CN-121374586-B - Multi-working-condition car coupler uncoupling robot operation self-adaptive control method

CN121374586BCN 121374586 BCN121374586 BCN 121374586BCN-121374586-B

Abstract

The invention belongs to the technical field of coupler uncoupling, in particular to a method for adaptively controlling the operation of a coupler uncoupling robot under multiple working conditions, which comprises the steps of acquiring a coupler uncoupling original data set of a time dimension and an action stage dimension and performing standardized processing; the method comprises the steps of carrying out multi-mode analysis on a standardized data set, constructing an execution parameter feature set of track response parameters, force feedback parameters and visual identification parameters, integrating and aligning the execution parameter feature set based on time and action stage dimensions, generating an execution action feature set, comparing the execution action feature set with a preset ideal strategy parameter set to obtain a simulation deviation data set, generating a compensation strategy based on deviation degree and deviation direction data, adjusting the execution action feature set through parameter compensation to form an adjusted action parameter set, generating a driving instruction by utilizing the adjusted action parameter set, and driving a robot to execute unhooking actions according to update track, force control and visual following strategies, so that self-adaptive adjustment and control precision improvement under multiple working conditions are realized.

Inventors

YU DESHUI
GUO LEI
CHANG MIN
Bo Xueliang
YANG YONGCHAO
WANG RUI
LI BAO
LI YANZHENG
YI LI
LIU JIANGHUI
SONG LEI
YAN WEI
ZHANG DONG
ZHAO BIN
WANG JING
Liang Laiwang
ZHAO QIANG
ZHANG PENG
GAO QINGBAO
ZHENG YOUCHENG

Assignees

华电内蒙古能源有限公司土默特发电分公司

Dates

Publication Date: 20260508
Application Date: 20251114

Claims (9)

1. An adaptive control method for operation of a coupler uncoupling robot under multiple working conditions is characterized by comprising the following steps: Acquiring an original coupler uncoupling data set containing a time dimension and an action stage dimension, and carrying out standardized processing on the original data set to obtain a standardized data set; performing multi-mode analysis on the standardized data set to construct an execution parameter feature set, wherein the execution parameter feature set comprises a track response parameter, a force feedback parameter and a visual identification parameter; Integrating and aligning the track response parameter, the force feedback parameter and the visual recognition parameter in the execution parameter feature set based on the time dimension and the action stage dimension to generate an execution action feature set; Comparing the execution action characteristic set with a preset ideal strategy parameter set item by item, and obtaining a simulation deviation data set by calculating quantization differences of the track response parameter, the force feedback parameter and the visual identification parameter with corresponding ideal values; evaluating the simulation deviation data set, obtaining deviation degree data by calculating a quantization index of the overall deviation, and obtaining deviation direction data by identifying characteristic dimensions causing main deviation; Generating a compensation strategy for track response parameters, force feedback parameters or visual identification parameters based on the deviation degree data and the deviation direction data, and calculating an action parameter set corresponding to the execution action feature set generated by the compensation strategy to obtain an adjusted action parameter set; and generating a driving instruction by using the adjusted action parameter set, wherein the driving instruction is used for driving the robot to execute unhooking action according to the updated track, force control and visual following strategy.
2. The method for adaptively controlling operation of a coupler extraction robot under multiple working conditions according to claim 1, wherein the integrating and aligning the trajectory response parameter, the force feedback parameter, and the visual recognition parameter in the execution parameter feature set based on the time dimension and the action phase dimension, and generating an execution action feature set includes: Establishing a unified time axis synchronous with a clock of a robot control system; Performing time stamp matching and sampling rate unification on the force feedback parameters and the visual identification parameters by taking the time sequence of the track response parameters as a reference to generate a time-space synchronous multi-mode data sequence; assigning phase identifiers to the time-space synchronized multimodal data sequences based on predefined sequential action phases, the sequential action phases including a positioning approach phase, a knuckle probe phase, a force control unlocking phase, and a separation evacuation phase; And according to the stage identification, dividing the time-space synchronous multi-mode data sequence into data subsets corresponding to each stage, and carrying out feature extraction and combination on the data subsets of each stage to generate an execution action feature set.
3. The method for adaptively controlling operation of a coupler extraction robot under multiple working conditions according to claim 2, wherein the feature extraction and combination of the data subsets of each stage to generate the execution action feature set comprises: Extracting displacement change rate, speed change rate and acceleration characteristics of the track response parameters from the data subsets of each stage to obtain track response parameter characteristics; extracting moment change characteristics and contact force change characteristics of the force feedback parameters from the data subsets of each stage to obtain force feedback parameter characteristics; extracting key point position change characteristics and relative pose characteristics of the visual identification parameters from the data subsets of each stage to obtain visual identification parameter characteristics; combining the track response parameter characteristics, the force feedback parameter characteristics and the visual identification parameter characteristics in a stage to form a stage characteristic vector comprehensively representing the action execution characteristics of the stage; and integrating the phase feature vectors of each phase in sequence to generate the executing action feature set.
4. The method for adaptively controlling operation of a coupler-decoupling robot under multiple working conditions according to claim 3, wherein said comparing the execution characteristic set with a preset ideal policy parameter set item by item, and obtaining a simulated deviation data set by calculating quantization differences between the trajectory response parameter, the force feedback parameter, and the visual recognition parameter and corresponding ideal values, comprises: Indexing and sequencing the track response parameters, the force feedback parameters and the visual identification parameters in the execution action feature set according to the time dimension and the action stage dimension; Respectively calculating absolute difference values and relative difference values of the track response parameters corresponding to the ideal strategy parameter set in each action stage for the track response parameters to generate a track response parameter deviation subset; Respectively calculating absolute difference values and relative difference values of stress feedback parameters with an ideal strategy parameter set in each action stage for the force feedback parameters to generate a force feedback parameter deviation subset; For the visual identification parameters, respectively calculating absolute position difference values and pose difference values of the visual identification parameters corresponding to the ideal strategy parameter set in each action stage, and generating a visual identification parameter deviation subset; Integrating the track response parameter deviation subset, the force feedback parameter deviation subset and the visual identification parameter deviation subset according to the time dimension and the action stage dimension to generate a simulation deviation data set.
5. The method for adaptively controlling operation of a coupler extraction robot under multiple working conditions according to claim 4, wherein said evaluating the simulated deviation data set to obtain deviation degree data by calculating a quantization index of the overall deviation comprises: Based on the track response parameter deviation subset, the force feedback parameter deviation subset and the visual recognition parameter deviation subset, respectively calculating deviation distribution characteristics of each parameter in the time dimension and the action stage dimension, and obtaining local deviation indexes of each parameter dimension; Normalizing the local deviation index, and performing weighted aggregation according to the time dimension and the action stage dimension to generate an overall deviation evaluation matrix; and calculating a deviation synthesis quantized value based on the integral deviation evaluation matrix to represent the integral execution deviation degree of the robot in the multi-mode execution parameter space, so as to obtain deviation degree data.
6. The method for adaptively controlling operation of a coupler extraction robot under multiple working conditions according to claim 5, wherein the step of obtaining deviation direction data by identifying characteristic dimensions causing main deviations comprises the steps of: identifying parameter dimensions with significant deviation between the time dimension and the action stage dimension based on the simulation deviation dataset and the corresponding local deviation index thereof, wherein the parameter dimensions comprise specific components or key features of track response parameters, force feedback parameters and visual identification parameters; Analyzing the deviation directions of parameter dimensions marked as remarkable deviation in each action stage, wherein the deviation directions comprise displacement, speed and acceleration deviation of track response parameters, moment and contact force deviation of force feedback parameters and key point position and pose deviation of visual identification parameters; and integrating the parameter dimension with obvious deviation obtained by analysis and the corresponding deviation direction thereof to generate deviation direction data.
7. The method for adaptively controlling operation of a coupler extraction robot under multiple working conditions according to claim 6, wherein the generating a compensation strategy for the trajectory response parameter, the force feedback parameter, or the visual recognition parameter based on the deviation degree data and the deviation direction data comprises: performing a compensation strategy generation based on predefined sequential action phases including a positioning approach phase, a knuckle probing phase, a force controlled unlocking phase, and a separation evacuation phase; and performing joint adjustment on at least two of the track response parameter, the force feedback parameter and the visual recognition parameter based on the deviation direction data through the compensation function model to generate the compensation parameter set.
8. The method for adaptively controlling operation of a coupler extraction robot under multiple working conditions according to claim 7, wherein the calculating the action parameter set corresponding to the action feature set generated by the compensation strategy to obtain the adjusted action parameter set includes: Mapping the compensation strategy to each stage of feature vector of the execution action feature set stage by stage according to the time dimension and the action stage dimension, so that each stage of feature vector corresponds to the compensation parameters of the track response parameter, the force feedback parameter and the visual identification parameter; Performing numerical correction on the track response parameters, the force feedback parameters and the visual identification parameters in the feature vectors of each stage of the execution action feature set according to the mapped compensation parameters; and integrating the motion parameter subsets corrected by each stage according to the motion stage sequence to generate a complete motion parameter sequence with cross-stage continuity and time sequence consistency as the adjusted motion parameter set.
9. The method for adaptively controlling operation of a car coupler extraction robot under multiple working conditions according to claim 8, wherein the generating a driving instruction by using the adjusted action parameter set, the driving instruction is used for driving the robot to execute the extraction according to the updated track, force control and visual following strategy, and the method comprises: generating a driving instruction by using the adjusted action parameter set, and driving the robot to execute the unhooking action; Comparing the actual execution state data with the state expected by the adjusted action parameter set to generate real-time deviation; And carrying out real-time checksum correction on the driving instruction sequence based on the real-time deviation so as to ensure that the robot accurately executes the unhooking action according to the updated track, force control and vision following strategy.

Description

Multi-working-condition car coupler uncoupling robot operation self-adaptive control method Technical Field The invention relates to the technical field of coupler uncoupling, in particular to a self-adaptive control method for operation of a coupler uncoupling robot under multiple working conditions. Background The coupler is a key mechanical device for realizing train marshalling connection in railway vehicles or rail transit equipment, and has the core functions of finishing mechanical coupling and uncoupling operations among vehicles through precise matching of components such as a coupler head, a coupler knuckle, a coupler lock iron and the like, enabling the coupler knuckle to be rotationally embedded into the coupler head of the other side to form rigid connection when the coupler head is connected, absorbing impact force in running of a train through a buffer device, enabling the coupler knuckle to return through an operating handle or an external force source when the coupler head is uncoupling, and realizing vehicle separation. In the prior art, by integrating visual recognition, force control feedback and a model reference self-adaptive algorithm, the data of the coupler model, the position and the connection state are acquired in real time by utilizing a sensor, the dynamics, the speed and the track of the uncoupling action are automatically adjusted by combining a dynamic environment sensing technology, and the parameters of an executing mechanism can be dynamically corrected through a self-adaptive control law under the condition of different vehicle types, motion states and working conditions, so that accurate uncoupling is realized, and the operation stability is ensured. The above scheme has some problems in practical application, although the prior art can finish unhooking operation of the coupler, under the influence of the change of multiple working conditions such as train running speed, connection angle, track impact vibration, environment temperature and humidity, the response delay, unbalanced force control or track deviation easily occur when the robot executes unhooking operation, the deviations can cause the fact that the coupler cannot be accurately abutted or the operation is incompletely executed, when the robot continuously works, the operation deviation can accumulate, the unhooking success rate is reduced, the operation consistency is reduced, the smooth proceeding of subsequent train grouping or unhooking operation is influenced, the operation delay or interruption is caused, the grouping or unhooking operation of the subsequent train is directly influenced due to the instability of unhooking operation, and the operation sequence is disordered, the operation delay or interruption are caused, so that the efficiency and the reliability of the whole grouping operation are reduced. Therefore, the invention provides an operation self-adaptive control method of the coupler uncoupling robot under multiple working conditions. Disclosure of Invention The embodiment of the application provides a method for adaptively controlling the operation of a coupler uncoupling robot under multiple working conditions, which realizes that the device can keep the stability, consistency and accuracy of uncoupling actions under complex working conditions. In order to achieve the above purpose, the application adopts the following technical scheme: The embodiment of the application provides a method for adaptively controlling the operation of a coupler uncoupling robot under multiple working conditions, which comprises the following steps: Acquiring an original coupler uncoupling data set containing a time dimension and an action stage dimension, and carrying out standardized processing on the original data set to obtain a standardized data set; performing multi-mode analysis on the standardized data set to construct an execution parameter feature set, wherein the execution parameter feature set comprises a track response parameter, a force feedback parameter and a visual identification parameter; Integrating and aligning the track response parameter, the force feedback parameter and the visual recognition parameter in the execution parameter feature set based on the time dimension and the action stage dimension to generate an execution action feature set; Comparing the execution action characteristic set with a preset ideal strategy parameter set item by item, and obtaining a simulation deviation data set by calculating quantization differences of the track response parameter, the force feedback parameter and the visual identification parameter with corresponding ideal values; evaluating the simulation deviation data set, obtaining deviation degree data by calculating a quantization index of the overall deviation, and obtaining deviation direction data by identifying characteristic dimensions causing main deviation; Generating a compensation strategy for track response parameters, force feedback par