Search

CN-122018310-A - Forging equipment fault prediction and self-adaptive control method and system based on Internet of things

CN122018310ACN 122018310 ACN122018310 ACN 122018310ACN-122018310-A

Abstract

The invention discloses a fault prediction and self-adaptive control method and system of forging equipment based on the Internet of things, and relates to the related field of forging equipment maintenance technology. The network layer uploads the fault prediction result of the forging equipment to the cloud platform layer, runs an intelligent algorithm to analyze the prediction result and the running state of the equipment, calculates optimal control parameters, generates an optimal control instruction and realizes self-adaptive control of the forging equipment. The problems of lag response, insufficient data analysis capability and low efficiency of the existing method are solved, and prediction precision and control efficiency are improved.

Inventors

  • MA JUN
  • GAO SHIHENG
  • YUAN YAPENG
  • HUANG XIAOLEI

Assignees

  • 徐州达一重锻科技有限公司

Dates

Publication Date
20260512
Application Date
20251229

Claims (10)

  1. 1. The forging equipment fault prediction and self-adaptive control method based on the Internet of things is characterized by comprising the following steps of: (1) Constructing an internet of things platform, uploading, storing and processing basic data of forging equipment, wherein the platform consists of an equipment layer, an edge computing layer, a network layer and a cloud platform layer; (2) Starting a plurality of sensors of an equipment layer of the animal networking platform, determining a data source according to a fault prediction target, and collecting various data in the operation of forging equipment in real time; (3) Converting the operation data of the forging equipment acquired by the sensor into multivariate time series data, preprocessing the multivariate time series data through an edge calculation layer of the platform of the Internet of things, inputting the preprocessed time series data into a pre-training time series abnormality detection model, and outputting a fault prediction result of the equipment; (4) The network layer of the Internet of things platform provides a network channel, a fault prediction result of the forging equipment is uploaded to the cloud platform layer, an intelligent algorithm is operated to analyze the prediction result and the running state of the equipment, and optimal control parameters are calculated; (5) Generating an optimal control instruction according to the calculated optimal control parameter, transmitting the optimal control instruction to an edge calculation layer, realizing self-adaptive control on forging equipment, and feeding back parameter configuration of an execution result optimal control algorithm.
  2. 2. The fault prediction and self-adaptive control method for forging equipment based on the Internet of things, as set forth in claim 1, is characterized in that in the step (1), the equipment layer is used for collecting the physical state and the operation parameters of the equipment, and consists of a plurality of sensors and a PLC (programmable logic controller), wherein the sensor types comprise a vibration sensor, a current sensor, a temperature sensor, a pressure transmitter and a displacement sensor, and the physical quantity is converted into a digital signal which is convenient for model processing; The edge computing layer performs preliminary processing and intelligent decision of data and consists of an edge intelligent gateway and edge computing software, wherein the gateway is provided with various interfaces for connecting various sensors and a PLC (programmable logic controller); the network layer establishes a bidirectional data transmission channel, accesses an industrial optical fiber private line, combines a commercial broadband to form a main and standby link, and automatically switches to a wireless network when the wired network is interrupted to ensure that data is not lost; The cloud platform layer provides mass data storage, calculation analysis and visual display, realizes self-adaptive control optimization, is a center for data aggregation, storage, deep analysis and application, comprises a plurality of time sequence databases and relational databases, stores various forging equipment data, provides big data and an AI platform, runs an intelligent optimization algorithm, creates an environment, and trains and evaluates a time sequence abnormality detection model.
  3. 3. The method for predicting and adaptively controlling the faults of forging equipment based on the Internet of things according to the claim 1 is characterized in that in the step (2), the measuring range, the precision and the frequency response of the sensor are set based on a fault prediction target, a data list is formulated according to a data source, a data point name and a sampling frequency are marked, a sensor output signal line is correctly connected to a data collector to supply power for all equipment, and configuration parameters of each accessed channel in an edge gateway are matched with the sensor, real-time data acquired by the data collector are received and synchronously uploaded to a cloud platform layer for storage.
  4. 4. The method for predicting and adaptively controlling the faults of the forging and pressing equipment based on the Internet of things according to claim 1 is characterized in that in the step (3), a pre-training time sequence abnormal detection model is obtained through a cloud platform layer in an Internet of things platform, normal samples in forging and pressing equipment operation historical data stored in the cloud platform layer are used as training data, specific sensor data are selected according to a fault prediction target, and multivariate time sequence data and key feature data are obtained through time alignment, preprocessing and key feature extraction, so that a data set for model training is formed; The cloud platform provides a model training environment, in the training environment, a time sequence abnormality detection model is built through code programming, super parameters of each part of the model are set, model parameters are initialized randomly, an optimizer type, an initial learning rate and a learning rate strategy for model training are configured, the model is trained according to a set training period and a set sample batch, model parameters are updated, the cloud platform automatically records loss function changes, the performance of the model is evaluated after each training period is finished, and when the loss function converges, the model parameters with optimal performance are saved to be a pre-training time sequence abnormality detection model.
  5. 5. The internet of things-based fault prediction and adaptive control method for forging equipment of claim 4, wherein the time sequence anomaly detection model is composed of a variable self-encoder and a long-term and short-term memory network: The variable self-encoder is a normal mode for generating a probability model and simulating multivariable time sequence data of forging equipment operation so as to extract local characteristics, and consists of an encoder and a decoder, wherein p sequential local windows of continuous readings are taken as input, q-dimensional low-dimensional embedding is estimated through the encoder, a sequential context relationship is modeled through a long-period memory network, and an original window is reconstructed through the decoder.
  6. 6. The internet of things-based forging equipment fault prediction and self-adaptive control method as set forth in claim 5, wherein the variation self-encoder takes normal forging equipment operation data as input features during training, and performs multi-layer nonlinear transformation, linear layer mapping and multi-scale convolution to map the input features to potential space and calculate a mean vector And logarithmic variance vector Obtaining input characteristics The corresponding normal distribution in the potential space: Wherein, the The probability distribution is represented by a graph of the probability distribution, A probabilistic representation of the input feature in a low-dimensional potential space, The mean value of the normal distribution is represented, The standard deviation of the normal distribution is indicated, To achieve back propagation, the probabilistic representation is further processed as: , Representing the multiplication by element, Obeying a normal distribution with a mean value of 0 and a variance of 1, and The original input characteristic structure is restored through the multi-layer nonlinear mapping of the decoder, the reconstruction characteristic is generated, the reconstruction characteristic output by the decoder is forced to approach the original input characteristic through calculating the reconstruction loss, and the model parameters of the variable self-encoder are optimized.
  7. 7. The internet of things-based fault prediction and self-adaptive control method for forging equipment according to claim 1, wherein in the step (4), the intelligent algorithm is an improved particle swarm optimization algorithm, and an optimal solution is found through simulation, and the method comprises the following steps: creating a group of particles, each particle representing a set of random control parameter combinations, initializing a particle speed, an individual optimal solution and a global optimal solution for improving a particle swarm optimization algorithm; For each particle, current parameters are injected into a digital twin model, the model is corrected according to the current fault prediction result, and the abnormal condition of the operation data of the forging and pressing equipment is simulated; simulating a complete stamping cycle in the digital twin model, wherein the complete flow comprises the steps of accelerating the sliding block to descend, contacting a workpiece, pressurizing, maintaining pressure and returning, so as to realize the working simulation of forging equipment; according to the simulation result, extracting position tracking curve, pressure curve, product thickness and adjusting time data, bringing the position tracking curve, pressure curve, product thickness and adjusting time data into an objective function, calculating the fitness value of the current particle, wherein the smaller the value is, the closer the current parameter combination is to an optimal solution; Each particle compares the fitness value of the current parameter combination with the historical optimal fitness value of the particle, updates the individual optimal solution of the particle, compares the fitness values of all particles of the whole particle group, and finds the current global optimal solution; Dynamically adjusting inertia weight according to the current iteration progress, updating the speed and the position of each particle to generate a new parameter combination, randomly selecting part of particles to perform variation if the global optimal solution is not improved by continuous multiple iterations, randomly perturbing the parameters of the particles, and calling out local optimal; when the maximum iteration times are reached, the global optimal solution at the moment is used as an optimal control parameter combination for adjusting the running state of the forging equipment, and the possibility of faults is reduced.
  8. 8. The forging equipment fault prediction and self-adaptive control system based on the internet of things, which is characterized in that the system is used for implementing the forging equipment fault prediction and self-adaptive control method based on the internet of things as set forth in any one of claims 1 to 7, and the system comprises: The Internet of things platform building module is used for building an Internet of things platform and uploading, storing and processing basic data of forging equipment, wherein the platform consists of an equipment layer, an edge computing layer, a network layer and a cloud platform layer; The forging equipment operation data acquisition module is used for starting various sensors of the equipment layer of the animal networking platform, determining a data source according to a fault prediction target and acquiring various data in the operation of the forging equipment in real time; The forging equipment fault prediction module is used for converting the forging equipment operation data acquired by the sensor into multivariable time series data, preprocessing the multivariable time series data through an edge calculation layer of the platform of the Internet of things, inputting the preprocessed time series data into the pre-training time series abnormality detection model, and outputting a fault prediction result of the equipment; The forging equipment control parameter optimization module is used for uploading a fault prediction result of the forging equipment to the cloud platform layer, analyzing the prediction result and the equipment running state by running the intelligent algorithm, and calculating the optimal control parameters; the control instruction generation and execution module is used for generating an optimal control instruction according to the calculated optimal control parameter, transmitting the optimal control instruction to the edge calculation layer, realizing self-adaptive control on forging equipment, and feeding back parameter configuration of an execution result optimal control algorithm.
  9. 9. Forging and pressing equipment fault prediction and self-adaptation control equipment based on thing networking, its characterized in that, forging and pressing equipment fault prediction and self-adaptation control equipment based on thing networking includes: The forging equipment fault prediction and self-adaption control method based on the internet of things comprises a memory, a processor and a forging equipment fault prediction and self-adaption control program based on the internet of things, wherein the forging equipment fault prediction and self-adaption control program based on the internet of things is stored in the memory and can run on the processor, and the forging equipment fault prediction and self-adaption control method based on the internet of things is realized when the forging equipment fault prediction and self-adaption control program based on the internet of things is executed by the processor.
  10. 10. A computer program product, characterized in that the computer program product comprises a forging equipment fault prediction and adaptive control program based on the internet of things, which when executed by a processor, implements the forging equipment fault prediction and adaptive control method based on the internet of things as claimed in any one of claims 1 to 7.

Description

Forging equipment fault prediction and self-adaptive control method and system based on Internet of things Technical Field The application relates to the technical field of forging equipment maintenance, in particular to a forging equipment fault prediction and self-adaptive control method and system based on the Internet of things. Background Forging equipment is used as core equipment in manufacturing industry, the running state of the forging equipment is directly related to production efficiency and product quality, and with the progress of industrial technology, fault prediction and health management of the forging equipment are becoming increasingly important. The traditional forging equipment periodic maintenance and post maintenance mode is difficult to adapt to the modern production rhythm, the periodic maintenance can generate excessive maintenance or insufficient maintenance, and the post maintenance is delayed in response; the intelligent transformation of manufacturing industry promotes the intellectualization of the operation and the maintenance of forging equipment, and the prior art evaluates the health state of the equipment and predicts the fault condition of a core component by monitoring data in real time, thereby realizing predictive maintenance, reducing unplanned shutdown and optimizing maintenance resources; In addition, control parameters (such as pressure, speed, position and the like) of the forging equipment are usually set in a debugging stage, are fixed once set, however, the equipment performance can change due to abrasion, aging and component heating in long-term operation, and the static control mode lacks a data analysis process and cannot adapt to the dynamic changes, so that the quality of products fluctuates, the qualification rate is reduced, operators often rely on experience to manually adjust the parameters, and scientificity and consistency are lacking. Disclosure of Invention Aiming at the technical problems in the prior art, the application provides the fault prediction and self-adaptive control method and system for the forging equipment based on the Internet of things, which are used for realizing predictive maintenance of the forging equipment by constructing an Internet of things platform, applying a time sequence anomaly detection model based on deep learning, monitoring the operation data of the forging equipment in real time, discovering potential faults of the equipment in time and carrying out self-adaptive control on the forging equipment through a dynamic optimization algorithm. The application provides a forging equipment fault prediction and self-adaptive control method based on the Internet of things, which comprises the following steps: (1) Constructing an internet of things platform, uploading, storing and processing basic data of forging equipment, wherein the platform consists of an equipment layer, an edge computing layer, a network layer and a cloud platform layer; (2) Starting a plurality of sensors of an equipment layer of the animal networking platform, determining a data source according to a fault prediction target, and collecting various data in the operation of forging equipment in real time; (3) Converting the operation data of the forging equipment acquired by the sensor into multivariate time series data, preprocessing the multivariate time series data through an edge calculation layer of the platform of the Internet of things, inputting the preprocessed time series data into a pre-training time series abnormality detection model, and outputting a fault prediction result of the equipment; (4) The network layer of the Internet of things platform provides a network channel, a fault prediction result of the forging equipment is uploaded to the cloud platform layer, an intelligent algorithm is operated to analyze the prediction result and the running state of the equipment, and optimal control parameters are calculated; (5) Generating an optimal control instruction according to the calculated optimal control parameter, transmitting the optimal control instruction to an edge calculation layer, realizing self-adaptive control on forging equipment, and feeding back parameter configuration of an execution result optimal control algorithm. Further, the internet of things platform adopts a classical cloud-pipe-side-end four-layer architecture, so that comprehensive acquisition, efficient transmission and intelligent processing of data are ensured. The device layer is provided with a plurality of sensors for sensing the running state of forging equipment, the edge computing layer performs preliminary processing and intelligent decision at a data source to realize low-delay response, the network layer establishes a safe, reliable and bidirectional data transmission channel, accesses an industrial optical fiber private line and combines a commercial broadband to form a main and standby link, and the cloud platform layer provides mass data storage,