Automatic detection system for defects of coated grain

Abstract

The automatic detection system for the defects of the coated grain comprises a negative pressure loading device, the manipulator is used for placing the coated grain to be detected into the groove and taking out the coated grain from the groove; the laser speckle detection device is used for detecting defects of the first end and the second end of the coated grain; the control unit is used for controlling the negative pressure cabin, the turntable, the manipulator and the laser speckle detection device; wherein the turntable is used for driving the coated grain on the sample table to rotate so as to align the first end and the second end to a laser speckle detection device respectively. The turntable can drive the coated grain to rotate, the two end faces of the coated grain can be detected by closing the hatch cover once, defect identification and unqualified product screening are automatically completed, and the detection efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the field of defect detection and identification, and in particular to an automatic defect detection system for coated drug columns. Background Art

[0002] Chinese patent ZL202210626459.7 discloses a method for identifying interface defect types of a solid propellant grain coating, wherein a flat plate sample with preset debonding defects, bubble defects and inclusion defects is prepared, the flat plate sample comprising a substrate and a rubber bonded to the substrate by an adhesive; a high-vibration-resistant electronic speckle interferometry real-time phase measurement system is used to perform defect detection on the flat plate sample under negative pressure loading to obtain the off-plane displacement corresponding to each defect; the off-plane displacement corresponding to each defect is normalized according to the thin plate deformation theory of elastic mechanics, and the relationship between the off-plane displacement corresponding to the debonding defect, bubble defect and inclusion defect and the negative pressure value is quantitatively analyzed as a detection standard; defect detection is performed on the solid propellant grain coating under negative pressure loading, the off-plane displacement corresponding to the defect is calculated and normalized, and the debonding defect of the solid propellant grain coating is identified by comparing it with the detection standard.

[0003] However, existing technologies including this patent require targeted inspection of each solid propellant grain with the assistance of manual operation. Traditional single-piece manual inspection is relatively inefficient and is prone to human errors, such as missed inspections.

[0004] Therefore, it is necessary to develop an automated detection system for coated drug column defects to solve one or more of the above-mentioned technical problems. Summary of the invention

[0005] In order to solve at least one of the above technical problems, according to one aspect of the present invention, a system for automatically detecting defects of coated drug pillars is provided, comprising:

[0006] The negative pressure loading device comprises: a negative pressure cabin; a turntable arranged in the negative pressure cabin; and a sample table arranged on the turntable and provided with a groove for placing the coated medicine column;

[0007] A manipulator, used for placing the coated medicine column to be tested into the groove and taking the coated medicine column out of the groove;

[0008] a laser speckle detection device, used to detect defects at the first end and the second end of the coated drug column; and

[0009] A control unit, used to control the negative pressure cabin, the turntable, the manipulator and the laser speckle detection device;

[0010] Wherein, the turntable is used to drive the coated medicine column on the sample stage to rotate so as to align the first end and the second end to the laser speckle detection device respectively.

[0011] According to another aspect of the present invention, the negative pressure cabin comprises a cabin body and an openable and closable cabin cover, and a closed space is formed when the cabin cover is closed on the cabin body.

[0012] According to another aspect of the present invention, the hatch is provided with a first photoelectric sensor, and the cabin is provided with a second photoelectric sensor. When the hatch is open, the first photoelectric sensor and the second photoelectric sensor form a first detection unit for detecting whether the manipulator leaves the negative pressure cabin.

[0013] According to yet another aspect of the present invention, the control unit is further configured to automatically control the closing of the cabin cover when the first detection unit detects that the manipulator has left the negative pressure cabin.

[0014] According to another aspect of the present invention, the automatic detection system for coated drug column defects further comprises a sample photoelectric sensor for detecting whether a coated drug column is placed on the sample stage.

[0015] According to yet another aspect of the present invention, the hatch cover and the cabin body are connected via a slide rail.

[0016] According to yet another aspect of the present invention, the groove is a V-shaped groove.

[0017] According to another aspect of the present invention, the control unit is used to control the negative pressure chamber to exhaust air to reach a set maximum air pressure value. When the air inlet valve lets air into the negative pressure chamber to a first negative pressure value, the laser speckle detection device collects a reference image of the first end. When the air continues to be let into the negative pressure chamber to a second negative pressure value, the laser speckle detection device collects a detection image of the first end. A first defect image is acquired based on the reference image and the detection image of the first end, and the difference between the first negative pressure value and the second negative pressure value is a first predetermined value.

[0018] According to another aspect of the present invention, the control unit is used to control the laser speckle detection device to collect a reference image of the second end when the air intake valve continues to intake air into the negative pressure cabin to a third negative pressure value, and to collect a detection image of the second end when the air intake valve continues to intake air into the negative pressure cabin to a fourth negative pressure value, and to obtain a second defect image based on the reference image and the detection image of the second end, and the difference between the third negative pressure value and the fourth negative pressure value is the first predetermined value.

[0019] According to another aspect of the present invention, the defect image can be automatically identified and its size can be automatically measured based on the YOLO neural network algorithm, and when the defect diameter of the first defect image or the second defect image is greater than a threshold, it is automatically determined to be unqualified.

[0020] The present invention can achieve one or more of the following technical effects:

[0021] 1. The turntable can drive the coated grain to rotate, and the single closing of the hatch cover can detect the two end surfaces of the coated grain, realizing automatic performance, reducing labor costs and human errors, and improving detection efficiency.

[0022] 2. The automatic detection manipulator leaves the negative pressure cabin, which can realize the automatic opening and closing of the cabin cover, automatic loading of samples, automatic triggering of measurements and automatic control of the turntable, so as to realize the mechanized batch and efficient detection of coated drug columns.

[0023] 3. Based on the YOLO neural network algorithm, the defect image can be automatically identified and the size can be automatically measured, and the product can be automatically determined whether it is qualified, reducing the influence of human judgment and truly realizing fully automated detection of coated drug columns. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0025] Figure 1 Schematic diagram of a negative pressure loading device according to a preferred embodiment of the present invention.

[0026] Figure 2 for Figure 1 Top view of the negative pressure loading device.

[0027] Figure 3 for Figure 1 Diagram of partial structure of the negative pressure loading device (rear part of the hatch cover).

[0028] Figure 4 Schematic diagram of a laser speckle detection device according to a preferred embodiment of the present invention.

[0029] Figure 5 This is an original speckle image, a filter image and an automatic defect recognition image for laser speckle detection according to a preferred embodiment of the present invention.

[0030] Description of main symbols:

[0031] 1. Hatch cover; 2. Slide rail; 3. Positioning rod; 4. Laser irradiation window; 5. Negative pressure cabin; 6. Negative pressure cabin support platform; 7. Turntable; 8. Positioning plate; 9. Sample stage; 10. Second photoelectric sensor; 11. Sample photoelectric sensor; 12. Sample stage screw slot; 13. Positioning plate screw slot;

[0032] 14. First photoelectric sensor; 15. Negative pressure cabin sealing plate; 16. Fixed sealing ring; 17. L-shaped fixing plate; 18. Flaw detector; 19. Laser; 20. Floor support frame DETAILED DESCRIPTION

[0033] The best implementation mode of the present invention is described below through preferred embodiments in conjunction with the accompanying drawings. The specific implementation mode here is to illustrate the present invention in detail and should not be understood as limiting the present invention. Various deformations and modifications can be made without departing from the spirit and essential scope of the present invention, which should be included in the protection scope of the present invention.

[0034] Example 1

[0035] According to a preferred embodiment of the present invention, see Figure 1-5 , provides an automated detection system for coated drug column defects, including:

[0036] The negative pressure loading device comprises: a negative pressure chamber 5; a turntable 7, which is arranged in the negative pressure chamber; and a sample table 9, which is arranged on the turntable and has a groove for placing the coated medicine column;

[0037] A manipulator, used for placing the coated medicine column to be tested into the groove and taking the coated medicine column out of the groove;

[0038] a laser speckle detection device 18, for detecting defects at the first end and the second end of the coated drug column; and

[0039] A control unit, used to control the negative pressure cabin, the turntable, the manipulator and the laser speckle detection device;

[0040] Wherein, the turntable is used to drive the coated medicine column on the sample stage to rotate so as to align the first end and the second end to the laser speckle detection device respectively.

[0041] According to another preferred embodiment of the present invention, the negative pressure cabin comprises a cabin body and an openable and closable cabin cover 1, and a closed space is formed when the cabin cover is closed on the cabin body.

[0042] According to another preferred embodiment of the present invention, the hatch is provided with a first photoelectric sensor 14, and the cabin is provided with a second photoelectric sensor 10. The first photoelectric sensor and the second photoelectric sensor form a first detection unit for detecting whether the manipulator leaves the negative pressure cabin when the hatch 1 is opened.

[0043] According to another preferred embodiment of the present invention, the control unit is further configured to automatically control the hatch cover 1 to close when the first detection unit detects that the manipulator leaves the negative pressure cabin 5 .

[0044] According to another preferred embodiment of the present invention, the automatic detection system for coated grain defects further comprises a sample photoelectric sensor 11 for detecting whether a coated grain is placed on the sample stage.

[0045] According to another preferred embodiment of the present invention, the hatch cover and the cabin body are connected via a slide rail 2 .

[0046] According to another preferred embodiment of the present invention, the groove is a V-shaped groove.

[0047] According to another preferred embodiment of the present invention, the control unit is used to control the negative pressure chamber to exhaust air to reach a set maximum air pressure value, and when the air inlet valve (not shown) lets air into the negative pressure chamber to a first negative pressure value, the laser speckle detection device collects a reference image of the first end, and when the air continues to be let into the negative pressure chamber to a second negative pressure value, the laser speckle detection device collects a detection image of the first end, and a first defect image is acquired according to the reference image and the detection image of the first end, and the difference between the first negative pressure value and the second negative pressure value is a first predetermined value.

[0048] According to another preferred embodiment of the present invention, the control unit is used to control the laser speckle detection device to collect a reference image of the second end when the air intake valve continues to intake air into the negative pressure cabin to a third negative pressure value, and the laser speckle detection device to collect a detection image of the second end when the air intake valve continues to intake air into the negative pressure cabin to a fourth negative pressure value, and obtain a second defect image based on the reference image and the detection image of the second end, and the difference between the third negative pressure value and the fourth negative pressure value is the first predetermined value.

[0049] According to another preferred embodiment of the present invention, when the defect diameter of the first defect image or the second defect image is greater than a threshold, the defect is automatically determined to be unqualified. Preferably, the defect image can be automatically identified and the size can be automatically measured based on the YOLO neural network algorithm.

[0050] Example 2

[0051] According to a preferred embodiment of the present invention, see Figure 1-5 , also provides an automated detection system for coated grain defects, including a negative pressure cabin 5 and a laser speckle detection device 18, a sample table 9 is installed on a turntable 7, a V-shaped groove in the middle of the sample table 9 is used to place the coated grain, and the sample table 9 rotates with the turntable 7 driven by the rotating mechanism, so that defects at both ends of the coated grain are detected.

[0052] Preferably, the laser speckle detection device 18 comprises a flaw detector fixed in front of the negative pressure cabin window and a floor support frame 20, and four laser sources 19 are installed on the side of the flaw detector. The laser source is a continuous laser emitter capable of emitting 532nm.

[0053] Preferably, a sample photoelectric sensor 11 is installed on the rear wall of the negative pressure cabin 5 behind the sample stage 9 to detect whether a coated drug column is placed in the sample stage.

[0054] Preferably, the negative pressure cabin 5 is also equipped with a first photoelectric sensor 14 and a second photoelectric sensor 10, and the cabin door slowly closes when the manipulator for placing the sample completely leaves the negative pressure cabin 5.

[0055] Preferably, two right-angle trapezoidal blocks with transverse through grooves are provided in the negative pressure chamber, which are arranged oppositely and in parallel to form a V-shaped groove to place the sample-coated drug column to be tested, and the trapezoidal blocks are fixed to the center of the bottom turntable with screws, and a rotating mechanism is installed at the bottom of the turntable.

[0056] Preferably, air inlets and outlets and a real-time air pressure detection gauge are provided on the side of the negative pressure cabin body.

[0057] Preferably, the laser speckle detection device 18 also includes a camera, which uses a high-resolution digital camera and an imaging lens. The camera and the lens can realize vertical movement (500-800mm), pitch (±30°) and rotation operation (±45°) together with the flaw detector to achieve azimuth adjustment.

[0058] Preferably, the bracket 20 can be raised and lowered, and the bracket feet are installed in triangular grooves with fixed relative positions.

[0059] Preferably, when the sample is placed inside the negative pressure chamber, the hatch cover 1 is slowly moved to a designated position via the top slide rail 2, and the sealing plate 15 inside the hatch cover is vertically lowered to form a closed space in the chamber.

[0060] Preferably, when the sample photoelectric sensor detects the sample and the manipulator exits the negative pressure chamber, the cabin cover automatically moves slowly and closes to perform a laser speckle detection operation.

[0061] Preferably, a first photoelectric sensor 14 is embedded and installed inside the rear side of the hatch cover 1, and a second photoelectric sensor 10 is also installed on the top of the rear side of the vacuum chamber, and the first photoelectric sensor and the second photoelectric sensor are arranged in parallel. When there is a foreign object blocking the hatch cover during closing, the hatch will immediately stop closing and return to the initial unclosed state, thereby stopping the detection operation.

[0062] Preferably, the rotating mechanism connected to the circular chassis under the fixed V-shaped groove is set to rotate 180°, that is, the interval time from the first detection end face to the second detection end face.

[0063] Preferably, after the second end face is detected and analyzed, the hatch automatically opens to take out the detected sample and wait for other samples to be detected.

[0064] Preferably, after the second sample is placed in, the low-frequency red light emitted by the sample photoelectric sensing device at the rear side of the cabin senses the sample and goes out, that is, the test object is detected to be placed in, and the manipulator moves out of the cabin. The first photoelectric sensing device 14 and the second photoelectric sensing device 10 at the rear side of the hatch are arranged in parallel and opposite to each other in space. When there is no obstruction by foreign matter in the middle, the hatch automatically closes slowly for the next round of detection. When the hatch is closed, the first photoelectric sensing device 14 displays a green light, and when it is obstructed by foreign matter, it displays a red light, and the hatch is reset without closing.

[0065] Preferably, the flaw detector includes a laser lighting device, a high-resolution camera, an imaging lens, and a high-vibration-resistant real-time phase processing device. The control unit is used to realize the indication of system power supply, the opening and closing of the laser, the automatic focusing of the zoom lens, and the data exchange between the phase shift generator and the computer.

[0066] Preferably, before defect measurement and / or deformation measurement, image calibration is automatically performed. The measurement results can be displayed on a computer and a defect phase diagram at the same time. At the same time, defect size and / or out-of-plane displacement can be measured.

[0067] Preferably, the negative pressure cabin is also provided with a positioning rod 3, a laser irradiation window 4, a negative pressure cabin support platform 6, a positioning plate 8, a sample stage screw slot 12, a positioning plate screw slot 13, a negative pressure cabin sealing plate 15, a fixed sealing ring 16, and an L-shaped fixing plate 17. The positioning rod 3 is fixed to the negative pressure cabin together with the slide rail 2, so that the hatch cover stops precisely above the negative pressure cabin during the closing process, and at the same time prevents the hatch cover from colliding with the rear photoelectric sensor 10 when moving forward and backward and from detaching from the slide rail 2 when opening. The laser irradiation window 4 is used to allow the laser outside the cabin to pass through the window and irradiate the sample detection surface. The positioning plate 8 is used to position the two parts combined into the sample stage 9 in the middle according to the actual sample size. The sample stage screw slot 12 is used to screw in the screws to fix the positioned sample stage 9 to the turntable 7. The positioning plate screw slot 13 is used to screw in the screws to fix the positioning plate to the turntable 7 to prevent it from moving and indirectly causing the position of the sample stage 9 to change. The negative pressure cabin sealing plate 15 is movably installed at the bottom of the hatch cover. When the hatch cover moves to the top of the negative pressure cabin, the negative pressure cabin sealing plate is pneumatically controlled to move down and fall on the fixed sealing ring 16, so that the negative pressure cabin forms a relatively closed space. The fixed sealing ring 16 is fixed to the negative pressure hatch to cooperate with the negative pressure cabin sealing plate to achieve a sealing effect. The L-shaped fixing plate 17 is fixed to the rear side of the negative pressure cabin to support the second photoelectric sensor 10 so that it is fixed in parallel with the first photoelectric sensor 14 in space.

[0068] Preferably, the implementation steps of the system are as follows:

[0069] Step 1: Place the medicine column on the sample table 9 in the negative pressure cabin by means of automated tooling (robot), and place the coated medicine column horizontally on the V-shaped sample table. After the robot completely leaves the area of ​​the first photoelectric sensor 14 and the second photoelectric sensor 10, the system automatically closes and locks the hatch 1.

[0070] Step 2: Turn on the vacuum pump and / or the air release valve to adjust the relative vacuum degree in the negative pressure chamber, automatically enter the real-time phase mode, refresh and select the high-quality phase reference image, display the phase image under the current loading state in real time and automatically save the corresponding defect image.

[0071] Step 3: After the negative pressure chamber is evacuated to reach the set maximum air pressure value, the air intake valve lets air into the negative pressure chamber to slowly reduce the air pressure value. The reference image is refreshed after, for example, 5 seconds of air intake.

[0072] Step 4: After refreshing the reference image, automatically collect, process and save the defect image in this state at an interval of, for example, 6 seconds, and stop the air intake at the same time; process the defect image and save it in a pre-set folder.

[0073] Step 5: Automatically control the tray in the cabin to rotate 180° to the next detection surface.

[0074] Step 6: The air intake valve slowly lets air into the negative pressure cabin, and refreshes the reference image after, for example, 5 seconds of air intake.

[0075] Step 7: After refreshing the reference image, automatically collect, process and save the defect image in this state at an interval of, for example, 9 seconds, process the defect image and save it in a pre-set folder; intake air until the atmospheric pressure value.

[0076] Step 8: After the inspection is completed, the hatch door is automatically opened, the coated grain is taken out through the automated tooling, the previously stored phase images of different areas of the coated grain are retrieved, the images are numbered according to the naming rules, the coated grain inspection images are stitched together, and the inspection is completed; after completion, the hatch cover is opened, the robot takes out the coated grain, and waits for the next coated grain to enter the cabin for inspection; the inspection report is automatically generated, and the original inspection data is uploaded, thereby completing the inspection of one coated grain.

[0077] The defect image is a defect image obtained by taking the air pressure when refreshing the reference image as the standard and then inhaling to generate a first pressure difference (first predetermined value) during the automatic detection process, and has nothing to do with the absolute air pressure value. Preferably, the defect image can be automatically identified and the size can be automatically measured based on the YOLO neural network algorithm. Preferably, when the defect diameter of the first defect image or the second defect image is greater than a threshold, the defect is automatically determined to be unqualified.

[0078] The present invention can achieve one or more of the following technical effects:

[0079] 1. The turntable can drive the coated grain to rotate, and the single closing of the hatch cover can detect the two end surfaces of the coated grain, realizing automatic performance, reducing labor costs and human errors, and improving detection efficiency.

[0080] 2. The automatic detection manipulator leaves the negative pressure cabin, which can realize the automatic opening and closing of the cabin cover, automatic loading of samples, automatic triggering of measurements and automatic control of the turntable, so as to realize the mechanized batch and efficient detection of coated drug columns.

[0081] 3. Based on the YOLO neural network algorithm, the defect image can be automatically identified and the size can be automatically measured ( Figure 5), and automatically determine whether the product is qualified, reducing the impact of human judgment, and truly realizing fully automated detection of coated drug columns.

[0082] Those skilled in the art should understand that the present invention is not limited to the above embodiments, and the above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, and these changes and improvements fall within the scope of the present invention to be protected. The scope of the present invention to be protected is defined by the attached claims and their equivalents.

Claims

1. An automatic detection system for coated drug column defects, comprising: The negative pressure loading device comprises: a negative pressure cabin; a turntable arranged in the negative pressure cabin; and a sample table arranged on the turntable and provided with a groove for placing the coated medicine column; A manipulator, used for placing the coated medicine column to be tested into the groove and taking the coated medicine column out of the groove; a laser speckle detection device, used to detect defects at the first end and the second end of the coated drug column; and A control unit, used to control the negative pressure cabin, the turntable, the manipulator and the laser speckle detection device; Wherein, the turntable is used to drive the coated medicine column on the sample stage to rotate so as to align the first end and the second end to the laser speckle detection device respectively.

2. The automatic detection system for coated drug column defects according to claim 1 is characterized in that The negative pressure cabin comprises a cabin body and an openable and closable cabin cover, and a sealed space is formed when the cabin cover is closed on the cabin body.

3. The automatic detection system for coated drug column defects according to claim 1 or 2, characterized in that The hatch is provided with a first photoelectric sensor, and the cabin is provided with a second photoelectric sensor. The first photoelectric sensor and the second photoelectric sensor form a first detection unit for detecting whether the manipulator leaves the negative pressure cabin when the hatch is opened.

4. The automatic detection system for coated drug column defects according to claim 3 is characterized in that The control unit is also used to automatically control the hatch cover to close when the first detection unit detects that the manipulator leaves the negative pressure cabin.

5. The automatic detection system for coated drug column defects according to claim 4 is characterized in that It also includes a sample photoelectric sensor for detecting whether a coated drug column is placed in the sample stage.

6. The automatic detection system for coated drug column defects according to claim 5 is characterized in that The hatch cover and the cabin body are connected by slide rails.

7. The automatic detection system for coated drug column defects according to claim 4 is characterized in that The groove is a V-shaped groove.

8. The automatic detection system for coated drug column defects according to claim 4 is characterized in that The control unit is used to control the negative pressure cabin to exhaust air to reach a set maximum air pressure value. When the air inlet valve lets air into the negative pressure cabin to a first negative pressure value, the laser speckle detection device collects a reference image of the first end. When the air continues to be let into the negative pressure cabin to a second negative pressure value, the laser speckle detection device collects a detection image of the first end. A first defect image is acquired based on the reference image and the detection image of the first end, and the difference between the first negative pressure value and the second negative pressure value is a first predetermined value.

9. The automatic detection system for coated drug column defects according to claim 8 is characterized in that The control unit is used to control the laser speckle detection device to collect a reference image of the second end when the air intake valve continues to intake air into the negative pressure cabin to a third negative pressure value, and to collect a detection image of the second end when the air intake valve continues to intake air into the negative pressure cabin to a fourth negative pressure value, and to obtain a second defect image according to the reference image and the detection image of the second end, and the difference between the third negative pressure value and the fourth negative pressure value is the first predetermined value.

10. The automatic detection system for coated drug column defects according to claim 9 is characterized in that Based on the YOLO neural network algorithm, the defects in the defect image can be automatically identified and the size can be automatically measured. When the defect diameter of the first defect image or the second defect image is greater than the threshold, it will be automatically judged as unqualified.

Images