Search

CN-121989556-A - State monitoring and maintaining system of piezoelectric spray head

CN121989556ACN 121989556 ACN121989556 ACN 121989556ACN-121989556-A

Abstract

The application provides a state monitoring and maintaining system of a piezoelectric spray head, which comprises an acoustic acquisition unit, a signal processing diagnosis unit and a maintenance execution unit, wherein the acoustic acquisition unit is used for acquiring acoustic signals generated by piezoelectric ceramic vibration when the piezoelectric spray head works in real time, the signal processing diagnosis unit is used for carrying out filtering, amplifying and analog-to-digital conversion processing on the acoustic signals, the processed acoustic signals are converted from a time domain to a frequency domain to extract real-time frequency spectrum characteristics, the real-time frequency spectrum characteristics are compared with characteristics in a reference voiceprint characteristic library, a health state diagnosis result of the piezoelectric spray head is output, and the maintenance execution unit is used for executing corresponding maintenance operation according to the health state diagnosis result. The application can sense the internal health state of the piezoelectric spray head in real time in an online and nondestructive manner with low cost, accurately predict faults, trigger intelligent maintenance as required, realize active prediction and ensure high stability, high yield and high continuity of the piezoelectric spray printing process.

Inventors

  • ZHANG QINGWEI

Assignees

  • 歌尔微电子股份有限公司

Dates

Publication Date
20260508
Application Date
20251208

Claims (10)

  1. 1. The state monitoring and maintaining system of the piezoelectric spray head is characterized by comprising an acoustic acquisition unit, a signal processing diagnosis unit and a maintenance execution unit; The signal processing diagnosis unit is respectively connected with the acoustic acquisition unit and the signal processing diagnosis unit, the acoustic acquisition unit is integrated on the piezoelectric spray head, and the signal processing diagnosis unit prestores a reference voiceprint feature library of the piezoelectric spray head under different health states; The acoustic acquisition unit is used for acquiring acoustic signals generated by piezoelectric ceramic vibration when the piezoelectric spray head works in real time; The signal processing diagnosis unit is used for carrying out filtering, amplifying and analog-to-digital conversion processing on the acoustic signals, converting the processed acoustic signals from a time domain to a frequency domain to extract real-time spectrum features, comparing the real-time spectrum features with features in the reference voiceprint feature library, and outputting a health state diagnosis result of the piezoelectric shower nozzle; and the maintenance execution unit is used for executing corresponding maintenance operation according to the health state diagnosis result.
  2. 2. The system according to claim 1, wherein the library of reference voiceprint features comprises reference spectral features corresponding to a normal state, a partially blocked state, a fully blocked state, a bubble-in-cavity state, and a degraded piezoelectric actuator performance state of the piezojet; The real-time spectrum features comprise at least one of energy value, spectrum peak value, frequency corresponding to the peak value, spectrum centroid and spectrum envelope shape of the specific frequency band.
  3. 3. The system according to claim 1, wherein the signal processing and diagnosing unit comprises a feature extraction module and an intelligent diagnosing module; the characteristic extraction module is used for carrying out frequency domain conversion on the processed acoustic signals and calculating and acquiring the real-time frequency spectrum characteristics; the intelligent diagnosis module is internally provided with a trained machine learning model, and the machine learning model is used for mapping the real-time frequency spectrum characteristics into the corresponding health states of the piezoelectric spray heads and generating the health state diagnosis results.
  4. 4. The system according to claim 1, wherein the maintenance execution unit comprises at least one of a negative pressure suction module, an ultrasonic cleaning module, and a cleaning liquid flushing module; the negative pressure suction module is used for applying negative pressure to the interior of the piezoelectric spray head to suck the blockage when the health state diagnosis result is a partial blockage state; the ultrasonic cleaning module is used for applying ultrasonic vibration to the piezoelectric nozzle to remove attachments when the health state diagnosis result shows that attachments exist on the inner wall of the nozzle; And the cleaning liquid flushing module is used for pumping cleaning liquid into the piezoelectric spray head for flushing when the result of the health state diagnosis is that the blockage or the attachment is difficult to clear by single negative pressure or ultrasonic wave.
  5. 5. The system for monitoring and maintaining the state of the piezojet according to claim 1 or 3, wherein the signal processing and diagnosing unit further comprises an early warning module; and the early warning module is used for generating and outputting early warning information when the health state diagnosis result is a hidden fault state.
  6. 6. The system of claim 5, wherein the signal processing and diagnosing unit further comprises a data recording module; The data recording module is used for recording the health state diagnosis result, the acoustic signal original data and the real-time frequency spectrum characteristics of the piezoelectric spray head in real time to form a historical data file for predicting the health state trend of the piezoelectric spray head.
  7. 7. The system according to claim 1 or 3, wherein the signal processing and diagnosing unit further comprises an environmental noise calibration module, the environmental noise calibration module is used for acquiring environmental noise signals of the piezojet in a non-working state of the piezojet, and noise reduction compensation is performed on acoustic signals of the piezojet, acquired by the acoustic acquisition unit, during working according to the environmental noise signals.
  8. 8. The system of claim 1, wherein the acoustic acquisition unit is integrated on the piezojet by a detachable fixing structure, the detachable fixing structure comprises an elastic clamping seat and a buffer pad, and the buffer pad is used for reducing interference of environmental vibration except for vibration of the piezojet on acoustic signals acquired by the miniature microphone.
  9. 9. The system of claim 6, further comprising a man-machine interaction unit electrically connected to the signal processing diagnostic unit for displaying real-time health status, historical data files and pre-warning information of the piezojet and supporting an operator to manually trigger or suspend maintenance operations of the maintenance execution unit.
  10. 10. The piezojet state monitoring and maintenance system of claim 3, wherein the training process of the machine learning model comprises: And acquiring a plurality of groups of acoustic sample signals of the piezoelectric spray head under different health states, preprocessing and extracting features of the acoustic sample signals to construct a training data set, and performing model training and parameter optimization on the training data set by adopting a random forest, a support vector machine or a convolutional neural network algorithm until the recognition accuracy of the model on the health states of the piezoelectric spray head reaches a preset threshold value.

Description

State monitoring and maintaining system of piezoelectric spray head Technical Field The application relates to the technical field of precision spray printing manufacturing equipment, in particular to a state monitoring and maintaining system of a piezoelectric spray head. Background Piezoelectric jet printing technology, which is one of the core technologies in the current precision manufacturing field, is widely applied to the leading edge fields of flexible electronic printing, biological cell printing, three-dimensional packaging, micro-nano structure manufacturing and the like by virtue of the unique advantages of non-contact liquid drop ejection. The technology controls the volume, the speed and the drop point of liquid drops through the accurate deformation of piezoelectric ceramics, the printing precision can reach the micrometer and even the nanometer level, and the technology can be adapted to various special media such as silver nanometer ink, bioactive materials, functional resins and the like, and becomes a key support for pushing high-end manufacture to upgrade to high precision, customization and low loss. In the actual production process, the piezoelectric spray head is used as a core execution component, and the health state of the piezoelectric spray head directly determines the stability of printing quality and the product yield. However, three core pain points still exist in the monitoring and maintenance means of the state of the spray head in the current industry, and the continuous requirement of high-precision manufacturing is difficult to meet: First, failure discovery is severely delayed, resulting in a large amount of scrap. In the prior art, the manual inspection and the passive mode of finished product spot inspection are relied on, and operators need to stop periodically to observe the printing effect, or detect defects caused by the problem of a spray head, such as broken lines, leakage points and liquid drop offset, through subsequent detection. However, the degradation of the nozzle performance (such as partial nozzle blockage, micro bubbles in the cavity, and aging of piezoelectric ceramics) is a progressive process, and when faults appear on the finished product, tens or even hundreds of unqualified products are always continuously produced, and particularly, obvious economic losses are caused in the scenes of expensive materials, such as biological printing, flexible electronics and the like, and complex manufacturing processes. Secondly, the state monitoring has a blind area, and a high-cost scheme is difficult to popularize. Part of high-end equipment adopts a machine vision system to monitor the liquid drop spraying process, and captures the liquid drop form through a high-speed camera to judge the state of the spray head, but the scheme can only observe the liquid drop behavior outside the spray head, and can not detect the key hidden trouble inside the spray head, such as the hidden problems of the adhesion of the inner wall material of a cavity, the energy attenuation of a piezoelectric actuator, tiny bubbles in a flow channel and the like, and the problems can be gradually deteriorated and finally cause dominant faults. Meanwhile, the vision system is required to be matched with a high-resolution camera, a special light source and a complex image algorithm, the cost of a single set can reach tens of thousands of yuan, and the vision system is difficult to comprehensively deploy in a multi-station production line, so that a plurality of stations are still in a non-monitoring state. Thirdly, the hidden fault early warning capability is lacking, and the maintenance mode is passive and inefficient. The main stream timing maintenance strategy in the current industry is to stop, disassemble and clean according to a fixed period (such as 8 hours per printing) no matter the actual health state of the spray head. On the one hand, the mode has excessive maintenance risk, frequent disassembly can damage the precise structure of the spray head (such as the aperture of the spray nozzle and the sealing element) and shorten the service life of the spray head, and on the other hand, the mode has insufficient maintenance problem, if the spray head has hidden faults such as micro-blockage, micro-bubbles and the like in advance in the maintenance period, the system cannot early warn in time, and still can continuously produce unqualified products until the faults are enlarged or the next maintenance is not found, so that the production continuity is interrupted. Disclosure of Invention The application aims to provide a state monitoring and maintaining system of a piezoelectric spray head, which can predict and self-maintain the health state of the piezoelectric spray head. The application provides a state monitoring and maintaining system of a piezoelectric spray head, which comprises an acoustic acquisition unit, a signal processing diagnosis unit and a maintenance execution unit, w