Search

CN-121980467-A - State monitoring method and system for metal die casting processing equipment

CN121980467ACN 121980467 ACN121980467 ACN 121980467ACN-121980467-A

Abstract

The application relates to the technical field of state monitoring, in particular to a state monitoring method and system of metal die casting processing equipment, wherein the method comprises the steps of collecting vibration data in the processing process of the equipment; determining final modal components of the K values in the search range of the K values, calculating quality indexes representing the decomposition effect of the K values, determining a target K value based on the quality indexes of the K values, identifying high-frequency modal components from the final modal components corresponding to the target K value, and realizing state monitoring according to the high-frequency modal components. According to the technical scheme, the state monitoring result of the metal die casting processing equipment can be accurately obtained.

Inventors

  • TAO HAITAO
  • CHEN XIAOYU

Assignees

  • 东莞市发斯特精密科技股份有限公司

Dates

Publication Date
20260505
Application Date
20260316

Claims (10)

  1. 1. A state monitoring method of metal die casting processing equipment is characterized by comprising the steps of collecting vibration data in the processing process of the equipment, determining final modal components of all K values in a K value searching range, calculating a quality index representing the decomposition effect of the K values, determining a target K value based on the quality index of all K values, identifying high-frequency modal components from the final modal components corresponding to the target K value, and realizing state monitoring according to the high-frequency modal components; The method comprises the steps of determining a penalty factor of a current K value by taking a minimum total bandwidth as a target, carrying out modal decomposition on vibration data based on the penalty factor and the current K value to obtain K first components, dividing each first component into a pure mode and a superposition mode according to modal purity, carrying out modal decomposition again after fusing the superposition modes until the number of the superposition modes is smaller than 2 or the iteration number of the modal decomposition reaches the maximum number of times, and obtaining a final modal component corresponding to the K value, wherein the quality index of the K value is the inter-class distance between a high-frequency component and a low-frequency component in the final modal component.
  2. 2. The method for monitoring the state of metal die casting processing equipment according to claim 1, wherein determining the penalty factor of the current K value with the aim of minimizing the total bandwidth of the modal components comprises iteratively adjusting the penalty factor in a preset range, and obtaining the penalty factor of the current K value when the sum of bandwidths of all the first components reaches the minimum.
  3. 3. The method for monitoring the state of metal die casting processing equipment according to claim 1 is characterized in that the dividing of each first component into a pure mode and a superposition mode according to the mode purity comprises the steps of carrying out Fourier transform on the first components to obtain center frequencies of the first components, setting a frequency window with preset width by taking the center frequencies as the center, calculating the proportion of energy in the frequency window to the total energy of the first components to obtain the mode purity of the first components, and responding to the mode purity being greater than a purity threshold, wherein the first components are pure modes, and otherwise, the first components are superposition modes.
  4. 4. The method for monitoring the state of the metal die casting machining equipment according to claim 1, wherein the mode decomposition is performed again after the superposition modes are fused, wherein the method comprises the steps of superposing all superposition modes on a time domain to generate a residual signal, taking the number of superposition modes as a residual K value of the residual signal, determining a penalty factor of the residual K value, and performing variation mode decomposition on the residual signal to update the first component.
  5. 5. The state monitoring method of the metal die casting machining equipment according to claim 1 is characterized in that the calculating method of the quality index of the K value comprises the steps of extracting center frequencies of final modal components, dividing the center frequencies into high-frequency clusters and low-frequency clusters by adopting a clustering algorithm, calculating frequency differences between the lowest frequencies of the high-frequency clusters and the highest frequencies of the low-frequency clusters, and taking the ratio of the frequency differences to the extreme differences of all the center frequencies as the quality index.
  6. 6. The method for monitoring the state of the metal die casting processing equipment according to claim 1 is characterized in that the monitoring of the state according to the high-frequency modal components is achieved by collecting start-stop time stamps of the equipment in each process stage, dividing the high-frequency modal components into signal segments corresponding to each process stage by utilizing the start-stop time stamps, calculating at least one health index for each signal segment to obtain diagnosis features, and comparing the diagnosis features with preset baseline features to judge the state of the equipment and locate the process stage in which abnormality occurs.
  7. 7. The method of claim 6, wherein each of the process stages comprises at least one of mold closing, mold locking, injection, dwell, cool down, mold opening, and ejection.
  8. 8. The method of claim 6, wherein the health indicator comprises at least one of signal energy and signal kurtosis.
  9. 9. The method for monitoring the state of a metal die casting machining apparatus according to claim 1, wherein before determining the final modal component of each K value, the method further comprises aligning the vibration data with a start-stop time stamp of each process stage on a time axis, and performing noise reduction processing on the aligned vibration data.
  10. 10. A state monitoring system of a metal die casting machining apparatus, characterized by comprising a processor and a memory, the memory storing computer program instructions which, when executed by the processor, implement a state monitoring method of a metal die casting machining apparatus according to any one of claims 1 to 9.

Description

State monitoring method and system for metal die casting processing equipment Technical Field The application relates to the technical field of state monitoring, in particular to a state monitoring method and system of metal die casting processing equipment. Background In the process of producing metal die castings such as mobile phone frames by using die casting equipment, the die casting equipment operates under severe working conditions of high temperature, high pressure and high speed, and because the mobile phone frames are small in size and thin in thickness and have high requirements on machining precision, the quality defects of products can be possibly caused by small vibration of the die casting equipment, so that the states of the metal die casting processing equipment are required to be monitored in real time, and the product quality of the metal die castings such as the mobile phone frames is guaranteed. At present, a patent application document with the application publication number of CN114137338A discloses a device running state monitoring method, a system and a storage medium, wherein the method comprises the following steps of obtaining a vibration signal in a device running process, decomposing the vibration signal into k modal components, wherein k values are automatically determined by sample entropy of the modal components and center frequency ratio of the modal components, determining Hilbert time spectrums corresponding to the k modal components, comparing the Hilbert time spectrums corresponding to the k modal components with preset Hilbert time spectrums, determining that the device fails when the difference value of the Hilbert time spectrums corresponding to the k modal components and the preset Hilbert time spectrums is smaller than the preset difference value, and determining that the device fails when the difference value of the Hilbert time spectrums corresponding to the k modal components and the preset Hilbert time spectrums is larger than the preset difference value. According to the method, the vibration signal is decomposed into k modal components, whether the equipment is in a fault state is judged by comparing the Hilbert time spectrum corresponding to the k modal components with the preset Hilbert time spectrum, wherein the k value is automatically determined by the sample entropy of the modal components and the center frequency ratio of the modal components, however, in the process of decomposing the vibration signal into the k modal components, some modal components can well represent one mode, some modal components can be in a mode superposition state due to over decomposition or under decomposition, the k value is directly determined by the sample entropy of the modal components and the center frequency ratio of the modal components, whether the modal components are in the mode superposition state cannot be distinguished, and further the decomposed modal components cannot accurately represent the fault characteristics of the equipment, so that the state monitoring result of the equipment is inaccurate. Disclosure of Invention In order to solve the technical problem that the state monitoring result of equipment is inaccurate, the application provides a state monitoring method and a state monitoring system for metal die casting processing equipment, which can accurately acquire the state monitoring result of the equipment. The application provides a state monitoring method of metal die casting processing equipment, which comprises the steps of collecting vibration data in the processing process of the equipment, determining final modal components of all K values in a K value searching range, calculating a quality index representing the decomposition effect of the K values, determining a target K value based on the quality index of all K values, identifying high-frequency modal components from the final modal components corresponding to the target K values, and realizing state monitoring according to the high-frequency modal components, wherein determining the final modal components of all K values comprises determining a penalty factor of the current K value by taking the minimum total bandwidth as a target, performing modal decomposition on the vibration data based on the penalty factor and the current K value to obtain K first components, dividing all the first components into pure modes and overlapped modes according to the modal purity, fusing the overlapped modes again until the number of the overlapped modes is smaller than 2 or the iteration number of the modal decomposition reaches the maximum, and obtaining final modal components corresponding to the K values, and the quality index of the K values is the inter-class distance between the high-frequency modal components and the low-frequency modal components in the final modal components. The mode purity is introduced to evaluate the decomposition quality of the first component, and the superposition mode is