CN116461931A - Belt longitudinal tearing detection and identification system and method based on deep learning - Google Patents
Belt longitudinal tearing detection and identification system and method based on deep learning Download PDFInfo
- Publication number
- CN116461931A CN116461931A CN202310497175.7A CN202310497175A CN116461931A CN 116461931 A CN116461931 A CN 116461931A CN 202310497175 A CN202310497175 A CN 202310497175A CN 116461931 A CN116461931 A CN 116461931A
- Authority
- CN
- China
- Prior art keywords
- belt
- detection
- time point
- damage
- region
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 306
- 238000000034 method Methods 0.000 title claims abstract description 21
- 238000013135 deep learning Methods 0.000 title claims abstract description 17
- 230000005540 biological transmission Effects 0.000 claims abstract description 87
- 238000004458 analytical method Methods 0.000 claims abstract description 63
- 238000005336 cracking Methods 0.000 claims description 45
- 238000013441 quality evaluation Methods 0.000 claims description 19
- 238000012545 processing Methods 0.000 claims description 12
- 238000001303 quality assessment method Methods 0.000 claims description 8
- 230000002159 abnormal effect Effects 0.000 claims description 5
- 238000012216 screening Methods 0.000 claims description 4
- 238000003708 edge detection Methods 0.000 claims description 3
- 230000007613 environmental effect Effects 0.000 claims description 3
- 230000007547 defect Effects 0.000 abstract description 5
- 230000007774 longterm Effects 0.000 abstract description 4
- 239000000463 material Substances 0.000 description 5
- 238000004519 manufacturing process Methods 0.000 description 3
- 239000002699 waste material Substances 0.000 description 3
- 230000002457 bidirectional effect Effects 0.000 description 2
- 229910000831 Steel Inorganic materials 0.000 description 1
- 230000005856 abnormality Effects 0.000 description 1
- 238000007792 addition Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000009546 growth abnormality Effects 0.000 description 1
- 238000009776 industrial production Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000010959 steel Substances 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B65—CONVEYING; PACKING; STORING; HANDLING THIN OR FILAMENTARY MATERIAL
- B65G—TRANSPORT OR STORAGE DEVICES, e.g. CONVEYORS FOR LOADING OR TIPPING, SHOP CONVEYOR SYSTEMS OR PNEUMATIC TUBE CONVEYORS
- B65G43/00—Control devices, e.g. for safety, warning or fault-correcting
- B65G43/02—Control devices, e.g. for safety, warning or fault-correcting detecting dangerous physical condition of load carriers, e.g. for interrupting the drive in the event of overheating
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/89—Investigating the presence of flaws or contamination in moving material, e.g. running paper or textiles
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Textile Engineering (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The invention relates to the technical field of longitudinal tearing of a belt, and particularly discloses a detection and identification system and method for longitudinal tearing of a belt based on deep learning, wherein the system comprises the following components: the belt historical cargo load analysis module, the belt detection module detection area division module, the belt breakage analysis module, the belt breakage growth coefficient analysis module, the early warning terminal and the cloud database are used for analyzing the historical transmission quality of the belt, so that the defect that the analysis strength of the belt historical condition is not deep enough in the prior art is overcome, the accuracy of the belt tearing expansion analysis result is further guaranteed, the risk of tearing the belt is timely ensured, early warning is given, long-term use of the belt is facilitated, the phenomenon that the regularity attention degree of the belt tearing area is low in the prior art is overcome, the safety of expanding the belt tearing area is guaranteed, the accuracy of the belt longitudinal tearing analysis result is further guaranteed, and the referential of the belt tearing growth coefficient analysis result is improved.
Description
Technical Field
The invention relates to the technical field of longitudinal tearing of a belt, in particular to a detection and identification system and method for longitudinal tearing of a belt based on deep learning.
Background
The belt is a transmission device widely applied to industrial production, and has the main effects of transmitting materials from one place to another place, and is applied to various industrial fields, such as industries of mines, ports, electric power, chemical industry and the like, and in the use process of the belt, the belt can be longitudinally torn due to various reasons, so that the materials leak on one hand, and further the problem of material waste exists, on the other hand, the safety of the belt device is difficult to ensure, the safety of related workers is difficult to ensure, the production efficiency and the safety are seriously influenced, and therefore, in order to improve the production efficiency and the safety, the occurrence of production accidents is reduced, and detection and identification of the belt are very necessary.
In the prior art, the belt can meet the current requirements to a certain extent through detection and identification, but certain defects exist, and the detection and identification are specifically shown as follows: (1) In the prior art, the longitudinal tearing analysis is carried out on the detection picture of the belt mostly, the analysis strength of the historical condition of the belt is not deep enough, the historical condition of the belt is also a part of factors influencing the tearing expansion of the belt, the neglect of the aspect of the prior art leads to the inaccuracy of the analysis result of the tearing expansion of the belt, the risk of tearing the belt is difficult to ensure and the timely early warning is given, the situation that the tearing and the growth of the belt are at great risk but not timely early warning is probably caused, and therefore unnecessary loss is caused, and the long-term use of the belt is not facilitated.
(2) The prior art has low attention to the regularity of the belt tearing area, the regularity of the belt tearing area influences the damage and the growth of the belt tearing to a certain extent, the neglect of the prior art on the aspect is difficult to ensure the safety of the expansion of the belt tearing area, and further the accuracy of the longitudinal belt tearing analysis result is difficult to ensure, so that powerful data support cannot be provided for the subsequent belt tearing growth, and the referential of the analysis result of the belt tearing growth coefficient is reduced.
Disclosure of Invention
In order to overcome the defects in the background art, the embodiment of the invention provides a belt longitudinal tearing detection and identification system and method based on deep learning, which can effectively solve the problems related to the background art.
The aim of the invention can be achieved by the following technical scheme: belt longitudinal tear detection and identification system based on deep learning includes: and the belt historical cargo load analysis module is used for acquiring cargo parameters corresponding to the belt in a set detection period from the cloud database, and further analyzing historical transmission quality evaluation coefficients corresponding to the belt.
The belt detection module is used for arranging all cameras on the belt target side along the belt transmission direction of the belt transmission device according to preset intervals, arranging all cameras on the belt designated side in the same way, and detecting images of the belt.
The detection area dividing module is used for dividing the belt detection area into belt detection subareas according to preset detection intervals according to the arrangement intervals of the cameras arranged on the target side.
The belt breakage analysis module is used for acquiring target edge images and appointed edge images of all belt detection subareas which all detection time points belong to, and further analyzing breakage coefficients corresponding to all belt detection subareas which all detection time points belong to.
And the belt breakage growth coefficient analysis module is used for analyzing each damaged belt subarea corresponding to each detection time point so as to analyze the breakage growth coefficient corresponding to each damaged belt subarea at each damage time point.
And the early warning terminal is used for carrying out corresponding early warning according to the damage coefficient of each belt detection sub-region at each detection time point and carrying out corresponding early warning according to the damage growth coefficient of each damaged belt sub-region corresponding to each damage time point.
And the cloud database is used for storing cargo parameters corresponding to the belt in a set detection period, storing a cracking gray value range and storing historical fault times corresponding to the belt.
Further, the cargo parameter includes a respective cargo weight for each transmission.
Further, the historical transmission quality evaluation coefficient corresponding to the belt comprises the following specific steps: extracting the weight G of the goods corresponding to each transmission from the goods parameters corresponding to the belt in the set detection period b Where b is the number of each transmission, b=1, 2.
And carrying out average value processing on the weight of the goods corresponding to each transmission of the belt, so as to obtain the weight average value G' of the goods corresponding to the belt.
Selecting the maximum cargo weight G corresponding to the belt transmission from the cargo weights corresponding to the belt transmissions max And a minimum cargo weight G min 。
Analyzing the weight deviation coefficient of the beltWherein G is b+1 For the weight of the goods corresponding to the b+1th transmission, G 'is the allowable weight error of the preset maximum weight of the goods corresponding to the minimum weight of the goods, G' is the allowable weight error of the preset two adjacent transmissions, c is the transmission times, lambda 1 、λ 2 、λ 3 The weight deviation is the weight coefficient of the influence to which the weight deviation corresponding to the maximum cargo weight and the minimum cargo weight belongs, wherein the weight deviation and the weight deviation are respectively the preset cargo weight deviation and the cargo weight deviation of two adjacent transmissions.
Counting the transmission times C corresponding to the belt, and further analyzing the historical transmission quality evaluation coefficient corresponding to the belt according to the predefined standard transmission times C' corresponding to the beltWhere e is a natural constant.
Further, the method for analyzing the breakage coefficient corresponding to each belt detection sub-region to which each detection time point belongs specifically comprises the following steps: and acquiring each gray value of the target edge image corresponding to each belt detection sub-region to which each detection time point belongs according to the target edge image of each belt detection sub-region to which each detection time point belongs.
Extracting a cracking gray value range corresponding to the belt from the cloud database, comparing each gray value of the target edge image of each belt detection sub-region to which each detection time point belongs with the cracking gray value range corresponding to the belt, screening each abnormal region corresponding to each belt detection sub-region to which each detection time point belongs, acquiring the corresponding area, selecting the region with the largest area from the areas as the target cracking region, and obtaining the target cracking region corresponding to each belt detection sub-region to which each detection time point belongs.
Acquiring the area S of the corresponding target cracking area of each belt detection sub-area to which each detection time point belongs jp Where j is the number of each detection time point, j=1, 2,..k, p is the number of each belt detection sub-region, p=1, 2,..q.
Analyzing the damage coefficient corresponding to each belt detection subarea to which each detection time point belongs during target edge detectionWherein S' is the preset area of the target cracking area.
Similarly, analyzing the breakage coefficient M 'corresponding to each belt detection sub-region to which each detection time point belongs during the detection of the designated edge' jp 。
Analyzing breakage coefficients corresponding to belt detection subareas to which each detection time point belongsWherein gamma is 1 、γ 2 Is a weight factor mu corresponding to the preset target cracking area and target cracking area regularity jp And the profile regularity of the target cracking area corresponding to the p belt detection sub-area to which the j detection time point belongs.
Further, the contour regularity of the target cracking region corresponding to each belt detection sub-region to which each detection time point belongs is determined by the specific analysis method: and acquiring the external contour of the target cracking area corresponding to each belt detection sub-area to which each detection time point belongs, and acquiring the corresponding line.
Building by taking central point corresponding to each belt detection subarea as originA three-dimensional coordinate system is established, each contour point is selected from the external contour lines of the target cracking area corresponding to each belt detection sub-area to which each detection time point belongs according to a set length, and the corresponding three-dimensional coordinate is obtainedWhere h is the number of each contour point, h=1, 2.
Analyzing the distance between each contour point corresponding to each belt detection sub-region to which each detection time point belongs and the adjacent contour point
Analyzing the contour regularity of the target cracking area corresponding to each belt detection sub-area to which each detection time point belongsWherein->The p belt detection sub-area which belongs to the j detection time point corresponds to the distance between the h+1th contour point and the h+2th contour point, g is the number of the contour points, and JL' is the error of the allowable distance deviation of the preset contour point distance.
Further, the method for analyzing each damaged belt subarea corresponding to each detection time point specifically comprises the following steps: comparing the damage coefficient corresponding to each belt detection sub-region to which each detection time point belongs with a preset damage coefficient threshold value, if the damage coefficient corresponding to a certain belt detection sub-region to which a certain detection time point belongs is larger than or equal to the damage coefficient threshold value, marking the belt detection sub-region as a damaged belt sub-region, marking the detection time point as a damaged time point, and further obtaining each damaged belt sub-region to which each damage time point belongs.
Further, the damage growth coefficient corresponding to each damaged belt subarea at each damage time point is specifically analyzed by the following method: according to the breakage coefficient corresponding to each belt detection subarea to which each detection time point belongsObtaining the damage coefficient corresponding to each damaged belt subarea to which each damaged time point belongs, and further summarizing the damage coefficients to obtain the damage coefficient sigma of each damaged belt subarea to which each damaged time point belongs m,f Where m is the number of each damaged belt sub-region, m=1, 2,..i, f is the number of each damaged time point, f=1, 2..t, and further obtain the interval J of each damage time point from its neighboring damage time points f,f-1 Wherein f > 1.
Analyzing the damage growth rate of each damaged belt subarea corresponding to each damage time pointWherein sigma m,f-1 The mth damaged belt subarea corresponds to the damage coefficient to which the f-1 th damaged time point belongs.
Comprehensively analyzing damage growth coefficients of damaged belt subareas corresponding to each damage time pointWhere ζ is a belt transport environment quality assessment factor, ω "is a predetermined allowable damage growth rate, χ 1 、χ 2 、χ 3 The method comprises the steps of respectively presetting a damage growth rate, historical transmission quality of the belt and a duty ratio factor corresponding to the environmental quality of the belt transmission.
Further, the belt transmission environment quality evaluation coefficient ζ is specifically analyzed by the following steps: acquiring belt transmission speed R corresponding to each detection time point j And acquiring the temperature WT corresponding to each belt detection subarea corresponding to each detection time point jp And extracting the historical failure times CS corresponding to the belt from the cloud database.
And carrying out average value processing on the temperature corresponding to each belt detection subarea corresponding to each detection time point, further obtaining a temperature average value corresponding to each detection time point, carrying out average value processing on the temperature average value corresponding to each detection time point, and further taking the result as a proper temperature WT' corresponding to the belt.
Analyzing the belt transmission environment quality evaluation coefficients corresponding to all detection time pointsWherein R 'is a predefined standard belt conveying speed, CS' is a preset number of belt allowed faults delta 1 、δ 2 、δ 3 The weight factor is an influence weight factor corresponding to the preset temperature, the transmission speed and the frequency of belt faults.
And carrying out average value processing on the belt transmission environment quality evaluation coefficients corresponding to the detection time points, and taking the results as the belt transmission environment quality evaluation coefficients.
The second aspect of the invention provides a belt longitudinal tear detection and identification method based on deep learning, comprising the following steps: s1, carrying out historical cargo load analysis on a belt: and acquiring cargo parameters corresponding to the belt in a set detection period from the cloud database, and further analyzing historical transmission quality assessment coefficients corresponding to the belt.
S2, belt detection: and arranging cameras on the belt transmission device along the belt transmission direction at preset intervals on the target edge of the belt, arranging the cameras on the designated edge of the belt in the same way, and detecting images of the belt.
S3, dividing detection areas: dividing the belt detection area into belt detection subareas according to preset detection intervals according to the arrangement intervals of the cameras arranged on the target side.
S4, belt breakage analysis: and acquiring target edge images and appointed edge images of the belt detection subareas which belong to the detection time points, and further analyzing breakage coefficients corresponding to the belt detection subareas which belong to the detection time points.
S5, analyzing a belt breakage growth coefficient: and analyzing each damaged belt subarea corresponding to each detection time point, so as to analyze the damage growth coefficient corresponding to each damaged belt subarea at each damage time point.
Compared with the prior art, the embodiment of the invention has at least the following advantages or beneficial effects: (1) According to the invention, the cargo parameters of the belt in the set detection period are obtained from the belt historical cargo load analysis module, so that the historical transmission quality of the belt is analyzed according to the cargo parameters, the defect that the analysis strength of the historical condition of the belt is not deep enough in the prior art is overcome, the accuracy of the analysis result of tearing and expanding the belt is further ensured, the risk of tearing the belt is timely ensured, early warning is given, the related loss is reduced, and the long-term use of the belt is facilitated.
(2) The invention carries out bidirectional detection on the belt in the belt detection module, the detection method is more accurate, the influence of the singleness of the data source on the accuracy of the belt tearing analysis result is avoided, and a powerful data foundation is laid for the subsequent belt tearing analysis.
(3) The belt tearing detection method and device provided by the invention have the advantages that the detection area division module divides the belt detection area, the belt tearing analysis is more accurate, and the area of the belt tearing risk can be more rapidly positioned.
(4) According to the invention, the area and the regularity of the belt tearing area are analyzed in the belt breakage analysis module, the phenomenon that the regularity attention of the belt tearing area is not high in the prior art is overcome, the safety of expanding the belt tearing area is ensured, and the accuracy of a longitudinal belt tearing analysis result is further ensured, so that powerful data support is provided for the subsequent analysis of the belt tearing growth, and the referential of the analysis result of the belt tearing growth coefficient is improved.
(5) According to the invention, the damage growth coefficient of the damaged subarea is analyzed in the belt damage growth coefficient analysis module, so that the safety of belt tearing growth is ensured, the phenomenon that the belt tearing growth is overlarge is avoided, and the belt is continuously detected under the condition that the damage coefficient of the current damaged subarea meets the requirement, thereby reducing the waste of manpower and material resources to a certain extent.
Drawings
The invention will be further described with reference to the accompanying drawings, in which embodiments do not constitute any limitation of the invention, and other drawings can be obtained by one of ordinary skill in the art without inventive effort from the following drawings.
Fig. 1 is a schematic diagram of the module connection of the present invention.
Fig. 2 is a flow chart of the method of the present invention.
Fig. 3 is a schematic view of a belt of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Referring to fig. 1, a first aspect of the present invention provides a deep learning based belt longitudinal tear detection and identification system, comprising: the system comprises a belt historical cargo load analysis module, a belt detection module detection area division module, a belt damage analysis module, a belt damage growth coefficient analysis module, an early warning terminal and a cloud database.
The belt detection module is connected with the detection area dividing module, the detection area dividing module is connected with the belt damage analysis module, the belt damage analysis module and the belt historical cargo carrying analysis module are connected with the belt damage growth coefficient analysis module, the cloud database is connected with the belt damage growth coefficient analysis module, the belt damage analysis module and the belt historical cargo carrying analysis module respectively, and the early warning terminal is connected with the belt damage growth coefficient analysis module and the belt damage analysis module respectively.
The belt historical cargo load analysis module is used for acquiring cargo parameters corresponding to the belt in a set detection period from the cloud database, and further analyzing historical transmission quality assessment coefficients corresponding to the belt.
In a specific embodiment of the invention, the cargo parameter comprises a weight of the corresponding cargo for each transmission.
In a specific embodiment of the present invention, the historical transmission quality evaluation coefficient corresponding to the belt comprises the following specific steps: extracting the weight G of the goods corresponding to each transmission from the goods parameters corresponding to the belt in the set detection period b Where b is the number of each transmission, b=1, 2.
And carrying out average value processing on the weight of the goods corresponding to each transmission of the belt, so as to obtain the weight average value G' of the goods corresponding to the belt.
Selecting the maximum cargo weight G corresponding to the belt transmission from the cargo weights corresponding to the belt transmissions max And a minimum cargo weight G min 。
Analyzing the weight deviation coefficient of the beltWherein G is b+1 For the weight of the goods corresponding to the b+1th transmission, G 'is the allowable weight error of the preset maximum weight of the goods corresponding to the minimum weight of the goods, G' is the allowable weight error of the preset two adjacent transmissions, c is the transmission times, lambda 1 、λ 2 、λ 3 The weight deviation is the weight coefficient of the influence to which the weight deviation corresponding to the maximum cargo weight and the minimum cargo weight belongs, wherein the weight deviation and the weight deviation are respectively the preset cargo weight deviation and the cargo weight deviation of two adjacent transmissions.
Counting the transmission times C corresponding to the belt, and further analyzing the historical transmission quality evaluation coefficient corresponding to the belt according to the predefined standard transmission times C' corresponding to the beltWhere e is a natural constant.
According to the invention, the cargo parameters of the belt in the set detection period are obtained from the belt historical cargo load analysis module, so that the historical transmission quality of the belt is analyzed according to the cargo parameters, the defect that the analysis strength of the historical condition of the belt is not deep enough in the prior art is overcome, the accuracy of the analysis result of tearing and expanding the belt is further ensured, the risk of tearing the belt is timely ensured, early warning is given, the related loss is reduced, and the long-term use of the belt is facilitated.
Referring to fig. 3, the belt detection module is configured to set each camera at a target edge of a belt according to a preset interval in a direction along which the belt is driven by the belt driving device, and set each camera at a designated edge of the belt in a similar manner, thereby performing image detection on the belt.
The invention carries out bidirectional detection on the belt in the belt detection module, the detection method is more accurate, the influence of the singleness of the data source on the accuracy of the belt tearing analysis result is avoided, and a powerful data foundation is laid for the subsequent belt tearing analysis.
The detection area dividing module is used for dividing the belt detection area into belt detection subareas according to preset detection intervals according to the arrangement intervals of the cameras arranged on the target side.
The belt tearing detection method and device provided by the invention have the advantages that the detection area division module divides the belt detection area, the belt tearing analysis is more accurate, and the area of the belt tearing risk can be more rapidly positioned.
The belt breakage analysis module is used for acquiring target edge images and appointed edge images of all belt detection subareas which all detection time points belong to, and further analyzing breakage coefficients corresponding to all belt detection subareas which all detection time points belong to.
In a specific embodiment of the present invention, the analyzing the breakage coefficient corresponding to each belt detection sub-area to which each detection time point belongs specifically includes: and acquiring each gray value of the target edge image corresponding to each belt detection sub-region to which each detection time point belongs according to the target edge image of each belt detection sub-region to which each detection time point belongs.
Extracting a cracking gray value range corresponding to the belt from the cloud database, comparing each gray value of the target edge image of each belt detection sub-region to which each detection time point belongs with the cracking gray value range corresponding to the belt, screening each abnormal region corresponding to each belt detection sub-region to which each detection time point belongs, acquiring the corresponding area, selecting the region with the largest area from the areas as the target cracking region, and obtaining the target cracking region corresponding to each belt detection sub-region to which each detection time point belongs.
The specific method for screening each abnormal region corresponding to each belt detection sub-region to which each detection time point belongs is as follows: if a gray value of a target edge image of a belt detection subarea to which a certain detection time point belongs is in a range of cracking gray values, marking the gray value as a cracking gray value, acquiring a region in which each cracking gray value of each belt detection subarea to which each detection time point belongs is located, and marking the region as each abnormal region corresponding to each belt detection subarea to which each detection time point belongs.
Acquiring the area S of the corresponding target cracking area of each belt detection sub-area to which each detection time point belongs jp Where j is the number of each detection time point, j=1, 2,..k, p is the number of each belt detection sub-region, p=1, 2,..q.
Analyzing the damage coefficient corresponding to each belt detection subarea to which each detection time point belongs during target edge detectionWherein S' is the preset area of the target cracking area.
Similarly, analyzing the breakage coefficient M 'corresponding to each belt detection sub-region to which each detection time point belongs during the detection of the designated edge' jp 。
Analyzing breakage coefficients corresponding to belt detection subareas to which each detection time point belongsWherein gamma is 1 、γ 2 Is a weight factor mu corresponding to the preset target cracking area and target cracking area regularity jp And the profile regularity of the target cracking area corresponding to the p belt detection sub-area to which the j detection time point belongs.
In a specific embodiment of the present invention, the contour regularity of the target cracking area corresponding to each belt detection sub-area to which each detection time point belongs is as follows: and acquiring the external contour of the target cracking area corresponding to each belt detection sub-area to which each detection time point belongs, and acquiring the corresponding line.
A three-dimensional coordinate system is established by taking the central point corresponding to each belt detection sub-region as an origin, and the three-dimensional coordinate system is set according to the set values from the external contour lines of the target cracking region corresponding to each belt detection sub-region which each detection time point belongs toEach contour point is selected according to the length of the frame, and the corresponding three-dimensional coordinates are obtainedWhere h is the number of each contour point, h=1, 2.
Analyzing the distance between each contour point corresponding to each belt detection sub-region to which each detection time point belongs and the adjacent contour point
It should be noted that, the distance between the h outline point and the h+1th outline point corresponding to the p-th belt detection sub-area to which the j-th detection time point belongs is analyzedThe specific calculation formula is as follows:wherein the method comprises the steps ofAnd the detection time point j is the three-dimensional coordinate of the (h+1) th contour point corresponding to the (p) th belt detection sub-area to which the detection time point j belongs.
Analyzing the contour regularity of the target cracking area corresponding to each belt detection sub-area to which each detection time point belongsWherein->The p belt detection sub-area which belongs to the j detection time point corresponds to the distance between the h+1th contour point and the h+2th contour point, g is the number of the contour points, and JL' is the error of the allowable distance deviation of the preset contour point distance.
According to the invention, the area and the regularity of the belt tearing area are analyzed in the belt breakage analysis module, the phenomenon that the regularity attention of the belt tearing area is not high in the prior art is overcome, the safety of expanding the belt tearing area is ensured, and the accuracy of a longitudinal belt tearing analysis result is further ensured, so that powerful data support is provided for the subsequent analysis of the belt tearing growth, and the referential of the analysis result of the belt tearing growth coefficient is improved.
The belt damage growth coefficient analysis module is used for analyzing each damaged belt subarea corresponding to each detection time point so as to analyze the damage growth coefficient corresponding to each damaged belt subarea at each damage time point.
In a specific embodiment of the present invention, the analyzing each damaged belt subarea corresponding to each detection time point specifically includes: comparing the damage coefficient corresponding to each belt detection sub-region to which each detection time point belongs with a preset damage coefficient threshold value, if the damage coefficient corresponding to a certain belt detection sub-region to which a certain detection time point belongs is larger than or equal to the damage coefficient threshold value, marking the belt detection sub-region as a damaged belt sub-region, marking the detection time point as a damaged time point, and further obtaining each damaged belt sub-region to which each damage time point belongs.
In a specific embodiment of the present invention, the damage growth coefficient corresponding to each damaged belt subarea at each damaged time point is specifically analyzed by: obtaining damage coefficients corresponding to the damaged belt subareas of the damaged time points according to the damage coefficients corresponding to the belt detection subareas of the damaged time points, and summarizing the damage coefficients to obtain damage coefficients sigma of the damaged belt subareas of the damaged time points m,f Where m is the number of each damaged belt sub-region, m=1, 2,..i, f is the number of each damaged time point, f=1, 2..t, and further obtain the interval J of each damage time point from its neighboring damage time points f,f-1 Wherein f > 1.
Analyzing the damage growth rate of each damaged belt subarea corresponding to each damage time pointWherein sigma m,f-1 For the mth damaged belt sub-region corresponding to the f-1 th damaged time pointIs a breakage coefficient of the steel sheet.
Comprehensively analyzing damage growth coefficients of damaged belt subareas corresponding to each damage time pointWhere ζ is a belt transport environment quality assessment factor, ω "is a predetermined allowable damage growth rate, χ 1 、χ 2 、χ 3 The method comprises the steps of respectively presetting a damage growth rate, historical transmission quality of the belt and a duty ratio factor corresponding to the environmental quality of the belt transmission.
In a specific embodiment of the present invention, the belt transmission environment quality evaluation coefficient ζ is specifically analyzed by: acquiring belt transmission speed R corresponding to each detection time point j And acquiring the temperature WT corresponding to each belt detection subarea corresponding to each detection time point jp And extracting the historical failure times CS corresponding to the belt from the cloud database.
The temperature sensor is used to obtain the temperature corresponding to each belt detection sub-area at each detection time point.
And carrying out average value processing on the temperature corresponding to each belt detection subarea corresponding to each detection time point, further obtaining a temperature average value corresponding to each detection time point, carrying out average value processing on the temperature average value corresponding to each detection time point, and further taking the result as a proper temperature WT' corresponding to the belt.
Analyzing the belt transmission environment quality evaluation coefficients corresponding to all detection time pointsWherein R 'is a predefined standard belt conveying speed, CS' is a preset number of belt allowed faults delta 1 、δ 2 、δ 3 The weight factor is an influence weight factor corresponding to the preset temperature, the transmission speed and the frequency of belt faults.
And carrying out average value processing on the belt transmission environment quality evaluation coefficients corresponding to the detection time points, and taking the results as the belt transmission environment quality evaluation coefficients.
According to the invention, the damage growth coefficient of the damaged subarea is analyzed in the belt damage growth coefficient analysis module, so that the safety of belt tearing growth is ensured, the phenomenon that the belt tearing growth is overlarge is avoided, and the belt is continuously detected under the condition that the damage coefficient of the current damaged subarea meets the requirement, thereby reducing the waste of manpower and material resources to a certain extent.
The early warning terminal is used for carrying out corresponding early warning according to the damage coefficient of each belt detection sub-region at each detection time point and carrying out corresponding early warning according to the damage growth coefficient of each damaged belt sub-region corresponding to each damage time point.
The damage coefficient of each belt detection subarea at each detection time point is compared with a preset damage coefficient warning value, if the damage coefficient of a certain belt detection subarea at a certain detection time point is greater than or equal to the damage coefficient warning value, the belt damage abnormality early warning is carried out at the detection time point, and similarly, the damage growth abnormality early warning is carried out according to the damage growth coefficient corresponding to each damage time point of each damaged belt subarea.
The cloud database is used for storing cargo parameters corresponding to the belt in a set detection period, storing a cracking gray value range and storing historical fault times corresponding to the belt.
Referring to fig. 2, a second aspect of the present invention provides a belt longitudinal tear detection and identification method based on deep learning, including: s1, carrying out historical cargo load analysis on a belt: and acquiring cargo parameters corresponding to the belt in a set detection period from the cloud database, and further analyzing historical transmission quality assessment coefficients corresponding to the belt.
S2, belt detection: and arranging cameras on the belt transmission device along the belt transmission direction at preset intervals on the target edge of the belt, arranging the cameras on the designated edge of the belt in the same way, and detecting images of the belt.
S3, dividing detection areas: dividing the belt detection area into belt detection subareas according to preset detection intervals according to the arrangement intervals of the cameras arranged on the target side.
S4, belt breakage analysis: and acquiring target edge images and appointed edge images of the belt detection subareas which belong to the detection time points, and further analyzing breakage coefficients corresponding to the belt detection subareas which belong to the detection time points.
S5, analyzing a belt breakage growth coefficient: and analyzing each damaged belt subarea corresponding to each detection time point, so as to analyze the damage growth coefficient corresponding to each damaged belt subarea at each damage time point.
The foregoing is merely illustrative of the structures of this invention and various modifications, additions and substitutions for those skilled in the art of describing particular embodiments without departing from the structures of the invention or exceeding the scope of the invention as defined by the claims.
Claims (9)
1. Belt longitudinal tear detects identification system based on degree of depth study, its characterized in that includes:
the belt historical cargo load analysis module is used for acquiring cargo parameters corresponding to the belt in a set detection period from the cloud database, and further analyzing historical transmission quality assessment coefficients corresponding to the belt;
the belt detection module is used for arranging cameras on the target edge of the belt according to preset intervals along the belt transmission direction of the belt transmission device, arranging the cameras on the designated edge of the belt in the same way, and detecting images of the belt;
the detection area dividing module is used for dividing the belt detection area into belt detection subareas according to preset detection intervals according to the arrangement intervals of the cameras arranged on the target edge;
the belt breakage analysis module is used for acquiring target edge images and appointed edge images of all belt detection subareas to which all detection time points belong, and further analyzing breakage coefficients corresponding to all belt detection subareas to which all detection time points belong;
the belt damage growth coefficient analysis module is used for analyzing each damaged belt subarea corresponding to each detection time point so as to analyze the damage growth coefficient corresponding to each damaged belt subarea at each damage time point;
the early warning terminal is used for carrying out corresponding early warning according to the damage coefficient of each belt detection sub-region at each detection time point and carrying out corresponding early warning according to the damage growth coefficient of each damaged belt sub-region corresponding to each damage time point;
and the cloud database is used for storing cargo parameters corresponding to the belt in a set detection period, storing a cracking gray value range and storing historical fault times corresponding to the belt.
2. The deep learning based belt longitudinal tear detection and identification system of claim 1, wherein: the cargo parameter includes a cargo weight corresponding to each transmission.
3. The deep learning based belt longitudinal tear detection and identification system of claim 2, wherein: the historical transmission quality evaluation coefficient corresponding to the belt comprises the following specific steps:
extracting the weight G of the goods corresponding to each transmission from the goods parameters corresponding to the belt in the set detection period b Where b is the number of each transmission, b=1, 2, c;
carrying out average value processing on the weight of goods corresponding to each transmission of the belt, and further obtaining a weight average value G' of the goods corresponding to the belt;
selecting the maximum cargo weight G corresponding to the belt transmission from the cargo weights corresponding to the belt transmissions max And a minimum cargo weight G min ;
Analyzing the weight deviation coefficient of the beltWherein G is b+1 For the weight of the goods corresponding to the b+1th transmission, G 'is the allowable weight error of the preset maximum weight of the goods corresponding to the minimum weight of the goods, G' is the allowable weight error of the preset two adjacent transmissions, c is the transmission times, lambda 1 、λ 2 、λ 3 The weight deviation is the weight coefficient of the influence to which the weight deviation corresponding to the maximum cargo weight and the minimum cargo weight is the preset cargo weight deviation of two adjacent transmissions;
counting the transmission times C corresponding to the belt, and further analyzing the historical transmission quality evaluation coefficient corresponding to the belt according to the predefined standard transmission times C' corresponding to the beltWhere e is a natural constant.
4. The deep learning based belt longitudinal tear detection and identification system of claim 3, wherein: the method for analyzing the damage coefficient corresponding to each belt detection sub-region to which each detection time point belongs comprises the following specific steps:
acquiring each gray value of the target edge image corresponding to each belt detection sub-region to which each detection time point belongs according to the target edge image of each belt detection sub-region to which each detection time point belongs;
extracting a cracking gray value range corresponding to the belt from a cloud database, comparing each gray value of a target edge image of each belt detection sub-region to which each detection time point belongs with the cracking gray value range corresponding to the belt, screening each abnormal region corresponding to each belt detection sub-region to which each detection time point belongs, acquiring the corresponding area, selecting the region with the largest area from the areas as a target cracking region, and obtaining the target cracking region corresponding to each belt detection sub-region to which each detection time point belongs;
acquiring the area S of the corresponding target cracking area of each belt detection sub-area to which each detection time point belongs jp Where j is the number of each detection time point, j=1, 2,..k, p is the number of each belt detection sub-region, p=1, 2,..q;
analyzing the damage coefficient corresponding to each belt detection subarea to which each detection time point belongs during target edge detectionS' is the area of a preset allowable target cracking area;
similarly, analyzing the damage coefficient Mj' p corresponding to each belt detection sub-region to which each detection time point belongs when the specified edge is detected;
analyzing breakage coefficients corresponding to belt detection subareas to which each detection time point belongsWherein gamma is 1 、γ 2 Is a weight factor mu corresponding to the preset target cracking area and target cracking area regularity jp And the profile regularity of the target cracking area corresponding to the p belt detection sub-area to which the j detection time point belongs.
5. The deep learning based belt longitudinal tear detection and identification system of claim 4, wherein: the contour regularity of the corresponding target cracking area of each belt detection sub-area to which each detection time point belongs is as follows:
acquiring the external contour of a target cracking area corresponding to each belt detection sub-area to which each detection time point belongs, and acquiring corresponding lines;
establishing a three-dimensional coordinate system by taking the central point corresponding to each belt detection sub-region as an origin, selecting each contour point from the external contour lines of the target cracking region corresponding to each belt detection sub-region to which each detection time point belongs according to a set length, and acquiring the corresponding three-dimensional coordinateWhere h is the number of each contour point, h=1, 2, g;
analyzing the distance between each contour point corresponding to each belt detection sub-region to which each detection time point belongs and the adjacent contour point
Analyzing the contour regularity of the target cracking area corresponding to each belt detection sub-area to which each detection time point belongsWherein->The p belt detection sub-area which belongs to the j detection time point corresponds to the distance between the h+1th contour point and the h+2th contour point, g is the number of the contour points, and JL' is the error of the allowable distance deviation of the preset contour point distance.
6. The deep learning based belt longitudinal tear detection and identification system of claim 1, wherein: the method for analyzing each damaged belt subarea corresponding to each detection time point comprises the following specific steps: comparing the damage coefficient corresponding to each belt detection sub-region to which each detection time point belongs with a preset damage coefficient threshold value, if the damage coefficient corresponding to a certain belt detection sub-region to which a certain detection time point belongs is larger than or equal to the damage coefficient threshold value, marking the belt detection sub-region as a damaged belt sub-region, marking the detection time point as a damaged time point, and further obtaining each damaged belt sub-region to which each damage time point belongs.
7. The deep learning based belt longitudinal tear detection and identification system of claim 4, wherein: the damage growth coefficient corresponding to each damaged belt subarea at each damage time point is specifically analyzed by the following steps:
obtaining damage coefficients corresponding to the damaged belt subareas of the damaged time points according to the damage coefficients corresponding to the belt detection subareas of the damaged time points, and summarizing the damage coefficients to obtain damage coefficients sigma of the damaged belt subareas of the damaged time points m,f Wherein m is the number of each damaged belt sub-region, m=1, 2, l, f is the number of each damage time point, f=1, 2,..t, and further each damage time point is obtained together with the sameInterval J of adjacent damaged time points f,f-1 Wherein f > 1;
analyzing the damage growth rate of each damaged belt subarea corresponding to each damage time pointWherein sigma m,f-1 The damage coefficient of the mth damaged belt subarea corresponding to the f-1 damage time point is determined;
comprehensively analyzing damage growth coefficients of damaged belt subareas corresponding to each damage time pointWhere ζ is a belt transport environment quality assessment factor, ω "is a predetermined allowable damage growth rate, χ 1 、χ 2 、χ 3 The method comprises the steps of respectively presetting a damage growth rate, historical transmission quality of the belt and a duty ratio factor corresponding to the environmental quality of the belt transmission.
8. The deep learning based belt longitudinal tear detection and identification system of claim 7, wherein: the belt transmission environment quality evaluation coefficient xi comprises the following specific analysis methods:
acquiring belt transmission speed R corresponding to each detection time point j And acquiring the temperature WT corresponding to each belt detection subarea corresponding to each detection time point jp Extracting historical fault times CS corresponding to the belt from a cloud database;
carrying out average value processing on the temperature corresponding to each belt detection subarea corresponding to each detection time point so as to obtain a temperature average value corresponding to each detection time point, carrying out average value processing on the temperature average value corresponding to each detection time point, and taking the result as a proper temperature WT' corresponding to the belt;
analyzing the belt transmission environment quality evaluation coefficients corresponding to all detection time pointsWherein R' is a predefined standard beltThe transmission speed CS' is the preset allowable fault times delta of the belt 1 、δ 2 、δ 3 The method is an influence weight factor corresponding to the preset temperature, transmission speed and belt failure times;
and carrying out average value processing on the belt transmission environment quality evaluation coefficients corresponding to the detection time points, and taking the results as the belt transmission environment quality evaluation coefficients.
9. The belt longitudinal tearing detection and identification method based on deep learning is characterized by comprising the following steps of: comprising the following steps:
s1, carrying out historical cargo load analysis on a belt: acquiring cargo parameters corresponding to the belt in a set detection period from a cloud database, and further analyzing historical transmission quality assessment coefficients corresponding to the belt;
s2, belt detection: arranging cameras on the belt transmission device along the belt transmission direction at preset intervals on the belt target side, arranging the cameras on the belt designated side in the same way, and detecting images of the belt;
s3, dividing detection areas: dividing a belt detection area into belt detection subareas according to preset detection intervals according to the arrangement intervals of the cameras arranged on the target side;
s4, belt breakage analysis: acquiring target edge images and appointed edge images of all belt detection subareas to which all detection time points belong, and further analyzing breakage coefficients corresponding to all belt detection subareas to which all detection time points belong;
s5, analyzing a belt breakage growth coefficient: and analyzing each damaged belt subarea corresponding to each detection time point, so as to analyze the damage growth coefficient corresponding to each damaged belt subarea at each damage time point.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310497175.7A CN116461931A (en) | 2023-05-05 | 2023-05-05 | Belt longitudinal tearing detection and identification system and method based on deep learning |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310497175.7A CN116461931A (en) | 2023-05-05 | 2023-05-05 | Belt longitudinal tearing detection and identification system and method based on deep learning |
Publications (1)
Publication Number | Publication Date |
---|---|
CN116461931A true CN116461931A (en) | 2023-07-21 |
Family
ID=87175333
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310497175.7A Pending CN116461931A (en) | 2023-05-05 | 2023-05-05 | Belt longitudinal tearing detection and identification system and method based on deep learning |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116461931A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117333727A (en) * | 2023-12-01 | 2024-01-02 | 山东瀚邦胶带有限公司 | Conveyer belt wearing and tearing area discernment analysis system based on visual image processing technique |
-
2023
- 2023-05-05 CN CN202310497175.7A patent/CN116461931A/en active Pending
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117333727A (en) * | 2023-12-01 | 2024-01-02 | 山东瀚邦胶带有限公司 | Conveyer belt wearing and tearing area discernment analysis system based on visual image processing technique |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN116461931A (en) | Belt longitudinal tearing detection and identification system and method based on deep learning | |
US10865052B2 (en) | Monitoring system for conveyor belt | |
CN113724258B (en) | Conveyor belt tearing detection method and system based on image processing | |
CN106770321B (en) | A kind of plastic part defect inspection method based on multi-threshold section | |
CN111591715A (en) | Belt longitudinal tearing detection method and device | |
CN100465836C (en) | Method and arrangement for working profile blanks | |
CN113129282A (en) | Belt abnormality determination method, device, equipment and storage medium | |
CN106956917A (en) | A kind of buffer storage and method for Real-time Balancing multitask production line | |
CN116177146A (en) | Conveying belt detection system and method based on AI and laser monitoring technology | |
CN115394060A (en) | Intelligent AI video monitoring system for coal mine fully-mechanized coal mining face based on internet | |
CN116664552B (en) | Elevator door abnormity detection method based on video image analysis | |
CN113762283A (en) | Method and device for monitoring deviation of conveying belt | |
CN115504187B (en) | Intelligent speed regulation and protection system and control method for mining belt | |
CN116189093A (en) | Slag soil vehicle dust on-line monitoring method and system | |
CN105021630A (en) | Automatic detection method for surface damages of conveyor belt | |
Martinez et al. | Online vision-based inspection system for thermoplastic hot plate welding in window frame manufacturing | |
JP6741527B2 (en) | Apparatus for measuring the number of laminated plywood and method for measuring the number thereof | |
Kleta et al. | Image processing and analysis as a diagnostic tool for an underground infrastructure technical condition monitoring | |
Onita et al. | Quality control in porcelain industry based on computer vision techniques | |
US11796420B2 (en) | Bearing maintenance timing acquisition method | |
CN114897846A (en) | Dot matrix target regularity detection method based on machine vision | |
CN107578185B (en) | Health detection method and device for industrial building | |
CN117489412B (en) | Real-time early warning method based on 5G mine safety | |
US10024657B2 (en) | Method for classifying metal tubes | |
CN110910357B (en) | Welding spot defect detection method, system and medium based on machine vision |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |