CN-121979077-A - Intelligent control method and system for land leveling operation process

CN121979077ACN 121979077 ACN121979077 ACN 121979077ACN-121979077-A

Abstract

The invention discloses an intelligent control method and system for a land leveling operation process, which comprises the steps of collecting real-time states, obtaining arc length coordinates through path grabbing, establishing an index relation between the arc length coordinates and a planned path point elevation reference, determining a dynamic front window scale according to operation speed and hydraulic operation characteristic parameters, extracting a predicted target sequence in front of the arc length coordinates, dynamically constructing on-line optimization of a future interval control quantity sequence based on the predicted target sequence, the operation speed and the hydraulic operation characteristic, rolling to output a current control instruction, performing hydraulic execution and closed-loop tracking, carrying soil monitoring and mode switching, performing space compensation on hydraulic hysteresis in advance by establishing a self-adaptive front view prediction window based on real-time positions and the operation speed, performing rolling optimization compensation control, monitoring the soil shoveling state in real time through a lattice type carrying soil sensor, and realizing intelligent switching and management of leveling and soil unloading modes based on the real-time, and realizing autonomous operation with high precision, high efficiency and low energy consumption.

Inventors

WEI XINHUA
SUN LEI
WEI XIAOYANG
LIU CHENGLIANG
SONG QI
SUN YONG

Assignees

江苏大学
镇江金泰智能科技有限公司

Dates

Publication Date: 20260505
Application Date: 20260211

Claims (10)

1. An intelligent control method for a land leveling operation process is characterized by comprising the following steps: s1, collecting real-time state in a control period Internal acquisition planning path point elevation reference Land leveler position Speed of operation Displacement of hydraulic cylinder And earth-carrying lattice signal Forming a tape time index The real-time state set of the filter is filtered and preprocessed; S2, establishing coordinate mapping, namely obtaining arc length coordinates through path grabbing And establish arc length coordinates With planned route point elevation references The index relation of the control target is updated along with the path position without depending on measurement and control co-points; S3, constructing a forward-looking prediction window according to the operation speed And hydraulic actuation characteristic parameter 、 Determining the dynamic front window scale, and extracting a predicted target sequence in front of arc length coordinates Expanding the control target into future interval constraint; s4, rolling optimization and compensation control, wherein the control is based on a predicted target sequence Dynamic construction of future interval control quantity sequence by operation speed and hydraulic actuation characteristics On-line optimization of (c) and scroll-outputting current control instruction To implement feed-forward compensation for spatial misalignment due to hydraulic hysteresis; S5, hydraulic execution and closed-loop tracking, namely receiving a current control instruction by a hydraulic execution module And according to the current control instruction Action, using displacement feedback Closed loop tracking of desired trajectories to promote execution robustness; s6, carrying soil monitoring and mode switching, namely according to the carrying soil lattice signal Estimating soil loading rate And generates a job mode flag Work mode flag Indicating full load, switching to terrain-like/soil-unloading control and limiting continued soil entry, as a duty mode flag Returning to the leveling control when the load is not fully loaded, and circularly executing S1 to S6 to finish continuous automatic leveling operation.
2. The intelligent control method for land leveling operation process according to claim 1, wherein the step S1 is specifically as follows: Collecting planned route point elevation reference in control period k, namely sampling period is Ts Land leveler position Speed of operation Displacement of hydraulic cylinder And earth-carrying lattice signal ; In order to restrain GNSS transient shaking, hydraulic shaking burrs and lattice contact shaking, sliding median filtering is carried out on a grader position/operation speed/hydraulic cylinder displacement sequence/soil carrying rate sequence: Is provided with Representing any acquired original discrete data sequence, And representing a corresponding filtering output result, wherein the sliding median filtering calculation mode is as follows: Wherein, the The sliding window half width is represented for determining the number of samples involved in the median calculation. Through the sliding median filtering processing, the influence of single-point mutation and high-frequency noise on the state quantity can be effectively inhibited, and abnormal sampling values are prevented from directly participating in subsequent control and optimization calculation.
3. The intelligent control method for land leveling operation process according to claim 2, wherein the step S2 is specifically as follows: The planned path is composed of a sequence of discrete path points provided by an external system, the path points being noted as Represent the first Plane coordinates of the path points, where Indexing for path points; giving the target elevation value corresponding to each path point Characterizing a design elevation of the planned path at the location; in order to facilitate the subsequent indexing along the path position, performing arc length parameterization on the discrete path points, and defining the first path point by taking the path starting point as the arc length zero point Arc length coordinates corresponding to the path points are Wherein For a pair of The method comprises the following steps: In the formula, Representing Euclidean distance, thereby obtaining a path point arc length sequence Describing the along-path position of the planned path; During the control period At the moment, the grader position obtained according to S1 and subjected to filtering processing In the sequence of path points Searching for the closest path point index to it Wherein The method meets the following conditions: And the arc length coordinate corresponding to the path point is used as the arc length coordinate representation of the current position on the planning path and is recorded as Wherein, the method comprises the steps of, For characterising the period of control of the grader The lower path along position; in obtaining arc length coordinates Then, establishing an index relation between the arc length coordinates and the planned path target elevation reference, and correspondingly matching the arc length and the elevation of the discrete path points Regarding as a path target elevation reference table, calculating arc length position through interpolation operator Corresponding target elevation reference values, noted as : Wherein, the An interpolation operation is represented for obtaining a continuous target elevation reference between discrete path points.
4. The intelligent control method for land leveling operation process according to claim 3, wherein S3 is specifically: During the control period At the moment, the job speed has been obtained by S1 Obtaining hydraulic actuation characteristic parameters by system identification or calibration, wherein Indicating a net lag time for the hydraulic system to begin producing a significant response from the control command input to the grader blade displacement, The parameters are used for describing the time response characteristic of the hydraulic execution system; Construction of a forward looking time scale from hydraulic actuation characteristics To cover the time correction coefficients of different oil temperature, load and earth resistance conditions, the time scale is seen before The definition is as follows: Wherein, the The representation of the "effective time scale required for the grader to complete the main height change under the action of the hydraulic system from the current moment"; on the time scale of obtaining the front view After that, mapping the time scale to the space scale, namely setting To control period At the working speed, the forward looking distance The definition is as follows: Wherein, the Indicating the arc length of the path that the grader will continue to advance at the current speed before the hydraulic system generates effective height response, limiting the forward looking distance to ensure the engineering feasibility, setting the minimum and maximum allowable forward looking distances as And (3) with The forward looking distance after clipping is recorded as : Wherein, the Representing a clipping operator for constraining the forward looking distance to be always within an achievable range; Before obtaining the forward looking distance Then, the arc length coordinates of the current position obtained in S2 are used As a front window start point, a front view prediction window interval is constructed in front of arc length coordinates: Sampling step length according to preset arc length in the interval Discrete sampling is carried out on the paths to obtain a forward looking arc length sampling point sequence Wherein And is also provided with To meet the requirements of Is the largest integer of (2); arc length based sampling point sequence From the planned path to the elevation reference function Extracting corresponding predicted target elevation sequences : Wherein, the Is shown in the control period Lower, the first Target elevation reference values corresponding to the front view sampling positions.
5. The intelligent control method for land leveling operation process according to claim 4, wherein S4 is specifically: During the control period At the moment, the predicted target elevation sequence has been obtained by S3 Wherein Representing the arc length coordinates in the current position Is the starting point, within the forward looking prediction window Target elevation reference values corresponding to the arc length sampling positions are set for describing the dynamic response of the hydraulic execution system to the control command in the future interval Is shown in the control period Before the interval optimization is introduced, the current height of the grader shovel is compared with the elevation according to basic control logic and is lifted to the elevation position, or the non-optimized control instruction is obtained by the previous period control structure, Representing displacement by hydraulic cylinders And (3) mapping the equivalent height state of the land scraper, and expressing a discrete prediction model of the hydraulic actuation characteristic as: Then the first time period of the first time period, Wherein, the Representing the discrete number of steps of hysteresis resulting from the pure lag time conversion, And (3) with For discrete model parameters determined by the inertial time constant, The model is used for predicting the change trend of the height of the land leveling shovel in each step in the future under the action of a given control sequence; Based on the discrete predictive model, future The control instruction in each control period is an optimized variable, and an interval control quantity sequence is constructed: and output by prediction And corresponding to the predicted target elevation The interval error between the two is taken as a main optimization target, a control smoothing term is introduced simultaneously to restrict and control the change rate, and an interval optimization objective function is constructed : Wherein, the And (3) with The interval error weight and the control smoothing weight are respectively used for balancing the target tracking precision and the control feasibility; In the optimization process, physical feasible region constraint is applied to the control command to ensure that the optimization result can be reliably executed by an actual hydraulic system, wherein the constraint at least comprises control command amplitude constraint and control increment constraint, and the general form is as follows: Wherein, the And (3) with The minimum and maximum allowable values of the hydraulic valve control input are respectively, Maximum variation amplitude allowed for the control command; At each control period In the method, the section optimization problem is solved on line based on the current state information and the predicted target sequence, and an optimal control quantity sequence is obtained ; By adopting a rolling time domain control strategy, only the first control quantity in the optimal sequence is used as a control instruction of the current period to be output: The optimization control instruction For replacing non-optimised control instructions And reconstructing the prediction model and the optimization problem by using the updated state information in the next control period so as to form a continuous rolling interval optimization and compensation control process, and simultaneously Will be input to the subsequent hydraulic execution and closed loop tracking steps.
6. The intelligent control method for land leveling operation process according to claim 5, wherein S6 is specifically: During the control period At moment, the system acquires soil-loaded dot matrix signals in the land leveling shovel through S1 Wherein The discrete trigger state set output by the lattice switch sensor array is used for representing whether soil body contact exists at different space positions inside the flat shovel or not, and the matrix signal is applied to soil carrying lattice Counting closing points, wherein the counting of the dot matrix closing quantity is as follows Total point number of Thereby constructing soil loading rate : Based on soil loading rate Construction of operation mode flag Wherein the operation mode mark is used for representing the control mode of the current operation, and when the soil loading rate is estimated When exceeding the preset full load judgment threshold value, the operation mode mark Switching to full load mode when the soil load rate is estimated When the operation mode is lower than the corresponding release threshold value, the operation mode mark is displayed By introducing a hysteresis interval between full load and non-full load judging thresholds, frequent mode switching caused by fluctuation of the soil loading rate near a critical state is avoided; when the operation mode is marked When the indication is not full load mode, the system maintains a leveling control mode, generates a control instruction according to the rolling optimization result of S4, drives the grader to finish leveling operation, and works as a mode mark When the full load mode is indicated, the system is switched to a terrain-like or soil unloading control mode, the continuous soil feeding behavior is limited, and a control target is adjusted to enable the land leveling shovel to run along the set terrain or strategy in the soil unloading stage, so that traction resistance sudden increase, hydraulic load weighting and control performance degradation caused by continuous soil shoveling under a high load state are avoided; In full load mode, the system continuously monitors the soil load rate estimate When the soil loading rate decreases along with the soil unloading process and returns to the non-full load judging section, the operation mode mark The system is controlled to form closed loop coupling with the load state through the soil loading monitoring and mode switching mechanism, and the system can avoid ineffective soil feeding and overload operation while ensuring the leveling precision; by circularly executing S1 to S6, continuous automatic leveling operation of the grader under the complex terrain and load change conditions is realized.
7. The system for realizing the intelligent control method for the land leveling operation process according to claims 1-6 is characterized by comprising a data acquisition module, a path grabbing module, a front window module, a central processing module, a soil load monitoring module and a hydraulic execution module; The data acquisition module is respectively in signal connection with the central processing module and the soil carrying amount monitoring module, the front window module and the path grabbing module are respectively in signal connection with the central processing module, and the central processing module is in signal connection with the hydraulic execution module; The data acquisition module is used for acquiring planned path point position elevation reference, land leveler position, course angle, operation speed, hydraulic cylinder displacement and soil-carrying lattice signals in a control period; The path grabbing module is used for projecting the real-time position to a planned path, obtaining arc length coordinates and establishing a corresponding relation with path references according to the arc length coordinates; the front window module is used for adaptively determining the length of a front prediction interval according to the operation speed and the hydraulic actuation characteristic parameters, forming a dynamic front window in front of the arc length coordinate, extracting a prediction target sequence from the dynamic front window, and sending the prediction target sequence to the central processing module; the soil loading quantity monitoring module is used for acquiring a soil loading lattice signal and sending the soil loading lattice signal to the data acquisition module, and the central processing module acquires the soil loading lattice signal through the acquisition module and judges whether the soil loading quantity is subjected to mode switching or not; The hydraulic execution module is used for executing the control instruction sent by the central processing module; The soil loading amount monitoring module sends the soil loading lattice signals to the acquisition module, the acquisition module sends the acquisition data to the central processing module, the path grabbing module and the front window module acquire the data through the central processing module to process the data, the data are fed back to the central processing module, and the central processing module judges whether the soil loading amount is subjected to mode switching according to the soil loading lattice signals acquired by the acquisition module in real time.
8. The system of claim 7, wherein the data acquisition module comprises a dual antenna Beidou RTK, a vehicle-mounted IMU and a displacement sensor arranged at an oil cylinder, which are respectively connected with the central processing module in a signal manner.
9. The system of claim 7, wherein the soil load monitoring module comprises soil load sensors arranged on the inner side of the grader blade, the soil load sensors being arranged in a 3 x 5 lattice array and being uniformly distributed in a matrix along the width direction and the height direction of the grader blade.
10. The system of claim 7, wherein the hydraulic execution module comprises an unloading valve, a proportional reversing valve, a hydraulic pump and a hydraulic cylinder, the hydraulic pump continuously pumps oil from the oil tank in a running state when the grader operates, the proportional reversing valve opens oil, the proportional reversing valve controls the hydraulic cylinder to lift or descend by controlling the oil inlet direction, the unloading valve is closed, the unloading valve is opened when the current action of the hydraulic cylinder needs to be maintained, the proportional reversing valve is closed, hydraulic oil directly flows back to the oil tank, and the hydraulic cylinder is kept motionless.

Description

Intelligent control method and system for land leveling operation process Technical Field The invention relates to an intelligent control method and system for a land leveling operation process, and belongs to the technical field of intelligent farmland leveling equipment and control in high-standard farmland construction. Background In high-standard farmland construction, land flatness is an important index. The improvement of the flatness of the farmland is a primary condition for guaranteeing the management effect of paddy rice slurry and reducing the harm of weeds in paddy fields, and is an important means for improving the irrigation water utilization rate and the irrigation water uniformity of dry fields. The engineering construction makes clear demands on the field scale and the field height difference that the field area of the plain area should be more than 100 mu, the field area of the hilly and mountain area should be more than 30 mu, and the allowable height difference of paddy fields is required to be within 3 cm and the allowable height difference of dry fields is required to be within 5 cm. The existing farmland leveling equipment is mainly divided into land leveling equipment based on a laser technology and land leveling equipment based on a Global Navigation Satellite System (GNSS), the working flow of the farmland leveling equipment is usually in two parts of 'measurement-control', namely, the actual height of a land leveling shovel is obtained through laser target height measurement or RTK height measurement, the land leveling shovel is driven to lift by an output control quantity control hydraulic system to finish land leveling operation after being compared with a target elevation, and the advanced land leveling machines at home and abroad at present are already provided with a terrain drawing function, so that personnel operation can be guided according to a terrain map, and part of tractors can realize unmanned operation, namely path tracking under a planned path. However, the hydraulic control system of the existing land leveller still has outstanding problems in engineering application, namely on one hand, most of the hydraulic control system of the existing land leveller is on the other hand, the hydraulic control system generally adopts the same point of measurement and control, hydraulic action has hysteresis and inertia and the machine body continuously moves forward, so that the machine body is driven away from a target position when the control action is generated, action hysteresis-space dislocation is formed, undercutting or overcutting is caused, cm-level differential indexes are difficult to stably meet, in addition, the real-time monitoring of the soil carrying state of the land leveller is lacked in the leveling process, and after full loading, continuous soil shoveling can cause traction resistance increase, oil consumption increase, efficiency decrease and attitude disturbance induction, thereby further deteriorating elevation control effect. Disclosure of Invention The invention aims to provide an intelligent control method and system for a land leveling operation process based on the existing topographic map and path tracking technology, and aims to overcome the defects in the prior art. The technical scheme is that the intelligent control method for the land leveling operation process comprises the following steps: s1, collecting real-time state in a control period Internal acquisition planning path point elevation referenceLand leveler positionSpeed of operationDisplacement of hydraulic cylinderAnd earth-carrying lattice signalForming a tape time indexThe real-time state set of the filter is filtered and preprocessed; S2, establishing coordinate mapping, namely obtaining arc length coordinates through path grabbing And establish arc length coordinatesWith planned route point elevation referencesThe index relation of the control target is updated along with the path position without depending on measurement and control co-points; S3, constructing a forward-looking prediction window according to the operation speed And hydraulic actuation characteristic parameter、Determining the dynamic front window scale, and extracting a predicted target sequence in front of arc length coordinatesExpanding the control target into future interval constraint; s4, rolling optimization and compensation control, wherein the control is based on a predicted target sequence Dynamic construction of future interval control quantity sequence by operation speed and hydraulic actuation characteristicsOn-line optimization of (c) and scroll-outputting current control instructionTo implement feed-forward compensation for spatial misalignment due to hydraulic hysteresis; S5, hydraulic execution and closed-loop tracking, namely receiving a current control instruction by a hydraulic execution module And according to the current control instructionAction, using displacement feedbackClosed loop trackin