TW201204482A - Method for detecting abnormal rolling at tail end of steel strip - Google Patents

Abstract

A method for detecting abnormal rolling at tail end of steel strip includes an equipment setup step, a production processing step, a computation step and a determination step. In the steel strip production line, the equipment setup step is to configure a sensing unit, a signal capturing unit and a computation processing unit at a rolling section of a rolling machine for steel strip rolling. The production processing step utilizes the sensing unit to measure the vibration when the steel strip passes the rolling section. The signal capturing unit captures the vibration measured by the sensing unit to obtain the vibration data and transmit to the computation processing unit. The computation step utilizes the computation processing unit for the computation on the vibration data to obtain a first abnormal determination feature. When the first abnormal determination feature is greater than a first abnormal determination threshold value, it is determined that there is an abnormal rolling at the tail end of steel strip; otherwise, it is determined there is no abnormality.

Description

201204482 VI. Description of the Invention: [Technical Field] The present invention relates to a method for detecting a steel strip process abnormality, which is a method for detecting an abnormality of rolling end of a steel strip. Special [Prior Art] The current steel strip production process, Daxie Indium production re-penetration 屮..., A 疋 疋 steel embryo is heated along the heating line and sent along the production line, after rough rolling, the fine emulsion is delayed to obtain the required thickness Steel belt,

After that, it is coiled into steel coils for subsequent operations such as transportation, re-salting, and the like. σ If in the dry-stretching process of the production line, due to the abnormal deviation of the steel section due to the deviation of the steel strip, when the rolling mill is in the machine, the tail end of the steel strip will impact the rolling of the dry-casting machine, so that the roll The surface is overstressed at the abnormal end of the milk-extending steel strip, which causes the unevenness of the rolling light surface to form unevenness. When the steel strip is rolled, it causes the effect of similar stamp image transfer, and will be formed in the The defect on the surface of the roll is transferred to the strip surface of the subsequent rolling strip which contacts the roll, so that the strip which is rolled by the roll subsequently forms a defect in the contact surface of the roll, so that the rolled product is rejected. The defect of the surface of the transfer belt after the abnormal rolling at the end of the steel strip causes the surface defect of the roll is called the roll transfer (Tai1 Pinch). The existing method for solving the roll transfer is mostly through various improvements in the production process of the steel strip. Process control of steel strip wedge, bend, offset and other defects further reduces the incidence of dry transfer. For example, Kiyota T et al. improve the strip wedge shape by controlling the amount of steel strip traverse. The disadvantage is that the reaction Slow speed and rolling Printing has a low correlation and can not effectively improve the roll transfer 3 201204482 •, while US Patent No. US 5722279 achieves the purpose of improving roll transfer by controlling the wedge shape and traverse of the steel strip, but requires special measurement. Tension equipment must be greatly modified for existing process equipment; US Patent No. 6148653 improves the end of the steel strip by adjusting the rolling process parameters to improve the roll transfer. But the disadvantage is also necessary Large-scale addition of additional equipment and modification of existing process equipment; US patent number us 731〇982 is to control the steel strip wedge shape by controlling the rolling process parameters, thereby reducing the incidence of pro-rolling transfer, but also to achieve the same effect Process equipment is being modified. It is not easy to improve, and the modification of the equipment is based on the current method, which is to prevent the occurrence of abnormalities at the end of the steel strip, thereby reducing the incidence of rolling transfer, and the improvement scheme requires a majority. The measuring equipment and the large-scale modification of the existing equipment are immediately applied to the cause of the pro-rolling transfer, and the additional cost is high.

The π clothing mad S and street parameters were adjusted to significantly reduce the occurrence of roll transfer 201204482, and the technology must be simple to implement and reduce the cost of introducing existing equipment and additional equipment. SUMMARY OF THE INVENTION It is therefore an object of the present invention to provide a method for automatically detecting the occurrence of an abnormality in the end of a steel strip and which is simple and inexpensive to detect the abnormality of the end of the steel strip. Therefore, the present invention detects a method for rolling an abnormality at the end of a steel strip, comprising a

Equipment erection steps, production travel steps, - operation steps, and - decision steps. The "X preparation sighs step in a rolling strip steel strip production line to roll rolling the actual rolling strip of the rolling tool - a rolling unit, a sensing unit, a connecting unit of the sensing unit And the operation processing single TL of the connection unit, the sensing unit measures the vibration of the steel strip passing through the dry extension, and the signal extraction unit extracts the vibration measured by the sensing unit into vibration After the data is transmitted, the vibration data is transmitted to the operation processing unit to perform arithmetic processing. The production travel step is to produce a steel strip on the production line and the steel strip is rolled by the rolling H to roll the rolling tool. Meanwhile, the sensing unit measures the vibration of the steel strip passing through the rolling section, and the signal is 撷The unit is synchronized to obtain the vibration data obtained by the sensing and the vibration data is transmitted to the operation processing unit. The ~/ operation step uses the wavelet processing to filter the vibration data by the wavelet transform to obtain the "wavelet reconstruction data, and then calculate the wavelet reconstruction data to obtain the first-abnormal determination feature for judging the rolling end abnormality of the steel strip. 201204482 The determining step defines a -first-abnormal determination inter-value, and when the _th abnormality determining feature is not greater than the first abnormality determining threshold, determining that the strip rolling is normal' when the first-abnormal determination characteristic is greater than the _ When the value is abnormally judged, it is determined that the end of the steel strip is abnormally rolled. The utility model has the advantages that the vibration signal of the steel strip passing through the rolling implement is measured by the sensing unit mounted on the Zhayan machine, and the vibration signal is wavelet-filtered and filtered, and the abnormal signal is extracted by using the window function. Take the actual signal to reflect whether the abnormal end of the steel strip is abnormal and its characteristics, and achieve the purpose of detecting the abnormal end of the steel strip. The above and other technical contents, features, and advantages of the present invention will be apparent from the following detailed description of the preferred embodiments. Before the present invention is described in detail, it is noted that in the following description, similar elements are denoted by the same reference numerals. Referring to Fig. 1', a method of detecting a rolling end abnormality of a steel strip is disclosed. The first preferred embodiment comprises an apparatus erection step u, a production running step u, an arithmetic step 13, and a determining step μ. Referring to FIG. 2 'Firstly, the equipment erection step 11 is carried out, and in the production line 2 of the Zhayan steel strip, a sensing unit 41 and a connection are arranged on the rolling section 21 of the rolling strip 3 actually rolling the steel strip. The signal intercepting unit 42 of the sensing unit 4, and the signal processing unit 43 connected to the signal capturing unit 42, wherein the sensing unit 41 is mounted on the rolling implement 3 and the sensing unit 41 The vibration of the steel strip 1 〇〇 passes through the rolling section 21, 201204482. The signal acquisition unit A 42 operates the vibration of the sensing sheet a 41 jin 4 and transmits the vibration data to the arithmetic processing. Unit 43 performs arithmetic processing. Preferably, the sensing unit 41 is mounted on the roll bearing housing 31 of the rolling tool 3. In the embodiment, the sensing unit 41 is an accelerating gauge. The arithmetic processing unit 43 is an industrial computer. The production travel step 12 is then carried out after the required measuring equipment is erected, and the rolling strip 3 is rolled while the steel strip 100 is produced on the production line 2 and the steel strip (10) is passed through the zigzag section 21, The sensing unit 41 measures the vibration of the steel strip 100 as it passes through the rolling section 21, and the signal capturing unit 42 synchronously captures the vibration measured by the sensing unit 41 to obtain vibration data, and then The vibration data is transmitted to the arithmetic processing unit 43. In detail, the vibration data is converted by the signal acquisition unit 42 taking the vibration generated by the sensing unit 41 in the process of measuring the amount of steel 〇〇 该 该 该 该 该 该For digitizing the data, preferably, the sensing unit 4 can be triggered by a level trigger to capture only the vibration at the end of the steel strip 100, or can be provided by a control computer that controls the rolling tool 3. The vibration of the tail end of the steel strip pair is taken out by the feedback signal, and the timing of the vibration signal is not emphasized by the present invention, so it will not be described here. Referring to Appendix 3, after the data is obtained by the arithmetic processing unit 43, the operation is performed (4) 13' U calculation processing unit (2), the vibration data is wave-transformed to remove noise, and in detail, the vibration data is The first layer decomposition by wavelet transform will obtain two frequency bands 201204482, and two sets of wavelet coefficients corresponding to the two frequency bands. Then, after the second layer decomposition of the j-wave coefficients of the corresponding low frequency bands, two frequency bands are obtained again. , and corresponding to the two sets of wavelet coefficients of the two bands, the third layer of the fourth layer, the fifth layer and the like, after the smaller, the smaller the wavelet coefficient in each layer is removed

Reconstruct the signal and get the sentence ·" + 士 $ 士丨 B 付 忒 忒 wavelet reconstruction resources, you can achieve the effect of filtering out noise. Preferably, the Daubechies (_) wavelet function is used to decompose, where N=5. It should be noted that each rolling tool has its own characteristics due to the gap between the components, the size difference of the components, the installation location of the machine, etc., so that the vibration data measured in each rolling tool will have some The difference is that the frequency band of the miscellaneous gas is not the same as the same. Therefore, the noise cannot be removed in a manner of removing the specific frequency band. In the present embodiment, the wavelet is used to remove the noise of the external noise, the machine itself, and the noise contained in the vibration data. The noise of the machine itself is affected by the difference, and the vibration signal of the steel strip passing through the rolling machine is indeed obtained. / After referring to Annex 2, 4', the window function (window functic) n) is performed on the wavelet reconstruction data, and the "abnormal signal section" is extracted, and the size is shown in Figure 2, and then the abnormal signal section is The data takes the root mean square and obtains a first-abnormal determination feature, and defines that the largest of the abnormal signal segment data is the maximum amplitude, the ratio of the maximum amplitude of the wave height branch to the abnormality.进行 After the obtaining of the first-abnormal determination feature, the determining step is performed, when the first-abnormal determination feature is not greater than—the preset first-abnormal inter-material value two' determines the steel strip rolling normal' when the first abnormality When the judgment feature is larger than the first-abnormal determination threshold, the determination of the end rolling of the steel strip is abnormal, and the maximum vibrating field represents the severity of the abnormality at 201204482, and the larger the value, the more serious the abnormality. Preferably, before determining whether the rolling of the steel strip is abnormal, the contour of the abnormal signal section is extracted by the window function, that is, the abnormal signal section; The key signal of the different f, such as the ellipse + [: in the Annex 2, 4; ^ indicates 'the correctness of the obtained first-abnormal determination feature is greatly increased, when the wave height exceeds the preset value, the abnormal signal zone The segment covers the key signal of the abnormality of the steel π tail cake. When the wave height is lower than the preset value, the abnormal signal section has no critical vibration signal covering the abnormal end of the steel strip. As can be seen from the above description, the first preferred embodiment of the present invention for detecting the abnormality of the end rolling of the steel strip is provided by the sensing unit 4 mounted on the roll bearing housing 31 of the rolling implement 3! With the vibration of the rolling process by the rolling machine 3, the vibration signal (filtered noise) indicated by the oval frame in the attachments 2 and 4 occurs when the tail end of the steel strip passes through the rolling machine 3, and then is converted by wavelet. The frequency band containing the noise is filtered out, and the key signal covering the tail end of the steel strip is taken out by a window function, and the first abnormality determining feature is extracted to determine whether the end of the steel strip is abnormally rolled. Referring to FIG. 3, a second preferred embodiment of the present invention for detecting the abnormality of the rolling end of the steel strip includes a device erection step 51, a production travel step Μ, a nose step 5 3 ' and a determination step 5 4. Referring to FIG. 2, firstly, the equipment erection step 51 is performed to set a sensing unit 41 and a rolling section 21 of a rolling strip 3 in a rolling strip steel strip production line 2 The signal intercepting unit 42 connected to the sensing unit 41 and the processing unit 0404482 processing unit 43 connected to the signal extracting unit 42 are provided with the φ, 4 α , ' Ύ s sensing unit 41 mounted on the rolling implement 3 When the sensing unit 41 selects 1 or more, the steel strip 100 passes the vibration number of the rolling section 21 to obtain the vibration measured by the sensing unit 41. After the vibration data is generated, the vibration data is transferred to the arithmetic processing unit 43 for arithmetic processing. Preferably, the vibration data is mounted on the roll bearing 31 of the rolling machine 3, In the present embodiment, the sensing unit 41 is an acceleration gauge, and the arithmetic processing unit 43 is an industrial computer. Then, the production traveling step 52 is performed, and the belt 1 (^() is produced on the production line 2, and the steel strip (10) is passed through the rolling section 21, and the rolling apparatus is used to take the amount of the sensing unit 41. When the steel strip 1 〇〇 passes through the milk extension 21 = vibration 'and the signal operation unit A 42 synchronously operates the vibration of the sensing unit 41 to obtain vibration data, and then the vibration data is obtained. The operation is performed to the operation processing unit 43. Then, the operation step 53 is performed, and the vibration processing data of the sub-i of the arithmetic processing unit 43 is first decomposed by wavelet transform, and then the smaller wavelet coefficients in each layer are removed, and then reconstructed. The wavelet reconstruction data is obtained by the signal to achieve the effect of filtering out the noise. Preferably, the Daubechies (dbN) wavelet function is used for wavelet decomposition, wherein N=5. After that, the wavelet reconstruction data is calculated. Degree to obtain a first inappropriate decision feature, wherein, the degree =

Where yi is the average of the plurality of data of the wavelet reconstruction data, and m is 10 201204482 the total number of data of the wavelet reconstruction data. Finally, the determining step 54 is performed. When the first abnormality determining feature is not at the preset first-abnormal threshold, the rolling of the steel strip is normal, and the first abnormality determining characteristic of the field is greater than the first When the abnormality is judged, it is determined that the end of the steel strip is abnormally rolled. After determining the abnormality of the rolling end of the steel strip, the wavelet reconstruction data is subjected to the window function operation, and an abnormal signal section is extracted, and then the data of the section is taken as the root mean square to obtain a second. The abnormality determination feature defines the largest one of the abnormal signal segment data as the maximum amplitude, and the wave height rate is the ratio of the maximum amplitude to the abnormality determination feature. When the second abnormality determining feature is not greater than a predetermined second abnormality determining interval value, determining that the steel strip rolling is normal, and determining that the second abnormality is determined by the second abnormality determining value, determining the steel strip tail end milk The extension does occur 'and' with the maximum amplitude representing the severity of the anomaly, and the greater the value, the more severe the anomaly. ▲ Preferably, before determining whether the rolling of the steel strip is abnormal, the wave function is used to confirm the appropriateness of the abnormal signal section by the window function, that is, whether the abnormal signal section covers the abnormal end of the steel strip The key signal makes the correctness of the (4) two abnormality determination feature of the 乂 increase greatly. When the wave height exceeds the preset value, the abnormal signal section covers the steel belt ^ ^, the key signal of the state 咿 鳊 anomaly, When the wave rate is lower than the preset value, the abnormality „ ^ w ^ hang. The hole section has no key vibration signal covering the abnormality of the steel strip tail μ. It can be seen from the above description that the present invention detects the end rolling of the steel strip The second preferred embodiment of the abnormality performs the first determination by the first abnormality determining feature. 11 201204482 When an abnormality occurs, a second determination is made to determine that an abnormality has occurred, thereby reducing the sense of cause. Abnormality such as measurement unit malfunction = no need to make a second judgment, and greatly reduce the calculation time of the fixed = brother.

In summary, the method of the present invention_the steel strip end end rolling abnormality is set by the sensing unit 41 (four), and the wavelet transform is performed, and the tail end rolling of the steel strip can be automatically performed unless the specific 'band noise is selected. The occurrence of an abnormality is provided for the on-site personnel to immediately adjust the rolling process parameters, or to immediately change the rolling contact of the rolling implement 3 to reduce the incidence of rolling transfer, and the tail end rolling of the steel strip of the present invention The abnormal method was operated by the on-site interpersonal, and after the actual monitoring image comparison, the abnormal detection of the end rolling of the steel strip was indeed detected: 'Additionally, there is no need to add additional equipment and equipment to be greatly increased. By reducing the cost of introduction, it is indeed possible to achieve the object of the present invention. However, the above description is only a preferred embodiment of the present invention, and when it is not possible to limit the scope of the practice of the present invention,

The simple equivalent changes and modifications 7 made by the scope and description of the invention are still within the scope of the invention. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a flow chart illustrating a first preferred embodiment of a method for detecting a rolling end abnormality of a steel strip according to the present invention; FIG. 2 is a top plan view illustrating the first preferred embodiment. A mounting position of a sensing unit of the embodiment; and FIG. 3 is a first-class schematic diagram illustrating a second preferred embodiment of the method for detecting the abnormality of the rolling end of the steel strip at the end of the present invention. 12 201204482 [A brief description of the attachment] Attachment 1 is a time-frequency diagram of the wavelet data converted from the time-frequency diagram 'Describe a rolling end abnormality of a steel strip' after the wavelet transform is decomposed by the wavelet transform; Annex 2 is a vibration data map, indicating a Abnormal vibration data of the end of the steel strip; 'External 3 is the wavelet transform time-frequency diagram, which shows the time-frequency diagram of the positive vibration data of a steel strip at the end of the rolling process after wavelet transformation; and the description includes a steel strip tail End rolling

Attachment 4 is the vibration data of the vibration data map.

13 201204482 [Description of main component symbols] 11 · · • · · Equipment installation step 41 - • Sensing unit 12 ·. • • • Production travel step 42 ·... • Signal acquisition unit 13 · · • • Operation step 43 - • arithmetic processing unit 14 · · • • • decision step 51 – • equipment erection step 2... • • • production line 52 – • production travel step 21 · · • * · rolling stage 53 β ... · • operation steps 3... • * · Rolling machine 54 · · * • Decision step 31 · · • •• Rolling bearing housing 100 ... • Steel strip

14

Claims (1)

201204482 VII. Scope of application for patents: 1. The method for measuring the abnormality of the end of the steel strip at the end of the test includes: - Equipment erection step 'In the production line of the --Yuyan steel strip, the actual delay of the rolling tool J; L ^ is earlier than L 4, S is further provided with a sensing unit, a signal intercepting unit connected to the sensing unit, and a signal processing unit for connecting the signal to the single unit, wherein the sensing unit The unit measures the vibration of the steel strip passing through the rolling section, and the signal extracting unit operates the sensing list After a kilogram of measured vibration into vibration data, the vibration data is transmitted to the arithmetic processing unit for arithmetic processing; a production traveling step, the steel strip is produced on the production line, and the steel strip is passed through the dry extension to be rolled pure At the same time, the sensing unit measures the vibration of the steel strip as it passes through the rolling section, and the signal M is synchronized. Negatively taking the vibration measured in the X sensing unit to obtain the vibration data, and transmitting the vibration data to the operation processing unit; - an operation step, the operation processing unit uses the wavelet transform to calculate the wavelet data to obtain the __ After the wavelet reconstruction data, the wavelet reconstruction data is used to obtain a first abnormality determining feature for judging the rolling end abnormality of the steel strip; and the '-determining step' defines a first abnormality determining value, and when the first abnormality determining feature is used When the value is not greater than the first abnormality determination threshold, it is determined that the steel strip is extended normally. When the first-abnormal determination characteristic is greater than the first-abnormal determination threshold, it is determined that the end of the steel strip is abnormally rolled. 2. The method for detecting the abnormality of the tail end of the steel strip according to the scope of the patent scope of the present invention, wherein the device is installed in the step of installing the sensor 15 201204482 in the drying of the rolling tool Light bearing box. 3. According to the method of applying for the patent model, the wavelet transform operation of the k=T line of the abnormal vibration data of the end of the measured strip at the end of the test strip is obtained by the phase-layer decomposition of the phase M h The frequency band, and the wavelet coefficients of the two sets of wavelet coefficients in the corresponding two bands are into the low-frequency virtual phase library. After the second layer is decomposed, the two bands are obtained again, and the two sets of wavelet coefficients are followed by K-layer decomposition The smaller wavelet coefficients are removed, and the wavelet reconstruction data is obtained by reconstructing the signal. 4. According to the third paragraph of the patent application scope, the detection of the end rolling anomaly of the steel strip is performed. The operation step is to perform a window on the wavelet reconstruction data: after the number operation, the abnormal signal section is extracted, and then the abnormality is The information of the signal area ί takes the root mean square to obtain the first-abnormal determination feature, and defines that the largest of the different hfl number segments is the maximum amplitude, and the wave height rate is the ratio of the maximum amplitude to the first abnormality determination feature. According to the 4th item of Zhongqing Patent Gu (4), the end rolling irregularity of the steel strip The method, 纟中,1海#|! The step also determines that when the wave height exceeds a preset value, the maximum amplitude represents the severity of the abnormality, and the larger the value, the more severe the abnormality. ^ According to the third aspect of the patent scope, the method for detecting the abnormality of the rolling end of the steel strip is as follows, wherein the operation step is to calculate the twist of the wavelet reconstruction data and serve the first abnormality prediction feature. And when the first abnormality determining feature is greater than the first abnormality determining threshold, and then extracting an abnormal signal section from the wavelet reconstruction data, and then extracting the abnormal signal section, and then the data of the abnormality 16 201204482 signal section Taking a root mean square, a second abnormality determining feature is obtained, and the largest one of the abnormal signal segment data is defined as a maximum amplitude, and the wave height rate is a ratio of the maximum amplitude to the second abnormality determining feature; and the tail end of the steel strip is abnormal. The determining step further defines a second abnormality determination threshold, and when the second abnormality determining feature is not greater than the second abnormality determining threshold, determining that the steel=rolling is normal, and when the second abnormality determining feature is greater than the second abnormality When judging the threshold value, it is determined that the end of the steel strip is abnormally rolled. According to the method value of Ψ ~, it is abnormally ugly. 17