger a chain of overload failures. More seriously, the constraint state of the working equipment degenerates from a statically indeterminate support to a statically determinate or even kinematic system with uncertain boundaries. The anti-overturning moment margin is weakened, and the equipment is prone to rollover or overturning accidents under the disturbance of subsequent working actions.

[0005] To address the aforementioned safety hazards, most existing leveling and detection technologies employ a control mode that combines static threshold judgment with independent leveling of a single leg. This mode disrupts the original overall mechanical balance when a single leg extends or retracts, causing interference between the legs and easily leading to repeated vehicle body swaying and a seesaw effect. This not only results in lengthy leveling times but can also lead to leveling failure under extreme tilt angle conditions. Summary of the Invention

[0006] The purpose of this invention is to provide a data-driven dynamic leveling method and equipment for hydraulic outriggers. This method uses data-driven pressure curve comparison to identify outrigger slack, and based on the multi-body spatial kinematics model and cost function of the equipment, determines the optimal leveling action combination through multi-leg coordination to solve the problems of vehicle body swaying and leveling failure caused by single-leg independent leveling mode.

[0007] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0008] On the one hand, the present invention provides a data-driven dynamic leveling method for hydraulic outriggers, comprising the following steps: S1: Collect the posture parameters of each leg joint of the working equipment under full working conditions, and establish a dynamic benchmark model including the optimal solution target array and the normalized standard pressure curve. S2: Real-time capture of the pressure sequence of the target outrigger, normalize the pressure sequence and draw a real-time pressure curve. Use a sliding window correlation coefficient or dynamic time warping algorithm to compare the real-time pressure curve with the normalized standard pressure curve for consistency. Based on the comparison results and combined with the trend of the tilt angle of the working equipment, determine that the corresponding outrigger has a false leg phenomenon when the preset judgment conditions are met. S3: After determining that a false leg phenomenon has occurred, the current posture parameters are matched with the historical safety reference posture parameters in the dynamic reference model. The most similar historical safety reference posture parameters are used as the reference posture parameters for leveling. With the reference posture parameters as a reference, the target center coordinates for this leveling are calculated according to the current position and geometric constraints of the working equipment, and the allowable posture error range is determined. Based on the built-in multi-body spatial kinematics model of the working equipment, the influence weight of different outrigger linkage schemes on the center of gravity of the working equipment is deduced. The optimal action combination with the least disturbance to the center of gravity of the working equipment is selected through a cost function. The optimal action combination includes the displacement compensation of each support cylinder. S4: Output the basic valve opening signal of the feedforward loop according to the displacement compensation amount of each support cylinder, and output the dynamic compensation signal of the feedback loop according to the pose deviation and pose deviation change rate between the current working equipment posture and the target center coordinates. After superimposing the basic valve opening signal and the dynamic compensation signal, independently adjust the opening degree of the electro-hydraulic proportional valve of the corresponding outrigger to control the movement of the support cylinder and realize the continuous dynamic adaptive leveling of the working equipment.

[0009] Furthermore, the establishment of the dynamic benchmark model, which includes an optimal solution objective array and a normalized standard pressure curve, specifically includes:

[0010] After manually leveling the equipment to a horizontal position, control the equipment to perform at least three full-condition operation actions, and simultaneously record the outrigger pressure, cylinder length, extension margin, and equipment tilt angle; perform time alignment and deviation standardization on the recorded outrigger pressure to generate a normalized standard pressure curve that reflects the normal load transfer law; extract the multi-dimensional parameter matrix that meets the horizontal position requirements, and construct the optimal solution target array.

[0011] Furthermore, the specific steps of the consistency morphology comparison are as follows: when using the sliding window correlation coefficient, calculate the correlation coefficient between the real-time pressure curve and the normalized standard pressure curve;

[0012] When using the dynamic time warping algorithm for consistency comparison, a two-dimensional distance matrix is ​​constructed between the real-time pressure curve and the normalized standard pressure curve. The optimal matching path is then searched in the two-dimensional distance matrix using dynamic programming recursion to calculate the morphological distance between the two sets of curves. The morphological distance is the cumulative distance value obtained by accumulating the distances between each matching point along the optimal matching path.

[0013] Furthermore, the method for determining the "false leg" phenomenon is as follows: a preset lower limit threshold for correlation coefficient and a threshold for exceeding the morphological distance limit are used; when the correlation coefficient falls below the lower limit threshold or the morphological distance exceeds the threshold for exceeding the morphological distance limit, and the real-time pressure curve exhibits at least one of abnormal burr characteristics, step characteristics, or flat-top characteristics, combined with the trend of the working equipment tilting towards the corresponding outrigger, it is determined that the corresponding outrigger has a "false leg" phenomenon.

[0014] Furthermore, the process of constructing the multibody spatial kinematic model of the operating equipment includes:

[0015] A fuselage coordinate system is established with the large rotation center of the working equipment as the origin, and the fixed position vector of each outrigger hinge point is determined in the fuselage coordinate system. A nonlinear mapping function between the extension and retraction length of each support cylinder and the position of the working equipment is established by the closed-loop vector chain method. In the nonlinear mapping function, the outrigger vector is determined by the product of the working equipment center position vector, the rotation matrix and the fixed position vector of the outrigger hinge point, and the outrigger grounding position vector.

[0016] Furthermore, the specific steps for calculating the influence weights of different outrigger linkage schemes on the center of gravity of the working equipment are as follows:

[0017] The partial derivative of the nonlinear mapping function is obtained to construct the Jacobian matrix to establish the mapping relationship between the small changes in the pose of the working equipment and the displacement increment of the support cylinder. By extracting each element in the Jacobian matrix, the sensitivity of the displacement of a single support cylinder to the center of gravity shift of the working equipment is obtained. The row vector of the Jacobian matrix composed of each sensitivity is the influence weight.

[0018] Furthermore, the cost function is: ; in, The cost function value, Let be the horizontal offset norm of the center of gravity spatial coordinates of the working equipment. This represents the displacement compensation amount for the k-th support cylinder participating in the leveling process. This is the weighting coefficient for center of gravity stability. To balance the stroke weighting coefficient, Let be the stroke penalty factor for the k-th supporting cylinder participating in the balancing; The total number of support cylinders involved in the leveling process; Under the condition of satisfying the physical stroke constraint of the support cylinder, the displacement increment control sequence that minimizes the cost function value is solved as the optimal action combination.

[0019] On the other hand, the present invention provides a working device for performing the aforementioned data-driven dynamic leveling method for hydraulic outriggers, comprising: The frame and multiple support mechanisms disposed around the frame, each support mechanism including a lifting cylinder, a swing leg and a support lower leg driven by the support cylinder; The electro-hydraulic direct control circuit uses an electro-hydraulic proportional valve that is directly connected to the support cylinder. The multi-source state sensing module includes a pressure sensor connected in parallel to the electro-hydraulic direct control circuit, a length sensor for detecting the stroke of the support cylinder, and an inclination sensor set at the large rotation center of the frame. The signal output terminals of the pressure sensor, the length sensor, and the inclination sensor are respectively connected to the signal input terminal of the main controller. The main controller is connected to the control signal input terminal of each of the electro-hydraulic proportional valves. The main controller is used to perform false leg determination and leveling actions based on the detection signals of each sensor.

[0020] Furthermore, the hydraulic circuit of the support mechanism is configured as a real-time sensing architecture without physical locking, and the inlet and outlet ports of each electro-hydraulic proportional valve are directly connected to the rodless chamber and rod chamber of the support cylinder; during non-leveling operations, the electro-hydraulic proportional valve remains in the neutral position; the pressure sensor is directly connected in parallel to the directly connected oil circuit to capture in real time the pressure fluctuation signal in the cylinder chamber caused by micro-settlement of the foundation or load transfer.

[0021] Furthermore, the drive signal output by the main controller is a composite control command, which is formed by superimposing the basic valve opening signal output by the feedforward loop and the dynamic compensation signal output by the feedback loop. The feedforward loop receives the support cylinder displacement pre-calculated by the main controller based on the multi-body spatial kinematic model of the working equipment, and outputs the basic valve opening signal. The feedback loop receives the pose deviation signal and pose deviation rate of change signal of the working equipment, incorporates a fuzzy PID controller, and outputs the dynamic compensation signal.

[0022] Compared with the prior art, the present invention has the following beneficial effects: This invention abandons the traditional single static pressure threshold judgment method and adopts a data-driven approach to establish a dynamic benchmark model. It compares the consistency of the real-time pressure curve with the normalized standard pressure curve through the sliding window correlation coefficient or dynamic time warping algorithm, and combines the trend of the tilt angle change of the working equipment to comprehensively judge the phenomenon of false leg. This overcomes the defect of traditional static pressure threshold judgment that is prone to misjudgment under dynamic load conditions, and eliminates the measurement interference caused by the dynamic transfer of load under multi-degree-of-freedom actions such as rotation, luffing, and lifting of multi-point support equipment.

[0023] This invention is based on the multibody spatial kinematics model of the work equipment and the Jacobian matrix to deduce the influence weight of different outrigger linkage schemes on the center of gravity. By using a cost function to select the optimal action combination that minimizes the disturbance to the center of gravity under the condition of satisfying the physical stroke constraint of the hydraulic cylinder, it overcomes the defect of the traditional single-leg sequential leveling mode in multi-point support statically indeterminate system that is prone to destroying the overall mechanical balance. It can prevent repeated shaking of the vehicle body, mechanism interference and leveling failure caused by blind extension and retraction of a single leg.

[0024] This invention employs an electro-hydraulic direct control circuit that directly connects an electro-hydraulic proportional valve to a support cylinder. During non-leveling periods, the proportional valve remains in the neutral position, and the pressure sensor is directly connected in parallel to the oil circuit to capture pressure fluctuations in the bearing chamber in real time. This overcomes the defect of pressure signal isolation under the traditional hydraulic locking structure and provides a delay-free pressure signal acquisition channel for data-driven real-time virtual leg detection.

[0025] This invention employs a control strategy that combines feedforward and feedback. The feedforward loop outputs a basic valve opening signal based on the displacement pre-calculated by the kinematic model to achieve rapid approximation. The feedback loop uses a fuzzy PID algorithm to fine-tune the pose deviation and rate of change in real time. The two are superimposed to output a composite control command, which overcomes the shortcomings of single feedback control in response lag or single feedforward control in anti-interference capability, and takes into account both the speed of leveling action and control stability. Attached Figure Description

[0026] Figure 1 The structural diagram of the work equipment provided by the present invention.

[0027] Figure 2 A schematic diagram of the hydraulic system of the right front outrigger of the work equipment provided by the present invention.

[0028] Figure 3 The present invention provides a flowchart of a data-driven dynamic leveling method for hydraulic outriggers.

[0029] Figure 4 The flowchart of the data-driven virtual leg detection method provided by the present invention is shown.

[0030] In the diagram: 101-Frame, 102-Lifting cylinder, 103-Swinging leg, 104-Tire, 105-Supporting lower leg, 106-Supporting cylinder, 107-Outrigger, 201-Pressure sensor, 202-Supporting up-and-down swinging cylinder, 203-Tire left-and-right swinging cylinder, 204-Tire up-and-down swinging cylinder, 205-Electro-hydraulic proportional valve, 206-Relief valve, 207-Hydraulic pump, 208-Stop valve, 209-Suction filter, 210-Return filter. Detailed Implementation

[0031] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0032] Example 1 like Figure 1 As shown, this work equipment is a wheel-mounted engineering chassis with high terrain adaptability. It mainly consists of a frame 101 as the load-bearing core, multiple support units symmetrically hinged to the periphery of the frame, and sensing and control modules integrated into each joint. Specifically, each support unit is driven by a lifting cylinder 102 to swing the leg 103 relative to the frame 101 to achieve the ground contact driving or ground lifting of the tire 104. During leveling operations, the support cylinder 106 drives the support leg 105 and its end foot 107 to extend and retract linearly along the axial direction, so that the foot 107 abuts against the unstructured foundation surface and forms a stable multi-point support state.

[0033] With the frame 101 as the core load-bearing structure, multiple independent operating support units are symmetrically distributed around its periphery. The root of each support unit is connected to the frame 101 via a hinge shaft, and its posture transformation is accomplished by a two-stage hydraulic drive mechanism. The first stage is the posture transformation stage, in which the lifting cylinder 102 extends and retracts to drive the swing leg 103 to rotate around the frame 101, realizing the ground contact or suspension of the tire 104; the second stage is the leveling support stage, with a support leg 105 hinged to the end of the swing leg 103, and a support cylinder 106 is installed inside the support leg 105. When entering the leveling operation mode, the support cylinder 106 directly drives its end foot 107 to extend towards the ground until it contacts the undulating unstructured foundation surface, thereby forming a stable multi-point physical support network for the operating equipment.

[0034] The hydraulic drive circuit of the support cylinder 106 adopts an electro-hydraulic direct control architecture. Specifically, the main controller is directly connected to the electro-hydraulic proportional valve configured in the oil circuit via a control bus, and the working inlet and outlet ports of the electro-hydraulic proportional valve form a direct physical connection with the rodless chamber and the rod chamber of the support cylinder 106. During the leveling operation, the electro-hydraulic proportional valve continuously changes its opening degree according to the analog signal issued by the main controller, driving the support cylinder 106 to flexibly extend and retract; while during the non-leveling operation, the electro-hydraulic proportional valve remains in the neutral position, so that the pressure of the bearing chamber of the entire support cylinder 106 can fluctuate in real time and without hysteresis with small changes in the external working load.

[0035] The hydraulic system of the right front outrigger, such as Figure 2 As shown, the hydraulic oil pump 207 draws hydraulic oil from the oil tank through the suction filter screen 209, and outputs high-pressure hydraulic oil to the high-pressure main oil circuit P through the shut-off valve 208. The high-pressure main oil circuit P is connected in parallel with the relief valve 206, which is used to regulate the system pressure and play an overload protection role.

[0036] The main controller is directly connected to the electro-hydraulic proportional valves 205 configured in each oil circuit via a control bus. Specifically, the working inlet and outlet ports of the electro-hydraulic proportional valve 205 are respectively connected to the rodless chamber and rod chamber of the corresponding oil cylinder (including the support cylinder 202 for vertical swing, the tire cylinder 203 for horizontal swing, and the tire cylinder 204 for vertical swing). The inlet end of each electro-hydraulic proportional valve 205 is connected to the high-pressure main oil circuit, and the return end of each electro-hydraulic proportional valve 205 is connected to the return oil circuit. The return oil circuit returns to the oil tank after passing through the return oil filter 210. The support cylinder 202 (i.e., the support cylinder 202 for vertical swing) Figure 1Taking the support cylinder 106 circuit as an example, during the leveling operation, the corresponding electro-hydraulic proportional valve 205 continuously changes its opening degree according to the analog signal sent by the main controller, driving the support swing cylinder 202 to flexibly extend and retract. During the non-leveling operation phase, the electro-hydraulic proportional valve 205 remains in the neutral position, allowing the pressure in the bearing chamber of the support swing cylinder 202 to fluctuate in real time without hysteresis, adapting to small changes in the external operating load. At this time, the pressure sensor 201, directly connected to the direct control oil circuit of the bearing chamber, can capture these pressure transients. Similarly, the tire left-right swing cylinder 203 and the tire up-down swing cylinder 204 are controlled by their respective electro-hydraulic proportional valves 205, and their control principle is the same as that of the support swing cylinder circuit described above.

[0037] To achieve data-driven dynamic adaptive leveling, the work equipment integrates a multi-source state sensing module, with the following spatial layout strategy: In the electro-hydraulic direct control circuit (i.e., the circuit where the electro-hydraulic proportional valve is directly connected to the support cylinder), a pressure sensor 201 is directly connected in parallel to pick up the transient changes in bearing pressure in the support cylinder 106 caused by foundation settlement or center of gravity shift, so as to form a pressure-time and pressure-slewing angle data stream; a length sensor is configured at each support cylinder 106 to provide real-time feedback on the absolute extension stroke and remaining available margin of the current outrigger; dual-axis tilt sensors are installed at the large slewing center of the frame 101 and on each support leg 105 to capture and calculate the roll angle, pitch angle, and actual spatial tilt posture of each outrigger in three-dimensional space in real time.

[0038] The main controller, serving as the computation and control center of the work equipment, forms a real-time control closed loop with the aforementioned multi-source state sensing module and electro-hydraulic proportional valves. During construction operations, the main controller frequently calls upon the data collected by the multi-source state sensing module. On one hand, it performs false leg detection, using a dynamic time warping algorithm or a sliding window correlation coefficient algorithm to compare pressure curves in real time to determine the false leg phenomenon. On the other hand, after determining that a false leg phenomenon has occurred and triggering leveling, the main controller uses the embedded multi-body spatial kinematic model and Jacobian matrix of the work equipment to deduce the coupling weight between the center of gravity offset of the work equipment and the stroke of each support cylinder 106 in real time. Through a feedback loop that integrates feedforward and fuzzy PID, it independently and smoothly adjusts the opening degree of each electro-hydraulic proportional valve, driving multiple outriggers to move synchronously and collaboratively without interference, achieving anti-tipping adaptive leveling in unstructured terrain.

[0039] Example 2 This embodiment provides a data-driven dynamic leveling method for hydraulic outriggers, applied to the work equipment described in Embodiment 1, such as... Figure 3 As shown, it includes the following steps: S1. Collect attitude parameters and establish a dynamic benchmark model: Collect the attitude parameters of each leg joint of the working equipment under full working conditions, and establish a dynamic benchmark model that includes the optimal solution target array and the normalized standard pressure curve.

[0040] S2. Determine if a false leg phenomenon occurs: Capture the pressure sequence of the target outrigger in real time, normalize the pressure sequence and draw a real-time pressure curve. Use a sliding window correlation coefficient or dynamic time warping algorithm to compare the real-time pressure curve with the normalized standard pressure curve for consistency. Based on the comparison results and combined with the trend of the tilt angle of the working equipment, determine that the corresponding outrigger has a false leg phenomenon when the preset judgment conditions are met.

[0041] S3. Determine the reference attitude parameters and solve for the optimal action combination: After determining that a false leg phenomenon has occurred, match the current attitude parameters with the historical safe reference attitude parameters in the dynamic reference model, and take the most similar historical safe reference attitude parameters as the reference attitude parameters for leveling. With the reference attitude parameters as a reference, calculate the target center coordinates for this leveling based on the current position and geometric constraints of the working equipment, and determine the allowable attitude error range. Based on the built-in multi-body spatial kinematic model of the working equipment, deduce the influence weight of different outrigger linkage schemes on the center of gravity of the working equipment, and select the optimal action combination that minimizes the disturbance to the center of gravity of the working equipment through a cost function. The optimal action combination includes the displacement compensation of each support cylinder.

[0042] S4. Performing actions to achieve leveling: Based on the displacement compensation of each support cylinder, output the basic valve opening signal of the feedforward loop, and based on the pose deviation and pose deviation change rate between the current working equipment posture and the target center coordinates, output the dynamic compensation signal of the feedback loop. After superimposing the basic valve opening signal and the dynamic compensation signal, independently adjust the opening of the electro-hydraulic proportional valve of the corresponding outrigger, control the movement of the support cylinder, and realize the continuous dynamic adaptive leveling of the working equipment.

[0043] Step S1 in this embodiment includes the following sub-steps: S1-1: Acquiring Attitude Parameters. Before the work equipment enters the formal construction site, the operator first controls the movement of each outrigger and joint via the control handle to complete the initial manual leveling. After manually confirming that the work equipment has reached a horizontal posture and is stably supported, the work equipment is controlled to perform at least three full-condition operation actions (full-condition refers to the working conditions corresponding to typical actions such as rotation, luffing, and lifting performed by the work equipment in actual operation). During this process, the pressure of each outrigger, cylinder length, extension margin, and attitude parameters such as the center coordinates and tilt angle of the work equipment are recorded simultaneously.

[0044] S1-2: The main controller extracts and calculates the attitude parameters collected under multiple working conditions. The outrigger pressure extracted from the attitude parameters undergoes time alignment and normalization: the main controller identifies characteristic trigger points (such as the instant of slewing start or pressure mutation points) in multiple full-condition actions as time zero points, and aligns pressure sequences of different durations to a unified time length and rotation angle reference using an interpolation resampling algorithm; a deviation normalization method is used to map the collected raw pressure values ​​to the [0, 1] interval to eliminate absolute numerical interference caused by initial load differences or sensor range fluctuations, thereby ensuring the comparability of different stress states. The above processing generates a smooth normalized standard pressure curve, which is then fixed and saved. The normalized standard pressure curve refers to a dynamic pressure template curve generated by statistically averaging multiple sets of normal operating pressure data after alignment and normalization. This curve characterizes the mechanical characteristics of the outriggers under conditions of no false leg defects and a solid foundation, as they change with operating actions, including the normal load transfer peak and trough shapes.

[0045] S1-3: Extract the multi-dimensional parameter matrix that meets the horizontal attitude requirements and construct the optimal solution target array. The optimal solution target array refers to a multi-dimensional parameter matrix stored in the main controller, which includes the ideal outrigger pressure distribution, target cylinder stroke value, and spatial coordinates of the work equipment center for different typical working postures (such as specific combinations of amplitude angle and rotation direction) under manual leveling conditions. The optimal solution target array serves as the target reference for the adaptive leveling control system, providing a precise logical endpoint offset reference for each outrigger electro-hydraulic proportional valve after automatic leveling is initiated.

[0046] like Figure 4 As shown, the specific process of step S2 in this embodiment is as follows: S2-1: Real-time pressure sequences are extracted using a time sliding window, and morphological differences are calculated using correlation coefficients or dynamic time warping algorithms.

[0047] Once the equipment enters normal operating mode, the pressure sensors directly connected in parallel to the electro-hydraulic direct control circuit continuously collect the dynamic pressure of each outrigger at a high frequency, forming a continuous pressure-time and pressure-rotation angle data stream. To capture transient anomalies, the main controller uses a preset-length time sliding window to dynamically capture the current pressure curve segment in real time.

[0048] The implementation mechanism of the time sliding window is as follows: the main controller allocates a fixed-length first-in-first-out (FIFO) circular data buffer in the underlying memory. As construction work progresses, the main controller pushes the latest acquired pressure data points into the head of the buffer at a preset high-frequency sampling rate (e.g., 100Hz), and simultaneously discards the oldest data points at the tail. The duration of this time sliding window is typically set to cover a typical transient force change cycle (e.g., 2 to 3 seconds), thereby ensuring that a dynamic data sequence that best reflects the current local force trend can be continuously captured and maintained with extremely low memory and computing power overhead.

[0049] The main controller calls the internal dynamic time warping algorithm or sliding window correlation coefficient algorithm to compare the real-time pressure curve corresponding to the real-time captured pressure curve segment with the normalized standard pressure curve in the dynamic benchmark model to determine the consistency of the pressure curve. The morphological distance or correlation coefficient between the two is calculated in real time, which is used as a quantitative indicator to assess the current stress health status of the outrigger.

[0050] When using the sliding window correlation coefficient, the main controller extracts the real-time pressure curve within the current time sliding window and performs Pearson correlation calculation with the standard sequence corresponding to the normalized standard pressure curve of the same time span in the dynamic benchmark model. This involves calculating the ratio of the product of the covariance and the standard deviation of the two sequences, as shown in the formula: ; Where r is the correlation coefficient. Let i be the i-th real-time pressure data point in the real-time pressure curve. This is the average value of all real-time pressure data in the real-time pressure curve. For the i-th pressure data point in the normalized standard pressure curve, is the average value of all pressure data in the normalized standard pressure curve; n is the total number of real-time pressure data points in the real-time pressure curve.

[0051] The calculated correlation coefficient ranges from -1 to 1. The closer the value is to 1, the more consistent the current force increase / decrease trend of the outrigger is with the normalized standard pressure curve; if the value is less than the preset lower limit threshold of the correlation coefficient, it indicates that the force trend has deviated.

[0052] When using a dynamic time warping algorithm for consistency comparison, to address the stretching or compression of the action time axis caused by inconsistent driver lever operation speeds, a two-dimensional distance matrix between the real-time pressure curve and the normalized standard pressure curve is constructed using dynamic programming. The core dynamic programming recursive formula is: ; in, The i-th data point in the real-time pressure curve The j-th data point in the normalized standard pressure curve The Euclidean distance between the two sets of curves is calculated; D(i,j) is the minimum cumulative morphological distance when the i-th data point in the real-time pressure curve aligns with the j-th data point in the normalized standard pressure curve. Under the premise of satisfying boundary conditions and monotonic continuity constraints, a regular path is searched in the two-dimensional distance matrix to optimally match the characteristic parts of the two sets of curves (such as aligning unloading troughs with troughs). The Euclidean distance between all matching points is accumulated along this optimal path to obtain the morphological distance. This morphological distance is a non-negative scalar, which eliminates the misalignment interference caused by the speed of movement and quantifies the difference in the shapes of the two sets of curves. The smaller the distance value, the healthier the outrigger's stress; an abnormally large distance value directly exposes a physical unloading mutation in the outrigger.

[0053] S2-2: The main controller compares the calculated morphological distance or correlation coefficient with the preset safety threshold in real time. The safety threshold includes the lower limit threshold of the correlation coefficient and the threshold for exceeding the morphological distance limit. If the calculated morphological distance is less than the threshold for exceeding the morphological distance limit and the correlation coefficient is higher than the lower limit threshold of the correlation coefficient safety threshold, it indicates that the force fluctuation of the current outrigger is consistent with the standard operating mechanical state, and no intervention is made, and the data collection is continued in a loop. Conversely, if the detected morphological distance is greater than the threshold for exceeding the morphological distance limit, or the correlation coefficient falls below the lower limit threshold of the correlation coefficient safety threshold, and the real-time pressure curve deviates significantly from the normalized standard pressure curve, exhibiting abnormal burr characteristics, step characteristics, or flat-top characteristics, it is determined that the corresponding outrigger is in a dangerous trend of abnormal unloading or false leg.

[0054] In practical implementation, to balance different construction geological conditions and the versatility of multi-tonnage equipment, the threshold for exceeding the morphological distance limit and the lower limit of the correlation coefficient safety limit are not fixed absolute constants, but rather adaptive empirical ranges generated based on the self-weight of the operating equipment, the accuracy level of the sensors, and the initial calibration data. Preferably, the lower limit of the correlation coefficient safety limit is set between 0.80 and 0.85. When the correlation coefficient remains above this range, it indicates that the current force trend of the outrigger is highly positively correlated with the mechanical morphology represented by the normalized standard pressure curve, and is considered safe. The threshold for exceeding the morphological distance limit is set as 115% to 120% of the standard morphological distance (the benchmark value obtained by averaging the morphological distances between multiple sets of real-time pressure curves under normal operating conditions and the normalized standard pressure curve using a dynamic time warping algorithm) as the maximum safety tolerance zone. When the morphological distance measured in real time exceeds this upper limit, it is determined that the error has exceeded the allowable range of mechanical vibration and noise during normal operation.

[0055] Furthermore, when the real-time pressure curve deviates significantly from the specific geometric shape of the normalized standard pressure curve, it corresponds to different physical unloading hazards of multi-point support equipment in complex terrain, specifically including: (1) Abnormal burr characteristics: refers to non-periodic, high-frequency amplitude abrupt changes that occur within a time window. Its physical essence is characterized by: when the unstructured foundation experiences local micro-fractures and settlement, or when transient jamming and slippage occurs inside the support cylinder, the effective supporting force of the outrigger experiences instantaneous high-frequency fluctuations.

[0056] (2) Step-like characteristics: refers to the instantaneous, irreversible cliff-like drop in amplitude of the real-time pressure curve in the continuous time domain, and the pressure remains at a low level after the drop. Its physical essence is: the foundation under the outrigger suddenly collapses on a large scale, or the center of gravity of the working equipment shifts sharply due to the rotation and lifting, causing the outrigger to be completely lifted off the ground in an instant, and the load it bears is instantly stripped away.

[0057] (3) Flat-top characteristic: This refers to a dynamic pressure curve that should show continuous peaks or troughs under standard operating conditions (such as the rotation of the operating equipment), but whose derivative approaches zero within a certain period of time, showing a flat straight line that does not change with time. Its physical essence is that the outrigger is in a completely suspended state, its rodless chamber pressure has dropped to the basic back pressure of the hydraulic system and remains unchanged, and the outrigger has completely lost its ability to respond to and support the dynamic load of the operating equipment.

[0058] In practice, after initially determining that a certain outrigger has a tendency to become a "false leg," to avoid malfunctions caused by normal vibration of the equipment or accidental errors of a single sensor, real-time data from the tilt sensor located at the center of the large slewing area is retrieved for comprehensive verification. If the trend of the equipment's tilt angle changes precisely indicates that the equipment is tilting or settling in the direction of the abnormal outrigger, then the outrigger is determined to have a false leg phenomenon. When a false leg phenomenon is determined, the main controller will not only immediately trigger an audible and visual false leg alarm to alert the operator, but will also automatically send a leveling request with the false leg's orientation information on the internal control bus, directly guiding the equipment into the adaptive dynamic leveling execution phase.

[0059] The specific process of step S3 in this embodiment is as follows: S3-1: Match historical safety baseline attitude parameters from the dynamic baseline model and calculate the target center coordinates.

[0060] After receiving the leveling request and the virtual leg position information, the main controller immediately freezes the current dangerous action and reads the real-time attitude parameters of the current working equipment, including the tilt angle of the working equipment, the extension length of each outrigger cylinder (i.e., cylinder length), the allowable extension margin, the center coordinates of the working equipment, and the current load pressure.

[0061] Subsequently, the main controller retrieves the previously established dynamic benchmark model from memory and uses a feature matching algorithm to retrieve the historical safety benchmark attitude parameters that are closest to the current operating condition from the optimal solution target array. Using these historical safety benchmark attitude parameters as a reference, the target center coordinates (X, Y, Z axes) and the allowable attitude error range for this leveling are calculated and locked.

[0062] S3-2: Based on the multibody spatial kinematics model of the work equipment and the Jacobian matrix, the influence weight of each outrigger linkage on the center of gravity is derived and optimized through the cost function.

[0063] Based on the target center coordinates, current coordinates, and the remaining available stroke fed back by the hydraulic cylinder sensors of each outrigger, the main controller deduces the influence weight of different outrigger linkage schemes on the center of gravity of the working equipment, thereby selecting the multi-leg cooperative control strategy with the least disturbance to the working equipment and the shortest time consumption (such as diagonal outrigger differential compensation).

[0064] The multibody spatial kinematic model of the work equipment refers to a mathematical model describing the mapping relationship between the extension and retraction lengths of each support cylinder and the posture of the work equipment. Its construction mechanism is as follows: This model is a mathematical mapping relationship of the work equipment posture established based on spatial vectors or the Denavit-Hartenberg (DH) parameter method. In kinematic modeling, a global coordinate system is defined as a fixed reference coordinate system. The origin of the global coordinate system is located at the center of the great rotation of the work equipment. The X-axis is along the longitudinal direction of the work equipment, the Y-axis is along the lateral direction of the work equipment, and the Z-axis is along the vertical direction. These three coordinates form a right-hand rectangular coordinate system, used to describe the absolute position and posture of the work equipment and each outrigger in space. Based on this, local relative coordinate systems are established for each outrigger's hinge point with the frame and for each outrigger's contact point with the ground. Through the linkage geometric parameters and homogeneous transformation matrix, this model constructs a forward solution model between the extension and retraction lengths of each outrigger support cylinder (i.e., joint spatial variables) and the work equipment's center coordinates (X, Y, Z axis positions) and spatial tilt angles (roll and pitch angles), which can be expressed as: ; The left side of the equation represents the mission space pose matrix of the operating equipment. Let X be the X-axis coordinate of the center of the working equipment in the global coordinate system. Let Y be the Y-axis coordinate of the center of the working equipment in the global coordinate system. Let Z be the Z-axis coordinate of the center of the working equipment in the global coordinate system. The roll angle of the work equipment. The pitch angle of the work equipment. The yaw angle of the operating equipment; the variable on the right. q represents the actual extension / retraction length of each support cylinder, where q is the total number of support cylinders; It is a nonlinear mapping function established based on the rigid body geometry of the work equipment.

[0065] The nonlinear mapping function based on the rigid body geometry of the equipment is constructed using the closed-loop vector chain method. First, a fuselage coordinate system is established at the center of great rotation, and local coordinate systems are established at the hinge points of each outrigger and the ground contact points. For the g-th outrigger, its closed-loop vector equation in space is: ; in, Let g be the vector of the g-th leg; Let T be the position vector of the center of the work equipment in the global coordinate system, and T be the transpose. Let be the fixed position vector of the hinge point of the g-th outrigger in the fuselage coordinate system; Let R be the outrigger contact position vector of the g-th outrigger; R is the rotation matrix representing the attitude of the working equipment, derived from the roll angle. Pitch angle Yaw angle This is confirmed, and its specific details are as follows: ; in, , , These are the rotation matrices around the Z-axis, Y-axis, and X-axis of the global coordinate system, respectively.

[0066] Article g: Actual extension / retraction length of the support cylinder That is, the vector The modulus length, i.e. , Let be the transpose of the vector of the g-th leg.

[0067] By writing the above closed-loop vector equations for all outriggers (g=1,2,...,q), a system of nonlinear algebraic equations consisting of q equations is obtained. This system of equations constitutes the explicit expression of the nonlinear mapping function, which is used to describe the mapping relationship between the joint space (cylinder stroke) and the task space (work equipment pose).

[0068] Based on the aforementioned established nonlinear mapping function, the main controller calculates the partial derivatives of its spatial variables (i.e., the actual extension and retraction lengths of each support cylinder), thereby constructing a Jacobian matrix in real time that characterizes the current pose coupling relationship of the working equipment. This matrix establishes a small displacement increment mapping relationship between joint spatial variables and task spatial variables, and its increment mapping equation is expressed as: ; in, It is a matrix of minute changes in the pose of the work equipment (including three-dimensional position increments and attitude angle increments). This is the matrix of minute displacement increments for each supporting hydraulic cylinder involved in the linkage. The Jacobian matrix is ​​given by the real-time pose, and its internal elements are... , For the operation equipment, the a-th pose component The actual extension and retraction length of the g-th support cylinder The partial derivative of this internal element maps the sensitivity of the minute displacement of a single support cylinder to the spatial coordinate shift of the center of gravity of the working equipment.

[0069] From the perspective of differential geometry and physical kinematics, the Jacobian matrix is ​​a mapping matrix obtained by differentiating the closed-loop vector equation with respect to time. It maps the extension and retraction velocity of the outrigger cylinders to the spatial linear and angular velocities of the working equipment. For the g-th outrigger, the row vector of its corresponding Jacobian matrix can be explicitly expressed as: ; in, Let g be the row vector of the Jacobian matrix corresponding to the g-th leg. The transpose of the unit vector along the support direction of the g-th leg, and the unit vector along the support direction of the g-th leg. ; This refers to the cross product operation of spatial vectors. The complete Jacobian matrix J of the operating equipment is formed by the longitudinal combination of the row vectors of all the outriggers, i.e. .

[0070] Based on the aforementioned complete Jacobian matrix, the main controller can extract the influence weights of each outrigger on the platform's pose adjustment by calculating the transpose or pseudo-inverse of the complete Jacobian matrix J under the current working equipment posture in real time. In adaptive leveling control, based on these influence weights, it can predict which outrigger combinations will contribute the most to correcting the current tilt posture. Simultaneously, when encountering extremely uneven foundations, the main controller can predict and avoid dead points in the mechanism's motion during multi-leg coordinated operation by analyzing the singular values ​​of the complete Jacobian matrix in real time, thereby ensuring the stability and anti-tipping safety of the working equipment during dynamic leveling.

[0071] Once the strategy is selected, the main controller calculates the displacement of each target support cylinder participating in the leveling in real time, i.e., the control error deviation of each support cylinder, and obtains the deviation change rate by differentiating the control error deviation.

[0072] Based on this, a cost function is introduced to comprehensively evaluate different outrigger linkage schemes. This cost function comprehensively considers two core optimization indicators: minimizing the horizontal offset of the equipment's center of gravity and minimizing the overall leveling motion stroke. It is used to select the optimal solution from multiple possible combinations of leveling displacements. Its mathematical expression is: ; in, The cost function value, Let be the horizontal offset norm of the center of gravity spatial coordinates of the working equipment. This represents the displacement compensation amount for the k-th support cylinder participating in the leveling process. This is the weighting coefficient for center of gravity stability. To balance the stroke weighting coefficient, is the stroke penalty factor for the k-th support cylinder participating in the leveling; m is the total number of support cylinders participating in the leveling.

[0073] Under the condition of satisfying the physical stroke constraints of the hydraulic cylinders, the displacement increment control sequence that minimizes the cost function value is solved. During the actual solution calculation, the main controller has an adaptive stroke bottom-out protection mechanism: if the length sensor reports that the real-time stroke of a certain outrigger is approaching its physical limit, the main controller will nonlinearly increase its corresponding stroke penalty weight factor to increase the cost score for that outrigger to continue moving, thereby directly inducing the main controller to reduce or abandon the extension / retraction of that outrigger at the algorithm optimization level. Finally, using the gradient descent method or the Lagrange multiplier method, under the premise of strictly satisfying the physical stroke boundary constraints of each support cylinder, the main controller solves in real time for the displacement increment control sequence of each support cylinder that minimizes the aforementioned cost function (Cost value), and outputs the optimal coordinated action command combination to each electro-hydraulic proportional valve, which can achieve both horizontal leveling and maintain the anti-tipping stability of the equipment's center of gravity.

[0074] The specific process of step S4 in this embodiment is as follows: S4-1: Based on the displacement compensation amount in the optimal action combination, the basic valve opening signal is pre-calculated and output through the multi-body spatial kinematic model of the working equipment. The main controller first provides a leveling feedforward reference (i.e., the predicted basic valve opening signal) based on the displacement of the support cylinder pre-calculated by the multi-body spatial kinematic model of the working equipment, so as to drive the support cylinder to quickly approach the target position.

[0075] S4-2: Based on the pose deviation and the rate of change of the deviation, the parameters are tuned online by the fuzzy PID controller, and a dynamic compensation signal is output. The control error deviation and the rate of change of the deviation of each support cylinder calculated above are used as inputs and sent to the fuzzy PID controller. The fuzzy PID controller tunes the proportional, integral, and derivative control parameters online in real time according to the preset fuzzy inference rule matrix, and outputs a dynamic compensation signal.

[0076] S4-3: After superimposing the basic valve opening signal and the dynamic compensation signal, the opening degree of each electro-hydraulic proportional valve is independently adjusted. After integrating the feedforward reference and feedback PID calculation results, the main controller superimposes the basic valve opening signal and the dynamic compensation signal, and outputs a smooth and continuous analog voltage (or current) signal to the electro-hydraulic proportional valve of the corresponding outrigger, independently adjusting the opening degree of the electro-hydraulic proportional valve of the corresponding outrigger, driving the support cylinder to perform flexible coordinated extension and retraction actions in the absence of hydraulic lock.

[0077] S4-4: Continuously monitor the pose data until the target error range is met, then end the leveling process. During the process of the electro-hydraulic proportional valve driving the support cylinder, the main controller enters a high-frequency compensation feedback loop. The tilt sensor and length sensor continuously transmit the latest pose data, and the main controller continuously judges whether the current attitude parameters are within the target center coordinates of the leveling and fall within the allowable attitude error range. If the judgment result is negative, it indicates that the leveling is not yet in place or there is overshoot. The main controller substitutes the latest data into the above-mentioned fuzzy PID logic to continue the cyclic tuning and fine-tuning. If the judgment result is positive, it indicates that the working equipment has successfully returned to a safe horizontal pose. The main controller then cuts off the leveling control signal, each electro-hydraulic proportional valve returns to the neutral position to maintain system pressure, this adaptive dynamic leveling ends, the working equipment automatically unfreezes its actions, and switches back to normal operation. Figure 4 The virtual leg detection status is shown.

[0078] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A data-driven dynamic leveling method for hydraulic outriggers, characterized in that, Includes the following steps: S1: Collect the posture parameters of each leg joint of the working equipment under full working conditions, and establish a dynamic benchmark model including the optimal solution target array and the normalized standard pressure curve. S2: Real-time capture of the pressure sequence of the target outrigger, normalize the pressure sequence and draw a real-time pressure curve. Use a sliding window correlation coefficient or dynamic time warping algorithm to compare the real-time pressure curve with the normalized standard pressure curve for consistency. Based on the comparison results and combined with the trend of the tilt angle of the working equipment, determine that the corresponding outrigger has a false leg phenomenon when the preset judgment conditions are met. S3: After determining that a false leg phenomenon has occurred, the current posture parameters are matched with the historical safety reference posture parameters in the dynamic reference model. The most similar historical safety reference posture parameters are used as the reference posture parameters for leveling. With the reference posture parameters as a reference, the target center coordinates for this leveling are calculated according to the current position and geometric constraints of the working equipment, and the allowable posture error range is determined. Based on the built-in multi-body spatial kinematics model of the working equipment, the influence weight of different outrigger linkage schemes on the center of gravity of the working equipment is deduced. The optimal action combination with the least disturbance to the center of gravity of the working equipment is selected through a cost function. The optimal action combination includes the displacement compensation of each support cylinder. S4: Output the basic valve opening signal of the feedforward loop according to the displacement compensation amount of each support cylinder, and output the dynamic compensation signal of the feedback loop according to the pose deviation and pose deviation change rate between the current working equipment posture and the target center coordinates. After superimposing the basic valve opening signal and the dynamic compensation signal, independently adjust the opening degree of the electro-hydraulic proportional valve of the corresponding outrigger to control the movement of the support cylinder and realize the continuous dynamic adaptive leveling of the working equipment.

2. The data-driven dynamic leveling method for hydraulic outriggers according to claim 1, characterized in that, The establishment of the dynamic benchmark model, which includes an optimal solution target array and a normalized standard pressure curve, specifically includes: After manually leveling the equipment to a horizontal position, control the equipment to perform at least three full-condition operation actions, and simultaneously record the outrigger pressure, cylinder length, extension margin, and equipment tilt angle; perform time alignment and deviation standardization on the recorded outrigger pressure to generate a normalized standard pressure curve that reflects the normal load transfer law; extract the multi-dimensional parameter matrix that meets the horizontal position requirements, and construct the optimal solution target array.

3. The data-driven dynamic leveling method for hydraulic outriggers according to claim 1, characterized in that, The specific steps for the consistency pattern comparison are as follows: when using the sliding window correlation coefficient, calculate the correlation coefficient between the real-time pressure curve and the normalized standard pressure curve; When using the dynamic time warping algorithm for consistency comparison, a two-dimensional distance matrix is ​​constructed between the real-time pressure curve and the normalized standard pressure curve. The optimal matching path is then searched in the two-dimensional distance matrix using dynamic programming recursion to calculate the morphological distance between the two sets of curves. The morphological distance is the cumulative distance value obtained by accumulating the distances between each matching point along the optimal matching path.

4. The data-driven dynamic leveling method for hydraulic outriggers according to claim 3, characterized in that, The method for determining the "false leg" phenomenon is as follows: a preset lower limit threshold for the correlation coefficient and an excessive threshold for the morphological distance are used; when the correlation coefficient falls below the lower limit threshold or the morphological distance exceeds the excessive threshold, and the real-time pressure curve exhibits at least one of abnormal burr characteristics, step characteristics, or flat-top characteristics, combined with the trend of the working equipment tilting towards the corresponding outrigger, it is determined that the corresponding outrigger has a "false leg" phenomenon.

5. The data-driven dynamic leveling method for hydraulic outriggers according to claim 1, characterized in that, The process of constructing the multibody spatial kinematic model of the operating equipment includes: A fuselage coordinate system is established with the large rotation center of the working equipment as the origin, and the fixed position vector of each outrigger hinge point is determined in the fuselage coordinate system. A nonlinear mapping function between the extension and retraction length of each support cylinder and the position of the working equipment is established by the closed-loop vector chain method. In the nonlinear mapping function, the outrigger vector is determined by the product of the working equipment center position vector, the rotation matrix and the fixed position vector of the outrigger hinge point, and the outrigger grounding position vector.

6. The data-driven dynamic leveling method for hydraulic outriggers according to claim 5, characterized in that, The specific steps for deriving the influence weight of different outrigger linkage schemes on the center of gravity of the working equipment are as follows: The partial derivative of the nonlinear mapping function is obtained to construct the Jacobian matrix to establish the mapping relationship between the small changes in the pose of the working equipment and the displacement increment of the support cylinder. By extracting each element in the Jacobian matrix, the sensitivity of the displacement of a single support cylinder to the center of gravity shift of the working equipment is obtained. The row vector of the Jacobian matrix composed of each sensitivity is the influence weight.

7. The data-driven dynamic leveling method for hydraulic outriggers according to claim 1, characterized in that, The cost function is: ； in, The cost function value, Let be the horizontal offset norm of the center of gravity spatial coordinates of the working equipment. This represents the displacement compensation amount for the k-th support cylinder participating in the leveling process. This is the weighting coefficient for center of gravity stability. To balance the stroke weighting coefficient, Let be the stroke penalty factor for the k-th supporting cylinder participating in the balancing; The total number of support cylinders involved in the leveling process; Under the condition of satisfying the physical stroke constraint of the support cylinder, the displacement increment control sequence that minimizes the cost function value is solved as the optimal action combination.

8. A work equipment for performing the data-driven hydraulic outrigger dynamic leveling method according to any one of claims 1 to 7, comprising a frame and a plurality of support mechanisms disposed around the periphery of the frame, each support mechanism comprising a lifting cylinder, a swinging thigh, and a support lower leg driven by the support cylinder, characterized in that, Also includes: The electro-hydraulic direct control circuit uses an electro-hydraulic proportional valve that is directly connected to the support cylinder. The multi-source state sensing module includes a pressure sensor connected in parallel to the electro-hydraulic direct control circuit, a length sensor for detecting the stroke of the support cylinder, and an inclination sensor set at the large rotation center of the frame. The signal output terminals of the pressure sensor, the length sensor, and the inclination sensor are respectively connected to the signal input terminal of the main controller. The main controller is connected to the control signal input terminal of each of the electro-hydraulic proportional valves. The main controller is used to perform false leg determination and leveling actions based on the detection signals of each sensor.

9. The work equipment according to claim 8, characterized in that, The hydraulic circuit of the support mechanism is configured as a real-time sensing architecture without physical locking. The inlet and outlet ports of each electro-hydraulic proportional valve are directly connected to the rodless and rod chambers of the support cylinder. During non-leveling operations, the electro-hydraulic proportional valve remains in the neutral position. The pressure sensor is directly connected in parallel to the directly connected oil circuit to capture pressure fluctuation signals in the cylinder chamber caused by micro-settlement of the foundation or load transfer in real time.

10. The work equipment according to claim 8, characterized in that, The drive signal output by the main controller is a composite control command, which is formed by superimposing the basic valve opening signal output by the feedforward loop and the dynamic compensation signal output by the feedback loop. The feedforward loop receives the support cylinder displacement pre-calculated by the main controller based on the multi-body spatial kinematic model of the working equipment, and outputs the basic valve opening signal. The feedback loop receives the posture deviation signal and posture deviation rate of change signal of the working equipment, incorporates a fuzzy PID controller, and outputs the dynamic compensation signal.

Images