A drilling fluid identification system based on image recognition

Abstract

The present invention discloses a drilling fluid identification system based on image recognition, which relates to the field of drilling fluids. The first camera of this system acquires the first continuous frame image information and the second continuous frame image information of the drilling fluid; the second camera acquires the static image information of the drilling fluid; the first processing module obtains the color characteristics of the drilling fluid according to the static image information and processes the drilling fluid for the first time according to the color characteristics of the drilling fluid; the second processing module is used to obtain the bubble movement trajectory of the drilling fluid according to the first continuous frame image information of the drilling fluid and process the drilling fluid for the second time according to the bubble movement trajectory of the drilling fluid; the third processing module is used to obtain the particle sedimentation velocity according to the second continuous frame image information of the drilling fluid and process the drilling fluid for the third time according to the particle sedimentation velocity. This system penetrates particles of different particle sizes through a multi-band light source to obtain the color characteristics, suspended particle distribution and bubble dynamics of the drilling fluid, and identifies and processes the drilling fluid through multiple parameters to improve the accuracy of monitoring.

Description

Technical Field

[0001] The present invention relates to the field of drilling fluids, and particularly to a drilling fluid identification system based on image recognition. Background Art

[0002] The functions of drilling fluids in oil drilling are usually for lubrication, cooling, and carrying cuttings, etc. Traditional detection methods may rely on physical or chemical analysis, such as measuring density, viscosity, sand content, etc. These methods may require off-line laboratory analysis, which is time-consuming and cannot monitor in real time.

[0003] Traditional methods require a long time for chemical analysis and processing to obtain the real-time situation of drilling fluids, and cannot be adjusted in time. At the same time, existing drilling fluid image recognition systems usually can only detect a single parameter during monitoring, cannot grasp the situation of drilling fluids in time, and cannot be adjusted in time.

[0004] In addition, when detecting drilling fluids, the camera cannot be cleaned in time, and the captured image data is inaccurate, which will affect the accuracy of the monitored experimental data. Summary of the Invention

[0005] The purpose of the present invention is to provide a drilling fluid identification system based on image recognition, which obtains multiple drilling fluid detection parameters through image information acquired under different light sources, and adjusts and optimizes the drilling fluid to improve the accuracy of monitoring.

[0006] This purpose is achieved by the following technical solutions:

[0007] A drilling fluid identification system based on image recognition includes an annular multi-spectral light source structure, a first camera, a second camera, a first processing module, a second processing module, and a third processing module;

[0008] Among them, the annular multi-spectral light source structure includes several visible light sources, several near-infrared light sources, and several ultraviolet band light sources; the first camera is used to obtain the first continuous frame image information of the drilling fluid under visible light sources and the second continuous frame image information of the drilling fluid under near-infrared light sources; the second camera is used to obtain the static image information of the drilling fluid under ultraviolet band light sources;

[0009] During use, first, the second camera obtains the static image information of the drilling fluid under ultraviolet band light sources, and the first processing module obtains the color characteristics of the drilling fluid according to the static image information and processes the drilling fluid for the first time according to the color characteristics of the drilling fluid;

[0010] After the drilling fluid is processed for the first time, the first camera obtains first continuous frame image information of the drilling fluid under a visible light source, the second processing module obtains the bubble movement trajectory of the drilling fluid according to the first continuous frame image information of the drilling fluid, and processes the drilling fluid for the second time according to the bubble movement trajectory of the drilling fluid;

[0011] After the drilling fluid is processed for the second time, the first camera obtains second continuous frame image information of the drilling fluid under a near-infrared light source, the third processing module obtains the particle settling velocity based on the second continuous frame image information of the drilling fluid, and processes the drilling fluid for the third time based on the particle settling velocity.

[0012] When in use, the above operations are repeated continuously to monitor the drilling fluid in real time. Among them, the color change of the drilling fluid may reflect the contamination or deterioration of the drilling fluid. When the drilling fluid encounters formation fluid or pollutants, the color may change. For example, oil pollution will make the color darker, and hydrogen sulfide may cause it to turn black. At this time, the color characteristics can be used as a monitoring indicator to detect and deal with problems in time.

[0013] The presence of bubbles will reduce the density of drilling fluid, causing an imbalance in wellbore pressure, which may cause a kick or blowout. The return velocity of the drilling fluid must be greater than the settling velocity of the cuttings to effectively remove the cuttings from the wellbore. If the settling velocity is too high, the cuttings will settle at the bottom of the well, causing the drill to get stuck or the wellbore to get blocked.

[0014] Therefore, the inventors monitor the drilling fluid from multiple aspects by capturing the color characteristics, bubble dynamics and particle settling velocity to improve the monitoring accuracy.

[0015] Secondly, the color characteristics of drilling fluid, bubble dynamics, and particle settling speed are also mutually reinforcing. The color of drilling fluid usually reflects its composition and density. When the color of drilling fluid is dark or brown-black speckled oil spots appear (such as crude oil mixing), it may be accompanied by abnormal bubble movement trajectory (such as continuous and concentrated, not easy to break). In such cases, the bubble problem is usually treated by degassing devices or defoaming agents (such as silicones). Turbid or grayish-white color may be caused by excessive content of solid particles (such as bentonite, cuttings). At this time, the particle settling speed is accelerated, and solid control equipment (such as vibrating screens, centrifuges) need to be optimized to reduce the solid content; and uneven color will lead to increased filtration loss, indirectly affecting the stability of liquid column pressure, exacerbating bubble expansion and particle settling.

[0016] When bubbles slip and rise in the drilling fluid, the local density will be reduced, resulting in fluctuations in the pressure of the liquid column, which may cause well kicks or well losses. At this time, the density needs to be stabilized by using a surface degassing device or adjusting the amount of defoaming agent (such as controlling the concentration to 0.05%~0.2%).

[0017] When bubbles gather to form slug flow, it will hinder the sedimentation of particles. It is necessary to improve the flow pattern by increasing the plastic viscosity of the drilling fluid (such as adding a thickener) and adjust the shear force to balance the rock carrying and sedimentation efficiency.

[0018] There is also a synergistic effect between the bubbles and the particles. Tiny bubbles (such as nanoscale) can adsorb on the particle surface to form a "gas-solid complex", reducing the effective density of the particles and delaying the settling velocity.

[0019] Therefore, the present invention first monitors the color characteristics of the drilling fluid, adjusts the drilling fluid according to the color characteristics, then monitors the bubbles, and finally monitors whether the adjustment is accurate through the particle settling velocity. Meanwhile, during the use process, this system continuously cycles to identify the color characteristics of the drilling fluid and perform the first treatment on the drilling fluid, identify the bubble dynamics of the drilling fluid and perform the second treatment on the drilling fluid, identify the particle settling velocity of the drilling fluid and perform the third treatment on the drilling fluid. By repeating this cycle, the state of the drilling fluid can be obtained from different parameter perspectives and adjusted in a timely manner. Not only is the processing speed fast, but the adjustment accuracy is also high.

[0020] In the specific processing process, the first processing module includes a first preprocessing module and a color processing module;

[0021] The first preprocessing module uses Gaussian filtering or median filtering to remove the image noise of the image information, enhances the color contrast through histogram, and converts the original RGB image into the HSV or Lab color space;

[0022] The color processing module is used to divide the color intervals in the HSV or Lab space, statistically analyze the pixel distribution in each interval, generate a color histogram, obtain the main components and distribution characteristics of the drilling fluid color through the color histogram, and perform the first treatment on the drilling fluid according to the main components and distribution characteristics of the drilling fluid color.

[0023] Among them, the color distribution of the drilling fluid is quantified through the HSV or RGB color space to identify the main color tone (such as dark brown, gray-green, etc.), which reflects the solid content, pollutant type (such as rust, clay) or chemical additive concentration. For example: if the main color is reddish, it may be contaminated with rust; if the main color is turbid and dark, there may be too many solid particles.

[0024] If the solid content is too high (the color is deep and the distribution is concentrated), solid control equipment such as a centrifuge and a vibrating screen is used to reduce the solid content, and a diluent (such as clear water or low-solid drilling fluid) is added to adjust the density of the drilling fluid.

[0025] If the color deviates from the reference, when it is a red tone, a rust remover (such as citric acid) is added to treat the iron ion pollution; when it is black / gray, check the mixing situation of the cuttings and enhance the efficiency of the solid control equipment.

[0026] In the distribution characteristics of color, hue: reflects the type of drilling fluid base fluid (e.g., oil-based is dark, water-based is lighter); saturation: high saturation may indicate too high concentration of chemical additives or aggregation of pollutants; value: low value may indicate small particle size and high dispersion of solid-phase particles.

[0027] The second processing module includes a second preprocessing module, a first marking module, and a bubble processing module.

[0028] The second preprocessing module uses Gaussian filtering to eliminate the image noise of the first continuous frame image information, enhances the bubble contrast through histogram equalization, and distinguishes bubbles according to the gray level difference.

[0029] The first marking module is used to mark the centroid coordinates of each bubble respectively, and match the same target of each bubble in adjacent frames based on the Hungarian algorithm.

[0030] The bubble processing module obtains the acceleration of the bubbles according to the centroid displacement of the bubbles with the same target in adjacent frames, and statistically analyzes the average diameter and aggregation state of the bubbles according to the image analysis algorithm, and secondarily processes the drilling fluid according to the acceleration, average diameter, and aggregation state.

[0031] The third processing module includes a third preprocessing module, a second marking module, and a particle processing module.

[0032] The third preprocessing module uses Gaussian filtering to eliminate the image noise of the second continuous frame image information, enhances the particle contrast through histogram equalization, and distinguishes particles according to the gray level difference.

[0033] The second marking module is used to mark the centroid coordinates of each particle respectively, and match the same target of each particle in adjacent frames based on the Hungarian algorithm.

[0034] The particle processing module obtains the particle sedimentation velocity according to the centroid displacement of the particles with the same target in adjacent frames, and thirdly processes the drilling fluid according to the particle sedimentation velocity.

[0035] The regulation of the particle sedimentation velocity needs to comprehensively consider the density, viscosity, shear force, and solid-phase characteristics of the drilling fluid. Therefore, in this system, the particle sedimentation velocity is used as the last detection parameter, and the main purpose is to assist in detecting the effects of the first and second treatments of the drilling fluid. After the first and second treatments, basically, the particle sedimentation velocity in the second continuous frame image information is accurate. If the particle sedimentation velocity in the second continuous frame image information is not within the preset range, the drilling fluid is subjected to the third treatment.

[0036] After the third treatment of the drilling fluid, the static image information of the drilling fluid is obtained again by the second camera under the ultraviolet band light source, and the above operations are repeated for real-time dynamic adjustment to timely feedback the real-time situation of the drilling fluid.

[0037] After the above operation is repeated for the second time, if the particle sedimentation speed in the second consecutive frame image information of the second time is not within the preset range, an alarm is triggered and manual inspection and processing are carried out.

[0038] On the other hand, when the system identifies, the first camera and the second camera need to operate under different light sources. During the long-term monitoring process, the lenses of the first camera and the second camera need to be cleaned regularly, and the light sources need to be switched and adjusted. Therefore, the inventor of the present invention proposes an annular multi-spectral light source structure for the system.

[0039] The annular multi-spectral light source structure includes a first connecting ring and a circular plate arranged concentrically. The upper end surface of the first connecting ring is concentrically connected with the first camera, and the lens of the first camera is located inside the inner ring of the first connecting ring; the circular plate is located between the first connecting ring and the first camera. The circular plate includes a number of third connecting pieces that can form a circle. A first cleaning layer is provided on the surface of the third connecting piece that contacts the lens of the first camera. The second camera is arranged on the surface opposite to the first cleaning layer of the third connecting piece; a working piece is arranged between the third connecting piece and the first connecting ring, and the working piece drives the third connecting piece to rotate to close or open the inner ring of the first connecting ring; a second cleaning layer for cleaning the second camera is provided on the upper end surface of the first connecting ring.

[0040] When the third connecting piece closes the inner ring of the first connecting ring, the first cleaning layer of the third connecting piece contacts the lens of the first camera, and a number of third connecting pieces form a circle; the first cleaning layer on the upper end surface of the third connecting piece is used to clean the lens of the first camera. At the same time, when the third connecting piece rotates to close the inner ring of the first connecting ring, the second cleaning layer on the upper end surface of the first connecting ring cleans the second camera.

[0041] When the third connecting piece opens the inner ring of the first connecting ring, the third connecting piece exposes the lens of the first camera, facilitating the first camera's lens to obtain the first consecutive frame image information or the second consecutive frame image information. At the same time, the second camera is located outside the first connecting ring and is also exposed, facilitating the acquisition of the static image information of the drilling fluid.

[0042] Therefore, during the process of the third connecting piece closing the inner ring of the first connecting ring, the first camera and the second camera can be cleaned simultaneously. At the same time, after the third connecting piece opens the inner ring of the first connecting ring, both the first camera and the second camera can take pictures.

[0043] In addition, the inventor of the actuating member preferably selects a structure. The actuating member includes a second connecting ring sleeved outside the first connecting ring. The second connecting ring can rotate circumferentially on the first connecting ring. An actuating rod is hinged between the second connecting ring and the third connecting member. One end of the third connecting member is hinged to the first connecting ring. The rotation of the second connecting ring drives the third connecting member to close or open the inner ring of the first connecting ring through the actuating rod. A second cleaning layer for cleaning the second camera is also provided on the upper end surface of the second connecting ring, improving the cleaning efficiency.

[0044] During use, rotate the second connecting ring. The second connecting ring drives one end of the actuating rod to move. The other end of the actuating rod acts on the third connecting member, causing the third connecting member to rotate around its connection point with the first connecting ring, realizing the opening or closing of the inner ring of the first connecting ring by the third connecting member.

[0045] On the other hand, the ultraviolet band light source is arranged on the lower end surface of the first connecting ring. The visible light source is arranged on the lower end surface of the second connecting ring. The near-infrared light source is arranged on the lower end of the third connecting member.

[0046] Ultraviolet light needs to irradiate from the side to excite the fluorescence of the drilling fluid, and at the same time avoid direct irradiation on the camera to prevent overexposure. Therefore, the ultraviolet band light source is arranged on the lower end surface of the first connecting ring. The ultraviolet band light source irradiates from the side of the drilling fluid at an inclined angle of 30° - 45°. When acquiring images, it is at a relatively long distance from the second camera, which is convenient for adjusting the angle of the ultraviolet band light source.

[0047] At the same time, the near-infrared light source is arranged on the lower end of the third connecting member. The near-infrared light source is at a relatively long distance from the first camera when acquiring images. Therefore, it is possible to invert the particle size distribution through lateral light sources and multi-angle scattered data.

[0048] The visible light sources are annularly distributed on the lower end surface of the second connecting ring for uniform illumination, ensuring that the dynamics of the bubbles are clearly recorded and reducing shadow interference.

[0049] Therefore, during the process of providing different light sources for the first camera and the second camera, the annular multi-spectral light source structure of this system can not only perform shooting but also clean the first camera and the second camera.

[0050] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0051] The present invention relates to a drilling fluid identification system based on image recognition. The system uses a multi-band light source to penetrate particles of different particle sizes, capturing the color characteristics (ultraviolet-excited fluorescence), suspended particle distribution (near-infrared scattering), and bubble dynamics (visible light imaging) of the drilling fluid. By identifying and processing the drilling fluid through multiple parameters, the accuracy of monitoring is improved. Moreover, when the system performs identification, it first identifies the color characteristics of the drilling fluid. After the first processing, based on the result of the first processing, it then identifies the bubbles in the drilling fluid and performs the second processing. In practical applications, generally after two processes, the expected state can be achieved. Therefore, the system further identifies through another parameter, suspended particles. If the expectation is met, the above data is cleared and the identification is repeated. If the expectation is not met, the third processing is performed. After the processing, the identification is repeated without clearing the data. When the loop is repeated twice and the third processing still does not meet the expectation, an alarm is triggered and manual processing is required, which can further improve the accuracy of identification and process the drilling fluid in all directions and quickly.

[0052] In addition, the annular multi-spectral light source structure of the system matches multiple light sources. When in use, it can not only quickly adjust the light source, but also quickly and conveniently clean the lenses of the first camera and the second camera, improving the accuracy of identification, thereby enhancing the accuracy of monitoring and being more conducive to long-term use. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] The drawings described herein are used to provide a further understanding of the embodiments of the present invention, form a part of this application, and do not limit the embodiments of the present invention. In the drawings:

[0054] Figure 1 is a schematic structural diagram of the system;

[0055] Figure 2 is a schematic diagram of the positional relationship between the first camera and the first connecting ring;

[0056] Figure 3 is a schematic diagram of the circular structure formed by the third connecting member when it closes the inner ring of the first connecting ring;

[0057] Figure 4 is a schematic diagram of the upper end of the first connecting ring when the third connecting member opens the inner ring of the first connecting ring;

[0058] Figure 5 is a schematic diagram of the lower end of the first connecting ring when the third connecting member opens the inner ring of the first connecting ring.

[0059] Marks in the drawings and corresponding component names:

[0060] 1 - First camera, 2 - Second camera, 3 - First connecting ring, 4 - Second connecting ring, 5 - Third connecting member, 6 - Acting rod, 7 - Near-infrared light source, 8 - Ultraviolet band light source, 9 - Visible light source, 10 - Connecting rod. Detailed implementation mode

[0061] To make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below in combination with embodiments and the accompanying drawings. The illustrative embodiments of the present invention and their descriptions are only used to explain the present invention and are not intended to limit the present invention.

[0062] In the description of the present invention, it should be understood that the orientation or positional relationships indicated by the terms "front", "rear", "left", "right", "upper", "lower", "vertical", "horizontal", "high", "low", "inner", "outer", etc. are based on the orientation or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as limiting the protection scope of the present invention.

[0063] Embodiment 1

[0064] As Figure 1 shown, the system includes a ring-shaped multi-spectral light source structure, a first camera 1, a second camera 2, a first processing module, a second processing module, and a third processing module;

[0065] Among them, the ring-shaped multi-spectral light source structure includes a plurality of visible light sources 9, a plurality of near-infrared light sources 7, and a plurality of ultraviolet band light sources 8; the first camera 1 is used to obtain the first continuous frame image information of the drilling fluid under the visible light source 9 and obtain the second continuous frame image information of the drilling fluid under the near-infrared light source 7; the second camera 2 is used to obtain the static image information of the drilling fluid under the ultraviolet band light source 8; the first processing module is used to obtain the color characteristics of the drilling fluid according to the static image information and process the drilling fluid for the first time according to the color characteristics of the drilling fluid; the second processing module is used to obtain the bubble movement trajectory of the drilling fluid according to the first continuous frame image information of the drilling fluid and process the drilling fluid for the second time according to the bubble movement trajectory of the drilling fluid; the third processing module is used to obtain the particle sedimentation velocity according to the second continuous frame image information of the drilling fluid and process the drilling fluid for the third time according to the particle sedimentation velocity.

[0066] The acquisition of static image information by the second camera 2, the acquisition of the first continuous frame image information by the first camera 1, and the acquisition of the second continuous frame image information by the first camera 1 form a cycle. And within one cycle, the time for the second camera 2 to acquire static image information, the first camera 1 to acquire the first continuous frame image information, and the first camera 1 to acquire the second continuous frame image information increases in sequence. That is, within one cycle, first, the second camera 2 acquires static image information, and the first processing module obtains the color characteristics of the drilling fluid based on the static image information. After the drilling fluid is processed for the first time according to the color characteristics of the drilling fluid, the first camera 1 acquires the first continuous frame image information of the drilling fluid. The second processing module obtains the bubble movement trajectory of the drilling fluid based on the first continuous frame image information of the drilling fluid, and processes the drilling fluid for the second time according to the bubble movement trajectory of the drilling fluid. Finally, the first camera 1 acquires the second continuous frame image information of the drilling fluid, and the third processing module obtains the particle sedimentation velocity based on the second continuous frame image information of the drilling fluid, and processes the drilling fluid for the third time according to the particle sedimentation velocity.

[0067] When in use, it includes the following steps:

[0068] Step 1, the ultraviolet band light source 8 of the annular multi-spectral light source structure is turned on, and the second camera 2 acquires the static image information of the drilling fluid under the ultraviolet band light source 8. The first processing module processes the drilling fluid for the first time according to the color characteristics of the drilling fluid;

[0069] Step 2, after the first processing, the ultraviolet band light source 8 of the annular multi-spectral light source structure is turned off, and the visible light source 9 is turned on. The first camera 1 acquires the first continuous frame image information of the drilling fluid under the visible light source 9, and processes the drilling fluid for the second time according to the bubble movement trajectory of the drilling fluid;

[0070] Step 3, after the second processing is completed, the visible light source 9 of the annular multi-spectral light source structure is turned off, and the near-infrared light source 7 is turned on. The first camera acquires the second continuous frame image information of the drilling fluid under the near-infrared light source 7;

[0071] Steps 1 - 3 form a cycle. Within one cycle, the third processing module obtains the particle sedimentation velocity based on the second continuous frame image information of the drilling fluid. If the particle sedimentation velocity is within the preset range, then within this cycle, the color characteristics, bubble movement trajectory, and particle sedimentation velocity information obtained by the first processing module, the second processing module, and the third processing module are cleared, and the next cycle is entered. That is, if the particle sedimentation velocity is within the preset range, then the data of steps 1, 2, and 3 are cleared, the next cycle is entered, and steps 1 - 3 are repeated for cyclic real-time monitoring. During this process, steps 1 and 2 have already processed the drilling fluid well and no further processing is required. Therefore, the data of steps 1, 2, and 3 of this cycle are cleared, and the next cycle is entered.

[0072] In some embodiments, if within one cycle period, the third processing module obtains the particle settling velocity based on the second consecutive frame image information of the drilling fluid, and if the particle settling velocity is not within the preset range, it indicates that steps 1 and 2 do not properly process the drilling fluid. Therefore, the data of this cycle period is retained.

[0073] In the next cycle period, the first processing module first processes the drilling fluid according to the color characteristics, bubble movement trajectories, and particle settling velocity obtained in the previous cycle period and the color characteristics obtained in this cycle period; that is, within the first cycle period, in step 3, if the particle settling velocity is not within the preset range, then it enters the second cycle period, and steps 1 - 3 are repeated for the second time. When repeating step 1 for the second time, the first processing module first processes the drilling fluid according to the color characteristics, bubble movement trajectories, and particle settling velocity obtained in the first cycle period and the color characteristics obtained in the second cycle period; when repeating steps 2 and 3 for the second time, the steps are the same as those in the first cycle period. When the particle settling velocity when repeating step 3 for the second time is still not within the preset range, an alarm is triggered and manual identification and monitoring are carried out.

[0074] When the particle settling velocity when repeating step 3 for the second time is within the preset range, the data of steps 1 - 3 in the first cycle period and the second cycle period is cleared.

[0075] Steps 1 - 3 are repeated again for cyclic real - time monitoring.

[0076] Embodiment 2

[0077] Based on the above - mentioned embodiment, in step 1, the first processing module includes a first pre - processing module and a color processing module; step 1 also includes the following steps:

[0078] Step 11, the first pre - processing module uses Gaussian filtering or median filtering to remove the image noise of the image information, enhances the color contrast through histogram equalization, and converts the original RGB image into the HSV or Lab color space;

[0079] Step 12, the color processing module is used to divide color intervals in the HSV or Lab space, count the pixel distribution in each interval, generate a color histogram, and obtain the main components and distribution characteristics of the drilling fluid color through the color histogram;

[0080] Step 13, the drilling fluid is first processed according to the main components and distribution characteristics of the drilling fluid color. If the color is deep and the distribution is concentrated, indicating that the solid content is too high, then solid control equipment such as a centrifuge and a shale shaker is used to reduce the solid content, and diluents such as fresh water or low - solid drilling fluid are added to adjust the density of the drilling fluid; if the color deviates from the reference, when it is a red tone, a rust remover such as citric acid is added to treat iron ion pollution; when it is black / gray, the mixing situation of cuttings is checked and the efficiency of the solid control equipment is enhanced.

[0081] In some embodiments, in step 2, the second processing module includes a second preprocessing module, a first marking module, and a bubble processing module; step 2 further includes the following steps:

[0082] Step 21, the second preprocessing module uses Gaussian filtering to eliminate the image noise of the first consecutive frame image information, enhances the bubble contrast through histogram equalization, and distinguishes bubbles according to the gray-scale difference;

[0083] Step 22, the first marking module is used to mark the centroid coordinates of each bubble respectively, and match the same target of each bubble in adjacent frames based on the Hungarian algorithm;

[0084] Step 23, the bubble processing module obtains the acceleration of the bubbles according to the centroid displacement of the bubbles with the same target in adjacent frames, and statistically analyzes the average diameter and aggregation state of the bubbles according to the image analysis algorithm, and secondarily processes the drilling fluid according to the acceleration, average diameter, and aggregation state. When the bubble acceleration is too high (such as due to excessive pump displacement resulting in intensified turbulence), it is necessary to improve the shear thinning property of the drilling fluid. High molecular polymers (such as polyacrylamide) can be added to enhance the high-viscosity characteristics of the annulus fluid to reduce the flow resistance, while maintaining a low-viscosity state at the drill bit nozzle.

[0085] When the bubble acceleration is abnormal, it may reflect the mismatch between the flow rate and the wellbore size. The pump displacement can be reduced or the nozzle diameter can be adjusted to reduce the entrainment of turbulence on the bubbles, and large bubbles can be preferentially processed through a liquid-gas separator (such as a vacuum degasser).

[0086] Small-diameter bubbles (<1 / 16 inch): Use silicone-based defoamers to reduce the liquid film stability, and combine with a vacuum degasser to enhance the buoyancy efficiency of small bubbles.

[0087] Large-diameter bubbles (1 / 9 - 1 inch): Rapidly separate through a liquid-gas separator, and control the concentration of the foaming agent (such as reducing the dosage of the surfactant) to avoid excessive generation of large bubbles.

[0088] In the aggregation state, if it is high-density foam, increase the defoamer concentration or switch to a high-activity type (such as a mineral oil defoamer containing hydrophobic particles); if it is low-density intermittent foam: use a slow-release defoamer (such as polyethers), and reduce the addition frequency by extending the foam suppression time.

[0089] In some embodiments, in step 3, the third processing module includes a third preprocessing module, a second marking module, and a particle processing module, and step 3 further includes the following steps:

[0090] Step 31, the third preprocessing module uses Gaussian filtering to eliminate the image noise of the second consecutive frame image information, enhances the particle contrast through histogram equalization, and distinguishes particles according to the gray-scale difference;

[0091] Step 32: The second marking module is used to mark the centroid coordinates of each particle respectively, and match the same target of each particle in adjacent frames based on the Hungarian algorithm;

[0092] Step 33: The particle processing module calculates the centroid displacement of the particles of the same target in adjacent frames to obtain the particle settling velocity, and processes the drilling fluid for the third time according to the particle settling velocity. Specifically, by adding a water-soluble polymer (such as polyacrylamide PAM) or increasing the solid-phase dispersion degree, the particle settling velocity can be slowed down. For example, when the concentration of the PAM solution increases by 0.1%, the settling velocity decreases by an average of 0.05 mm. Harmful solids are removed by using equipment such as vibrating screens and hydrocyclone desanders, or fresh water is added for dilution to reduce the driving force for particle settlement.

[0093] Example 3

[0094] On the basis of the above embodiment, the annular multi-spectral light source structure includes a first connecting ring 3 and a circular plate arranged concentrically. The upper end surface of the first connecting ring 3 is concentrically connected to the first camera 1 through a connecting rod 10, as Figure 2 shown, the lens of the first camera 1 is located inside the inner ring of the first connecting ring 3;

[0095] The circular plate is located between the first connecting ring 3 and the first camera 1. The circular plate includes a number of third connecting members 5 that can form a circle. A first cleaning layer is provided on the surface of the third connecting member 5 that contacts the lens of the first camera 1, and the second camera 2 is arranged on the surface opposite to the first cleaning layer of the third connecting member 5; The acting member includes a second connecting ring 4 sleeved outside the first connecting ring 3. The second connecting ring 4 can rotate circumferentially on the first connecting ring 3. An acting rod 6 is hinged between the second connecting ring 4 and the third connecting member 5, and one end of the third connecting member 5 is hinged to the first connecting ring 3; The rotation of the second connecting ring 4 drives the third connecting member 5 to close or open the inner ring of the first connecting ring 3 through the acting rod 6. Second cleaning layers for cleaning the second camera 2 are provided on the upper end surfaces of the first connecting ring 3 and the second connecting ring 4.

[0096] When the third connecting member 5 closes the inner ring of the first connecting ring 3, as Figure 3 shown, a number of third connecting members 5 form a circle, and moreover, the first cleaning layer of the third connecting member 5 is in contact with the lens of the first camera 1, the second camera on the lower end surface of the third connecting member 5 is in contact with the second cleaning layers on the upper end surfaces of the first connecting ring 3 and the second connecting ring 4, and the first cleaning layer and the second cleaning layer respectively complete the cleaning of the first camera and the second camera.

[0097] When the third connecting member 5 opens the inner ring of the first connecting ring 3, as Figure 4As shown, the second connecting ring 4 rotates circumferentially. During the process of the second connecting ring 4, one end of the actuating rod 6 is driven to move, and the other end of the actuating rod 6 drives the third connecting member 5 to rotate. When the third connecting member 5 opens the inner ring of the first connecting ring 3, as Figure 5 shown, for the third connecting member 5, the ultraviolet band light source 8 is arranged on the lower end surface of the first connecting ring 3. The visible light source 9 is arranged on the lower end surface of the second connecting ring 4. The near-infrared light source 7 is arranged on the lower end of the third connecting member 5. When the third connecting member 5 opens the inner ring of the first connecting ring 3, the second camera 2 on the lower end surface of the third connecting member 5 is exposed, and the lens of the first camera 1 is exposed at the inner ring of the first connecting ring 3.

[0098] During use, different light sources are provided by the near-infrared light source 7, the ultraviolet band light source 8, and the visible light source 9, and the lenses of the first camera 1 and the second camera 2 are quickly and conveniently cleaned by rotating the second connecting ring 4, improving the use efficiency.

[0099] The "first", "second", "third", etc. used in this article are only used to distinguish the corresponding components for the sake of clear description, and are not intended to limit any order or emphasize importance, etc. In addition, the term "connected" used in this article, without special explanation, can be directly connected or indirectly connected through other components.

[0100] The specific embodiments described above have further elaborated on the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only the specific embodiments of the present invention and is not used to limit the protection scope of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A drilling fluid identification system based on image recognition, characterized in that: It comprises an annular multi-spectral light source structure, a first camera (1), a second camera (2), a first processing module, a second processing module and a third processing module; The annular multi-spectral light source structure comprises a plurality of visible light sources (9), a plurality of near-infrared light sources (7) and a plurality of ultraviolet light sources (8); The first camera (1) is used to obtain first continuous frame image information of the drilling fluid under a visible light source (9), and to obtain second continuous frame image information of the drilling fluid under a near-infrared light source (7); The second camera (2) is used to obtain static image information of the drilling fluid under the ultraviolet light source (8); The first processing module is used to obtain the color characteristics of the drilling fluid according to the static image information, and process the drilling fluid for the first time according to the color characteristics of the drilling fluid; The second processing module is used to obtain the bubble movement trajectory of the drilling fluid according to the first continuous frame image information of the drilling fluid after processing the drilling fluid for the first time, and process the drilling fluid for the second time according to the bubble movement trajectory of the drilling fluid; The third processing module is used for obtaining the particle settling velocity according to the second continuous frame image information of the drilling fluid after processing the drilling fluid for the second time, and processing the drilling fluid for the third time according to the particle settling velocity; The annular multi-spectral light source structure comprises a first connecting ring (3) and a circular plate which are arranged concentrically, wherein the upper end surface of the first connecting ring (3) is concentrically connected to the first camera (1), and the lens of the first camera (1) is located in the inner ring of the first connecting ring (3); the circular plate is located between the first connecting ring (3) and the first camera (1), and the circular plate comprises a plurality of third connecting members (5) which can form a circle, a first cleaning layer being arranged on the surface where the third connecting member (5) and the lens of the first camera (1) contact each other, and the second camera (2) is arranged on the surface of the third connecting member (5) opposite to the first cleaning layer; An operating member is provided between the third connecting member (5) and the first connecting ring (3), and the operating member drives the third connecting member (5) to rotate to close or open the inner ring of the first connecting ring (3); When the third connecting member (5) closes the inner ring of the first connecting ring (3), the first cleaning layer of the third connecting member (5) and the lens of the first camera (1) are in contact with each other, and a plurality of the third connecting members (5) form a circle; When the third connecting member (5) opens the inner ring of the first connecting ring (3), the lens of the first camera (1) is used to obtain first continuous frame image information or second continuous frame image information; the action member comprises a second connecting ring (4) sleeved outside the first connecting ring (3), the second connecting ring (4) being able to rotate circumferentially on the first connecting ring (3), an action rod (6) being hingedly connected between the second connecting ring (4) and the third connecting member (5), and one end of the third connecting member (5) being hingedly connected to the first connecting ring (3); the second connecting ring (4) rotates to drive the third connecting member (5) to close or open the inner ring of the first connecting ring (3) through the action rod (6); a second cleaning layer for cleaning the second camera (2) is provided on the upper end surfaces of the first connecting ring (3) and the second connecting ring (4); The first preprocessing module uses Gaussian filtering or median filtering to remove image noise from image information, enhances color contrast through histogram, and converts the original RGB image into HSV or Lab color space; The color processing module is used to divide the color intervals in HSV or Lab space, count the pixel distribution in each interval, generate a color histogram, and obtain the main components and distribution characteristics of the drilling fluid color through the color histogram; The drilling fluid is first treated according to the main components and distribution characteristics of the drilling fluid color. If the color is dark and the distribution is concentrated, the solid content is reduced and a diluent is added to adjust the density of the drilling fluid; If the color deviates from the reference, when it is red, add rust remover to treat iron ion pollution; when it is black / gray, check whether drill cuttings are mixed in; The second preprocessing module uses Gaussian filtering to eliminate image noise of the first continuous frame image information, enhances bubble contrast through histogram equalization, and distinguishes bubbles based on grayscale differences; The first marking module is used to mark the centroid coordinates of each bubble respectively, and match the same target of each bubble in adjacent frames based on the Hungarian algorithm; The bubble processing module obtains the acceleration of the bubble according to the displacement of the center of mass of the bubble of the same target in adjacent frames, and calculates the average diameter and aggregation state of the bubble according to the image analysis algorithm, and processes the drilling fluid for the second time according to the acceleration, average diameter and aggregation state; The third preprocessing module uses Gaussian filtering to eliminate image noise of the second continuous frame image information, enhances particle contrast through histogram equalization, and distinguishes particles according to grayscale differences; The second marking module is used to mark the centroid coordinates of each particle respectively, and match the same target of each particle in adjacent frames based on the Hungarian algorithm; The particle processing module calculates the centroid displacement of particles of the same target in adjacent frames, obtains the particle settling velocity, and processes the drilling fluid for the third time according to the particle settling velocity.

2. A drilling fluid identification system based on image recognition according to claim 1, characterized in that: The second camera (2) acquires static image information, the first camera (1) acquires first continuous frame image information, and the first camera (1) acquires second continuous frame image information, which constitutes a cycle period. In a cycle period, the time for the second camera (2) to acquire static image information, the time for the first camera (1) to acquire first continuous frame image information, and the time for the first camera (1) to acquire second continuous frame image information increases in sequence.

3. A drilling fluid identification system based on image recognition according to claim 2, characterized in that: Within a cycle, the third processing module obtains the particle settling velocity based on the second continuous frame image information of the drilling fluid. If the particle settling velocity is within a preset range, then within the cycle, the color features, bubble movement trajectories and particle settling velocity information obtained by the first processing module, the second processing module and the third processing module are cleared and enter the next cycle.

4. The drilling fluid identification system based on image recognition according to claim 2, characterized in that: In one cycle, the third processing module obtains the particle settling velocity according to the second continuous frame image information of the drilling fluid. If the particle settling velocity is within the preset range, the next cycle is entered. In the next cycle, the first processing module processes the drilling fluid for the first time according to the color features, bubble motion trajectory and particle settling velocity obtained in the previous cycle and the color features obtained in the current cycle; In the next cycle, the third processing module obtains the particle settling velocity according to the second continuous frame image information of the drilling fluid, and triggers an alarm if the particle settling velocity is within an unpredictable range.

5. The drilling fluid identification system based on image recognition according to claim 1, characterized in that: The ultraviolet band light source (8) is arranged on the lower end surface of the first connecting ring (3).

6. The drilling fluid identification system based on image recognition according to claim 1, characterized in that: The visible light source (9) is arranged on the lower end surface of the second connecting ring (4).

7. The drilling fluid identification system based on image recognition according to claim 1, characterized in that: The near-infrared light source (7) is arranged on the lower end of the third connecting member (5).

Images