Intelligent control method and system for flat ground operation process

Abstract

The invention discloses an intelligent control method and system for a flat ground operation process. The method comprises the following steps: collecting a real-time state; obtaining an arc length coordinate through path capturing, and establishing an index relationship between the arc length coordinate and a planned path point location target elevation reference; determining a dynamic front window scale according to the operating speed and the hydraulic actuation characteristic parameters, and extracting a prediction target sequence in front of the arc length coordinates; on the basis of the predicted target sequence, the operation speed and the hydraulic actuation characteristics, online optimization of a future interval control quantity sequence is dynamically constructed, and a current control instruction is output in a rolling mode; performing hydraulic execution and closed-loop tracking; carrying soil monitoring and mode switching; a self-adaptive foresight prediction window based on a real-time position and an operation speed is established, rolling optimization compensation control is implemented, and space compensation is performed on hydraulic lag in advance; the soil shoveling state is monitored in real time through the dot-matrix type soil loading sensor, intelligent switching and management of the leveling mode and the soil unloading mode are achieved based on the soil shoveling state, and high-precision, high-efficiency and low-energy-consumption autonomous operation is achieved.

Description

Technical Field

[0001] This invention relates to an intelligent control method and system for leveling operations, belonging to the field of intelligent equipment and control technology for farmland leveling in the construction of high-standard farmland. Background Technology

[0002] In the construction of high-standard farmland, land leveling is a crucial indicator. Improving farmland leveling is a primary condition for ensuring effective rice irrigation management and reducing weed damage in paddy fields. It is also an important means to improve the utilization rate and uniformity of irrigation water in dryland fields. The project construction sets clear requirements for the size of strip fields and the elevation difference between fields: the area of ​​strip fields in plain areas should reach more than 100 mu, and the area of ​​fields in hilly and mountainous areas should reach more than 30 mu. It also requires that the elevation difference between paddy fields be within 3 cm and the elevation difference between dryland fields be within 5 cm. Existing farmland leveling equipment is mainly divided into leveling equipment based on laser technology and leveling equipment based on the Global Navigation Satellite System (GNSS). Their workflow is usually divided into two parts: "measurement and control". That is, the actual height of the leveling shovel is obtained by laser target measurement or RTK measurement. After comparing it with the target elevation, the control quantity is output to control the hydraulic system to drive the leveling shovel to lift and lower to complete the leveling operation. At present, the more advanced leveling machines at home and abroad have been equipped with terrain mapping functions, which can guide the operation of personnel according to the terrain map. Some tractors can achieve unmanned operation, that is, path tracking under the planned path.

[0003] However, prominent problems still exist in engineering applications: On the one hand, the existing hydraulic control systems of graders are mostly switch-type or simple control, with low control accuracy and poor anti-interference ability; on the other hand, the "same point measurement and control" is generally used in grading operations, and the hydraulic action has lag and inertia while the machine continues to move forward, resulting in the machine having already left the target position when the control action is generated, forming "action lag-spatial misalignment", which leads to under-cutting or over-cutting, making it difficult to stably meet the centimeter-level elevation difference index; in addition, the grading process lacks real-time monitoring of the soil loading status of the grader shovel. Continuing to shovel soil after being fully loaded will cause increased traction resistance, increased fuel consumption, decreased efficiency and induce attitude disturbances, which will further deteriorate the elevation control effect. Summary of the Invention

[0004] Purpose of the invention: Based on existing topographic mapping and path tracking technologies, and addressing the shortcomings of existing technologies, this invention provides an intelligent control method and system for leveling operations. This invention improves leveling accuracy and operational efficiency by using forward-looking prediction to compensate for hydraulic lag and combining it with soil load monitoring to achieve mode switching.

[0005] Technical solution: An intelligent control method for leveling operations, comprising the following steps:

[0006] S1. Real-time status acquisition: During the control cycle Internal data collection planning path point target elevation reference Grader location Work speed Hydraulic cylinder displacement With soil-carrying array signal , forming a time index The real-time state set is then filtered and preprocessed.

[0007] S2. Coordinate Mapping Establishment: Obtain arc length coordinates through path capture. And establish arc length coordinates Reference elevation of target points along the planned route The index relationship allows the control target to be updated with the path position without depending on the measurement and control point;

[0008] S3. Forward Prediction Window Construction: Based on Operation Speed With hydraulic actuation characteristic parameters , Determine the dynamic forward window scale and extract the predicted target sequence in front of the arc length coordinate. This expands the control objective to a future interval constraint.

[0009] S4. Rolling Optimization and Compensation Control: Based on Predicted Target Sequence Dynamically constructing future interval control quantity sequences based on operating speed and hydraulic actuation characteristics Online optimization, and continuously output the current control commands. To implement feedforward compensation for spatial misalignment caused by hydraulic hysteresis;

[0010] S5, Hydraulic Actuation and Closed-Loop Tracking: The hydraulic actuation module receives the current control command. And in accordance with the current control instructions Actions, utilizing displacement feedback Closed-loop tracking of the expected trajectory improves execution robustness;

[0011] S6. Soil Load Monitoring and Mode Switching: Based on soil load array signals... Estimated soil bearing rate And generate job mode flags When the operation mode flag is displayed When the indicator is fully loaded, switch to terrain-following / unloading control and limit further soil intake. (This is indicated by the operation mode flag.) When the load is not full, return to leveling control and cycle through S1 to S6 to complete continuous automatic leveling operations.

[0012] In a preferred embodiment, S1 specifically includes:

[0013] Within the control period 𝑘, i.e., the sampling period 𝑇𝑠, the target elevation reference of the planned path points is collected. Grader location Work speed Hydraulic cylinder displacement With soil-carrying array signal ;

[0014] To suppress GNSS transient jitter, hydraulic vibration burrs, and dot matrix contact jitter, a sliding median filter was applied to the grader position / operating speed / hydraulic cylinder displacement sequence / soil load rate sequence:

[0015] set up This represents any acquired raw discrete data sequence. Let represent the corresponding filtered output result. The calculation method for moving median filtering is as follows:

[0016]

[0017] in, This represents the half-width of the sliding window, used to determine the number of samples involved in the median calculation.

[0018] The above-mentioned sliding median filtering process can effectively suppress the influence of single-point mutations and high-frequency noise on state variables, and avoid abnormal sampled values ​​from directly participating in subsequent control and optimization calculations.

[0019] In a preferred embodiment, S2 specifically comprises:

[0020] Suppose the planned path consists of a sequence of discrete path points provided by an external system, and these path points are denoted as... Indicates the first The planar coordinates of the path points, where For path point indexing; The target elevation value is given for each path point. This is used to characterize the design elevation of the planned path at that location;

[0021] To facilitate subsequent indexing along the path, discrete path points are parameterized by arc length, with the path start point as the zero point of the arc length, and the first arc length is defined as... The arc length coordinates corresponding to each path point are: ,in ,right have:

[0022]

[0023] In the formula, Representing the Euclidean distance, we obtain the sequence of arc lengths of the path points. , used to describe the locations along the planned path;

[0024] During the control cycle At time S1, the grader position is obtained and filtered. In the path point sequence Search for the nearest path point index in the search. ,in satisfy:

[0025]

[0026] And the arc length coordinates corresponding to this path point are used as the arc length coordinates of the current position on the planned path, denoted as . ;in, Used to characterize the grader during the control cycle The location of the item along the path;

[0027] Obtaining arc length coordinates Then, establish the index relationship between the arc length coordinates and the target elevation reference of the planned path; and map the arc length-elevation of discrete path points to each other. Treating the target elevation as a reference table, the arc length position is calculated using interpolation operators. The corresponding target elevation reference value is denoted as :

[0028]

[0029] in, This represents an interpolation operation used to obtain a continuous target elevation reference between discrete path points.

[0030] In a preferred embodiment, S3 specifically includes:

[0031] During the control cycle The operating speed has been obtained from S1 at this time. The hydraulic actuation characteristic parameters are obtained by system identification or calibration, among which... This represents the pure time lag between the input of a control command and the moment the shovel begins to move and produce a noticeable response in the hydraulic system. The inertial time constant represents the main dynamic process of the hydraulic system; the above parameters are used to characterize the time response characteristics of the hydraulic actuator system.

[0032] Construct a forward-looking time scale based on hydraulic actuation characteristics: Let... To cover time correction factors under different oil temperatures, loads, and soil resistance conditions, the forward time scale is... Defined as:

[0033]

[0034] in, This indicates "the effective time scale required for the shovel to complete the main height change under the action of the hydraulic system from the current moment";

[0035] In obtaining the forward-looking time scale Then, the time scale is mapped to the spatial scale, that is, let... To control the cycle At the lower operating speed, the forward sight distance... Defined as:

[0036]

[0037] in, This represents the arc length of the path the grader will continue to travel at its current speed before the hydraulic system generates an effective height response; to ensure project feasibility, the forward sight distance is limited, with the minimum and maximum allowable forward sight distances set as follows: and The forward sight distance after limiting is denoted as :

[0038]

[0039] in, This represents a limiting operator used to ensure that the forward look-ahead distance remains within a feasible range.

[0040] Obtaining forward sight distance Then, using the current position arc length coordinates obtained in S2 As the starting point of the forward-looking window, construct the forward-looking prediction window interval in front of the arc length coordinate:

[0041]

[0042] Within this interval, sampling is performed according to the preset arc length step size. Discrete sampling is performed on the path to obtain a sequence of sampling points for the forward arc length. ,in

[0043]

[0044] and To meet The largest integer;

[0045] Based on arc length sampling point sequence The elevation reference function of the planned path target Extract the corresponding prediction target high-order sequence :

[0046]

[0047] in, Indicates during the control cycle Next, the The target elevation reference value corresponding to each forward-looking sampling position.

[0048] In a preferred embodiment, S4 specifically comprises:

[0049] During the control cycle At that time, the high-order sequence of the predicted target has been obtained by S3. ,in Indicates the arc length coordinates of the current position. Starting from the first point, within the forward prediction window... The target elevation reference value corresponding to each arc length sampling position; simultaneously, to describe the dynamic response of the hydraulic actuator to control commands within a future interval, let... Indicates during the control cycle Before the introduction of interval optimization, based on the basic control logic, the current height of the leveling shovel is compared with the elevation, and it is raised or lowered to the elevation position, or the unoptimized control command obtained from the control structure of the previous cycle is used. Indicates displacement by hydraulic cylinder The equivalent height state of the flatbed shovel obtained by mapping is then represented by the discrete prediction model of the hydraulic actuation characteristics as follows:

[0050]

[0051] So,

[0052] in, This represents the number of discrete lag steps calculated from the pure lag time. and The parameters of the discrete model are determined by the inertial time constant. To control the sampling period, this model is used to predict the trend of shovel height variation in future steps under a given control sequence.

[0053] Based on the above discrete prediction model, with future... The control commands within a control cycle are used as optimization variables to construct an interval control quantity sequence:

[0054]

[0055] And with predicted output Corresponding predicted target elevation The interval error is the primary optimization objective, while a control smoothing term is introduced to constrain the rate of change of control, thus constructing the interval optimization objective function. :

[0056]

[0057] in, and These are the interval error weight and the control smoothing weight, used to balance target tracking accuracy and control feasibility.

[0058] During the optimization process, physical feasible region constraints are applied to the control commands to ensure that the optimization results can be reliably executed by the actual hydraulic system; the constraints include at least control command amplitude constraints and control increment constraints, and their general form is as follows:

[0059]

[0060] in, and These are the minimum and maximum permissible values ​​for the hydraulic valve control input, respectively. To control the maximum allowable range of change in the command;

[0061] In each control cycle Within this framework, the above-mentioned interval optimization problem is solved online based on the current state information and the predicted target sequence to obtain the optimal control quantity sequence. ;

[0062]

[0063] A rolling time-domain control strategy is adopted, which outputs only the first control variable in the optimal sequence as the control command for the current cycle:

[0064]

[0065] The optimized control command Used to replace unoptimized control commands And in the next control cycle, the updated state information is used to reconstruct the prediction model and optimization problem, thus forming a continuously rolling interval optimization and compensation control process; simultaneously This will serve as the input for subsequent hydraulic actuation and closed-loop tracking steps.

[0066] In a preferred embodiment, S6 specifically includes:

[0067] During the control cycle At that moment, the system acquired the soil-loading point array signal inside the grading shovel via S1. ,in This represents the discrete set of trigger states output by the dot-matrix switch sensor array, used to characterize whether soil contact exists at different spatial locations inside the grading shovel; for the soil-loaded dot-matrix signal The number of closed points in the lattice is counted. The total number of points is Thus, the soil carrying capacity is constructed. :

[0068]

[0069] Based on soil carrying capacity Construct operation mode flag The operation mode flag is used to characterize the control mode of the current operation; when the soil load ratio estimate is... When the preset full load threshold is exceeded, the operation mode flag is activated. Switch to full load mode; when the estimated soil load is... When the value is below the corresponding release threshold, the operation mode flag is activated. Switch to non-full load mode; by introducing a hysteresis interval between the full load and non-full load determination thresholds, frequent mode switching caused by fluctuations in the soil load rate near the critical state is avoided.

[0070] When the work mode flag When the indication is non-full load mode, the system maintains the leveling control mode, generates control commands according to the rolling optimization results of S4, and drives the grader to complete the leveling operation; when the operation mode flag is displayed... When the indication is full load mode, the system switches to terrain-following or unloading control mode to restrict the continued soil intake and adjust the control target so that the flatbed shovel runs along the predetermined terrain or strategy during the unloading stage, thereby avoiding the sudden increase in traction resistance, increased hydraulic load and deterioration of control performance caused by continuing to shovel under high load.

[0071] In full-load mode, the system continuously monitors the estimated soil load rate. When the soil loading rate decreases during the unloading process and returns to the non-full load judgment range, the operation mode indicator is activated. The system switches back to the non-full load mode and re-enables the leveling control mode. Through the above-mentioned soil monitoring and mode switching mechanism, the leveling control and load status are coupled in a closed loop, so that the system can avoid invalid soil entry and overload operation while ensuring leveling accuracy.

[0072] By cyclically executing S1 to S6, the grader can perform continuous automatic leveling operations under complex terrain and load variation conditions.

[0073] The system for implementing intelligent control methods in the leveling operation process includes a data acquisition module, a path capture module, a forward view window module, a central processing module, a soil load monitoring module, and a hydraulic execution module.

[0074] The data acquisition module is signal-connected to the central processing module and the soil load monitoring module, the forward window module and the path capture module are signal-connected to the central processing module, and the central processing module is signal-connected to the hydraulic execution module.

[0075] The data acquisition module is used to acquire the target elevation reference of the planned path points, the position of the grader, the heading angle, the operating speed, the displacement of the hydraulic cylinder and the soil-carrying point array signal within the control cycle;

[0076] The path capture module is used to project the real-time location onto the planned path, obtain the arc length coordinates, and establish a correspondence with the path reference accordingly.

[0077] The forward window module is used to adaptively determine the length of the forward prediction interval based on the operating speed and hydraulic actuation characteristic parameters, form a dynamic forward window in front of the arc length coordinate, extract the predicted target sequence from it, and send it to the central processing module.

[0078] The soil loading monitoring module is used to acquire soil loading matrix signals and send them to the data acquisition module. The central processing module acquires the soil loading matrix signals through the acquisition module and determines whether the soil loading mode is switched.

[0079] The hydraulic actuator module is used to execute control commands issued by the central processing module;

[0080] The soil load monitoring module sends the soil load matrix signal to the acquisition module, and the acquisition module sends the acquired data to the central processing module. The path capture module and the forward window module acquire the data through the central processing module, process it, and then feed it back to the central processing module. The central processing module determines whether to switch the soil load mode based on the soil load matrix signal acquired by the acquisition module in real time.

[0081] This invention integrates a data acquisition module, a path capture module, a forward-looking window module, a soil load monitoring module, and a hydraulic execution module, and combines them with the coordinated control of a central processing module to achieve intelligent predictive control and real-time judgment of soil load status during the leveling operation process. This solves the problems of motion lag and spatial misalignment caused by "simultaneous measurement and control" in the prior art, as well as the problems of increased traction resistance and deteriorated control effect caused by the lack of soil load monitoring, thereby improving the accuracy and efficiency of leveling operations.

[0082] In a preferred embodiment, to improve the measurement accuracy and reliability of position, attitude, and actuator status, the data acquisition module includes a dual-antenna BeiDou RTK mounted on the grader, an on-board IMU, and a displacement sensor mounted on the hydraulic cylinder. The dual-antenna BeiDou RTK, the on-board IMU, and the displacement sensor are respectively connected to the central processing module.

[0083] By employing a data acquisition scheme that combines dual-antenna Beidou RTK, vehicle-mounted IMU, and cylinder displacement sensor, the real-time and accurate perception of the grader's three-dimensional position and blade elevation is achieved, providing precise input data for intelligent control.

[0084] In a preferred embodiment, in order to achieve distributed and refined sensing of the soil loading status of the leveler, the soil loading monitoring module includes a soil loading sensor arranged inside the leveler. The soil loading sensor is arranged in a 3×5 dot matrix array and is evenly distributed in a matrix form along the width and height directions of the leveler.

[0085] By employing a 3×5 dot array of soil load sensors, real-time area array monitoring of soil accumulation along the width and height of the shovel face is achieved, providing a reliable basis for the central processing module to determine the soil load and switch operating modes.

[0086] In a preferred embodiment, to improve the control accuracy and response characteristics of the hydraulic system and reduce energy consumption, the hydraulic actuator module includes an unloading valve, a proportional directional valve, a hydraulic pump, and a hydraulic cylinder. When the grader is running, the hydraulic pump maintains its operating state and continuously pumps oil from the oil tank. The proportional directional valve opens to allow oil to enter, and at the same time, the proportional directional valve controls the hydraulic cylinder to rise or fall by controlling the direction of oil inflow. At this time, the unloading valve closes. When it is necessary to maintain the current action of the hydraulic cylinder, the unloading valve opens, the proportional directional valve closes, the hydraulic oil flows directly back to the oil tank, and the hydraulic cylinder remains unchanged.

[0087] By employing a control strategy that combines a proportional directional valve with an unloading valve, continuous and precise control of the hydraulic cylinder's lifting action is achieved, as well as rapid unloading and status locking of the system when the position needs to be maintained, thereby improving the dynamic performance of elevation control and system energy efficiency.

[0088] Beneficial Effects: This invention elevates the control of traditional graders from "instantaneous error correction at the same measurement and control point" to "predictive-optimized rolling control with interval constraints." By extracting the terrain / target elevation sequence ahead through a dynamic forward window, it explicitly incorporates future interval targets into the control decision, and continuously solves and outputs the control quantity sequence within each control cycle. This achieves feedforward compensation and overshoot suppression for response lag in the hydraulic actuators and spatiotemporal misalignment caused by the machine's forward movement. In contrast, existing technologies primarily rely on real-time measurement data for slope calculation and immediate adjustment, often failing to address the spatial positioning deviation caused by actuation lag at the mechanistic level.

[0089] This invention introduces a soil-loaded array of sensors for closed-loop monitoring of soil loading status. Based on soil load rate estimation and combined with hysteresis anti-shaking discrimination, it achieves full-load identification, thereby triggering a smooth switch and constraint management of the "leveling-terrain imitation / unloading" operation mode. This mechanism can effectively avoid the adverse effects of continued soil loading under full load, such as increased traction resistance, increased power load, and increased fuel consumption. It also directly incorporates the "operational capacity boundary (full-load threshold)" as a control constraint into closed-loop decision-making and safety management, rather than simply providing a general "overload / empty load" description at the path / construction level, thus significantly improving the continuity, reliability, and safety of operations under complex working conditions.

[0090] This invention constructs a complete closed-loop unmanned control system covering "state perception—range optimization—hydraulic execution—operation mode management," realizing intelligent autonomous control of the grader's leveling operation process. This system can maintain stable leveling accuracy and operating efficiency under disturbances and changes in operating conditions, reducing reliance on manual operation and parameter adjustment costs. Furthermore, it reduces energy consumption and fuel costs by minimizing ineffective cutting and excessive adjustments, providing key technical support for high-standard farmland leveling operations. Attached Figure Description

[0091] To more clearly illustrate the technical solutions in the embodiments of the present invention 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 embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0092] Figure 1 This is a flowchart of the method of the present invention;

[0093] Figure 2 This is a system structure block diagram of the present invention. Detailed Implementation

[0094] 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 some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0095] In the description of this invention, it should be understood that the terms "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and 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 invention.

[0096] In this invention, unless otherwise explicitly specified and limited, "above" or "below" the second feature can include direct contact between the first and second features, or contact between the first and second features through another feature between them. Furthermore, "above," "over," and "on top" of the second feature includes the first feature directly above or diagonally above the second feature, or simply indicates that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature includes the first feature directly below or diagonally below the second feature, or simply indicates that the first feature is at a lower horizontal level than the second feature.

[0097] like Figure 1 As shown, an intelligent control method for leveling operations includes the following steps:

[0098] S1. Real-time status acquisition: During the control cycle Internal data collection planning path point target elevation reference Grader location Work speed Hydraulic cylinder displacement With soil-carrying array signal , forming a time index The real-time state set is then filtered and preprocessed.

[0099] Within the control period 𝑘, i.e., the sampling period 𝑇𝑠, the target elevation reference of the planned path points is collected. Grader location Work speed Hydraulic cylinder displacement With soil-carrying array signal ;

[0100] To suppress GNSS transient jitter, hydraulic vibration burrs, and dot matrix contact jitter, a sliding median filter was applied to the grader position / operating speed / hydraulic cylinder displacement sequence / soil load rate sequence:

[0101] set up This represents any acquired raw discrete data sequence. Let represent the corresponding filtered output result. The calculation method for moving median filtering is as follows:

[0102]

[0103] in, This represents the half-width of the sliding window, used to determine the number of samples involved in the median calculation.

[0104] The above-mentioned sliding median filtering process can effectively suppress the influence of single-point mutations and high-frequency noise on state variables, and avoid abnormal sampled values ​​from directly participating in subsequent control and optimization calculations.

[0105] S2. Coordinate Mapping Establishment: Obtain arc length coordinates through path capture. And establish arc length coordinates Reference elevation of target points along the planned route The index relationship allows the control target to be updated with the path position without depending on the measurement and control point;

[0106] Suppose the planned path consists of a sequence of discrete path points provided by an external system, and these path points are denoted as... Indicates the first The planar coordinates of the path points, where For path point indexing; The target elevation value is given for each path point. This is used to characterize the design elevation of the planned path at that location;

[0107] To facilitate subsequent indexing along the path, discrete path points are parameterized by arc length, with the path start point as the zero point of the arc length, and the first arc length is defined as... The arc length coordinates corresponding to each path point are: ,in ,right have:

[0108]

[0109] In the formula, Representing the Euclidean distance, we obtain the sequence of arc lengths of the path points. , used to describe the locations along the planned path;

[0110] During the control cycle At time S1, the grader position is obtained and filtered. In the path point sequence Search for the nearest path point index in the search. ,in satisfy:

[0111]

[0112] And the arc length coordinates corresponding to this path point are used as the arc length coordinates of the current position on the planned path, denoted as . ;in, Used to characterize the grader during the control cycle The location of the item along the path;

[0113] Obtaining arc length coordinates Then, establish the index relationship between the arc length coordinates and the target elevation reference of the planned path; and map the arc length-elevation of discrete path points to each other. Treating the target elevation as a reference table, the arc length position is calculated using interpolation operators. The corresponding target elevation reference value is denoted as :

[0114]

[0115] in, This represents an interpolation operation used to obtain a continuous target elevation reference between discrete path points.

[0116] S3. Forward Prediction Window Construction: Based on Operation Speed With hydraulic actuation characteristic parameters , Determine the dynamic forward window scale and extract the predicted target sequence in front of the arc length coordinate. This expands the control objective to a future interval constraint.

[0117] During the control cycle The operating speed has been obtained from S1 at this time. The hydraulic actuation characteristic parameters are obtained by system identification or calibration, among which... This represents the pure time lag between the input of a control command and the moment the shovel begins to move and produce a noticeable response in the hydraulic system. The inertial time constant represents the main dynamic process of the hydraulic system; the above parameters are used to characterize the time response characteristics of the hydraulic actuator system.

[0118] Construct a forward-looking time scale based on hydraulic actuation characteristics: Let... To cover time correction factors under different oil temperatures, loads, and soil resistance conditions, the forward time scale is... Defined as:

[0119]

[0120] in, This indicates "the effective time scale required for the shovel to complete the main height change under the action of the hydraulic system from the current moment";

[0121] In obtaining the forward-looking time scale Then, the time scale is mapped to the spatial scale, that is, let... To control the cycle At the lower operating speed, the forward sight distance... Defined as:

[0122]

[0123] in, This represents the arc length of the path the grader will continue to travel at its current speed before the hydraulic system generates an effective height response; to ensure project feasibility, the forward sight distance is limited, with the minimum and maximum allowable forward sight distances set as follows: and The forward sight distance after limiting is denoted as :

[0124]

[0125] in, This represents a limiting operator used to ensure that the forward look-ahead distance remains within a feasible range.

[0126] Obtaining forward sight distance Then, using the current position arc length coordinates obtained in S2 As the starting point of the forward-looking window, construct the forward-looking prediction window interval in front of the arc length coordinate:

[0127]

[0128] Within this interval, sampling is performed according to the preset arc length step size. Discrete sampling is performed on the path to obtain a sequence of sampling points for the forward arc length. ,in

[0129]

[0130] and To meet The largest integer;

[0131] Based on arc length sampling point sequence The elevation reference function of the planned path target Extract the corresponding prediction target high-order sequence :

[0132]

[0133] in, Indicates during the control cycle Next, the The target elevation reference value corresponding to each forward-looking sampling position.

[0134] S4. Rolling Optimization and Compensation Control: Based on Predicted Target Sequence Dynamically constructing future interval control quantity sequences based on operating speed and hydraulic actuation characteristics Online optimization, and continuously output the current control commands. To implement feedforward compensation for spatial misalignment caused by hydraulic hysteresis;

[0135] During the control cycle At that time, the high-order sequence of the predicted target has been obtained by S3. ,in Indicates the arc length coordinates of the current position. Starting from the first point, within the forward prediction window... The target elevation reference value corresponding to each arc length sampling position; simultaneously, to describe the dynamic response of the hydraulic actuator to control commands within a future interval, let... Indicates during the control cycle Before the introduction of interval optimization, based on the basic control logic, the current height of the leveling shovel is compared with the elevation, and it is raised or lowered to the elevation position, or the unoptimized control command obtained from the control structure of the previous cycle is used. Indicates displacement by hydraulic cylinder The equivalent height state of the flatbed shovel obtained by mapping is then represented by the discrete prediction model of the hydraulic actuation characteristics as follows:

[0136]

[0137] So,

[0138] in, This represents the number of discrete lag steps calculated from the pure lag time. and The parameters of the discrete model are determined by the inertial time constant. To control the sampling period, this model is used to predict the trend of shovel height variation in future steps under a given control sequence.

[0139] Based on the above discrete prediction model, with future... The control commands within a control cycle are used as optimization variables to construct an interval control quantity sequence:

[0140]

[0141] And with predicted output Corresponding predicted target elevation The interval error is the primary optimization objective, while a control smoothing term is introduced to constrain the rate of change of control, thus constructing the interval optimization objective function. :

[0142]

[0143] in, and These are the interval error weight and the control smoothing weight, used to balance target tracking accuracy and control feasibility.

[0144] During the optimization process, physical feasible region constraints are applied to the control commands to ensure that the optimization results can be reliably executed by the actual hydraulic system; the constraints include at least control command amplitude constraints and control increment constraints, and their general form is as follows:

[0145]

[0146] in, and These are the minimum and maximum permissible values ​​for the hydraulic valve control input, respectively. To control the maximum allowable range of change in the command;

[0147] In each control cycle Within this framework, the above-mentioned interval optimization problem is solved online based on the current state information and the predicted target sequence to obtain the optimal control quantity sequence. ;

[0148]

[0149] A rolling time-domain control strategy is adopted, which outputs only the first control variable in the optimal sequence as the control command for the current cycle:

[0150]

[0151] The optimized control command Used to replace unoptimized control commands And in the next control cycle, the updated state information is used to reconstruct the prediction model and optimization problem, thus forming a continuously rolling interval optimization and compensation control process; simultaneously This will serve as the input for subsequent hydraulic actuation and closed-loop tracking steps.

[0152] S5, Hydraulic Actuation and Closed-Loop Tracking: The hydraulic actuation module receives the current control command. And in accordance with the current control instructions Actions, utilizing displacement feedback Closed-loop tracking of the expected trajectory improves execution robustness;

[0153] S6. Soil Load Monitoring and Mode Switching: Based on soil load array signals... Estimated soil bearing rate And generate job mode flags When the operation mode flag is displayed When the indicator is fully loaded, switch to terrain-following / unloading control and limit further soil intake. (This is indicated by the operation mode flag.) When the load is not full, return to leveling control and cycle through S1 to S6 to complete continuous automatic leveling operations;

[0154] During the control cycle At that moment, the system acquired the soil-loading point array signal inside the grading shovel via S1. ,in This represents the discrete set of trigger states output by the dot-matrix switch sensor array, used to characterize whether soil contact exists at different spatial locations inside the grading shovel; for the soil-loaded dot-matrix signal The number of closed points in the lattice is counted. The total number of points is Thus, the soil carrying capacity is constructed. :

[0155]

[0156] Based on soil carrying capacity Construct operation mode flag The operation mode flag is used to characterize the control mode of the current operation; when the soil load ratio estimate is... When the preset full load threshold is exceeded, the operation mode flag is activated. Switch to full load mode; when the estimated soil load is... When the value is below the corresponding release threshold, the operation mode flag is activated. Switch to non-full load mode; by introducing a hysteresis interval between the full load and non-full load determination thresholds, frequent mode switching caused by fluctuations in the soil load rate near the critical state is avoided.

[0157] When the work mode flag When the indication is non-full load mode, the system maintains the leveling control mode, generates control commands according to the rolling optimization results of S4, and drives the grader to complete the leveling operation; when the operation mode flag is displayed... When the indication is full load mode, the system switches to terrain-following or unloading control mode to restrict the continued soil intake and adjust the control target so that the flatbed shovel runs along the predetermined terrain or strategy during the unloading stage, thereby avoiding the sudden increase in traction resistance, increased hydraulic load and deterioration of control performance caused by continuing to shovel under high load.

[0158] In full-load mode, the system continuously monitors the estimated soil load rate. When the soil loading rate decreases during the unloading process and returns to the non-full load judgment range, the operation mode indicator is activated. The system switches back to the non-full load mode and re-enables the leveling control mode. Through the above-mentioned soil monitoring and mode switching mechanism, the leveling control and load status are coupled in a closed loop, so that the system can avoid invalid soil entry and overload operation while ensuring leveling accuracy.

[0159] By cyclically executing S1 to S6, the grader can perform continuous automatic leveling operations under complex terrain and load variation conditions.

[0160] like Figure 2 As shown, the system for realizing intelligent control of the leveling operation process includes a data acquisition module, a path capture module, a forward view window module, a central processing module, a soil load monitoring module, and a hydraulic execution module.

[0161] The data acquisition module is signal-connected to the central processing module and the soil load monitoring module, the forward window module and the path capture module are signal-connected to the central processing module, and the central processing module is signal-connected to the hydraulic execution module.

[0162] The data acquisition module is used to acquire the target elevation reference of the planned path points, the position of the grader, the heading angle, the operating speed, the displacement of the hydraulic cylinder and the soil-carrying point array signal within the control cycle;

[0163] The path capture module is used to project the real-time location onto the planned path, obtain the arc length coordinates, and establish a correspondence with the path reference accordingly.

[0164] The forward window module is used to adaptively determine the length of the forward prediction interval based on the operating speed and hydraulic actuation characteristic parameters, form a dynamic forward window in front of the arc length coordinate, extract the predicted target sequence from it, and send it to the central processing module.

[0165] The soil loading monitoring module is used to acquire soil loading matrix signals and send them to the data acquisition module. The central processing module acquires the soil loading matrix signals through the acquisition module and determines whether the soil loading mode is switched.

[0166] The hydraulic actuator module is used to execute control commands issued by the central processing module;

[0167] The soil load monitoring module sends the soil load matrix signal to the acquisition module, and the acquisition module sends the acquired data to the central processing module. The path capture module and the forward window module acquire the data through the central processing module, process it, and then feed it back to the central processing module. The central processing module determines whether to switch the soil load mode based on the soil load matrix signal acquired by the acquisition module in real time.

[0168] The data acquisition module includes a dual-antenna Beidou RTK mounted on the grader, an on-board IMU, and a displacement sensor mounted on the hydraulic cylinder. The dual-antenna Beidou RTK, the on-board IMU, and the displacement sensor are respectively connected to the central processing module.

[0169] The soil load monitoring module includes a soil load sensor arranged inside the leveling shovel. The soil load sensor is arranged in a 3×5 dot matrix array and is evenly distributed in a matrix form along the width and height directions of the leveling shovel.

[0170] The hydraulic actuator module includes an unloading valve, a proportional directional valve, a hydraulic pump, and a hydraulic cylinder. When the grader is running, the hydraulic pump continuously pumps oil from the oil tank to maintain its operating state. The proportional directional valve opens to allow oil to enter, and at the same time, the proportional directional valve controls the hydraulic cylinder to rise or fall by controlling the direction of oil inflow. At this time, the unloading valve closes. When it is necessary to maintain the current action of the hydraulic cylinder, the unloading valve opens, the proportional directional valve closes, and the hydraulic oil flows directly back to the oil tank, while the hydraulic cylinder remains unchanged.

[0171] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to in the method section.

[0172] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for intelligent control of leveling operations, characterized in that: Includes the following steps: S1. Real-time status acquisition: During the control cycle Internal data collection planning path point target elevation reference Grader location Work speed Hydraulic cylinder displacement With soil-carrying array signal , forming a time index The real-time state set is then filtered and preprocessed. S2. Coordinate Mapping Establishment: Obtain arc length coordinates through path capture. And establish arc length coordinates Reference elevation of target points along the planned route The index relationship allows the control target to be updated with the path position without depending on the measurement and control point; S3. Forward Prediction Window Construction: Based on Operation Speed With hydraulic actuation characteristic parameters , Determine the dynamic forward window scale and extract the predicted target sequence in front of the arc length coordinate. This expands the control objective to a future interval constraint. S4. Rolling Optimization and Compensation Control: Based on Predicted Target Sequence Dynamically constructing future interval control quantity sequences based on operating speed and hydraulic actuation characteristics Online optimization, and continuously output the current control commands. To implement feedforward compensation for spatial misalignment caused by hydraulic hysteresis; S5, Hydraulic Actuation and Closed-Loop Tracking: The hydraulic actuation module receives the current control command. And in accordance with the current control instructions Actions, utilizing displacement feedback Closed-loop tracking of the expected trajectory improves execution robustness; S6. Soil Load Monitoring and Mode Switching: Based on soil load array signals... Estimated soil bearing rate And generate job mode flags When the operation mode flag is displayed When the indicator is fully loaded, switch to terrain-following / unloading control and limit further soil intake. (This is indicated by the operation mode flag.) When the load is not full, return to leveling control and cycle through S1 to S6 to complete continuous automatic leveling operations.

2. The intelligent control method for leveling operations according to claim 1, characterized in that: Specifically, S1 is: Within the control period 𝑘, i.e., the sampling period 𝑇𝑠, the target elevation reference of the planned path points is collected. Grader location Work speed Hydraulic cylinder displacement With soil-carrying array signal ; To suppress GNSS transient jitter, hydraulic vibration burrs, and dot matrix contact jitter, a sliding median filter was applied to the grader position / operating speed / hydraulic cylinder displacement sequence / soil load rate sequence: set up This represents any acquired raw discrete data sequence. Let represent the corresponding filtered output result. The calculation method for moving median filtering is as follows: in, This represents the half-width of the sliding window, used to determine the number of samples involved in the median calculation. The above-mentioned sliding median filtering process can effectively suppress the influence of single-point mutations and high-frequency noise on state variables, and avoid abnormal sampled values ​​from directly participating in subsequent control and optimization calculations.

3. The intelligent control method for leveling operations according to claim 2, characterized in that: Specifically, S2 is: Suppose the planned path consists of a sequence of discrete path points provided by an external system, and these path points are denoted as... Indicates the first The planar coordinates of the path points, where For path point indexing; The target elevation value is given for each path point. This is used to characterize the design elevation of the planned path at that location; To facilitate subsequent indexing along the path, discrete path points are parameterized by arc length, with the path start point as the zero point of the arc length, and the first arc length is defined as... The arc length coordinates corresponding to each path point are: ,in ,right have: In the formula, Representing the Euclidean distance, we obtain the sequence of arc lengths of the path points. , used to describe the locations along the planned path; During the control cycle At time S1, the grader position is obtained and filtered. In the path point sequence Search for the nearest path point index in the search. ,in satisfy: And the arc length coordinates corresponding to this path point are used as the arc length coordinates of the current position on the planned path, denoted as . ;in, Used to characterize the grader during the control cycle The location of the item along the path; Obtaining arc length coordinates Then, establish the index relationship between the arc length coordinates and the target elevation reference of the planned path; and map the arc length-elevation of discrete path points to each other. Treating the target elevation as a reference table, the arc length position is calculated using interpolation operators. The corresponding target elevation reference value is denoted as : in, This represents an interpolation operation used to obtain a continuous target elevation reference between discrete path points.

4. The intelligent control method for leveling operations according to claim 3, characterized in that: Specifically, S3 is: During the control cycle The operating speed has been obtained from S1 at this time. The hydraulic actuation characteristic parameters are obtained by system identification or calibration, among which... This represents the pure time lag between the input of a control command and the moment the shovel begins to move and produce a noticeable response in the hydraulic system. The inertial time constant represents the main dynamic process of the hydraulic system; the above parameters are used to characterize the time response characteristics of the hydraulic actuator system. Construct a forward-looking time scale based on hydraulic actuation characteristics: Let... To cover time correction factors under different oil temperatures, loads, and soil resistance conditions, the forward time scale is... Defined as: in, This indicates "the effective time scale required for the shovel to complete the main height change under the action of the hydraulic system from the current moment"; In obtaining the forward-looking time scale Then, the time scale is mapped to the spatial scale, that is, let... To control the cycle At the lower operating speed, the forward sight distance... Defined as: in, This represents the arc length of the path the grader will continue to travel at its current speed before the hydraulic system generates an effective height response; to ensure project feasibility, the forward sight distance is limited, with the minimum and maximum allowable forward sight distances set as follows: and The forward sight distance after limiting is denoted as : in, This represents a limiting operator used to ensure that the forward look-ahead distance remains within a feasible range. Obtaining forward sight distance Then, using the current position arc length coordinates obtained in S2 As the starting point of the forward-looking window, construct the forward-looking prediction window interval in front of the arc length coordinate: Within this interval, sampling is performed according to the preset arc length step size. Discrete sampling is performed on the path to obtain a sequence of sampling points for the forward arc length. ,in and To meet The largest integer; Based on arc length sampling point sequence The elevation reference function of the planned path target Extract the corresponding prediction target high-order sequence : in, Indicates during the control cycle Next, the The target elevation reference value corresponding to each forward-looking sampling position.

5. The intelligent control method for leveling operations according to claim 4, characterized in that: Specifically, S4 is: During the control cycle At that time, the high-order sequence of the predicted target has been obtained by S3. ,in Indicates the arc length coordinates of the current position. Starting from the first point, within the forward prediction window... The target elevation reference value corresponding to each arc length sampling position; simultaneously, to describe the dynamic response of the hydraulic actuator to control commands within a future interval, let... Indicates during the control cycle Before the introduction of interval optimization, based on the basic control logic, the current height of the leveling shovel is compared with the elevation, and it is raised or lowered to the elevation position, or the unoptimized control command obtained from the control structure of the previous cycle is used. Indicates displacement by hydraulic cylinder The equivalent height state of the flatbed shovel obtained by mapping is then represented by the discrete prediction model of the hydraulic actuation characteristics as follows: So, in, This represents the number of discrete lag steps calculated from the pure lag time. and The parameters of the discrete model are determined by the inertial time constant. To control the sampling period, this model is used to predict the trend of shovel height variation in future steps under a given control sequence. Based on the above discrete prediction model, with future... The control commands within a control cycle are used as optimization variables to construct an interval control quantity sequence: And with predicted output Corresponding predicted target elevation The interval error is the primary optimization objective, while a control smoothing term is introduced to constrain the rate of change of control, thus constructing the interval optimization objective function. : in, and These are the interval error weight and the control smoothing weight, used to balance target tracking accuracy and control feasibility. During the optimization process, physical feasible region constraints are applied to the control commands to ensure that the optimization results can be reliably executed by the actual hydraulic system; the constraints include at least control command amplitude constraints and control increment constraints, and their general form is as follows: in, and These are the minimum and maximum permissible values ​​for the hydraulic valve control input, respectively. To control the maximum allowable range of change in the command; In each control cycle Within this framework, the above-mentioned interval optimization problem is solved online based on the current state information and the predicted target sequence to obtain the optimal control quantity sequence. ; A rolling time-domain control strategy is adopted, which outputs only the first control variable in the optimal sequence as the control command for the current cycle: The optimized control command Used to replace unoptimized control commands And in the next control cycle, the updated state information is used to reconstruct the prediction model and optimization problem, thus forming a continuously rolling interval optimization and compensation control process; simultaneously This will serve as the input for subsequent hydraulic actuation and closed-loop tracking steps.

6. The intelligent control method for leveling operations according to claim 5, characterized in that: Specifically, S6 is: During the control cycle At any given time, the system acquires the soil-loading point array signal inside the grading shovel via S1. ,in This represents the discrete set of trigger states output by the dot-matrix switch sensor array, used to characterize whether soil contact exists at different spatial locations inside the grading shovel; for the soil-loaded dot-matrix signal Perform lattice closure statistics; the number of lattice closures is counted as follows: The total number of points is Thus, the soil carrying capacity is constructed. : Based on soil carrying capacity Construct operation mode flag The operation mode flag is used to characterize the control mode of the current operation; when the soil load ratio estimate is... When the preset full load threshold is exceeded, the operation mode flag is activated. Switch to full load mode; when the estimated soil load is... When the value is below the corresponding release threshold, the operation mode flag is activated. Switch to non-full load mode; by introducing a hysteresis interval between the full load and non-full load determination thresholds, frequent mode switching caused by fluctuations in the soil load rate near the critical state is avoided. When the work mode flag When the indication is non-full load mode, the system maintains the leveling control mode, generates control commands according to the rolling optimization results of S4, and drives the grader to complete the leveling operation; when the operation mode flag is displayed... When the indication is full load mode, the system switches to terrain-following or unloading control mode to restrict the continued soil intake and adjust the control target so that the flatbed shovel runs along the predetermined terrain or strategy during the unloading stage, thereby avoiding the sudden increase in traction resistance, increased hydraulic load and deterioration of control performance caused by continuing to shovel under high load. In full-load mode, the system continuously monitors the estimated soil load rate. When the soil loading rate decreases during the unloading process and returns to the non-full load judgment range, the operation mode indicator is activated. The system switches back to the non-full load mode and re-enables the leveling control mode. Through the above-mentioned soil monitoring and mode switching mechanism, the leveling control and load status are coupled in a closed loop, so that the system can avoid invalid soil entry and overload operation while ensuring leveling accuracy. By cyclically executing S1 to S6, the grader can perform continuous automatic leveling operations under complex terrain and load variation conditions.

7. A system for implementing the intelligent control method for leveling operations as described in claims 1-6, characterized in that, It includes a data acquisition module, a path capture module, a front view window module, a central processing module, a soil load monitoring module, and a hydraulic execution module; The data acquisition module is signal-connected to the central processing module and the soil load monitoring module, the forward window module and the path capture module are signal-connected to the central processing module, and the central processing module is signal-connected to the hydraulic execution module. The data acquisition module is used to acquire the target elevation reference of the planned path points, the position of the grader, the heading angle, the operating speed, the displacement of the hydraulic cylinder and the soil-carrying point array signal within the control cycle; The path capture module is used to project the real-time location onto the planned path, obtain the arc length coordinates, and establish a correspondence with the path reference accordingly. The forward window module is used to adaptively determine the length of the forward prediction interval based on the operating speed and hydraulic actuation characteristic parameters, form a dynamic forward window in front of the arc length coordinate, extract the predicted target sequence from it, and send it to the central processing module. The soil loading monitoring module is used to acquire soil loading matrix signals and send them to the data acquisition module. The central processing module acquires the soil loading matrix signals through the acquisition module and determines whether the soil loading mode is switched. The hydraulic actuator module is used to execute control commands issued by the central processing module; The soil load monitoring module sends the soil load matrix signal to the acquisition module, which then sends the acquired data to the central processing module. The path capture module and the forward window module acquire and process the data through the central processing module, and then feed it back to the central processing module. The central processing module determines whether to switch the soil load mode based on the soil load matrix signal acquired in real time by the acquisition module.

8. The system of the intelligent control method for leveling operations according to claim 7, characterized in that, The data acquisition module includes a dual-antenna Beidou RTK mounted on the grader, an on-board IMU, and a displacement sensor mounted on the hydraulic cylinder. The dual-antenna Beidou RTK, the on-board IMU, and the displacement sensor are respectively connected to the central processing module.

9. The system of the intelligent control method for leveling operations according to claim 7, characterized in that, The soil load monitoring module includes a soil load sensor arranged inside the leveling shovel. The soil load sensor is arranged in a 3×5 dot matrix array and is evenly distributed in a matrix form along the width and height directions of the leveling shovel.

10. The system of the intelligent control method for leveling operations according to claim 7, characterized in that, The hydraulic actuator module includes an unloading valve, a proportional directional valve, a hydraulic pump, and a hydraulic cylinder. When the grader is running, the hydraulic pump continuously pumps oil from the oil tank to maintain its operating state. The proportional directional valve opens to allow oil to enter, and at the same time, the proportional directional valve controls the hydraulic cylinder to rise or fall by controlling the direction of oil inflow. At this time, the unloading valve closes. When it is necessary to maintain the current action of the hydraulic cylinder, the unloading valve opens, the proportional directional valve closes, the hydraulic oil flows directly back to the oil tank, and the hydraulic cylinder remains stationary.

Images