Collecting and subpackaging equipment for zinc powder

Abstract

The invention belongs to the field of zinc powder collecting and subpackaging, and particularly relates to zinc powder collecting and subpackaging equipment which comprises a dust collecting system, an automatic subpackaging system and a packaging and sealing system.The zinc powder collecting and subpackaging equipment has the beneficial effects that manual intervention is reduced through automatic operation, and the production efficiency is improved; an accurate metering control system ensures the quality consistency of subpackage of each batch of zinc powder; an efficient collecting and sealing technology is adopted, so that the loss of the zinc powder in the conveying and sub-packaging processes is reduced; the visual inspection system ensures the sealing quality, and material waste caused by packaging defects is reduced; the automatic fault detection and alarm system can find and process equipment abnormity in time, and equipment damage and production interruption are avoided; detection parameters are adjusted through a self-adaptive algorithm, and it is ensured that the equipment can keep stable performance in different material environments.

Description

Technical Field

[0001] The invention relates to the technical field of zinc powder collecting and packaging, and in particular to zinc powder collecting and packaging equipment. Background Art

[0002] Zinc powder collection and packaging device is a kind of equipment specially used for processing zinc powder, mainly used for collecting, storing and accurately packaging zinc powder. This device is widely used in chemical, metallurgical, coating and other industries, especially in the production of zinc-based coatings, galvanizing process and other fields.

[0003] In the related technology, existing equipment mostly adopts simple filtering or sealing methods for dust control, but the efficiency is low and it is easy to cause overflow or dust accumulation problems. In addition, the existing packaging equipment is lacking in automated operation and requires manual intervention to adjust the packaging volume or replace the packaging container, resulting in low production efficiency. In addition, the flexible adjustment capability under different specifications of packaging requirements is insufficient, which limits the applicability of the equipment. For this reason, we propose a collection and packaging equipment for zinc powder.

[0004] The above information disclosed in this Background section is only for understanding the background of the present inventive concept and therefore it may contain information that does not constitute prior art. Summary of the invention

[0005] In order to overcome the shortcomings of the prior art, the present invention provides a zinc powder collecting and packaging equipment to solve the problem that the existing equipment proposed in the above background technology mostly adopts a simple filtering or sealing method for dust control, but the efficiency is low and it is easy to cause overflow or dust accumulation problems, and the existing packaging equipment is lacking in automated operation, and manual intervention is required to adjust the packaging amount or replace the packaging container, resulting in low production efficiency. In addition, the flexible adjustment capability under different specifications of packaging requirements is insufficient, which limits the applicability of the equipment.

[0006] The technical solution adopted by the present invention to solve the technical problem is: a zinc powder collection and packaging device, including a dust collection system, an automatic packaging system and a packaging and sealing system, wherein:

[0007] Dust collection system: The dust collection system is the core part of the equipment, responsible for effectively collecting the dust generated during the production and packaging of zinc powder to prevent it from leaking into the environment. The dust collection system includes the following components:

[0008] High-efficiency filter: Use high-density filter material and HEPA filter element to capture tiny dust and ensure clean air. The filter element adopts modular design for easy cleaning and replacement;

[0009] Negative pressure dust suction pipe: The equipment is equipped with a negative pressure pipe system to directly suck the dust in the working area into the filtration system;

[0010] Dust collection bin: Dust is transported to a special dust collection bin through a pipeline. The bin is equipped with a vibration device to ensure the rapid sedimentation and storage of dust and prevent accumulation and blockage.

[0011] Automatic packaging system: The automatic packaging system is used to accurately and quantitatively package zinc powder, reduce manual operations, and improve production efficiency. The automatic packaging system consists of the following parts:

[0012] Silo and vibrating screen: The zinc powder is first screened by a vibrating screen to remove impurities and then enter the automatic filling silo. A vibrating device is installed at the bottom of the silo to keep the zinc powder flowing evenly and prevent agglomeration.

[0013] Quantitative controller: According to the set packaging quantity, the quantitative controller automatically controls the output of zinc powder. The quantitative controller uses a high-precision electronic scale and flow meter to monitor the packaging accuracy to ensure that the weight and volume of each packaging are consistent;

[0014] Filling port and conveying device: The filling port is connected to the automated conveying system to ensure that the zinc powder enters the packaging bag accurately and stably. The conveying system adopts a spiral conveying device with an anti-blocking design to ensure smooth operation under high-density material processing;

[0015] Packaging sealing system: The packaging sealing system is used to complete the packaging of zinc powder and ensure that the packaging is intact to prevent leakage or contamination. The packaging sealing system includes:

[0016] Automatic packaging bag placement device: automatically selects packaging bags of appropriate specifications according to packaging requirements and positions them below the dispensing port;

[0017] Automatic sealing device: After the subpackaging is completed, the packaging bag is sealed by a sealing device, using heat sealing technology to ensure that the bag mouth is firmly sealed to prevent zinc powder from leaking or getting wet;

[0018] Package inspection and coding system: After each package is completed, the equipment uses a visual inspection system to check the sealing quality to ensure that each package meets the standards. In addition, the equipment automatically codes the production date and batch number to facilitate subsequent tracking and management.

[0019] As an optimized technical solution, the visual inspection system includes the following structure:

[0020] Imaging system: including high-resolution industrial cameras, lenses and adaptive lighting equipment. The illumination equipment uses one or more of LED light sources and ultraviolet light sources. The lighting equipment can eliminate reflection interference and enhance the contrast of defects. The camera is fixed above the production line to capture high-definition images of the sealing area.

[0021] Data processing module: using edge computing devices, using smart cameras or central processing units (CPU / GPU) for image analysis; a deep learning algorithm is set in the data processing module for image classification and defect detection, and unsealed, foreign body inclusion and uneven defects are classified through pattern recognition, and the deep learning algorithm uses a convolutional neural network;

[0022] Control module: The control module is synchronized with the production line, receives sensor trigger signals, and coordinates the camera to capture images; when defects are detected, it sends instructions to the exclusion system to remove the defective products. The exclusion system uses a robotic arm or airflow device;

[0023] Human-machine interface (HMI): Provides real-time display of sealing quality status, defect statistics, and system diagnostic information. The system is integrated with the packaging line through industrial communication protocols. Inspection results and historical data are stored in a local database or in the cloud for quality traceability and optimization analysis.

[0024] As an optimized technical solution, the steps for testing the sealing quality are as follows:

[0025] Image acquisition: When the packaged product passes through the conveyor belt to the inspection area, the photoelectric sensor triggers the camera to take a high-definition image of the sealing part, and the adaptive algorithm adjusts the inspection parameters according to different materials;

[0026] Image preprocessing: denoising, graying and edge enhancement to ensure clear target features;

[0027] Defect analysis: The deep learning model extracts features from the image. Through the training sample library, the model identifies sealing integrity, foreign matter inclusions, or material wrinkles.

[0028] Judgment and action execution: The test results are classified as "qualified" or "unqualified". If unqualified, the system sends a signal to the sorting mechanism to remove the defective products;

[0029] Data storage and reporting: Test results are uploaded to the database and quality statistics reports are generated for production optimization.

[0030] As an optimized technical solution, the steps to adjust the detection parameters according to different materials are as follows:

[0031] Material identification and feature extraction: During the production process, when the packaged product passes the inspection point, the system first identifies the material type and classifies it by image color, reflectivity, and surface texture features. Based on these features, the system will automatically select the appropriate image processing method and parameters;

[0032] Image preprocessing and adaptive thresholding: Based on the different optical properties of the material, including transparency, reflectivity, and texture, the algorithm optimizes the input image by dynamically adjusting image preprocessing parameters, including contrast enhancement, grayscale, and edge detection.

[0033] Feature matching and defect detection: After feature extraction, the algorithm will automatically match the defect pattern on the material surface, including scratches, bubbles, and wrinkles. For different materials, the shape, size, and contrast of the defects are different. Therefore, the system will judge the defect type of the material based on the trained model. The model uses the CNN model.

[0034] Feedback and optimization: Once a defect is detected, the system will optimize parameters based on the current detection results;

[0035] Dynamic adjustment and automatic update: As the production process progresses, the algorithm will adjust parameters based on real-time feedback, including detection accuracy and error rate. The process is dynamic and the algorithm can continuously learn and optimize to adapt to material changes in different production batches.

[0036] As an optimized technical solution, a control and monitoring system is also included, and the control and monitoring system includes the following structural components:

[0037] Central control system: The PLC controller manages the operation of each module in a unified manner, adjusts the equipment operating parameters in real time, and the operation interface is displayed on the touch screen, which is convenient for operators to set and adjust parameters;

[0038] Automatic fault detection and alarm system: This system can monitor the operating status of the equipment in real time, including motor load, conveyor belt speed, and filter status. Once an abnormality is found, the system will sound an alarm and automatically shut down to prevent equipment damage;

[0039] Data acquisition and storage module: real-time recording of equipment production data and operating parameters to facilitate future data analysis and quality control.

[0040] As an optimized technical solution, the real-time monitoring of the equipment operation status of the automatic fault detection and alarm system relies on sensor technology, data processing algorithms and edge artificial intelligence (EdgeAI) framework, aiming to capture motor load, conveyor belt speed, filter status parameters, analyze the difference between normal and abnormal states through algorithms, and trigger alarms and shutdown protection equipment when necessary. The following are its specific steps:

[0041] Data collection:

[0042] Use a variety of sensors, including vibration sensors, temperature sensors, and current sensors to collect equipment operation data in real time, and convert the collected data into signal format for further analysis;

[0043] Data preprocessing: denoising, normalizing and feature extraction of raw data, using time-frequency domain analysis methods, including short-time Fourier transform and continuous wavelet transform, to convert time series data into spectrum or feature vector;

[0044] Fault detection: Using machine learning or deep learning algorithms, convolutional neural networks or recursive neural networks can be used to analyze processed data in real time, identify abnormal patterns, and apply statistical models to determine whether the device is in a fault state based on the changing trend of the device behavior;

[0045] Fault diagnosis and alarm: If an abnormality is detected, the system automatically classifies the fault type and sends the information to the control center. The system triggers an alarm mechanism, including sound, light or network notification, and may perform an automatic shutdown;

[0046] Recording and Analysis: Failure data is stored for subsequent analysis to optimize models and improve prediction accuracy.

[0047] As an optimized technical solution, in the process of automatic fault detection, the steps of time-frequency domain analysis are as follows:

[0048] Data collection: Sensors monitor the operating status of equipment in real time and collect time series data, including motor vibration signals or temperature changes;

[0049] Signal preprocessing: Remove noise or interference signals through filters to enhance feature information.

[0050] Time-frequency transformation: Use short-time Fourier transform (STFT) to analyze non-stationary signals, divide the time series data into segments and apply Fourier transform to each segment to generate the time-frequency spectrum of the signal;

[0051] Use continuous wavelet transform (CWT) to convolve the signal with the original signal through a series of scaled and translated mother wavelet functions to obtain time-frequency distribution information. Continuous wavelet transform is suitable for processing complex signals containing transient features;

[0052] Feature extraction: Extract key features from the spectrum or wavelet coefficients, including the rate of change of frequency or the time location of transient events;

[0053] Anomaly detection: Classify the extracted features and identify abnormal patterns through machine learning algorithms or rule-based methods. Machine learning algorithms use support vector machines or deep learning.

[0054] Alarm and response: The system triggers an alarm and performs predefined response operations after detecting an anomaly.

[0055] As an optimized technical solution, an environmental protection and recycling system is also included, and the environmental protection and recycling system includes the following components:

[0056] Dust recovery module: The dust collected by the dust collection bin and filtration system inside the equipment can be put back into production through the automated recovery system, reducing resource waste and lowering environmental pressure;

[0057] Waste gas treatment module: For harmful gases in the production process, the equipment is equipped with a waste gas purification device, which uses activated carbon adsorption or catalyst decomposition technology to treat harmful gases to meet environmental emission standards.

[0058] The beneficial effects of the present invention are:

[0059] Through automated operation, human intervention is reduced and production efficiency is improved; the precise metering control system ensures the quality consistency of each batch of zinc powder packaging; the use of efficient collection and sealing technology reduces the loss of zinc powder during transmission and packaging; the visual inspection system ensures the sealing quality and reduces material waste caused by packaging defects; the automatic fault detection and alarm system can promptly detect and handle equipment abnormalities to avoid equipment damage and production interruptions; the detection parameters are adjusted through adaptive algorithms to ensure that the equipment can maintain stable performance in different material environments; the combination of time-frequency domain analysis and machine learning algorithms can achieve real-time monitoring and fault prediction of complex operating conditions; the system can dynamically adjust the packaging process to meet the needs of different zinc powder characteristics; it reduces the leakage of zinc powder, reduces pollution to the environment, reduces manual operations, and reduces the risk of workers being exposed to hazardous substances. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 The present invention provides a system connection block diagram of a zinc powder collecting and packaging device. DETAILED DESCRIPTION

[0061] The present invention is further described below in conjunction with the accompanying drawings and specific implementation methods. It should be noted that, under the premise of no conflict, the various embodiments or technical features described below can be arbitrarily combined to form a new embodiment.

[0062] It should be noted that the terms "center", "up", "down", "left", "right", "vertical", "horizontal", "inside", "outside", etc. indicate directions or positional relationships based on the directions or positional relationships shown in the accompanying drawings, or are the directions or positional relationships in which the inventive product is usually placed when in use. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore should not be understood as a limitation on the present invention.

[0063] In the description of the present invention, it is also necessary to explain that, unless otherwise clearly specified and limited, the terms "set", "install", "connect", and "connect" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0064] Embodiment 1:

[0065] A zinc powder collecting and packaging device in this embodiment includes a dust collection system, an automatic packaging system and a packaging and sealing system, wherein:

[0066] Dust collection system: The dust collection system is the core part of the equipment, responsible for effectively collecting the dust generated during the production and packaging of zinc powder to prevent it from leaking into the environment. The dust collection system includes the following components:

[0067] High-efficiency filter: Use high-density filter material and HEPA filter element to capture tiny dust and ensure clean air. The filter element adopts modular design for easy cleaning and replacement;

[0068] Negative pressure dust suction pipe: The equipment is equipped with a negative pressure pipe system to directly suck the dust in the working area into the filtration system;

[0069] Dust collection bin: Dust is transported to a special dust collection bin through a pipeline. The bin is equipped with a vibration device to ensure the rapid sedimentation and storage of dust and prevent accumulation and blockage.

[0070] Automatic packaging system: The automatic packaging system is used to accurately and quantitatively package zinc powder, reduce manual operations, and improve production efficiency. The automatic packaging system consists of the following parts:

[0071] Silo and vibrating screen: The zinc powder is first screened by a vibrating screen to remove impurities and then enter the automatic filling silo. A vibrating device is installed at the bottom of the silo to keep the zinc powder flowing evenly and prevent agglomeration.

[0072] Quantitative controller: According to the set packaging quantity, the quantitative controller automatically controls the output of zinc powder. The quantitative controller uses a high-precision electronic scale and flow meter to monitor the packaging accuracy to ensure that the weight and volume of each packaging are consistent;

[0073] Filling port and conveying device: The filling port is connected to the automated conveying system to ensure that the zinc powder enters the packaging bag accurately and stably. The conveying system adopts a spiral conveying device with an anti-blocking design to ensure smooth operation under high-density material processing;

[0074] The steps and working principle for achieving precise quantitative packaging are as follows:

[0075] Process steps:

[0076] Material conveying: The material is conveyed to the weighing unit through a feeding system (such as a vibrating conveyor belt or a screw conveyor). The conveying speed and material flow rate are dynamically adjusted by the control system to avoid excessive or insufficient feeding;

[0077] Dynamic weighing: In a weighing unit, the material is weighed by a high-precision sensor (such as a strain gauge sensor), which converts the weight of the material into an electrical signal and sends the data to the control system;

[0078] Weight adjustment: The control system adjusts the feed in real time based on the preset target weight, and achieves accurate approach to the target value through the coordination of fine flow and coarse flow (large particles are fed quickly and small particles are replenished slowly);

[0079] Quantitative control and packaging: When the set weight is reached, the gate of the hopper or weighing bucket opens, and the material is accurately placed into the packaging container. The system automatically closes the gate to avoid overpackaging;

[0080] Packaging and inspection: After the container is sealed, the quality inspection system (such as visual inspection or sensors) checks the packaging quality and packaging weight. Any product that exceeds the tolerance range will be automatically rejected;

[0081] Working principle:

[0082] The system manages all processes through PLC (Programmable Logic Controller) or HMI (Human Machine Interface). PLC receives weight data, feeding status and other information in real time, and controls the conveying, weighing and packaging equipment;

[0083] PID (proportional-integral-derivative) control algorithm is used to adjust the feed rate to achieve fast and stable weight convergence;

[0084] Output of load cell:

[0085] W = S·K;

[0086] Where: W is the measured weight, S is the electrical signal output of the sensor, and K is the sensitivity coefficient of the sensor;

[0087] PID control model:

[0088]

[0089] Where: u(t) is the control output (adjusting the feed valve or conveyor belt speed), e(t) is the error between the target weight and the current weight, K p , K i, K d are the proportional, integral and differential coefficients respectively;

[0090] Error tolerance check: The system compares the weight W of each package measured and the target value W target :

[0091] |W measured -W target |≤ΔW;

[0092] Among them, ΔW is the allowable error range;

[0093] Packaging sealing system: The packaging sealing system is used to complete the packaging of zinc powder and ensure that the packaging is intact to prevent leakage or contamination. The packaging sealing system includes:

[0094] Automatic packaging bag placement device: automatically selects packaging bags of appropriate specifications according to packaging requirements and positions them below the dispensing port;

[0095] Automatic sealing device: After the subpackaging is completed, the packaging bag is sealed by a sealing device, using heat sealing technology to ensure that the bag mouth is firmly sealed to prevent zinc powder from leaking or getting wet;

[0096] Package inspection and coding system: After each package is completed, the equipment uses a visual inspection system to check the sealing quality to ensure that each package meets the standards. In addition, the equipment automatically codes the production date and batch number to facilitate subsequent tracking and management.

[0097] As an optimized technical solution, the visual inspection system includes the following structures:

[0098] Imaging system: including high-resolution industrial cameras, lenses and adaptive lighting equipment. The illumination equipment uses one or more of LED light sources and ultraviolet light sources. The lighting equipment can eliminate reflection interference and enhance the contrast of defects. The camera is fixed above the production line to capture high-definition images of the sealing area.

[0099] Data processing module: using edge computing devices, using smart cameras or central processing units (CPU / GPU) for image analysis; a deep learning algorithm is set in the data processing module for image classification and defect detection, and unsealed, foreign body inclusion and uneven defects are classified through pattern recognition, and the deep learning algorithm uses a convolutional neural network;

[0100] Control module: The control module is synchronized with the production line, receives sensor trigger signals, and coordinates the camera to capture images; when defects are detected, it sends instructions to the exclusion system to remove the defective products. The exclusion system uses a robotic arm or airflow device;

[0101] Human-machine interface (HMI): Provides real-time display of sealing quality status, defect statistics, and system diagnostic information. The system is integrated with the packaging line through industrial communication protocols. Inspection results and historical data are stored in a local database or in the cloud for quality traceability and optimization analysis.

[0102] As an optimized technical solution, the steps for testing the sealing quality are as follows:

[0103] Image acquisition: When the packaged product passes through the conveyor belt to the inspection area, the photoelectric sensor triggers the camera to take a high-definition image of the sealing part, and the adaptive algorithm adjusts the inspection parameters according to different materials;

[0104] Image preprocessing: denoising, graying and edge enhancement to ensure clear target features;

[0105] Defect analysis: The deep learning model extracts features from the image. Through the training sample library, the model identifies sealing integrity, foreign matter inclusions, or material wrinkles.

[0106] Judgment and action execution: The test results are classified as "qualified" or "unqualified". If unqualified, the system sends a signal to the sorting mechanism to remove the defective products;

[0107] Data storage and reporting: Test results are uploaded to the database and quality statistics reports are generated for production optimization;

[0108] Convolutional Neural Network (CNN): used for image feature extraction and classification:

[0109] Convolution formula:

[0110] Z=∑ i,j X[i,j]·K[i,j]+b;

[0111] Among them, X[i,j] is the input image, K[i,j] is the convolution kernel, and b is the bias;

[0112] Edge detection: used to determine the sealing boundary, using the Sobel operator:

[0113] G x =∑ i, jX[i,j]·S x [i,j].

[0114] G y =∑ i, jX i,j ]·S y [i,j],;

[0115]

[0116] Anomaly detection: The model generates a probability distribution based on the training data to detect sealing defects that deviate from the standard distribution.

[0117] As an optimized technical solution, the steps to adjust the detection parameters according to different materials are as follows:

[0118] Material identification and feature extraction: During the production process, when the packaged product passes the inspection point, the system first identifies the material type and classifies it by features such as image color, reflectivity, and surface texture. For example, a transparent film may have a low reflectivity, while a metal foil has a high reflectivity. Based on these features, the system will automatically select the appropriate image processing method and parameters;

[0119] Image preprocessing and adaptive thresholding: Based on the different optical properties of the material (such as transparency, reflectivity, texture, etc.), the algorithm optimizes the input image by dynamically adjusting image preprocessing parameters (such as contrast enhancement, grayscale, edge detection). For example, for transparent films, the contrast of the image may need to be enhanced to highlight defects; for metal foils, the brightness may need to be adjusted or different filters may need to be used to reduce the interference of reflections.

[0120] Feature matching and defect detection: After feature extraction, the algorithm will automatically match the defect pattern (such as scratches, bubbles, wrinkles, etc.) on the material surface. For different materials, the shape, size and contrast of the defect are different. Therefore, the system will judge the defect type of the material based on the trained model (such as CNN);

[0121] Feedback and optimization: Once a defect is detected, the system will optimize parameters based on the current test results. For example, if a large number of false positives (misidentification due to low contrast) are detected on a transparent film, the system will automatically enhance the contrast or adjust the detection threshold to improve the accuracy of subsequent tests.

[0122] Dynamic adjustment and automatic update: As the production process progresses, the algorithm will adjust parameters based on real-time feedback (such as detection accuracy, error rate, etc.). This process is dynamic, and the algorithm can continuously learn and optimize to adapt to material changes in different production batches;

[0123] Feature extraction: Feature extraction is to obtain relevant information from the material surface through image processing algorithms. You can use the Sobel operator for edge detection, or use a convolutional neural network (CNN) to extract high-level features:

[0124]

[0125] G y =∑ i,j I(i,j)·S y (i,j);

[0126] Gx , G y : Gradients in horizontal and vertical directions;

[0127] S x , S y : Sobel operator template;

[0128] I(i,j): image pixel;

[0129] Adaptive threshold adjustment: For different material surfaces, the contrast and brightness of the image vary greatly. Therefore, the adaptive algorithm can automatically adjust the threshold based on the statistical characteristics of the input image (such as mean, variance):

[0130] T = μ + α·σ;

[0131] T: adaptive threshold;

[0132] μ: mean value of image pixels;

[0133] σ: standard deviation of image pixels;

[0134] α: Adaptive coefficient, depends on the material type;

[0135] Model training and prediction: Use convolutional neural networks (CNN) or other machine learning algorithms (such as support vector machines SVM) to train the characteristics of different materials. The model automatically optimizes the detection strategy based on the material category:

[0136]

[0137] Model prediction output (i.e., test results);

[0138] W, b: weight and bias;

[0139] X: input features;

[0140] σ: activation function (e.g. ReLU).

[0141] In this embodiment, a control and monitoring system is also included, and the control and monitoring system includes the following structural components:

[0142] Central control system: The PLC controller manages the operation of each module in a unified manner, adjusts the equipment operating parameters in real time, and the operation interface is displayed on the touch screen, which is convenient for operators to set and adjust parameters;

[0143] Automatic fault detection and alarm system: This system can monitor the operating status of the equipment in real time, including motor load, conveyor belt speed, and filter status. Once an abnormality is found, the system will sound an alarm and automatically shut down to prevent equipment damage;

[0144] Data acquisition and storage module: real-time recording of equipment production data and operating parameters to facilitate future data analysis and quality control.

[0145] Embodiment 2:

[0146] In this embodiment, the real-time monitoring of the equipment operation status of the automatic fault detection and alarm system relies on sensing technology, data processing algorithms, and edge artificial intelligence (EdgeAI) framework, aiming to capture motor load, conveyor belt speed, filter status parameters, analyze the difference between normal and abnormal states through algorithms, and trigger alarms and shutdown protection devices when necessary. The following are the specific steps:

[0147] Data collection:

[0148] Use a variety of sensors, including vibration sensors, temperature sensors, and current sensors to collect equipment operation data in real time, and convert the collected data into signal format for further analysis;

[0149] Data preprocessing: denoising, normalizing and feature extraction of raw data, using time-frequency domain analysis methods, including short-time Fourier transform and continuous wavelet transform, to convert time series data into spectrum or feature vector;

[0150] Fault detection: Using machine learning or deep learning algorithms, convolutional neural networks or recursive neural networks can be used to analyze processed data in real time, identify abnormal patterns, and apply statistical models to determine whether the device is in a fault state based on the changing trend of the device behavior;

[0151] Fault diagnosis and alarm: If an abnormality is detected, the system automatically classifies the fault type and sends the information to the control center. The system triggers an alarm mechanism, including sound, light or network notification, and may perform an automatic shutdown;

[0152] Recording and analysis: Failure data is stored for subsequent analysis to optimize models and improve prediction accuracy;

[0153] Use dimensionality reduction techniques based on principal component analysis (PCA) or autoencoders to find key features;

[0154] Define the threshold T: If the residual It is judged as abnormal;

[0155] Using Bayesian networks combined with time series data, the conditional probability of failure of different components is calculated:

[0156]

[0157] Dynamically adjust warnings based on probability values;

[0158] Data is initially processed locally on the device, reducing communication latency and bandwidth requirements;

[0159] Deep learning models such as convolutional neural networks (CNNs) improve real-time performance through feature extraction;

[0160] Multi-sensor fusion methods (such as Kalman filtering) comprehensively analyze information from multiple data sources:

[0161]

[0162] K k is the Kalman gain, z k is the measurement value and H is the measurement matrix.

[0163] Specifically, in the process of automatic fault detection, the steps of time-frequency domain analysis are as follows:

[0164] Data collection: Sensors monitor the operating status of equipment in real time and collect time series data, including motor vibration signals or temperature changes;

[0165] Signal preprocessing: Remove noise or interference signals through filters to enhance feature information.

[0166] Time-frequency transformation: Use short-time Fourier transform (STFT) to analyze non-stationary signals, divide the time series data into segments and apply Fourier transform to each segment to generate the time-frequency spectrum of the signal;

[0167] Use continuous wavelet transform (CWT) to convolve the signal with the original signal through a series of scaled and translated mother wavelet functions to obtain time-frequency distribution information. Continuous wavelet transform is suitable for processing complex signals containing transient features;

[0168] Feature extraction: Extract key features from the spectrum or wavelet coefficients, including the rate of change of frequency or the time location of transient events;

[0169] Anomaly detection: Classify the extracted features and identify abnormal patterns through machine learning algorithms or rule-based methods. Machine learning algorithms use support vector machines or deep learning.

[0170] Alarm and response: The system triggers an alarm and performs predefined response operations after detecting an anomaly;

[0171] The environmental protection and recycling system includes the following components:

[0172] Dust recovery module: The dust collected by the dust collection bin and filtration system inside the equipment can be put back into production through the automated recovery system, reducing resource waste and lowering environmental pressure;

[0173] Waste gas treatment module: For harmful gases in the production process, the equipment is equipped with a waste gas purification device, which uses activated carbon adsorption or catalyst decomposition technology to treat harmful gases to meet environmental emission standards;

[0174] Short-time Fourier transform (STFT):

[0175]

[0176] in:

[0177] x(τ) is the input signal, w(τ-t) is the window function, and f is the frequency. By adjusting the window size, the time resolution and frequency resolution can be balanced;

[0178] Continuous Wavelet Transform (CWT):

[0179]

[0180] in:

[0181] x(t) is the signal, ψ is the mother wavelet function, a is the scale parameter (controls the frequency resolution), b is the translation parameter (controls the time positioning), and the wavelet transform can detect high-frequency transient events with high time resolution;

[0182] The detection of different materials (such as transparent films or metal foils) can be adapted by changing the mother wavelet function or the frequency band range. Transparent films require higher frequency resolution to detect tiny defects, while metal foils focus on low-frequency characteristics. By combining short-time Fourier transform and continuous wavelet transform, the short-term spectral information and multi-scale characteristics of the signal can be obtained simultaneously, significantly improving the accuracy of fault detection.

[0183] The above-mentioned embodiments are only preferred embodiments of the present invention and cannot limit the scope of protection of the present invention in turn. Various combinations, modifications or equivalent substitutions of the technical solutions of the present invention by technicians in this field do not depart from the spirit and scope of the technical solutions of the present invention, and should be included in the scope of the claims of the present invention.

Claims

1. A zinc powder collecting and packaging device, characterized in that: It includes dust collection system, automatic filling system and packaging sealing system, including: Dust collection system: The dust collection system is the core part of the equipment, responsible for effectively collecting the dust generated during the production and packaging of zinc powder to prevent it from leaking into the environment. The dust collection system includes the following components: High-efficiency filter: Use high-density filter material and HEPA filter element to capture tiny dust and ensure clean air. The filter element adopts modular design for easy cleaning and replacement; Negative pressure dust suction pipe: The equipment is equipped with a negative pressure pipe system to directly suck the dust in the working area into the filtration system; Dust collection bin: Dust is transported to a special dust collection bin through a pipeline. The bin is equipped with a vibration device to ensure the rapid sedimentation and storage of dust and prevent accumulation and blockage. Automatic packaging system: The automatic packaging system is used to accurately and quantitatively package zinc powder, reduce manual operations, and improve production efficiency. The automatic packaging system consists of the following parts: Silo and vibrating screen: The zinc powder is first screened by a vibrating screen to remove impurities and then enter the automatic filling silo. A vibrating device is installed at the bottom of the silo to keep the zinc powder flowing evenly and prevent agglomeration. Quantitative controller: According to the set packaging quantity, the quantitative controller automatically controls the output of zinc powder. The quantitative controller uses a high-precision electronic scale and flow meter to monitor the packaging accuracy to ensure that the weight and volume of each packaging are consistent; Filling port and conveying device: The filling port is connected to the automated conveying system to ensure that the zinc powder enters the packaging bag accurately and stably. The conveying system adopts a spiral conveying device with an anti-blocking design to ensure smooth operation under high-density material processing; Packaging sealing system: The packaging sealing system is used to complete the packaging of zinc powder and ensure that the packaging is intact to prevent leakage or contamination. The packaging sealing system includes: Automatic packaging bag placement device: automatically selects packaging bags of appropriate specifications according to packaging requirements and positions them below the dispensing port; Automatic sealing device: After the subpackaging is completed, the packaging bag is sealed by a sealing device, using heat sealing technology to ensure that the bag mouth is firmly sealed to prevent zinc powder from leaking or getting wet; Package inspection and coding system: After each package is completed, the equipment uses a visual inspection system to check the sealing quality to ensure that each package meets the standards. In addition, the equipment automatically codes the production date and batch number to facilitate subsequent tracking and management.

2. A zinc powder collecting and packaging device according to claim 1, characterized in that: The visual inspection system comprises the following structure: Imaging system: including high-resolution industrial cameras, lenses and adaptive lighting equipment. The illumination equipment uses one or more of LED light sources and ultraviolet light sources. The lighting equipment can eliminate reflection interference and enhance the contrast of defects. The camera is fixed above the production line to capture high-definition images of the sealing area. Data processing module: using edge computing devices, using smart cameras or central processing units (CPU / GPU) for image analysis; a deep learning algorithm is set in the data processing module for image classification and defect detection, and unsealed, foreign body inclusion and uneven defects are classified through pattern recognition, and the deep learning algorithm uses a convolutional neural network; Control module: The control module is synchronized with the production line, receives sensor trigger signals, and coordinates the camera to capture images; when defects are detected, it sends instructions to the exclusion system to remove the defective products. The exclusion system uses a robotic arm or airflow device; Human-machine interface (HMI): Provides real-time display of sealing quality status, defect statistics, and system diagnostic information. The system is integrated with the packaging line through industrial communication protocols. Inspection results and historical data are stored in a local database or in the cloud for quality traceability and optimization analysis.

3. The zinc powder collecting and packaging equipment according to claim 1, characterized in that: The steps for testing sealing quality are as follows: Image acquisition: When the packaged product passes through the conveyor belt to the inspection area, the photoelectric sensor triggers the camera to take a high-definition image of the sealing part, and the adaptive algorithm adjusts the inspection parameters according to different materials; Image preprocessing: denoising, graying and edge enhancement to ensure clear target features; Defect analysis: The deep learning model extracts features from the image. Through the training sample library, the model identifies sealing integrity, foreign matter inclusions, or material wrinkles. Judgment and action execution: The test results are classified as "qualified" or "unqualified". If unqualified, the system sends a signal to the sorting mechanism to remove the defective products; Data storage and reporting: Test results are uploaded to the database and quality statistics reports are generated for production optimization.

4. The zinc powder collecting and packaging equipment according to claim 3, characterized in that: The steps to adjust the test parameters according to different materials are as follows: Material identification and feature extraction: During the production process, when the packaged product passes the inspection point, the system first identifies the material type and classifies it by image color, reflectivity, and surface texture features. Based on these features, the system will automatically select the appropriate image processing method and parameters; Image preprocessing and adaptive thresholding: Based on the different optical properties of the material, including transparency, reflectivity, and texture, the algorithm optimizes the input image by dynamically adjusting image preprocessing parameters, including contrast enhancement, grayscale, and edge detection. Feature matching and defect detection: After feature extraction, the algorithm will automatically match the defect pattern on the material surface, including scratches, bubbles, and wrinkles. For different materials, the shape, size, and contrast of the defects are different. Therefore, the system will judge the defect type of the material based on the trained model. The model uses the CNN model. Feedback and optimization: Once a defect is detected, the system will optimize parameters based on the current detection results; Dynamic adjustment and automatic update: As the production process progresses, the algorithm will adjust parameters based on real-time feedback, including detection accuracy and error rate. The process is dynamic and the algorithm can continuously learn and optimize to adapt to material changes in different production batches.

5. The zinc powder collecting and packaging equipment according to claim 1, characterized in that: It also includes a control and monitoring system, which includes the following structural components: Central control system: The PLC controller manages the operation of each module in a unified manner, adjusts the equipment operating parameters in real time, and the operation interface is displayed on the touch screen, which is convenient for operators to set and adjust parameters; Automatic fault detection and alarm system: This system can monitor the operating status of the equipment in real time, including motor load, conveyor belt speed, and filter status. Once an abnormality is found, the system will sound an alarm and automatically shut down to prevent equipment damage; Data acquisition and storage module: real-time recording of equipment production data and operating parameters to facilitate future data analysis and quality control.

6. The zinc powder collecting and packaging equipment according to claim 5, characterized in that: The real-time monitoring of the equipment operating status of the automatic fault detection and alarm system relies on sensor technology, data processing algorithms and edge artificial intelligence (EdgeAI) framework, which aims to capture motor load, conveyor belt speed, filter status parameters, analyze the difference between normal and abnormal conditions through algorithms, and trigger alarms and shutdown protection equipment when necessary. The following are its specific steps: Data collection: Use a variety of sensors, including vibration sensors, temperature sensors, and current sensors to collect equipment operation data in real time, and convert the collected data into signal format for further analysis; Data preprocessing: denoising, normalizing and feature extraction of raw data, using time-frequency domain analysis methods, including short-time Fourier transform and continuous wavelet transform, to convert time series data into spectrum or feature vector; Fault detection: Using machine learning or deep learning algorithms, convolutional neural networks or recursive neural networks can be used to analyze processed data in real time, identify abnormal patterns, and apply statistical models to determine whether the device is in a fault state based on the changing trend of the device behavior; Fault diagnosis and alarm: If an abnormality is detected, the system automatically classifies the fault type and sends the information to the control center. The system triggers an alarm mechanism, including sound, light or network notification, and may perform an automatic shutdown; Recording and Analysis: Failure data is stored for subsequent analysis to optimize models and improve prediction accuracy.

7. The zinc powder collecting and packaging equipment according to claim 6, characterized in that: In the process of automatic fault detection, the steps of time-frequency domain analysis are as follows: Data collection: Sensors monitor the operating status of equipment in real time and collect time series data, including motor vibration signals or temperature changes; Signal preprocessing: Remove noise or interference signals through filters to enhance feature information. Time-frequency transformation: Use short-time Fourier transform (STFT) to analyze non-stationary signals, divide the time series data into segments and apply Fourier transform to each segment to generate the time-frequency spectrum of the signal; Use continuous wavelet transform (CWT) to convolve the signal with the original signal through a series of scaled and translated mother wavelet functions to obtain time-frequency distribution information. Continuous wavelet transform is suitable for processing complex signals containing transient features; Feature extraction: Extract key features from the spectrum or wavelet coefficients, including the rate of change of frequency or the time location of transient events; Anomaly detection: Classify the extracted features and identify abnormal patterns through machine learning algorithms or rule-based methods. Machine learning algorithms use support vector machines or deep learning. Alarm and response: The system triggers an alarm and performs predefined response operations after detecting an anomaly.

8. The zinc powder collecting and packaging equipment according to claim 1, characterized in that: It also includes an environmental protection and recycling system, which includes the following components: Dust recovery module: The dust collected by the dust collection bin and filtration system inside the equipment can be put back into production through the automated recovery system, reducing resource waste and lowering environmental pressure; Waste gas treatment module: For harmful gases in the production process, the equipment is equipped with a waste gas purification device, which uses activated carbon adsorption or catalyst decomposition technology to treat harmful gases and meet environmental emission standards.

Images