CN-121225339-B - Intelligent coal unloading control system and method for coal unloading machine

CN121225339BCN 121225339 BCN121225339 BCN 121225339BCN-121225339-B

Abstract

The invention relates to the technical field of intelligent control of coal unloading machines, in particular to an intelligent coal unloading machine coal unloading control system and method, an environment fusion sensing unit fuses coal pile point cloud data of a three-dimensional laser scanner and RGB-D camera texture information, a real-time digital twin model is constructed through hierarchical semantic segmentation, the surface morphology, the spatial distribution and the residual coal layer thickness of a cabin cleaning stage of a coal pile are mapped, an autonomous decision planning unit adopts a hierarchical optimization strategy, the coal unloading grid is divided by a global layer through a heuristic space segmentation algorithm, a grab bucket is planned, a spiral progressive cabin cleaning track of a self-adaptive cut-in point is generated by a local layer based on a coal pile curvature gradient, a dynamic posture adjustment unit is used for dynamically correcting the cut-in angle, the excavation depth and the swing amplitude of the grab bucket by collecting excavation resistance moment data and combining a coal pile thickness to build a coal quality hardness grading model, and a closed-loop vibration suppression control unit is used for generating vibration suppression signals by adopting an active damping controller of a Lyapunov stability theory to realize optimal energy consumption of the grab bucket.

Inventors

XU LEI
LIN YUEHUI
DONG ZHAOHUA
ZHI YI
Xia Xianjie
WANG LONG
YU JUNJUN
LIANG YUZHE
LIU JIA

Assignees

华能山东发电有限公司烟台发电厂

Dates

Publication Date: 20260508
Application Date: 20251107

Claims (9)

1. An intelligent coal unloader coal unloading control system, which is characterized by comprising: The environment fusion sensing unit (1) acquires three-dimensional point cloud data of a coal pile in a carriage in real time through a three-dimensional laser scanner, and adopts a three-dimensional point cloud reconstruction and semantic segmentation fusion algorithm to construct a real-time digital twin model of the carriage coal unloading operation environment, and maps the surface morphology, the spatial distribution and the thickness change of a residual coal layer in a cabin cleaning stage of the coal pile; The specific method for constructing the real-time digital twin model by the environment fusion sensing unit (1) comprises the following steps: Carrying out space-time alignment on point cloud data acquired by a three-dimensional laser scanner and texture information acquired by an RGB-D camera through a multi-source heterogeneous data space-time fusion mechanism, utilizing a hierarchical semantic segmentation strategy, firstly segmenting a carriage structure boundary through a point cloud density clustering algorithm, and then adopting a convolutional neural network to identify a coal pile and a residue semantic region, wherein a point cloud topology analysis module detects the surface concave-convex characteristics of the coal pile based on normal vector consistency, a spatial interpolation algorithm reconstructs geometrical morphology of the shielded region, a thickness inversion model dynamically calculates residual coal seam thickness according to a mapping relation between the point cloud reflection intensity and the coal quality density, and finally generating a real-time digital twin model through a dust interference compensation filter to provide environment topology representation for an autonomous decision planning unit (2); An autonomous decision planning unit (2) generates a global layer coal unloading sequence and a local layer cabin cleaning path through a layering optimization strategy based on the real-time digital twin model, wherein the global layer adopts a heuristic space segmentation algorithm to decompose a carriage into a coal unloading grid unit and plan grab bucket traversing priority, and a spiral progressive cabin cleaning track of a self-adaptive cut-in point is generated on the local layer based on coal pile curvature gradient analysis; The dynamic posture adjusting unit (3) analyzes the coal pile form change data in real time in the process of executing coal unloading operation of the grab bucket, and dynamically corrects the cutting angle, the excavation depth and the swing amplitude of the grab bucket through a mechanical feedback driving strategy; The closed-loop vibration suppression control unit (4) combines the grab bucket pose vibration spectrum acquired by the sensor, adopts an active damping controller based on Lyapunov stability theory to generate a vibration suppression compensation signal, and regulates and controls the acceleration curve of the steel wire rope hoisting mechanism through the variable frequency driver to realize stable tracking of the grab bucket running track and optimal energy consumption.
2. The intelligent coal unloading control system of the coal unloading machine according to claim 1, wherein the implementation mode of mapping the surface morphology, the spatial distribution and the thickness change of the residual coal layer in the cabin cleaning stage of the coal pile by the environment fusion sensing unit (1) is as follows: Aiming at the surface morphology mapping of the coal pile, a moving least square method is adopted to fit a point cloud curvature field, a coal pile fluctuation degree thermodynamic diagram characterized by Gaussian curvature is generated, a coal pile volume distribution topological grid is constructed on the basis of Dirony triangulation aiming at space distribution mapping, the coal pile aggregation degree is quantized by calculating the point cloud centroid offset in each grid unit, a high-frequency scanning mode is started in the cabin cleaning stage aiming at the thickness variation mapping of the residual coal layer in the cabin cleaning stage by combining a thickness inversion model, the elevation difference between a cabin bottom datum plane and the coal pile bottom surface is dynamically extracted by utilizing a multi-layer point cloud differential algorithm, and the real-time digital twin model is updated in real time by using the thickness contour cloud diagram.
3. The intelligent coal unloader coal unloading control system according to claim 1, wherein the specific implementation flow of the hierarchical optimization strategy of the autonomous decision planning unit (2) is as follows: at a global layer, decomposing a carriage into a plurality of coal unloading grid units by adopting a self-adaptive region growing algorithm based on a coal pile space distribution thermodynamic diagram, and generating a grab bucket traversal priority sequence by taking the volume density of the coal piles in the grid as a weight; And starting spiral progressive cabin cleaning track planning on a high-curvature area according to Gaussian curvature thermodynamic diagram at a local layer, wherein the spiral progressive cabin cleaning track planning comprises the steps of generating an equal-pitch spiral path along the gradient descending direction of a coal pile by taking curvature extreme points as initial access points, adaptively adjusting the spiral stepping distance through curvature radius, and triggering global sequence re-planning when the coal pile density of the local area is suddenly changed through cooperation of a dynamic priority arbiter at the global layer and the local layer.
4. The intelligent coal unloader coal unloading control system according to claim 3, wherein the heuristic space segmentation algorithm in the autonomous decision-making unit (2) specifically comprises: And a grid decomposition strategy of compartment space structure constraint is introduced, a compartment three-dimensional boundary frame is established according to compartment side wall point cloud data, and then optimal grid division granularity is calculated based on coal pile volume distribution entropy values, wherein a dynamic weight distribution module gives comprehensive weight coefficients to each grid unit, the coefficients are jointly determined by the effective volume of a coal pile in a grid, the Manhattan distance of the current position of a grab bucket and a coal pile stability factor, and finally an improved ant colony algorithm is adopted, and the shortest path coverage sequence of the grab bucket is generated by taking the comprehensive weight coefficients as heuristic information.
5. The intelligent coal unloading control system of the coal unloader, according to claim 3, wherein the spiral progressive cleaning track of the autonomous decision planning unit (2) specifically comprises: The curvature driving type track generator is designed based on a coal pile curvature gradient field, a small-pitch high-density spiral line is adopted in a high-curvature area to avoid collapse risk, a large-pitch spiral line is adopted in a low-curvature area to accelerate coverage, a cutting-in angle optimization module dynamically adjusts an initial cutting angle of a grab bucket according to a curvature direction vector, so that a grab bucket cutting edge is always perpendicular to a coal pile cutting plane, and a track smoothing module fits discrete spiral point columns through Bezier curves to generate continuous guided grab bucket running tracks.
6. The intelligent coal unloading control system of the coal unloading machine according to claim 1, wherein the mechanical feedback driving strategy of the dynamic posture adjusting unit (3) specifically comprises the following steps: The six-dimensional force sensor embedded in the hinged position of the grab bucket is used for collecting excavation resistance moment data in real time, a coal hardness grading model is built by combining the coal pile thickness distribution diagram, when the resistance moment exceeds a threshold value, the angle adjustment module increases the front inclination angle of the grab bucket according to the curvature direction of the current cutting point to crush hard coal cores, the depth controller dynamically limits the excavation depth to avoid overload based on the linear relation between the inversion result of the coal pile thickness and the resistance moment, and the swing strategy generator triggers a resonant excavation mode to increase the swing amplitude to loosen a coal seam when the periodic fluctuation of the resistance moment is detected, and restricts the swing amplitude boundary through the spiral track curvature data.
7. The intelligent coal unloader coal unloading control system according to claim 1, wherein the active damping controller of the closed-loop vibration suppression control unit (4) comprises: A Lyapunov stability theory and online spectrum analysis fusion framework is adopted, wavelet packet decomposition is carried out through vibration signals acquired by an IMU sensor, longitudinal vibration main frequency and swing angle transverse vibration main frequency of a steel wire rope are extracted, a self-adaptive observer builds a state space equation based on a double-frequency coupling vibration model, a Lyapunov energy function constructor takes the sum of vibration kinetic energy and potential energy as a reference, a compensation moment which enables the derivative of an energy function to be negatively fixed is generated, and the compensation moment is superimposed to a torque command of a variable frequency driver after passing through a bandwidth limiting filter, so that exponential convergence of vibration energy is realized.
8. The intelligent coal unloading control system of the coal unloading machine according to claim 7, wherein the energy consumption optimization implementation mode of the closed-loop vibration suppression control unit (4) is as follows: The variable frequency driver adopts an acceleration curve smooth optimization algorithm to decompose vibration suppression compensation signals into cooperative control instructions of a steel wire rope winch motor and a rotary motor, dynamically adjusts acceleration torque slope according to grab potential energy change rate aiming at a winch mechanism, enables potential energy recovery mode at a descending section, predicts centripetal force change aiming at the rotary mechanism based on spiral track curvature data, compensates centrifugal vibration in advance, and finally coordinates double-motor working points through a power equalizer.
9. A method for implementing a coal discharge control system comprising an intelligent coal discharge machine according to any one of claims 1-8, characterized by the steps of: S1, acquiring point cloud data of a carriage coal pile in real time through a three-dimensional laser scanner, fusing RGB-D camera texture information, constructing a real-time digital twin model by adopting a hierarchical semantic segmentation strategy, and dynamically mapping the surface morphology, the spatial distribution and the residual coal layer thickness change in a cabin cleaning stage of the coal pile; S2, generating a global coal unloading sequence and a local cabin cleaning path by adopting a hierarchical optimization strategy based on the real-time digital twin model, wherein a global layer decomposes a carriage into a coal unloading grid unit and plans grab bucket traversing priority by a heuristic space segmentation algorithm, and a local layer generates a spiral progressive cabin cleaning track of a self-adaptive access point based on coal pile curvature gradient analysis; s3, analyzing the coal pile form change data in real time in the process of executing coal unloading operation of the grab bucket, dynamically correcting the grab bucket cutting angle, the excavation depth and the swing amplitude through a mechanical feedback driving strategy, and adaptively adjusting the action parameters according to a coal hardness grading model; S4, combining the grab bucket pose vibration spectrum, generating a vibration suppression compensation signal by adopting an active damping controller based on Lyapunov stability theory, and regulating and controlling an acceleration curve of the steel wire rope hoisting mechanism through a variable frequency driver to realize stable tracking of the grab bucket running track and optimal energy consumption.

Description

Intelligent coal unloading control system and method for coal unloading machine Technical Field The invention relates to the technical field of intelligent control of coal unloading machines, in particular to an intelligent coal unloading machine coal unloading control system and method. Background The intelligent control of the coal unloader is an important technology, is particularly applied to a grab bucket operation accurate control link of the coal unloader in a coal loading and unloading scene, and has the core that high efficiency, safety and energy conservation of the coal unloading operation are realized through dynamic sensing and intelligent decision, high requirements of large-scale coal transportation on the coal unloading efficiency and equipment safety are adapted, the current coal unloader depends on a preset fixed track or manual operation, and dynamic change of the coal pile form in a carriage is difficult to adapt, so that the operation efficiency and the safety are limited; The existing coal unloader control system cannot realize accurate sensing of dynamic characteristics of a coal pile, intelligent adaptation of grab bucket actions and integrated control of cooperative optimization of vibration and energy consumption, so that coal unloading efficiency is low, safety risk is high, energy consumption is overlarge, a traditional system lacks real-time accurate mapping of concave-convex shapes and spatial distribution of the surface of the coal pile and residual coal seam thickness in a cabin cleaning stage, grab bucket operation is driven only according to a fixed track, collapse or skip unloading of the coal pile is easily caused by improper cutting-in angles and uncontrolled excavation depth, vibration of the grab bucket due to mechanical inertia and reaction force of the coal pile in operation is not effectively restrained, the position and posture deviation of the grab bucket influence operation precision, abrasion of steel wire ropes and energy consumption are aggravated, meanwhile, operation suitability is further reduced due to the fact that the limitation makes the coal unloader difficult to cope with complex and changeable coal unloading working conditions, efficiency, safety and energy-saving requirements are not considered, and in order to solve the problem, the intelligent coal unloader control system and the method are provided. Disclosure of Invention The invention aims to provide an intelligent coal unloading control system and method for a coal unloader, which are used for solving the problems in the background technology. In order to achieve the above object, there is provided an intelligent coal unloading control system for a coal unloader, comprising: the environment fusion sensing unit acquires three-dimensional point cloud data of a coal pile in a carriage in real time through a three-dimensional laser scanner, a real-time digital twin model of a carriage coal unloading operation environment is constructed by adopting a three-dimensional point cloud reconstruction and semantic segmentation fusion algorithm, and the surface morphology, the spatial distribution and the thickness change of a residual coal layer in a cabin cleaning stage of the coal pile are mapped; The autonomous decision planning unit generates a global layer coal unloading sequence and a local layer cabin cleaning path through a layering optimization strategy based on the real-time digital twin model, wherein the global layer adopts a heuristic space segmentation algorithm to decompose a carriage into a coal unloading grid unit and plan grab bucket traversing priority, and a spiral progressive cabin cleaning track of a self-adaptive cut-in point is generated on the local layer based on coal pile curvature gradient analysis; The dynamic posture adjusting unit analyzes the coal pile form change data in real time in the process of executing coal unloading operation of the grab bucket, and dynamically corrects the cutting angle, the excavation depth and the swing amplitude of the grab bucket through a mechanical feedback driving strategy; The closed-loop vibration suppression control unit combines the grab bucket pose vibration spectrum acquired by the sensor, adopts an active damping controller based on Lyapunov stability theory to generate a vibration suppression compensation signal, and regulates and controls the acceleration curve of the steel wire rope hoisting mechanism through the variable frequency driver to realize stable tracking of the grab bucket running track and optimal energy consumption. The second object of the present invention is to provide a method for implementing the intelligent coal unloading control system of the coal unloading machine, which comprises the following steps: S1, acquiring point cloud data of a carriage coal pile in real time through a three-dimensional laser scanner, fusing RGB-D camera texture information, constructing a real-time digital twin model by adopting a hie