Production process of heat-shrinkable tool holder

Abstract

The application discloses a production process of a heat-shrinkable tool holder and belongs to the technical field of tool holder production. The specific steps of the production process are as follows: (1) receiving a tool holder engineering drawing to determine tool holder parameter information; (2) collecting real-time image information of a stainless steel processing process and performing image optimization; (3) performing cascade analysis on production images and interrupting an abnormal production process; (4) polishing the surface of a completed heat-shrinkable tool holder and testing the deviation precision of the tool holder; and (5) performing risk analysis on operation logs and block-storing tool holder information of the batch. The application can realize automatic control of tool holder production, does not need manual monitoring of production by workers, reduces the production difficulty of tool holders, effectively improves the production amount of tool holders, reduces the workload of workers, improves the use experience of workers, reduces the energy consumption of data processing, effectively improves the data processing efficiency, guarantees the data security, and prevents malicious tampering of data.

Description

TECHNICAL FIELD

[0001] The present application relates to the production technology field of tool holder, and particularly relates to a production process of heat-shrinkable tool holder. BACKGROUND

[0002] High-speed machining is being increasingly widely applied in the fields of aerospace, mechanical manufacturing, automobile and die & mould industry, etc. As a key component for connecting the tool and the spindle in the high-speed machining tool system, the tool holder directly affects the quality and efficiency of high-speed machining. At present, the tool holder has become a hotspot of attention and research by domestic and foreign experts and scholars. With the rapid development of advanced manufacturing technology and equipment technology, the speed of the spindle of machine tool is getting higher and higher, and the requirements for the machining precision and surface quality of parts are becoming increasingly demanding. Therefore, it is particularly important to invent a production process of heat-shrinkable tool holder.

[0003] The existing production process of heat-shrinkable tool holder needs manual monitoring of production by workers, and the production of tool holder is difficult, which increases the workload of workers. In addition, the existing production process of heat-shrinkable tool holder has high energy consumption in data processing, reduces the data processing efficiency, and has poor data security. Therefore, the present application provides a production process of heat-shrinkable tool holder. SUMMARY

[0004] The present application aims at solving the defects in the prior art and provides a production process of heat-shrinkable tool holder.

[0005] In order to achieve the above-mentioned purpose, the present application adopts the following technical scheme:

[0006] The production process of heat-shrinkable tool holder has the following specific steps:

[0007] (1) receiving the tool holder engineering drawing to determine the parameter information of the tool holder;

[0008] (2) collecting the image information of the stainless steel processing process in real time and optimizing the image;

[0009] (3) performing cascade analysis on the manufacturing image and interrupting the abnormal manufacturing process;

[0010] (4) polishing the surface of the manufactured heat-shrinkable tool holder and testing the deviation precision of the tool holder;

[0011] (5) performing risk analysis on the operation log and block-storing the batch tool holder information.

[0012] As a further scheme of the present application, the image optimization in step (1) has the following specific steps:

[0013] Step one: the collected image information is extracted frame by frame to obtain picture data, then the picture data is divided into blocks according to the display ratio, then the high frequency components in the data are analyzed and extracted by Fourier transform, and the data are smoothed by Gaussian filtering;

[0014] Step two: the average value of the gray value of each picture data is calculated, then the gray value of each group of pixels in the divided picture data is compared with the calculated average value, and all the pixels with a gray value greater than the average value are constructed into a segmentation target, and all the pixels with a gray value less than the average value are constructed into a background of the segmentation image.

[0015] As a further scheme of the application, the Fourier transform in step one is specifically transformed as follows:

[0016]

[0017]

[0018] Wherein, u and v are frequency variables, x and y are the coordinates of each pixel point of the picture data, formula (1) is the Fourier forward transform, and formula (2) is the Fourier inverse transform.

[0019] As a further scheme of the application, the cascade analysis in step (3) is specifically as follows:

[0020] Step ①: obtain the past handle manufacturing information and integrate it into a sample data set, then calculate the standard deviation of the sample data set to remove abnormal sample data, standardize the remaining data, and then convert each group of processed data to a specified interval by normalization method;

[0021] Step ②: a convolutional neural network is constructed, and the sample data set is input into the neural network input layer to obtain the linear combination output by the hidden nodes of the output layer, and the least squares recursive method is used to define the energy function of the convolutional neural network multi-round learning, when the energy function is less than the target error, the training process is ended and the analysis neural model is output;

[0022] Step ③: the analysis neural model calculates the interval time of each image information actual video frame, and establishes a motion model through Kalman filtering theory, simultaneously obtains the motion state of the tracking target in real time through the established motion model, then collects the motion state of the stainless steel in the current video frame, and constructs a prediction equation to estimate the motion state of each stainless steel in the next video frame;

[0023] Step 4: Scale normalization is performed on each picture data by image pyramid, feature extraction is performed on each group of picture data, and the extracted features are sent to a bidirectional feature pyramid for feature fusion, and classification regression is performed on the fusion results to obtain a detection frame, the relevant picture data is enlarged and cropped according to the produced detection frame, and then the cropped stainless steel image is analyzed, if there is a manufacturing deviation, the manufacturing process is interrupted and feedback is given to the relevant staff.

[0024] As a further scheme of the present application, the standard deviation calculation formula in step 1 is as follows:

[0025]

[0026]

[0027] Where, v n is the data deviation of the sample data set, s is the standard deviation, if the deviation v n of any data x i satisfies |v n |>3σ, it is judged that the data is abnormal data and is rejected.

[0028] The interval time calculation formula in step 3 is as follows:

[0029]

[0030]

[0031] In the formula, Δt k+1 represents the interval time between two groups of video frames, represents the delay time between the down-sampled video frames and the original video stream, represents the consumption time of the tracking algorithm processing video frames.

[0032] As a further scheme of the present application, the risk analysis in step (5) is as follows:

[0033] Step I: Deploy relevant log collection plug-ins on different system monitoring platforms or obtain operation logs recorded in different system monitoring platforms through syslog servers, and filter out log information meeting the preset conditions of the staff;

[0034] Step II: Process the remaining operation logs into uniform format data, match the processed user operation behavior and abnormal behavior characteristics in the log, generate corresponding alarm information according to the matching result, calculate the risk score of each alarm information and output the calculation result, then feedback the alarm information to the relevant staff and interrupt the relevant operation process.

[0035] As a further scheme of the present application, the block storage in step (5) has the following specific steps:

[0036] First step: the handle information and manufacturing information are pre-processed into a unified format, and are processed into qualified blocks, when entering the network, each node in the blockchain network generates a local public-private key pair as its identification code in the network, when a node waits for the local role to become a candidate node, it broadcasts the leadership application to other nodes in the network and sends it;

[0037] Second step: after the application is passed, the candidate node becomes the leader node, and the other nodes become the follower nodes, then the leader node broadcasts the block record information, the follower nodes broadcast the received information to other follower nodes after receiving the information and record the repetition times, and use the information with the most repetition times to generate the block header, and send a verification application to the leader node;

[0038] Third step: after the verification is passed, the leader node sends an addition command and enters a sleep period, which cannot apply to become a leader node again during the sleep period, until the end of the sleep period, the follower nodes receive the confirmation information, add the newly generated blocks to the blockchain, and return to the candidate identity.

[0039] Compared with the prior art, the present application has the following advantages:

[0040] 1、The production process of the heat-shrinkable handle obtains the past handle manufacturing information and integrates it into a sample data set, then calculates the standard deviation of the sample data set to remove abnormal sample data, standardizes the remaining data, obtains a related analysis neural model, then the analysis neural model calculates the interval time of each image information actual video frame, establishes a motion model through Kalman filtering theory, and constructs a prediction equation to estimate the motion state of each stainless steel in the next video frame, scales the image pyramid to normalize each picture data, extracts features from each group of picture data, and then sends the extracted features to a bidirectional feature pyramid for feature fusion, and classifies and regresses the fusion result to obtain a detection frame, according to the detection frame of the production, the related picture data is enlarged and cropped, then the cropped stainless steel image is analyzed, if there is manufacturing deviation, the manufacturing process is interrupted and feedback to the relevant staff, which can realize automatic control of handle production, without manual monitoring of production by staff, reduce the difficulty of handle production, effectively improve the handle production capacity, and reduce the workload of staff, improve the user experience of staff.

[0041] 2. The manufacturing process of this heat-shrinkable tool holder preprocesses the tool holder information and manufacturing information into a unified format and processes them into blocks that meet certain conditions. When joining the network, each node in the blockchain network generates a local public-private key pair as its own identifier in the network. When a node is waiting to become a candidate node, it broadcasts a leader application to other nodes in the network and sends it. After the application is approved, the candidate node becomes the leader node, and other nodes become follower nodes. Then, the leader node broadcasts the block record information. After receiving the information, the follower nodes broadcast the received information to other follower nodes and record the number of repetitions. They use the information with the most repetitions to generate a block header and send a verification application to the leader node. After the verification is approved, the leader node sends an add command and enters a dormant period. During the dormant period, it cannot apply to become a leader node again until the dormant period ends. After receiving the confirmation information, the follower nodes add the newly generated blocks to the blockchain and return the candidate identity. This can reduce data processing energy consumption, effectively improve data processing efficiency, and at the same time ensure data security and prevent data from being maliciously tampered with. Attached Figure Description

[0042] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof.

[0043] Figure 1 This is a flowchart illustrating the manufacturing process of a heat-shrinkable tool holder proposed in this invention. Detailed Implementation

[0044] Example 1

[0045] Reference Figure 1 A manufacturing process for heat-shrinkable tool holders, the specific steps of which are as follows:

[0046] Receive the tool holder engineering drawing to determine the parameter information of the tool holder.

[0047] Real-time acquisition of images of the stainless steel processing process and image optimization.

[0048] Specifically, the acquired image information is extracted frame by frame to obtain image data. Then, the images are divided into blocks according to their proportions. After that, the high-frequency components in each block of image data are analyzed and extracted using Fourier transform, and smoothed using Gaussian filtering. The average gray value of each image data is calculated. Then, the gray value of each pixel in each block of image data is compared with the calculated average value. Pixels with gray values ​​greater than the average value are considered as segmentation targets, and pixels with gray values ​​less than the average value are considered as the background of the segmented image.

[0049] It should be further explained that the specific formula for the Fourier transform is as follows:

[0050]

[0051]

[0052] wherein u and v are frequency variables, x and y are coordinates of each pixel point of the picture data, formula (1) is Fourier transform, and formula (2) is inverse Fourier transform.

[0053] Cascade analysis is performed on the production image, and the abnormal production process is interrupted.

[0054] Specifically, the past handle production information is obtained and integrated into a sample data set, and then the standard deviation of the sample data set is calculated to remove abnormal sample data, the remaining data is standardized, and each group of processed data is converted to a specified interval by a normalization method, a convolutional neural network is constructed, and the sample data set is input into the neural network input layer to obtain the linear combination of the hidden node output of the output layer, and the least squares recursive method is used to define the energy function of the convolutional neural network multi-round learning, when the energy function is less than the target error, the training process is ended and the analysis neural model is output, the analysis neural model calculates the interval time of each image information actual video frame, and a motion model is established through Kalman filtering theory, and the motion state of the tracking target is obtained in real time through the constructed motion model, then the motion state of the stainless steel in the current video frame is collected, and a prediction equation is constructed to estimate the motion state of each stainless steel in the next video frame, the image pyramid is used to perform scale normalization processing on each picture data, and feature extraction is performed on each group of picture data, then the extracted features are sent to a bidirectional feature pyramid for feature fusion, and classification regression is performed on the fusion result to obtain a detection frame, the relevant picture data is enlarged and cropped according to the detection frame, then the cropped stainless steel image is analyzed, if there is a production deviation, the production process is interrupted and feedback is given to the relevant staff.

[0055] The specific calculation formula of the standard deviation is as follows:

[0056]

[0057]

[0058] wherein v n is the data deviation of the sample data set, s is the standard deviation, if the deviation v n of any data x i satisfies |v n |>3σ, the data is judged as abnormal data and is removed.

[0059] The specific calculation formula of the interval time is as follows:

[0060]

[0061]

[0062] wherein, Δt k+1 represents the interval time between two groups of video frames, represents the delay time between the down-sampled video frames and the original video stream, represents the consumption time of the tracking algorithm processing the video frames.

[0063] Embodiment 2

[0064] Referring to Figure 1 A production process of a heat-shrinkable tool holder, the specific steps of which are as follows:

[0065] Polish the surface of the finished heat-shrinkable tool holder and test the deviation accuracy of the tool holder.

[0066] Risk analysis is performed on the operation log, and the batch tool holder information is stored in blocks.

[0067] Specifically, the relevant log collection plug-ins are deployed on the monitoring platforms of different systems, or the operation logs recorded in the monitoring platforms of different systems are obtained through a syslog server, and log information meeting the preset conditions of the staff is screened out, the remaining operation logs are processed into data in a unified format, the user operation behavior and abnormal behavior characteristics recorded in the processed logs are matched, and corresponding alarm information is generated according to the matching result, and the risk score of each alarm information is calculated and the calculation result is output, then the alarm information is fed back to the relevant staff, and the relevant operation process is interrupted.

[0068] Specifically, the tool holder information and manufacturing information are pre-processed into a unified format and processed into blocks meeting the conditions. When entering the network, each node in the blockchain network generates a local public-private key pair as its identifier in the network. When a node waits for the local role to become a candidate node, it broadcasts a leadership application to other nodes in the network and sends it. After the application is approved, the candidate node becomes the leader node, and the other nodes become the follower nodes. Then the leader node broadcasts the block record information, and the follower nodes broadcast the received information to other follower nodes and record the repetition times after receiving the information, and use the information with the most repetition times to generate a block header, and send a verification application to the leader node. After verification, the leader node sends an add command and enters a sleep period, during which it cannot apply to become a leader node again. Until the end of the sleep period, the follower nodes receive confirmation information, add the newly generated blocks to the blockchain, and return to the candidate identity.

Claims

1. A manufacturing process for a heat-shrinkable tool holder, characterized in that, The specific steps of this production process are as follows: (1) Receive the tool holder engineering drawing to determine the parameter information of the tool holder; (2) Real-time acquisition of image information of stainless steel processing and image optimization; (3) Perform cascade analysis on the produced images and interrupt the abnormal production process; (4) Grind the surface of the finished heat-shrink tool holder and test the deviation accuracy of the tool holder; (5) Perform risk analysis on the operation log and store the batch of tool holder information in blocks; The specific steps of the cascade analysis described in step (3) are as follows: Step 1: Obtain past tool holder manufacturing information and integrate it into a sample dataset. Then, calculate the standard deviation of the sample dataset to remove outlier data. Standardize the remaining data and then normalize each group of processed data to a specified range. Step 2: Construct a set of convolutional neural networks and input the sample dataset into the input layer of the neural network to obtain a linear combination of the output layers as hidden node outputs. Define the energy function of the convolutional neural network for multiple rounds of learning using the least squares recursive method. When the energy function is less than the target error, the training process ends and the analysis neural model is output. Step 3: Analyze the neural model to calculate the interval time of each video frame of the image information, and then establish a motion model through Kalman filtering theory. At the same time, the motion state of the tracked target is obtained in real time through the constructed motion model. Then, the motion state of the stainless steel in the current video frame is collected, and a prediction equation is constructed to estimate the motion state of each stainless steel in the next video frame. Step 4: Scale normalize each image data using an image pyramid, extract features from each group of image data, and then feed the extracted features into a bidirectional feature pyramid for feature fusion. Simultaneously, perform classification and regression on the fusion results to obtain detection boxes. Based on the generated detection boxes, enlarge and crop the relevant image data. Then, analyze the cropped stainless steel image. If there are any production deviations, interrupt the production process and provide feedback to the relevant staff.

2. The manufacturing process of a heat-shrinkable tool holder according to claim 1, characterized in that, The specific steps of image optimization described in step (1) are as follows: Step 1: Extract the acquired image information frame by frame to obtain image data. Then, divide the image data into blocks according to the proportion of each image data. After that, analyze and extract the high-frequency components in each block of image data through Fourier transform, and smooth the data through Gaussian filtering. Step 2: Calculate the average grayscale value of each image data. Then, compare the grayscale value of each group of pixels in each segmented image data with the calculated average value. Pixels with grayscale values ​​greater than the average value are considered as segmentation targets, and pixels with grayscale values ​​less than the average value are considered as the background of the segmented image.

3. The manufacturing process of a heat-shrinkable tool holder according to claim 2, characterized in that, The specific Fourier transform formula described in step one is as follows: （1） （2） Where u and v are frequency variables, x and y are the coordinates of each pixel in the image data, formula (1) is the forward Fourier transform, and formula (2) is the inverse Fourier transform.

4. The manufacturing process of a heat-shrinkable tool holder according to claim 1, characterized in that, The specific formula for calculating the standard deviation mentioned in step ① is as follows: （1） （2） in, For the data bias of the sample dataset, Let be the standard deviation, if any data deviation satisfy If the data is abnormal, it will be removed. The specific formula for calculating the interval time mentioned in step ③ is as follows: （3） （4） In the formula, This represents the time interval between two sets of video frames. This represents the time delay between the downsampled video frame and the original video stream. This represents the time consumed by the tracking algorithm to process video frames.

5. The manufacturing process of a heat-shrinkable tool holder according to claim 1, characterized in that, The specific steps of the risk analysis described in step (5) are as follows: Step 1: Deploy relevant log collection plugins on the monitoring platforms of different systems or obtain the operation logs recorded on the monitoring platforms of different systems through the syslog server, and filter out the log information that meets the preset conditions of the staff; Step II: Process the remaining operation logs into a unified data format, then match the user operation behavior recorded in the processed logs with the abnormal behavior characteristics, and generate corresponding alarm information based on the matching results. At the same time, calculate the risk score of each alarm information and output the calculation result. Then, feed the alarm information back to the relevant staff and interrupt the relevant operation process.

6. The manufacturing process of a heat-shrinkable tool holder according to claim 1, characterized in that, The specific steps of block storage described in step (5) are as follows: Step 1: Preprocess the handle information and manufacturing information into a unified format and process them into blocks that meet the conditions. When joining the network, each node in the blockchain network generates a local public-private key pair as its own identifier in the network. When a node is waiting for its local role to become a candidate node, it broadcasts a leadership application to other nodes in the network and sends it. Step 2: After the application is approved, the candidate node becomes the leader node and the other nodes become follower nodes. Then the leader node broadcasts the block record information. After receiving the information, the follower nodes broadcast the received information to other follower nodes and record the number of repetitions. They then use the information with the most repetitions to generate the block header and send a verification request to the leader node. Step 3: After verification, the leader node sends an add command and enters a dormant period. During the dormant period, it cannot apply to become a leader node again until the dormant period ends. After the follower nodes receive the confirmation information, they add the newly generated blocks to the blockchain and return the candidate identity.

Images