KR101105790B1 - Failure recognition system - Google Patents

Abstract

The present invention relates to a failure determination method for detecting a case where a failure occurs as well as a post-action by detecting a failure.

The failure determination method according to the present invention includes a data setting step (S510) for setting preliminary failure data for which future failure is expected, for each measurement means; A data measuring step (S520) of extracting actual data measured from the facility by using one or more measuring means; A data comparison step (S530) for comparing the actual data measured in the data measurement step (S520) and the preliminary bad data set in the data setting step (S510); A post-action step (S600) of stopping the facility or expressing an abnormal phenomenon when the actual data match the preliminary bad data as a result of the comparison in the data comparing step (S530); And the like. According to the present invention as described above, there is an advantage that can prevent the loss of material costs and damage to the equipment due to defects.

Defective, good, judgment, advance, spare

Description

Failure recognition method {Failure recognition method}

The present invention relates to a failure determination method, and more particularly, to a failure determination method for detecting a case where a failure has occurred as well as a case where a failure is expected and performing a follow-up action.

In the related art, in order to determine the quality of a product formed by a machining operation such as a press, the operator visually judges the product. In other words, by visually recognizing each product by the individual judgment of the operator to determine the defect and good quality.

Therefore, in the case of the worker's judgment, because the product must be inspected one by one, there is a problem that the work efficiency is reduced. In addition, inspection standards may change depending on the worker's situation. That is, there is a problem that the quality judgment is subjective because the standard of defective and good products may change according to the condition of the worker on that day.

Recently, a method of recognizing the shape of a product by artificial intelligence or neural network theory has emerged. However, even in the case of artificial intelligence or neural network theory, all the relevant criteria should be programmed and specified basically, such as setting the product and assigning a variable for it.

In other words, even when judging the shape of a product by conventional artificial intelligence or neural network theory, input neurons of many neural networks are required, so that the learning time of such neural networks is excessively excessive and computational time is required.

In addition, since the shape determination is possible only for a specific corresponding product, it is cumbersome to modify the program each time and set the standard of the corresponding product in order to determine the shape of a plurality of products.

In addition, since there is no separate learning system for learning by judging good quality and defects, in order to add a bad recognition image to a new or existing bad determination method, it is necessary to add images one by one to recognize good quality and badness. There is a problem in that it is possible to determine the difference between two images only when the image exactly matches.

In addition, in the conventional method, there is a problem of waste of materials and damage of equipment by taking action after a failure has already occurred.

An object of the present invention is to solve the problems of the prior art as described above, and to provide a defect determination method capable of preventing a large quantity of defects in advance through a measuring means (sensor).

Another object of the present invention is to provide a defect determination method capable of determining an abnormal precursor phenomenon through digital data as well as determining defects or abnormal symptoms through an image or sound.

The failure determination method according to the present invention for achieving the above object includes a data setting step of setting, by measurement means, preliminary defective data for which future failure is expected; A data measuring step of extracting actual data measured from the facility by using one or more measuring means; A data comparing step of comparing the actual data measured in the data measuring step with the preliminary bad data set in the data setting step; And a post-action step of stopping the facility or expressing an abnormal phenomenon to the outside when the actual data matches the preliminary bad data as a result of the comparison in the data comparing step.

According to the failure determination method of the present invention as described above, various advantages arise.

First, because the quality can be determined by learning without a separate operation for a number of products, the work efficiency is improved, the quality function is improved. That is, according to the present invention, even when the type of the product is different, the corresponding type is automatically determined by learning to determine whether the product is defective based on the quality standard of the corresponding product. Thus, according to the present invention, since it is possible to learn and judge closer to human beings, quality judgment work becomes more effective.

Second, in the learning process, the number of times that an image sensor reads randomly from several images and the maximum distance that the vectors within the maximum distance range are considered to occur at the same time. It is a necessary parameter in the process of inference and inference, and provides an optimal value for Sigma that specifies the range of cases that occur simultaneously as the standard deviation of the normal distribution. Therefore, there is an advantage that the time required for learning is shortened and the accuracy of the good judgment of the product is improved.

Third, in the present invention, by comparing the sound generated during the production of the product with the normal state to grasp the abnormal symptoms. Therefore, since the failure phenomenon is also determined by the image and sound, there is an effect that the failure determination ability is doubled.

Fourth, in the present invention, a bad precursor phenomenon is identified by a large number of digital data. In other words, by grasping the shape of the numerical value, such as the hydraulic pressure of the equipment, it is determined that the bad rolling phenomenon. Therefore, since the defect is prevented in advance, there is an advantage that it is possible not only to waste materials but also to prevent failures such as damage to equipment.

Hereinafter, the defect determination method by this invention is demonstrated.

The present invention relates to a method of learning and reasoning in a system for determining quality by a recognition engine based on HTM.

The Hierarchical Temporal Memory (HTM) model mimics the mechanisms of the neocortex that govern human intelligence. In other words, Hawkins is a computational model of the neocortex proposed by modeling the behavior of neurons and synapses in the human brain.

This Hierarchical Temporal Memory (HTM) model is an artificial neural network based on conventional heuristic search (Artificial intelligence) or simple neurons (Artificial neural) network (ANN)).

In HTM, a basic unit of the neocortex consisting of six layers is defined as a node, which is a basic unit of the network, that is, a neocortex column as a unit node (which is approximately 1,000 units). Consisting of neurons and 1,000,000 synapses) and hierarchical networks, using which the world's spatial-temporal pattern information (especially temporal information) ) So that you can make intelligent judgments efficiently.

Therefore, the recognition of the world is more efficient and more robust than the existing AI or ANN.

More specifically, HTM, in hierarchical sense, is organized into hierarchical nodes, each node having a learning and memory function and an internal algorithm for performing the same. The lower level node receives a large amount of input from the outside, processes it, and sends the processing result to the next higher level.

Ii) In the sense of temporal, during training, the object is represented in the form of changing over time, e.g. while training a picture application, the image appears from top to bottom and from left to right as if the image is It is represented as moving over time. And this temporal factor is very important, and the algorithm is expected to expect inputs that change over time.

I) In the sense of memory, an application operates in two stages. That is, it trains memory and uses it for inference. During training, the HTM network learns to recognize patterns in the input it is accepting, where it is trained separately at each level of the hierarchy.

In this way, when the network is fully trained, each level in the hierarchy will systematically have all the objects in its world in memory.

Next, while inferring and recognizing, the HTM network receives the new object and determines that it is most likely one of the objects it already knows.

This hierarchical temporal memory (HTM) is a new computing paradigm that software-models the neocortex of the human brain to enable computers to solve problems that are easy for humans but hard for machines to solve.

With HTM, you can discover patterns in the incoming sensor data, and in some simple applications (image recognition, diagram recognition, wave recognition), the performance excellence has already been demonstrated.

The present invention, by using such HTM, can be utilized in the judgment for quality assurance of the manufacturing plant. That is, the system can determine the quality situation in or between processes using data collected from various sensors (image, temperature, humidity, dust).

As described above, in a recognition engine based on HTM, when a plurality of images are mixed, the plurality of images are classified and judged by similar types. This is as if children were trained and recognized by pictures or in real life, and similar shapes would be recognized as trains and planes, regardless of their size.

1 is a block diagram showing a schematic configuration of a failure determination method according to the present invention.

As shown in the drawing, the defect determination method according to the present invention includes an image determination method (S10) for determining whether a defect is caused by an image, and a sound determination method for determining a defect state by the presence or absence of an abnormal sound generated during product production. (S400), and the failure pre-recognition method (S500) for grasping the abnormal phenomenon of the facility and predicting the occurrence of failure in advance. In addition, when a defect is detected or expected or an abnormal symptom is detected by at least one of the image determination method (S10), the sound determination method (S400), and the failure preliminary recognition method (S500), a post-action step (S600). Is taken.

Detailed configuration of each method and the post-action step (S600) will be described in detail below.

2 is a block diagram showing a schematic configuration of the image determination method S10.

As shown in this, the image determination method (S10), the learning step (S50) to obtain the information on the goods and defects by learning, and the setting step of setting the reference information of the system for determining the quality of the product (S100) ), The product inspection step (S150) for inspecting the product according to the criteria set in the setting step (S100), and the image of the product measured in the product inspection step (S150) to specify the type of product and the exact product shape Product recognition step (S160) to recognize, and the product quality determination step (S170), etc. to determine the defective and good quality based on the information acquired in the learning step (S50) the image finally recognized in the product recognition step (S160) Is done.

And, if a product failure occurs by the product quality determination step (S170), the post-action step (S600) for controlling the equipment according to the control method set in the setting step (S100) and at the same time notifying the fact to the outside. Proceeds.

The learning step (S50) is shown in FIG. That is, FIG. 3 is a flowchart specifically showing the learning step S50.

As shown therein, the learning step (S50), the item selection process (S52) for selecting the article to learn, the type determination process (S54) for determining whether to continue learning or new learning and If the result of the determination in the type determination process (S54) is a new learning, a category formation process (S56) for forming a new category, and the result of the determination in the type determination process (S54), if the existing learning continues the existing category A category selection process (S58) for selecting a, a ratio selection process (S60) for selecting a ratio of the number of data to be learned for each category, and a folder selection for selecting a folder for storing image data (learned) for each category Process (S62), the input process (S64) for receiving and storing the image, the test process (S66) for testing using the actual product, and the verification process (S68) for checking the results in the test process (S66) And award In the result determination process (S70) for determining whether the test result confirmed in the previous confirmation process (S68) satisfies the required learning result value, and learning is satisfied if the determination result in the result determination process (S70) is satisfied. It consists of a storage process (S72) for storing the result.

4A shows an example of a main screen for performing the learning step S50. As shown in this figure, good and bad, learning data, test data, etc. are collectively managed.

The type judging process S54 is a process of determining whether the content to be learned is about a new product or whether to learn additionally about a product that has been previously learned.

Therefore, according to the determination result in the type determination process (S54), the category forming process (S56) or category selection process (S58) is performed.

The category forming process (S56) is a process of forming a new category in order to keep the learned image data of the new product.

In this process, the product is largely divided into a good product category and a bad category, and the bad category is composed of various categories such as bad 1, bad 2, ..., etc., depending on the type of bad.

For example, in detail, Defect 1 is a type in which a part of the product is torsional, Defect 2 is a type in which a part of the product is cut and chipped, and 'Other' is not included in any type but is not good This category belongs.

4B is a screen showing an example of a type of category formed by the category forming process (S56).

The category selecting process (S58) is a process of selecting a category of a corresponding product from among a plurality of existing categories, and selecting and specifying a category to store a newly input learning image.

After the category forming process S56 is completed, a ratio selection process S60 is next performed.

The ratio selection process (S60) is a process of determining the ratio of each category to perform the learning. That is, the process of selecting the ratio of the number of images to be trained. Here, the ratio of the number of images and data to be learned about good quality and defects is determined by Set the ratio similar to the ratio. As such, it is desirable that the ratio of the basic data between good and bad for learning to be the same as the ratio occurring in the actual field.

In the ratio selection process (S60), it is most preferable that the ratio of the number of image data data to be learned for each type of good: bad 1: bad 2 is 4: 1: 1. This ratio selection process (S60) will be described in detail below. 4C illustrates an example of a screen for selecting a ratio between good and bad images according to the ratio selection process S60.

The input process (S64) is a process of judging good or bad by inputting an already stored image, or a process of receiving good and defective items as an image through actual testing and storing them.

In other words, in the case where an image of a product has been scanned and stored in a computer in advance, it is a process of determining whether a good image or a defective image is obtained by loading such image data.

If the image is not stored in advance, the product image is read while actually sending the product to a line, and the image is classified and stored according to the type. That is, in the state where the image sensor is operated, a good or defective product is actually flowed in a line, and an image is received and stored through the image sensor.

When the input process (S64) as described above proceeds with the information about the defects, and the information about the goods, the test process (S66) for actually testing the product in accordance with the learning information on the good (良 不) proceeds. .

The test process (S66), in fact, to flow the product (Line), to determine whether the value read through the sensor is good or bad. In this process, the test is performed while adjusting various parameters. In other words, a test is performed by adjusting an optimal parameter value. In FIG. 4D, an example of a screen for setting the parameter value is shown. A specific example of the test process S66 will be described below.

In addition, the checking step S68 is a step of checking whether the result of the test in the test step S66 is properly performed. That is, it is a process of confirming whether good quality and poor quality are judged properly according to the form learned above.

4E shows an example of a screen for checking a test data value generated through the test process S66 as described above.

The result determination process (S70) is a process of determining whether the test result confirmed in the verification process (S68) is a required result value. That is, for example, when at least 900% accuracy is required, it is a process of determining whether the test result in the checking process S68 has an accuracy of 90% or more.

The storing process S72 is a process of storing the learning result when the determination result in the result judging process S70 is satisfied, and applying it to a future site. That is, the results are compared and then the learning results determined to be the most accurate are transmitted to the failure determination method and stored, so that they can be used in the field application.

 On the other hand, if the result of the determination in the result determination process (S70), the result is not satisfied, the adjustment process (S74) to check and adjust the cause proceeds. In other words, if the test result does not have the required accuracy, the error type is analyzed to determine why there is a decision error, and actions are taken to correct the error.

For example, if an error occurs due to an incorrect selection of the corresponding item, the process returns to the item selection process (S52) and selects the correct item again to proceed with learning.

If it is determined that the error is concentrated in some of the types of the good or the defective product because the ratio of the good product and the defective product is wrong, the process returns to the ratio selection process S60 to adjust the ratio and learn again.

5 shows an experimental procedure for finding an optimal number of data in the ratio selection process S60. That is, FIG. 5 is a chart and a graph showing the percentage (%) of the defective 1, the defective 2 and the good when the ratio of the number of data between the good and the bad 1 and the bad 2 is different.

As shown in FIG. 1, at the level 1, defective ratio 1: defective 2: good quality = 120: 120: 120, the ratio of good quality is poor 1: poor 2: good quality = 60.86%: 41.25%: 62.50%, and at Level 2, the ratio of good or bad when bad 1: bad 2: good = 60: 60: 120 is bad 1: bad 2: good = 70.00%: 27.50%: 80.00% The ratio of good and bad in the case of bad 1: bad 2: good quality = 30: 30: 120 is bad 1: bad 2: good quality = 51.67%: 16.25%: 80.38%.

Therefore, as a result of comparing the ratio of good and defective goods with the ratio of actual good and bad goods, in the ratio selection process (S60), when classifying types of good goods and bad goods 1 and bad goods, image data ( Data) The ratio of number is good: good: bad 1: bad 2 = 4: 1: 1.

In other words, despite the fact that the ratio of good products is high and the ratio of defective items is very low in the actual field, the level of the level 1 and the level 2 in the learning experiment is too high. Therefore, in such a case, there is a closed end in which many good products are judged as defective goods and frequent interruption of the production line occurs. Therefore, the ratio of the number of image data (Data) is good quality: bad quality 1: bad quality 2 = 4: 1: 1, and can improve the accuracy of good quality judgment.

6A to 8B, the results of looking at the accuracy according to the change of the parameter are shown while performing the test process S66. In other words, it shows experimentally which parameter value is best to use for more accurate judgment.

6A and 6B illustrate a change in the number of iterations, which means the number of times that an image sensor is randomly read from a plurality of images in the test process S66. It shows the test result that tested the accuracy. That is, FIG. 6A is an experimental result value when the 'iteration number' is equal to all of the levels 1 to 3, and FIG. 6B shows the 'iteration number' equal to the level 1 and the level 2. This is the result when only Level 3 is changed.

According to these experimental results, in Fig. 6a, the level 1 to level 3 are all so weak that the change in the 'iteration number' is meaningless, and in Fig. 6b, the change in the 'iteration number' is the accuracy of the good judgment. It can be seen that it does not affect Accuracy. In conclusion, it is judged that the high and low number of iterations does not affect the accuracy of good and bad judgments.

Therefore, if the number of iterations is unnecessary, the performance of the computer hits the limit and the learning time becomes long. Therefore, the number of iterations is preferably 5,000 or more. More specifically, the minimum number of iterations is about 5,000.

7A to 7C show experimental results obtained by obtaining an optimum value of 'MaxDistance', which is a parameter of the HTM recognition engine used in the present invention. Here, 'MaxDistance' is the maximum distance from the stored values of the input vectors while learning the Euclidean distance, so that the vectors within the maximum distance range are generated simultaneously. Means the distance considered to be. In other words, 'MaxDistance' means the maximum distance for comparing the image to be compared with the original image from one reference sensor.

This 'MaxDistance' can be compared more accurately than if the value is set small. However, if the value of the maximum distance (MaxDistance) is set too low, the performance of the computer hit the limit can be impossible to learn and the learning time is excessively excessive. Therefore, it is desirable to find the smallest value within the range that does not affect the accuracy of judging good or bad.

In consideration of such peripheral requirements, looking at the experimental results, it is preferable that the 'MaxDistance' is about 1400 to 1900. In other words, the value of 'MaxDistance' is relatively small and the ratio of good products is high, so the level of accuracy is considered to be quite high. Among them, the value of 'MaxDistance' is 1,500. It can be seen that this is more optimal.

8A and 8B are experimental data comparing an accuracy result according to a change of a 'sigma' value. Sigma is a necessary parameter in the process of inference and specifies the range of cases that occur simultaneously as the standard deviation of the normal distribution. In other words, 'Sigma' is a parameter that sets the noise level within the comparison range, and since the accuracy varies greatly depending on the noise level of the image to be compared, a suitable sigma is included. Setting the value is important and you need to find the Sigma value with the highest accuracy.

As shown in the experimental results, the 'sigma' value greatly influences the accuracy. In addition, the 'Sigma' value can be seen that the highest accuracy (Accuracy) before and after the square root of the 'MaxDistance'. That is, when the 'MaxDistance' in the experiment is 1,500, it can be seen that the highest accuracy (Accuracy) around 38.7 corresponding to the square root of 1,500.

On the other hand, the setting step (S100), the facility information input process (S110) for inputting the product production equipment information for producing the product and the sensor information for measuring the environmental conditions such as pressure or temperature, and the product inspection in some way Inspection condition setting process (S120) for setting a condition in terms of time or quantity whether to execute, and loading the information learned in the learning step (S50) that is a criterion of defective and good quality for each of a plurality of products (Loading The quality standard setting process (S130) and the equipment control method setting process (S140) for setting a control method for how to control the operation of the equipment in the event of a failure occurs.

The facility information input process (S110) is a process of inputting equipment for producing a product and other sensor information. That is, the quality judgment system according to the present invention is to determine whether the defective image by determining the image of the product in the process of producing the product, it should be linked with the facility. Therefore, information on these facilities must be entered.

Further, in the quality judgment system according to the present invention, various sensors can be installed in order to determine the product quality during the process. That is, not only an image sensor, but also sensors for sensing temperature, humidity, dust, and the like may be installed. In this case, information on these various sensors is input in advance.

In addition, the information on the facility and the information on the various sensors should be matched with each other.

In the inspection condition setting process (S120), it is set how to inspect the product. Here, it is common to set the conditions in time or in terms of quantity. That is, for example, it is possible to set a quantitative inspection method to execute the inspection once every 10 seconds or once every minute or to inspect one of ten or twenty products.

The quality standard setting process (S130) is a process of setting the information to determine the good or bad of the product. That is, it is a process of setting the criterion by inputting information on the quality and defective state of the product learned in the learning step (S100).

The facility control method setting process (S140) is a process of setting the information on how to control the facility in the event of a failure and to notify the outside of the failure condition when such a failure occurs. In this case, when a failure occurs continuously, the equipment is stopped, and the status is set to notify the remote user through a messenger (SMS) at the same time as a notification to the outside through a buzzer (Buzzer). That is, it is set to transmit a bad state to the user's mobile phone through the messenger (SMS).

Here, the continuous failure state may be variously changed depending on the degree of quality requirements of the product or the user's will. In other words, if three defects are continuously reported, stop the equipment and notify the outside at the same time, or if three defects are continuously, only the function of notifying the outside is activated, and in case of ten consecutive defects Various settings are possible, such as to stop the equipment.

The product inspection step (S150), the image reading process (152) for taking an image of the product to read the image (Image) and the image editing process for editing the image read in the image reading process (152) (154) And an image storing process 156 for storing the image manipulated by the image editing process 154.

9 is a detailed process of the product inspection step (S150) is shown sequentially.

As shown in the drawing, the image reading process 152 of the product inspection step S150 is a process of reading an image of a product photographed by the photographing apparatus 200 into a BMP file. That is, a plurality of products are simultaneously input as one image, and at this time, the input is at a resolution of 1024 × 768.

And, the image editing process 154 of the product inspection step (S150), converts the image read by the image reading process 152 into a binary image, grasp the position of each product and at the same time cut out by region A first operation S154 ′ and a second operation S154 ″ for facilitating image determination by adjusting the brightness and contrast of the image passed through the first operation S154 ′.

The first operation S154 ′ is a crop process of cutting a plurality of products one by one as shown, and the second operation S154 ″ increases contrast by adjusting brightness and contrast. Contrast process.

The image storing process 156 of the product inspection step S150 is a process of storing the BMP image files which have undergone the image editing process 154, respectively. That is, as shown, the process of storing the image to each BMP image file having a resolution of 128 × 128.

FIG. 10 shows the result of change in accuracy according to the enhancement of contrast performed in the second operation S154 ″.

As shown in FIG. 2, when the contrast is enhanced, the recognition rate is increased as a whole. In other words, the recognition rate improved from 26.0% to 56.0 for defective 1, 20.9% to 23.0% for defective 2, and 90.0% to 92.5% for good products.

Therefore, in order to increase the accuracy, it is necessary to make the outline clear, and thus a method of increasing the contrast in a manner to make the outline clear is used.

11 schematically shows the configuration of the product recognition step (S160).

As shown in the drawing, the product recognition step (S160) loads an image file of 128 × 128 pixels input in the product inspection step (S150) and reads it as a 16 × 16 node. A second level process (S164) of reading a first level process (S162) and a result of the conversion into a 16x16 node (Node) by the first level process (S162) as an 8x8 node (S164). And a third level process S166 of reading the result of the transformation into an 8x8 node by the second level process S164 as a 4x4 node, and the third level process. And a fourth level process S168 for recognizing the entire image area converted into 4x4 nodes at a time (S166).

In this case, as the level is increased, that is, the number of nodes decreases as the level is increased. That is, in the first level process S162 to the third level process S166, four nodes in the child node are combined into one node in the parent node. In the fourth level process S168, a total of 16 nodes are combined into one node so that the entire image area can be recognized at once.

When the image is finally determined as described above, the type of the product is determined. Accordingly, the learned image of the corresponding product is compared with the input image to determine whether the product having the input image is defective or good quality. This step is the product quality determination step (S170).

The post-action step (S600) is a process of processing when a defect occurs in the product, the action is taken according to the contents set in the facility control method setting process (S140).

Meanwhile, the facility information input process (S110), the inspection condition setting process (S120), the quality standard setting process (S130), and the facility control method setting process (S140) are performed simultaneously or sequentially regardless of the order. In other words, the steps may be performed regardless of the order, and may be simultaneously performed.

12 is a flowchart illustrating a state of use of the image determination method S10.

As shown in this, when the system is started, basic information about equipment and items is loaded (S300). In addition, the defect number C representing the quantity of the product found to be defective is set to zero (S310).

Next, it is determined whether an image is input (S320). If an image is not input, it is continuously detected whether an image is input, and if an image is input, the image processing step (S330) is performed. Passing The image processing step S330 is a step corresponding to the product inspection step S150 described above.

Then, it is determined what the corresponding product is to match the setting criteria of the corresponding product, that is, the criterion of defective and good goods determined by the learning in the learning step (S50) (S340). This process corresponds to the product recognition step (S160).

As such, when the input image is matched with the learning image for the defective determination of the corresponding product, it is determined whether the product is defective (S350). In addition, this step (S350) corresponds to the product quality determination step (S170) described above.

As a result of the determination in step S350, if the product is a good product, the process returns to step S310 as shown.

On the other hand, if the product is defective, 1 is added to the defective number C (S360), and then it is determined whether the defective number C reaches the set reference number N (S370). That is, it is determined whether the user has reached the continuous failure reference number N set in the facility control method setting process S140 (S370).

As a result of the determination in the step S370, when the continuous failure number C does not reach the reference number N, the process returns to the above setting step S320.

On the other hand, when the continuous failure number (C) reaches the reference number (N), the post-action step (S600) proceeds. That is, at this time, the facility is stopped according to the user's setting, and the messenger SMS is transmitted to the user at the same time as a notification to the outside through the buzzer.

When the defective state is delivered to the user by the above process, the user directly checks the defective state of the product and checks the equipment.

13 is a block diagram showing the configuration of the sound determination method (S400).

As shown in this, the sound judging method (S400), measuring the sound generated during the production of a good product sound quality setting step (S410) for setting a data value of good sound, and at the time of actual product production by sound measuring means Sound determination step of measuring the sound generated (S420) and the actual sound measured by the sound measuring means compared with the data value set in the sound quality setting step (S410) sound judgment step ( S430) and the like.

The sound quality setting step (S410) is a step of measuring the sound when the actual good quality is produced, and inputting and setting digital data into a computer in advance. In this step, the vibration period (frequency) of sound (sound wave), the magnitude and intensity of the amplitude, and the like are input.

The sound measurement step (S420) is a process of measuring the sound generated in the actual production process through a measuring means such as a sound sensor. For example, by installing a super-directional microphone on the upper inside of the press equipment, and measures the sound generated in the process of forming the product by the press. Then, the sound measured by the measuring means (microphone) is converted into a digital value and stored as a sound file.

14 is a graph showing the measured sound in the form of a wave. As shown, the sound can be expressed in frequency (period) or amplitude, so that it is possible to distinguish between good and bad conditions based on these factors.

In the sound determination step (S430), by monitoring the change of the actual sound measured by the measuring means, the abnormal phenomenon is detected by the abnormality of the vibration period or amplitude of the sound wave.

In detail, in the sound determination step (S430), the data value set in the good quality sound setting step (S410) and the data value measured in the sound measurement step (S420) are compared with each other, and the actual measured value is set to good quality sound. In step S410, it is determined whether or not the sound data range of a good product is set in advance.

For example, if the maximum and minimum amplitudes are set in a limited manner, a deviation from this range results in a bad state. In addition, even with the limitation of the frequency (period), it is possible to judge good quality and defectiveness.

In the sound judging step (S430) as described above, if the actually measured sound data is out of the data range set in the good sound quality setting step (S410), a failure occurs or a failure is foreseen, and the post-action step (S600). Proceeds.

FIG. 15 is a block diagram showing each component of the failure dictionary recognition method S500.

As shown in this, the failure pre-recognition method (S500), a data setting step (S510) for setting the preliminary bad data expected to be a future failure, and a data measurement step (S520) for extracting the actual data measured from the facility And a data comparison step S530 for comparing the actual data measured in the data measuring step S520 with the preliminary bad data set in the data setting step S510.

The data setting step (S510) is a process of setting preliminary bad data for each measuring means. That is, the measuring means is provided with one or a plurality, and each of the measuring means to set the preliminary bad data.

The measuring means includes a sensor for measuring the pressure of the pump, a sensor for measuring the pressure applied to the accelerator, and the like. That is, the measuring means is mainly composed of various kinds of sensors to detect the change in quality by quantifying the pressure of each part.

For example, the measuring means may include a hydraulic pump coefficient (PUMP_PRESSURE_CALCULATE), a pump pressurized pressure (PUMP_PRESSURE_INPUT), a pump current pressurized pressure (PUMP_PRESSURE_INPUT_PRESENT), an accelerator pressure coefficient (ACC_PRESSURE_CALCULATE), an accelerator pressurized pressure (ACC_PRESSURE_INPUT), a main pump pressure (main pump pressure) MAIN_PRESSURE_INPUT), the main pump pressure maximum value (Main_Pressure_Peak_Value), the Vring pressure maximum value (Vring_Pressure_Peak_Value), etc., the sensor is measured to digitize (digitize) the measured value of each part.

In addition, the preliminary bad data set in the data setting step (S510) may be of various types or forms.

Specifically, the preliminary bad data is a case in which at least six data values measured by the measuring means are continuously increasing or decreasing. That is, since the six data values continuously increase or decrease, there is a high probability of failure. Therefore, the six data values are set as defective precursors.

Second, the preliminary defective data is a case where at least nine data values measured by the measuring means are located to one side of the average value.

Third, the preliminary defective data is a case in which at least 13 data values measured by the measuring means repeat the increase and the decrease.

Fourth, the preliminary defective data is a case where at least one or more data values measured by the measuring means are outside the standard deviation value 3σ range.

Fifth, the preliminary defective data is a case in which at least four or more data values of the four data values measured by the measuring means deviate from the standard deviation value 2σ range to one side of the average value.

Sixth, the preliminary defective data is a case in which at least six or more data values of the six data values measured by the measuring means deviate from the standard deviation value 1σ region to one side of the average value.

Of course, in addition to the preliminary bad data exemplified above, it is possible to set various other bad states.

In addition, the preliminary defect data as described above is based on the data derived by the experiment. That is, after measuring the value of the hydraulic pump coefficient (PUMP_PRESSURE_CALCULATE), after 5 points of digital data value increases or decreases continuously, 6 points of continuous points occur after 3σ (SIGMA: Standard Deviation) deviation points occur. When five points were in the area | region exceeding 1 (sigma) to one side of a center line, defect generate | occur | produced continuously, respectively.

In addition, as a result of measuring the main pump pressure (MAIN_PRESSURE_INPUT), after 6 points of digital data value increases or decreases continuously, 8 points are continuously positioned on one side of the center line (average), and 13 points are repeatedly increased or decreased. After that, after three of four consecutive points were in the area | region exceeding 2 (sigma) to one side of a center line, defect generate | occur | produced continuously.

FIG. 16 is a graph showing data values obtained by measuring the Vring pressure maximum value (Vring_Pressure_Peak_Value).

As shown therein, the solid red lines on the upper and lower sides are 'bad lines'. In other words, when the measured data values deviate above and below these defective lines, they become defective.

In addition, the red dotted line inside the "bad limit line" is a line which shows a 3σ upper limit and a lower limit, and the lime green dotted line means a 2σ upper limit and a lower limit. In addition, the green dotted line in a center part is an average value, and the blue dotted line above and below an "average value" shows an upper limit value and a lower limit value of 1σ.

The data measuring step S520 is a process of measuring actual data using one or more measuring means. That is, measuring means such as a pressure sensor is installed in a pump or an accelerator to measure actual data values.

The data comparison step S530 is a process of comparing the measured actual data with preliminary bad data. In this case, when the measured real data value coincides with the value of the preliminary bad data which is already set, since the failure is expected to occur, the post-action step S600 is performed.

That is, as a result of the comparison of the data comparison step S530, if the actual data matches the preliminary bad data, a post-action step S600 for stopping the facility or expressing an abnormal phenomenon to the outside is performed.

And, since the measuring means is provided with a plurality, the post-action step (S600) as described above, the actual data measured by at least one measuring means of the plurality of measuring means, preliminary bad data corresponding to the measuring means Is also performed if That is, even if the value of the actually measured data coincides with any one of the six preliminary bad data mentioned above, the follow-up step S600 is performed.

Each process performed in the post-action step S600 is illustrated in FIG. 17.

As shown in this, the post-action step (S600), the facility control process (S610) for controlling the equipment for producing the product, the external notification process (S620) for notifying the abnormal phenomenon to the outside, and measured by the measuring means It consists of an inspection history inquiry process (S630) for querying the value of the data.

Specifically, when a product defect occurs by the image determination method (S10) or the sound determination method (S400), or when a defect is foreseen by the failure pre-recognition method (S500), such a fact is notified to outside. At the same time the post-action step (S600) for controlling the equipment is in progress. Of course, in the post-action step (S600), the data value measured (inspected) by the measuring means is displayed to the outside, and the inspection history inquiry process (S630) for inquiring the inspection history by the measuring means is also performed.

In addition, the control such as the stop of the equipment performed in the post-action step (S600), it is preferable to control the equipment according to the control method set in the setting step (S100).

The facility control process (S610) is a process of stopping the facility when a product defect occurs continuously for a predetermined number of times or when a product defect is predicted by a sound or a plurality of measuring means.

Specifically, in the facility control process (S610), when the failure occurs continuously (eg, when three failures occur continuously) by the image determination method (S10), or the value of the sound data is good If the condition is out of the state or the actual measurement data value in the failure pre-recognition method (S500) coincides with the preliminary defect data, the equipment is stopped.

The external notification process (S620) is a process of transmitting a current situation to a remote user through an instant messenger (SMS) when an abnormality of the product occurs more than a predetermined number of consecutive times. That is, it notifies to the outside through a buzzer or the like and notifies the remote user even by using a communication means. In this case, it may be possible to inform the outside using the facility PLC communication or other messengers.

The inspection history inquiry process (S630) is a process in which the value of the data measured so far is displayed to the outside, and the user inquires the history of the measurement data. That is, the user inquires whether the pressure value measured by the measuring means or the actual data matches the preliminary bad data.

The scope of the present invention is not limited to the above-exemplified embodiments, and many other modifications based on the present invention will be possible to those skilled in the art within the above technical scope.

For example, the preliminary defective data exemplified above is an example of a case where the preliminary defective data is applied to a pressing process, and the pattern may vary depending on a production process or a product, or a production facility.

1 is a block diagram showing a schematic configuration of a failure determination method according to the present invention.

Fig. 2 is a block diagram schematically showing the configuration of an image determination method constituting a failure determination method according to the present invention.

Figure 3 is a flow chart showing a preferred embodiment of the learning step constituting the failure determination method according to the present invention.

4A to 4E are screen configuration diagrams showing an example of each process of the learning step constituting the image determination method according to the embodiment of the present invention.

5 is an experimental result table showing the accuracy (Accuracy) results according to the number of data of good and bad in the image determination method of the embodiment of the present invention.

6A and 6B are diagrams and graphs showing experimental results of testing accuracy according to a change in the number of iterations in the image determination method according to the embodiment of the present invention.

7A to 7C are diagrams and graphs showing experimental results of obtaining an optimum value of 'MaxDistance' in the image determination method of an embodiment of the present invention.

8A and 8B are diagrams and graphs showing accuracy results according to 'Sigma' change in the image determination method of the embodiment of the present invention.

9 is a view showing a detailed process of the product inspection step constituting the image determination method of the embodiment of the present invention.

10 is a chart showing the results of the change in accuracy according to the enhancement of contrast in the image determination method of the embodiment of the present invention.

11 is a view for explaining the product recognition step constituting the image determination method of the embodiment of the present invention.

12 is a flowchart showing the use state of the image determination method of the embodiment of the present invention.

Fig. 13 is a block diagram showing the configuration of the sound judging method of the embodiment of the present invention.

14 is a graph showing a wave form in the form of sound measured in the sound determination method of the embodiment of the present invention.

Fig. 15 is a block diagram showing each structure of the failure dictionary recognition method of the embodiment of the present invention.

FIG. 16 is a graph of data values obtained by measuring a Vring pressure maximum value (Vring_Pressure_Peak_Value), showing a distribution of actual data values in a failure pre-recognition method constituting an embodiment of the present invention. FIG.

Figure 17 is a block diagram showing each process of the post-action step constituting the embodiment of the present invention.

Explanation of symbols on the main parts of the drawings

S10. Image Learning Method S50. Learning phase

S100. Setting step S150. Product inspection stage

S160. Product recognition step S170. Product Quality Judgment Stage

S400. Sound judgment method S410. Good sound setting step

S420. Sound measurement step S430. Sound judgment stage

S500. Bad dictionary recognition method S510. Data setting step

S520. Data measurement step S530. Data comparison step

S600. Follow-up step S610. Facility Control Process

S620. External notification process S630. Inspection history inquiry process

Claims (13)

An image judging method (S10) for judging whether the computer is defective by the image; A sound judging method (S400) for determining, by the computer, a defective state by the presence or absence of an abnormal sound generated during product production; A failure pre-recognition method (S500) for the computer to identify an abnormality of the facility and predict a failure in advance; Post-action step (S600) that is taken when a defect is detected or expected or there is an abnormal symptom by at least one of the image determination method (S10), the sound determination method (S400), and the failure prior recognition method (S500). It includes; The image determination method (S10), A learning step (S50) of acquiring the information of the good and the defective by the computer learning, a setting step (S100) of setting the reference information of the system for determining whether the product is defective or not, and the reference set in the setting step (S100) In accordance with the product inspection step (S150) and the computer to inspect the product according to the product identification step (S160) to recognize the product type and the correct product shape by specifying the image of the product measured in the product inspection step (S150), And a product quality judging step (S170) of determining, by the computer, the defect and the good product based on the information acquired in the learning step (S50) of the image finally recognized in the product recognition step (S160): The sound determination method (S400), Computer-generated sound setting step (S410) for setting the data value of good sound by the computer to measure the sound generated during the production of good quality, and sound measurement step for the computer to measure the sound generated during the actual product production by sound measuring means (S420) And a sound judging step (S430) of detecting an abnormal phenomenon by comparing the actual sound measured by the sound measuring means with a data value set in the good sound quality setting step (S410). In the sound determination step (S430), The computer monitors the change of the actual sound measured by the measuring means, and detects the abnormality by the abnormality of the vibration period or amplitude of the sound wave: The failure dictionary recognition method (S500), A data setting step (S510) of setting preliminary bad data for which a future failure is expected for each measurement means by a computer, a data measuring step (S520) of extracting actual data measured from a facility by a computer using at least one measuring means, and A data comparison step (S530) of comparing the actual data measured in the data measuring step (S520) and the preliminary bad data set in the data setting step (S510) with a computer; The preliminary bad data, At least six data values measured by the measuring means continuously increase or decrease, or at least nine data values measured by the measuring means are located to one side of the average value, or at least 13 measured by the measuring means. 4 data values repeat the increase and decrease, or at least one data value measured by the measuring means is outside the standard deviation 3σ range, or 4 measured by the measuring means. At least four or more data values out of the standard deviation value 2σ range to one side of the mean value, or at least six or more data values out of six data values measured by the measuring means One side of this average value is outside the standard deviation value 1σ range; The post-action step (S600), A facility control process (S610) for controlling a facility producing a product, an external notification process (S620) for notifying abnormalities to the outside, and an inspection history inquiry process (S630) for querying a value of data measured by a measuring means; Includes; The facility control process (S610), Three or more defects are continuously generated by the image determination method (S10), out of the state when the value of sound data is good, or the actual measured data value in the defect pre-recognition method (S500) is a preliminary defect. If it matches the data, the computer automatically stops the installation; The external notification process (S620), If the product defect occurs continuously for a certain number of times, the computer automatically generates an alarm and transmits the current status to a remote user through SMS (SMS): The inspection history inquiry process (S630), The computer displays the value of the data measured so far to the outside, and the user checks the history of the displayed measurement data to check whether the measured pressure value or the actual data matches the preliminary bad data. The defect determination method characterized by the above-mentioned. delete delete delete delete delete delete delete delete delete delete delete delete

Images