Search

CN-121973171-A - Industrial robot health abnormality precursor identification and operation and maintenance control method under silver-colored economic scene

CN121973171ACN 121973171 ACN121973171 ACN 121973171ACN-121973171-A

Abstract

The invention belongs to the technical field of smart city information, and particularly relates to a method for identifying and controlling abnormal health precursors of an industrial robot in a silver-colored economic scene. Covering the robot abnormal precursor recognition, scene demand adaptation and operation and maintenance closed-loop control full flow. The technical bottlenecks of large health, heavy faults, light precursors, coarse adaptation, no closed loop of the traditional industrial robot are broken through by adopting core algorithms such as abnormal precursor recognition, requirement layering adaptation, operation and maintenance closed loop control and the like, the precise recognition of the abnormal precursor of the robot, layering adaptation of the requirements of the silver-colored scene and the closed loop control of the operation and maintenance full flow are realized, the operation safety, the requirement adaptation precision and the operation and maintenance high efficiency of the robot in the silver-colored scene are obviously improved, and the fault occurrence rate and the operation and maintenance cost are reduced.

Inventors

  • SU PING

Assignees

  • 北京知筹汇智科技有限公司

Dates

Publication Date
20260505
Application Date
20260331

Claims (10)

  1. 1. The industrial robot health abnormality precursor identification and operation and maintenance control method under the silver-colored economic scene is characterized by comprising the following steps of: S1, abnormal precursor identification processing, namely acquiring operation time sequence data and abnormal associated data of a robot, and constructing an abnormal precursor identification model of the robot through an abnormal precursor feature extraction algorithm and a precursor identification precision calibration algorithm to realize accurate identification and identification precision calibration of the abnormal precursor; S2, silver hair demand layering adaptation processing, namely constructing a demand layering adaptation model based on precursor identification results and silver hair scene demand characteristics through a silver hair demand layering clustering algorithm and a robot running mode layering adaptation algorithm, and realizing layering accurate adaptation of the robot and the silver hair scene demands; s3, operation and maintenance closed loop control processing is carried out, precursor identification data, demand adaptation data and operation and maintenance full-flow data are integrated, an operation and maintenance closed loop control model is built through an operation and maintenance task layered scheduling algorithm and an operation and maintenance closed loop efficiency self-optimization algorithm, and closed loop control and efficiency optimization of the robot operation and maintenance full-flow are realized; The abnormal precursor recognition accuracy calibration algorithm in step S1 includes an accuracy accounting formula: The constraint is that , For the accuracy of the precursor recognition, For a properly identified number of abnormal precursors, In order to identify the number of errors, Intelligent training scene for identifying precursor total number and intelligent auxiliary silvery Community silvery care scene 。
  2. 2. The method according to claim 1, wherein the abnormal precursor feature extraction algorithm in the step S1 comprises the following substeps of fusing the operation time sequence data of the robot with the abnormality related data, constructing a feature extraction model by adopting an improved transducer model, capturing weak features and time sequence change rules of the abnormal precursor, realizing accurate extraction of the abnormal precursor features, and distinguishing precursor differences of different types of abnormalities.
  3. 3. The method according to claim 1, wherein the precursor recognition accuracy calibration algorithm in the step S1 includes the substeps of establishing a precursor recognition error evaluation system, combining the differentiated requirements of the silver-colored scenes on recognition accuracy, dynamically adjusting feature extraction weights and recognition thresholds, correcting the recognition accuracy, eliminating false recognition signals, and ensuring that the recognition results meet the control requirements of different silver-colored scenes.
  4. 4. The method according to claim 1, wherein the silver-colored demand hierarchical clustering algorithm in step S2 includes the substep of adopting an improved DBSCAN clustering algorithm, and combining the urgency, importance and adaptation difficulty of silver-colored scene demands to perform the layering (core demands, regular demands and auxiliary demands) on the demands, and determining the priority and adaptation standard of the demands of each layer.
  5. 5. The method according to claim 1, wherein the robot operation mode hierarchical adaptation algorithm in the step S2 comprises the following substeps of establishing a precursor identification result-requirement hierarchy-operation mode association model, and matching corresponding robot operation modes for different hierarchy requirements based on an abnormal precursor identification result and a requirement hierarchy result to realize hierarchical accurate adaptation of requirements.
  6. 6. The method of claim 1, wherein the operation and maintenance task layered scheduling algorithm in the step S3 includes the substeps of constructing an operation and maintenance task layered scheduling model, layering the operation and maintenance tasks in combination with severity of abnormal precursors and priority of demand levels, and dynamically distributing operation and maintenance resources by adopting an improved particle swarm algorithm to realize layered and accurate scheduling of the operation and maintenance tasks.
  7. 7. The method according to claim 1, wherein the operation and maintenance closed-loop performance self-optimization algorithm in step S3 includes the substeps of establishing an operation and maintenance closed-loop performance evaluation model, quantifying core indexes such as precursor identification accuracy, demand adaptation degree, operation and maintenance response speed, closed-loop completion rate, and the like, dynamically optimizing operation and maintenance closed-loop flow and scheduling strategies, and ensuring optimal operation and maintenance closed-loop performance.
  8. 8. The method according to any one of claims 1-7, wherein the technological parameters of abnormal precursor identification and operation and maintenance closed-loop control are precursor identification response time less than or equal to 7ms, required layered adaptation response time less than or equal to 9ms, operation and maintenance closed-loop response time less than or equal to 22ms, and adaptation of the intelligent auxiliary equipment for silver hair is intelligent, community silver hair care, household old care and other types of silver hair scenes.
  9. 9. The method according to any one of claims 1-7, wherein the method is applicable to 2-10 axis industrial robots, and is adapted to multiple types of robots such as smart assistive silver-coat production robots, community silver-coat caretaking robots, home-based senior-support robots, and the like, supporting single independent management and control and multiple clustered closed-loop operation and maintenance.
  10. 10. The industrial robot big health abnormal precursor identification and operation and maintenance closed-loop control system under the silver-colored economic scene is characterized by comprising an abnormal precursor identification module, a silver-colored demand layering adaptation module, an operation and maintenance closed-loop control module and a control center, wherein the control center is in bidirectional communication with three functional modules, and the method as claimed in any one of claims 1-9 is executed to realize the industrial robot big health abnormal precursor identification, scene demand layering adaptation and operation and maintenance closed-loop control.

Description

Industrial robot health abnormality precursor identification and operation and maintenance control method under silver-colored economic scene Technical Field The invention belongs to the technical field of information, and particularly relates to a method for identifying and controlling abnormal health precursors of an industrial robot in a silver-colored economic scene. Background The current silver hair economy is rapidly developed, the application of the industrial robot in the silver hair scene is more and more extensive, and the identification capability of abnormal precursors, the layering precision of demand adaptation and the closed-loop control level of operation and maintenance are directly related to the service quality of the silver hair scene, the safety of the old people and the operation cost. However, in the actual industrial application of industrial robot big health management and control and silver hair scene adaptation, 3 specific and not effectively solved real operability problems still exist for three big subdivision scenes of abnormal precursor recognition, requirement layering adaptation and operation and maintenance loop control, the real operability problems are all process specific pain points, and no repetition exists between the real operability problems and the existing and previous industrial robot big health related technologies (including hidden danger tracing, full life cycle prejudgement, personalized adaptation, light weight cooperation, degradation quantification and other directions), and the real operability problems are as follows: 1. The method has the advantages that the abnormal precursor is difficult to identify, no special precursor extraction and precision calibration algorithm exists, the prior art can only perform fault identification and maintenance after a robot generates a fault, weak abnormal precursors before the fault is generated cannot be identified, a special abnormal precursor identification model is not built, no precursor identification precision calibration algorithm exists, the identification precision cannot be corrected according to the identification requirements of different silver-colored scenes, misidentification and missing identification are easy to occur, and especially the situations that the intelligent silver-colored auxiliary tool has extremely high requirements on the abnormal precursor identification precision are not adapted, and product quality problems or potential safety hazards are easy to be caused. 2. The demand adaptation is not layered, a layering and mode adaptation algorithm is not required, the demand of the silver-colored scene has obvious layering difference (such as the situation that the middle-aged and old people take care of silver-colored hair, the aged emergency assistance is the core demand, the daily companion is the conventional demand, and the environment is clean as the auxiliary demand), the prior art adopts a unified demand adaptation mode, lacks a silver-colored demand layering clustering algorithm, cannot accurately layer the demand, and an inorganic robot operation mode layering adaptation algorithm cannot match the corresponding operation mode according to the demand layering, so that the core demand is not fully adapted, the auxiliary demand is excessively adapted, and the resource waste is serious. 3. The operation and maintenance is free of closed-loop management and control, the operation and maintenance is free of task layered scheduling and efficiency self-optimization algorithm, the operation and maintenance process is complex, the operation and maintenance process is involved in a plurality of links such as precursor identification, fault early warning, operation and maintenance execution, effect verification and the like, the prior art adopts a 'fragmentation' operation and maintenance mode, the operation and maintenance task is lack of the operation and maintenance task layered scheduling algorithm, the operation and maintenance tasks cannot be distributed according to the severity degree of abnormal precursors and the demand priority, the operation and maintenance closed-loop efficiency self-optimization algorithm is free of operation and maintenance closed-loop process cannot be optimized, the operation and maintenance process is disjointed, the response is slow, the efficiency is low, the closed-loop management and control of 'identification-early warning-operation and maintenance-verification' cannot be formed, and the operation and maintenance cost is high. The existing industrial robot big health related technology focuses on fault maintenance, static adaptation, rough synergy or degradation quantification, does not aim at the specificity of the silvery economy, forms core algorithm innovation in abnormal precursor identification, silvery demand layering adaptation and operation and maintenance closed loop management and control directions, has obvious technical blank especially in th