Search

CN-121994411-A - Motor rotor dynamic balance detection system and detection method

CN121994411ACN 121994411 ACN121994411 ACN 121994411ACN-121994411-A

Abstract

The invention discloses a motor rotor dynamic balance detection system and method, the method comprises multidimensional sensing deployment and full-working-condition signal redundancy acquisition pretreatment, a multi-working-condition phase-locking reference is built by fusion signals, environmental error compensation and state marking are carried out, multi-source interference is separated based on the reference extraction characteristics, high confidence coefficient data is obtained, a counterweight is calculated by combining working conditions and the data, intelligent optimization is carried out to generate a target instruction, the instruction is automatically executed, the verification and multi-working-condition rechecking are carried out, cloud archiving data, self-learning and full-flow fault tolerance are carried out on a model, and detection results are output and related to motor full-life-cycle data. The invention improves the suitability of complex working conditions and the robustness of the system, realizes the automation and the intellectualization of detection, can support the full life cycle operation and maintenance of the motor by using detection data, and is suitable for the mass production and the on-site operation and maintenance detection of the motor.

Inventors

  • WANG JIANHUI

Assignees

  • 湖南迈克尔实验仪器有限公司

Dates

Publication Date
20260508
Application Date
20260324

Claims (10)

  1. 1. The motor rotor dynamic balance detection method is characterized by comprising the following steps of: S1, carrying out multidimensional sensing deployment on a motor rotor to complete multi-type signal redundancy acquisition and pretreatment under all working conditions; S2, establishing a multi-working-condition phase-locked reference by fusing the acquired multi-dimensional signals, and completing error compensation and state marking of the reference by combining environmental parameters; s3, carrying out feature extraction on the signal based on the compensated phase-locked reference, and simultaneously separating multi-source interference to obtain high-confidence feature data; s4, combining working conditions and high confidence characteristic data to complete weight calculation, and optimizing the weight calculation through an intelligent optimizing algorithm to generate a target weight instruction; s5, automatically executing a target counterweight instruction, checking a correction result, carrying out dynamic balance rechecking on the motor rotor under multiple working conditions, and carrying out iterative correction if the dynamic balance rechecking is not up to standard; S6, uploading the detection whole-flow data to a cloud to finish archiving and model self-learning, and simultaneously implementing full-link fault-tolerant processing on the detection whole-flow data; And S7, outputting a motor rotor dynamic balance detection result, and correlating the detection data with the full life cycle data of the motor.
  2. 2. The method for detecting dynamic balance of a motor rotor according to claim 1, wherein in the step S1, the multi-dimensional sensing arrangement is used for constructing a measuring point network of a core measuring point, an auxiliary measuring point and a redundant measuring point, each measuring point is provided with a multi-axis sensing element, corresponding sensing elements are respectively configured at a motor rotating shaft, a load end and an environment end, the multi-type signals comprise vibration signals, angular position signals, rotating speed signals, torque load signals, environment signals and current signals, the all-condition comprises a constant speed section, a variable speed sweeping frequency section, a variable load section and a start-stop section, and the preprocessing comprises electromagnetic shielding, filtering denoising and signal validity marking.
  3. 3. The method for detecting dynamic balance of a motor rotor according to claim 1, wherein in the step S2, a multi-working condition phase locking reference is established by fused multi-dimensional signals, namely, a rotating speed and a load signal are extracted, the change rate of the rotating speed and the load signal is used as a phase locking correction factor, a multi-order mechanical frequency parallel tracking channel is established to complete phase locking, a basic phase reference is formed after debouncing treatment, the error compensation is to call a preset environment error compensation model, and the basic phase reference is corrected by combining temperature, humidity and dust environment parameters, and the state mark comprises quality scoring of the reference, and state marks of locking, tracking and losing.
  4. 4. The motor rotor dynamic balance detection method according to claim 1 is characterized in that in step S3, an equal-angle self-adaptive sampling sequence is firstly constructed to resample an angle domain of a signal, then a lightweight decomposition algorithm is adopted to decompose the resampled signal to extract a characteristic complex amplitude value, the separation of multi-source interference is to remove the influence of electromagnetic, load and environment interference on characteristic data by taking current, load and environment related signals as references, and the high-confidence characteristic data is obtained through cross-measuring point consistency verification and then marking.
  5. 5. The method for detecting dynamic balance of a motor rotor according to claim 1 is characterized in that in step S4, the working condition is integrated with rotating speed, load, environmental parameters and measuring point position information to form a multi-dimensional vector, the balance weight calculation is achieved through a multi-objective intelligent optimizing model, the model loads physical constraint conditions of manufacturing and assembling and finishes manufacturability processing to generate an initial balance weight calculation, the intelligent optimizing algorithm is used for checking balance weight effects through virtual correction and pre-modeling, and model parameters are optimized in an iterative mode when a threshold value is not reached, and model parameters of the same-batch or same-model motors in a cloud are multiplexed at the same time, so that the calculation efficiency is improved.
  6. 6. The method for detecting dynamic balance of a motor rotor according to claim 1, wherein in the step S5, the automatic execution of the target weight command is performed by converting the weight command into a numerical control operation parameter through linkage of a numerical control device and a rotor positioning system, automatic weight removal or weight increase operation is performed, verification of the correction result is performed by a quality detection sensing element to accurately verify the correction position and the quality value, correction is performed again if verification is not passed, the multi-condition re-detection comprises a target condition, a limit condition and a variable load condition, and re-detection data is fed back to a weight resolving link to perform rapid iterative correction if re-detection fails.
  7. 7. The motor rotor dynamic balance detection method according to claim 1 is characterized in that in the step S6, the cloud side files the detected working condition labels, target weight instructions, rechecking results and sensing data into a motor dynamic balance large database, the model self-learning is to conduct clustering analysis on cloud side same-batch same-model motor data and optimize model parameters for weight calculation through big data training, and the full-link fault tolerance processing comprises redundant signal switching during signal distortion, rapid recovery during phase reference loss, breakpoint continuous measurement during flow interruption and parameter compensation during deviation correction.
  8. 8. The method for detecting the dynamic balance of the motor rotor according to claim 1, wherein in the step S7, the detection result is output in a detection report form, the report comprises measurement point distribution, working condition coverage, vibration indexes, counterweight instructions and retest result information, the whole life cycle data of the motor comprise factory detection data, field operation data and maintenance data, and a digital twin body of the motor is constructed through data association to provide data support for monitoring the dynamic balance state and fault early warning of the motor.
  9. 9. The motor rotor dynamic balance detection system for realizing the motor rotor dynamic balance detection method according to any one of claims 1-8 is characterized by comprising a multi-dimensional sensing acquisition module, a multiplexing Kuang Suoxiang reference module, an intelligent characteristic resolving module, an adaptive counterweight optimizing module, an automatic correction executing module, a cloud data cooperative module and a full-link fault-tolerant module, wherein each module is networked through an industrial bus or a wireless communication mode, and a software layer adopts a cooperative mode of real-time edge-end resolving and cloud cooperative optimization.
  10. 10. The motor rotor dynamic balance detection system according to claim 9, wherein the multi-dimensional sensing acquisition module is used for realizing redundancy acquisition and pretreatment of all measuring points and multiple signals; The multi-working condition phase-locked reference module receives the signals of the sensing acquisition module, generates a load-rotating speed bi-dimensional anti-jitter phase reference and completes error compensation; The intelligent feature resolving module completes signal feature extraction and multisource disturbance suppression separation based on a phase reference and outputs high-confidence feature data; the self-adaptive counterweight optimization module completes counterweight calculation and quick optimization through a multi-objective intelligent optimizing model, and generates objective counterweight instructions; the automatic correction execution module executes the counterweight instruction, finishes correction result verification and multi-working condition rechecking, and triggers iterative correction if the correction result is not up to standard; the cloud data collaboration module is used for realizing cloud archiving and model self-learning and parameter multiplexing of detection data; The full-link fault-tolerant module provides the functions of signal exception handling, flow breakpoint continuous measurement, reference quick recovery and deviation compensation for each module, and ensures the stable operation of the whole flow of the system.

Description

Motor rotor dynamic balance detection system and detection method Technical Field The invention relates to the technical field of motor testing and dynamic balance, in particular to a motor rotor dynamic balance detection system and a motor rotor dynamic balance detection method, which are suitable for mass production detection and on-site operation and maintenance detection of various motors and can cover motor rotor dynamic balance detection and correction in constant speed, variable load and extreme environments. Background In the process of manufacturing and assembling the motor, the dynamic balance performance of the rotor is directly related to the vibration level, the running efficiency and the service life of the motor. The existing motor rotor dynamic balance test method generally depends on a test weight method or a vibration measurement mode under a single steady-state working condition, and the unbalanced state is judged by adding a test weight on the surface of a rotor and gradually adjusting the position and the quality of the test weight. Although dynamic balance correction can be realized under laboratory conditions in the traditional method, the testing process is long in time consumption, relies on manual experience, and is difficult to meet the efficiency requirement of mass production. With the development of the motor to high speed and high power, the traditional dynamic balance method gradually shows defects in the aspects of test precision and stability. The existing motor rotor dynamic balance detection technology has the problems of narrow working condition adaptability, low model resolving and parameter optimizing efficiency, long single detection period during mass production, weak system anti-interference and fault tolerance capability, signal distortion or flow interruption needing full-flow retest, low degree of automation of physical correction and detection closed loop, high manual dependency and difficult detection consistency, and can only cover constant speed or simple variable speed no-load working conditions, and single easy-missing key vibration signal is arranged at the measuring points. In order to solve the above-mentioned drawbacks, a detection system and method that is adaptive to complex working conditions, high in robustness and high in efficiency is needed, and the dual requirements of mass production detection and industrial field operation and maintenance detection of the production vehicle are considered. Disclosure of Invention The invention aims to provide a motor rotor dynamic balance detection system and a detection method, which are used for solving the technical problems of narrow working condition adaptation, low resolving efficiency, weak anti-interference fault tolerance and low degree of automation in the traditional detection technology. The technical scheme for solving the technical problems is as follows: A motor rotor dynamic balance detection method comprises the following steps: S1, carrying out multidimensional sensing deployment on a motor rotor to complete multi-type signal redundancy acquisition and pretreatment under all working conditions; S2, establishing a multi-working-condition phase-locked reference by fusing the acquired multi-dimensional signals, and completing error compensation and state marking of the reference by combining environmental parameters; s3, carrying out feature extraction on the signal based on the compensated phase-locked reference, and simultaneously separating multi-source interference to obtain high-confidence feature data; s4, combining working conditions and high confidence characteristic data to complete weight calculation, and optimizing the weight calculation through an intelligent optimizing algorithm to generate a target weight instruction; s5, automatically executing a target counterweight instruction, checking a correction result, carrying out dynamic balance rechecking on the motor rotor under multiple working conditions, and carrying out iterative correction if the dynamic balance rechecking is not up to standard; S6, uploading the detection whole-flow data to a cloud to finish archiving and model self-learning, and simultaneously implementing full-link fault-tolerant processing on the detection whole-flow data; And S7, outputting a motor rotor dynamic balance detection result, and correlating the detection data with the full life cycle data of the motor. In a further scheme, in step S1, the multi-dimensional sensing is deployed to construct a measuring point network of a core measuring point, an auxiliary measuring point and a redundant measuring point, multi-axis sensing elements are arranged at each measuring point, corresponding sensing elements are respectively configured at a motor rotating shaft, a load end and an environment end, the multi-type signals comprise vibration signals, angular position signals, rotating speed signals, torque load signals, environment signals and current signals, th