Image sensor for online detection of oil quality

Abstract

The invention provides an image sensor for online detection of oil quality, and the image sensor comprises a sensor body which is provided with an oil inlet, an oil outlet and a detection chamber, the detection chamber is arranged between the oil inlet and the oil outlet, and the detection chamber comprises a cavity and a transparent protection plate arranged at the top of the cavity; the magnetic adsorption assembly comprises an annular electromagnet arranged above the transparent protection plate; the image acquisition assembly is arranged on the upper surface of the transparent protection plate and comprises an imaging lens, a CMOS sensor and an annular light source; the processing circuit is used for preprocessing and analyzing the acquired optical image; and the control circuit is used for controlling the coil electrifying preset time so as to adsorb the abrasive particles and control image acquisition. According to the invention, real-time and accurate detection of the abrasive particles in the oil can be realized, and the morphology feature information of the abrasive particles can be obtained.

Description

Technical Field

[0001] The present application relates to the technical field of oil detection, and in particular to an image sensor for online detection of oil quality. Background Art

[0002] In the field of oil monitoring for industrial equipment, the state of abrasive particles in the oil is an important indicator for reflecting the operating status of the equipment and predicting equipment failure. At present, the mainstream oil abrasive particle detection method mainly detects the number and size distribution of abrasive particles in the oil through high-precision abrasive particle detection devices and back-end microprocessing circuit systems. Although this method has a certain detection accuracy, due to the limitations of its detection principle, the above-mentioned traditional detection method can only obtain the number and size information of abrasive particles, and cannot obtain the morphological characteristics of abrasive particles, and the morphological characteristics of abrasive particles are of great significance for judging the wear type and degree of wear of equipment.

[0003] Although the offline ferrography analysis method can obtain the morphological characteristics of wear particles, it requires manual observation using a microscope, which not only has low analysis efficiency, but also is easily affected by human factors. As industrial equipment develops towards intelligence and automation, the traditional offline ferrography analysis method can no longer meet the real-time and accuracy requirements of oil wear particle detection in industrial sites.

[0004] Therefore, how to achieve real-time and accurate detection of abrasive particles in oil at industrial sites and obtain the morphological characteristics of the abrasive particles is a technical problem that needs to be solved urgently. Summary of the invention

[0005] The purpose of the present application is to provide an image sensor for online detection of oil quality, so as to improve the accuracy of detection of abrasive particles in oil.

[0006] The embodiment of the present application provides an image sensor for online detection of oil quality, comprising: a sensor body, the sensor body comprising an oil inlet, an oil outlet and a detection chamber, the detection chamber being arranged between the oil inlet and the oil outlet, the detection chamber comprising a cavity and a transparent protective plate arranged on the top of the cavity; a magnetic adsorption component, the magnetic adsorption component comprising an annular electromagnet arranged above the transparent protective plate, the annular electromagnet comprising an annular iron core and a coil wound on the annular iron core; an image acquisition component, the image acquisition component being arranged on the upper surface of the transparent protective plate, comprising an imaging lens, a CMOS sensor, a device and a ring light source, the optical axis of the imaging lens is perpendicular to the upper surface of the transparent protective plate; a processing circuit, the processing circuit is electrically connected to the image acquisition component, the processing circuit includes an image processing unit and a feature analysis unit, the image processing unit is used to pre-process the acquired optical image, and the feature analysis unit is used to analyze the pre-processed wear particle image; and a control circuit, the control circuit is electrically connected to the magnetic adsorption component and the image acquisition component, and is used to control the coil to be energized for a preset time to generate a magnetic field to adsorb ferromagnetic wear particles in the oil, and control the image acquisition component to capture an image after the preset time.

[0007] In the above image sensor, the transparent protective plate is made of one of the following materials: quartz glass, sapphire, polymethyl methacrylate, epoxy resin, and polycarbonate.

[0008] In the above image sensor, a slope is arranged at the bottom of the detection chamber, and the bottom surface of the oil outlet of the detection chamber is not higher than the lowest end of the slope.

[0009] In the above image sensor, the inner diameter of the annular core is larger than the outer diameter of the imaging lens, and the outer diameter of the annular core is larger than the outer diameter of the annular light source.

[0010] In the above-mentioned image sensor, an optical coupling agent is provided between the imaging lens and the upper surface of the transparent protective plate, and the refractive index of the optical coupling agent matches the refractive index of the transparent protective plate.

[0011] In the above image sensor, a flushing pipe is further provided at the bottom of the detection chamber. The flushing pipe includes a plurality of nozzles evenly distributed along the circumference of the detection chamber. The flushing pipe is connected to an external high-pressure oil pump.

[0012] In the above image sensor, a hydrophobic layer is disposed on the lower surface of the transparent protective plate, and a polytetrafluoroethylene anti-adhesion coating is formed on the hydrophobic layer.

[0013] In the above-mentioned image sensor, the image acquisition component is fixed to the upper surface of the transparent protective plate through a support seat, the support seat is an annular structure, and its inner diameter is larger than the outer diameter of the imaging lens and its outer diameter is smaller than the inner diameter of the annular iron core. An annular groove is provided at the bottom of the support seat, and the annular light source is embedded in the annular groove. The support seat is provided with a lens seat for mounting the imaging lens.

[0014] In the above-mentioned image sensor, the annular light source includes a plurality of LED light sources evenly distributed along the circumferential direction, the light emission wavelength of the LED light source is within the near-infrared spectrum, and the light emitted by the annular light source is incident on the transparent protective plate at an incident angle of 20 to 40 degrees.

[0015] In the above-mentioned image sensor, the image processing unit is used to preprocess the optical image captured by the image acquisition component, including: image enhancement of the optical image to improve the clarity of the image, noise removal of the optical image to eliminate interference information, and image segmentation of the optical image to separate the image of the abrasive particles; the feature analysis unit is used to analyze the preprocessed abrasive particle image, including: identifying the size, quantity and morphological characteristics of the ferromagnetic abrasive particles fixed on the lower surface of the transparent protective plate, and identifying the size, quantity and morphological characteristics of the non-ferromagnetic abrasive particles located on the lower surface of the transparent protective plate, wherein the morphological characteristics include the shape, surface texture and edge characteristics of the abrasive particles.

[0016] The image sensor for online detection of oil quality provided by the present application, by arranging a transparent protective plate on the top of the detection chamber and arranging a magnetic adsorption component including an annular electromagnet above the transparent protective plate, can use the magnetic field force to adsorb and fix ferromagnetic abrasive particles in the oil on the lower surface of the transparent protective plate, so that the abrasive particles remain in a stable state; by arranging an image acquisition component including an imaging lens, a CMOS sensor and an annular light source on the upper surface of the transparent protective plate, and making the optical axis of the imaging lens perpendicular to the upper surface of the transparent protective plate, it is possible to obtain a clear optical image of the adsorbed and fixed abrasive particles, thereby avoiding the technical defect that the traditional detection method cannot obtain the morphological characteristics of the abrasive particles; the power-on time of the coil is controlled by the control circuit, so that the abrasive particles are adsorbed and fixed within a preset time, and the image is acquired after the preset time, thereby ensuring the timing and controllability of the detection process and improving the accuracy of the detection; the collected optical image is preprocessed by the image processing unit in the processing circuit, and the preprocessed abrasive particle image is analyzed by the feature analysis unit, so that not only the quantity and size information of the abrasive particles can be obtained, but also the morphological characteristics of the abrasive particles can be obtained, thereby overcoming the limitation that the traditional electromagnetic induction detection method can only obtain the quantity and size information of the abrasive particles. In addition, the image sensor adopts an online detection method, which does not require manual sampling and microscope observation, avoiding the problems of low efficiency and susceptibility to human factors in traditional offline iron spectrum analysis methods, and can realize real-time and accurate detection of wear particles in oil. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 Schematic diagram of an image sensor for online detection of oil quality provided in an embodiment of the present application.

[0018] Figure 2 yes Figure 1 A schematic diagram of the cross-sectional structure of an image sensor for online detection of oil quality is shown.

[0019] Figure 3 yes Figure 2 The image sensor for online detection of oil quality is shown in a top view of the whole consisting of a transparent protective plate, a magnetic adsorption component, an image acquisition component and a support base.

[0020] Figure 4 yes Figure 3 Schematic diagram of the AA' section of the image sensor for online detection of oil quality. DETAILED DESCRIPTION

[0021] The specific implementation methods of the present application are described in detail below with reference to the accompanying drawings.

[0022] The terms "first", "second" and similar words do not indicate any order, quantity or importance, but are only used to distinguish different technical features. The term "plurality" and similar words mean two or more, unless otherwise clearly defined.

[0023] In order to solve the technical problem of low accuracy in detecting abrasive particles in the oil quality detection process of existing image sensors, the present application provides an image sensor 20 for online detection of oil quality.

[0024] The image sensor 20 of the present application can be applied to an oil circulation system of industrial equipment, which includes a main oil tank, an auxiliary oil tank, an oil pump group, a filter device, a cooling device and a pipeline. The main oil tank is used to store working oil, the auxiliary oil tank is used to replenish and adjust the oil, the oil pump group is used to drive the oil to circulate in the oil circulation system, and the filter device and the cooling device are used to purify and cool the circulating oil respectively.

[0025] The key monitoring points of the oil circulation system are equipped with multiple detection units, each of which is equipped with the image sensor 20 of the present application. The detection units are set at the main oil tank outlet, before and after the filter device, and at the cooling device outlet. The changes in the oil quality at these locations have an important indicative effect on the operating status of the equipment.

[0026] Each detection unit includes an image sensor 20, a shunt detection branch and a mounting assembly. The shunt detection branch is arranged in parallel with the main oil circuit, and the shunt detection branch includes an oil inlet pipe section 10 and an oil outlet pipe section 30, and the image sensor 20 is arranged between the oil inlet pipe section 10 and the oil outlet pipe section 30. The oil inlet pipe section 10 is connected to the main oil circuit through an electric regulating valve, and the oil outlet pipe section 30 is connected to the main oil circuit through a back pressure valve. A flow regulating device is provided on the oil inlet pipe section 10, and the flow regulating device includes a throttle valve and a flow stabilizing valve arranged in series, which is used to adjust the oil in a turbulent state to a laminar state.

[0027] The mounting assembly includes a mounting base and a vibration-damping layer. The vibration-damping layer is made of rubber material and is disposed between the mounting base and the image sensor 20 to absorb external vibrations and prevent the vibrations from affecting the clarity of image acquisition.

[0028] On the flow shunt detection branch, a flow sensor is arranged on the oil inlet pipe section 10 , downstream of the flow regulating device, for real-time monitoring of the oil flow entering the image sensor 20 .

[0029] In actual application, the oil enters the oil inlet pipe section 10 of the shunt detection branch from the main oil circuit through the electric regulating valve, and the oil is changed from a turbulent state to a laminar state through the regulation of the throttle valve and the steady flow valve of the flow regulating device. After the oil in the laminar state enters the image sensor 20, a stable flow field is formed in the detection chamber 202 of the image sensor 20, wherein the detection chamber 202 is the area below the transparent protective plate 204, which corresponds to the shooting range of the image acquisition component.

[0030] like Figure 1 and Figure 2 As shown, the image sensor 20 for online detection of oil quality provided by the present application includes: The sensor body comprises an oil inlet 201, an oil outlet 203 and a detection chamber 202, wherein the detection chamber 202 is arranged between the oil inlet 201 and the oil outlet 203; the detection chamber 202 comprises a cavity and a transparent protective plate 204 arranged on the top of the cavity, wherein the transparent protective plate 204 is made of one of the following materials: quartz glass, sapphire, polymethyl methacrylate (PMMA), epoxy resin, polycarbonate (PC); wherein the refractive index of quartz glass and sapphire is between 1.45 and 1.55, the refractive index of PMMA is 1.49, the refractive index of epoxy resin is 1.46, and the refractive index of PC is 1.48, and the refractive index of the above materials is close to the refractive index of engine oil (about 1.47) to reduce the reflection and refraction of light. A plurality of semi-cylindrical protrusions are arranged on the inner wall of the detection chamber 202 at intervals along the oil flow direction. The radius of the semi-cylindrical protrusions is 2-5 mm, and is used to slow down the oil flow rate.

[0031] Magnetic adsorption component 205, such as Figure 3 As shown, the magnetic adsorption component 205 includes an annular electromagnet arranged above the transparent protective plate 204, and the annular electromagnet includes an annular iron core and a coil wound on the annular iron core, the inner diameter of the annular iron core is larger than the outer diameter of the imaging lens 2061, and the outer diameter of the annular iron core is larger than the outer diameter of the annular light source 2062; when the coil is energized, an annular magnetic field is generated in the annular iron core, and the annular magnetic field passes through the transparent protective plate 204 to adsorb ferromagnetic abrasive particles in the oil and fix the ferromagnetic abrasive particles on the lower surface of the transparent protective plate 204; when the coil is de-energized, the annular magnetic field disappears, and the ferromagnetic abrasive particles fixed on the lower surface of the transparent protective plate 204 detach from the lower surface of the transparent protective plate 204 under the action of gravity and flow out of the detection chamber 202 with the oil.

[0032] An image acquisition component 206 is arranged on the upper surface of the transparent protective plate 204, and is used to acquire an optical image of the ferromagnetic abrasive particles fixed on the lower surface of the transparent protective plate 204 through the transparent protective plate 204; the image acquisition component 206 includes an imaging lens 2061, a CMOS sensor and a ring light source 2062; the optical axis of the imaging lens 2061 is perpendicular to the upper surface of the transparent protective plate 204, and an optical coupling agent is provided between the imaging lens 2061 and the upper surface of the transparent protective plate 204, and the refractive index of the optical coupling agent matches the refractive index of the transparent protective plate 204, so as to reduce the reflection loss of light at the interface between the transparent protective plate 204 and the imaging lens 2061.

[0033] The processing circuit 207 is electrically connected to the image acquisition component 206, and the processing circuit 207 includes an image processing unit and a feature analysis unit; the image processing unit is used to pre-process the optical image acquired by the image acquisition component 206, including: image enhancement of the optical image to improve the clarity of the image, noise removal of the optical image to eliminate interference information, and image segmentation of the optical image to separate the image of the abrasive particles; the feature analysis unit is used to analyze the pre-processed abrasive particle image, including: identifying the size, quantity and morphological characteristics of the ferromagnetic abrasive particles fixed on the lower surface of the transparent protective plate 204, and identifying the size, quantity and morphological characteristics of the non-ferromagnetic abrasive particles and bubbles located on the lower surface of the transparent protective plate 204; wherein the morphological characteristics include the shape, surface texture and edge characteristics of the abrasive particles and bubbles.

[0034] The control circuit 208 is electrically connected to the magnetic adsorption component 205 and the image acquisition component 206, and is used to control the coil of the electromagnet unit to be energized for a preset time to generate a magnetic field when the pressure value detected by the pressure sensor is stable within a preset range, and the ferromagnetic abrasive particles in the oil are adsorbed on the lower surface of the transparent protective plate 204 during the preset time. The control circuit 208 is used to control the image acquisition component 206 to capture an image after the preset time is over, and is used to control the coil of the electromagnet unit to be de-energized after the image acquisition is completed, so that the ferromagnetic abrasive particles fixed on the lower surface of the transparent protective plate 204 are detached under the action of gravity and flow out of the detection chamber 202 with the oil.

[0035] The preset time is determined according to the oil flow rate, so that the volume of oil passing through the detection chamber 202 within the preset time is a preset value.

[0036] The communication module is electrically connected to the processing circuit 207 and is used to send the abrasive information identified by the processing circuit 207 to the remote monitoring terminal, wherein the abrasive information includes the type, size, quantity and morphological characteristics of the abrasive, and the volume of oil corresponding to the abrasive information.

[0037] The technical solution is provided with a transparent protective plate 204 on the top of the detection chamber 202, and an electromagnet unit is provided above the transparent protective plate 204. When the electromagnet unit is powered on, a circular magnetic field is formed on the lower surface of the transparent protective plate 204, and the circular magnetic field can adsorb the ferromagnetic abrasive particles in the oil on the lower surface of the transparent protective plate 204; when the electromagnet unit is powered off, the ferromagnetic abrasive particles detach from the lower surface of the transparent protective plate 204 and flow with the oil. An image acquisition component 206 is provided above the transparent protective plate 204, and the image acquisition component 206 acquires an optical image of the ferromagnetic abrasive particles. The image processing unit preprocesses the optical image, including image enhancement, noise removal and image segmentation, and the feature analysis unit analyzes the preprocessed abrasive particle image to identify the size, quantity and morphological characteristics of the ferromagnetic abrasive particles fixed on the lower surface of the transparent protective plate 204, wherein the morphological characteristics include the shape, surface texture and edge characteristics of the abrasive particles, thereby realizing accurate detection of the abrasive particles in the oil.

[0038] The annular core includes an inner ring and an outer ring, the inner ring and the outer ring are coaxially arranged and connected by a plurality of radial connectors, and coils are wound around the inner ring and the outer ring; the coil of the inner ring and the coil of the outer ring are connected in series, and the winding direction of the coil of the inner ring is opposite to the winding direction of the coil of the outer ring, so that the magnetic field generated by the inner ring and the magnetic field generated by the outer ring form a uniformly distributed magnetic field on the lower surface of the transparent protective plate 204; the turns ratio of the coil of the inner ring and the coil of the outer ring is 1:1.2 to 1:1.5, so as to compensate for the difference in magnetic field strength caused by the difference in magnetic path length; the number of radial connectors is 4 to 8, and the radial connectors are uniformly distributed along the circumferential direction; a magnetic shielding layer is provided in the annular space between the inner ring and the outer ring, and the magnetic shielding layer is used to prevent the magnetic fields of the inner ring and the outer ring from interfering with each other.

[0039] The processing circuit 207 is also used to analyze the magnetic field distribution characteristics of the annular magnetic field on the lower surface of the transparent protective plate 204, establish a magnetic field intensity distribution model for different areas on the lower surface of the transparent protective plate 204, and establish a distribution density model of ferromagnetic abrasive particles in different areas on the lower surface of the transparent protective plate 204 based on the magnetic field distribution model, wherein the distribution density model characterizes the corresponding relationship between the distribution density of ferromagnetic abrasive particles in different areas on the lower surface of the transparent protective plate 204 and the distribution density of ferromagnetic abrasive particles at the middle position of the annular shape under the same oil abrasive content conditions; the processing circuit 207 is also used to establish a correction model between the distribution density of ferromagnetic abrasive particles at the middle position of the annular shape and the actual content of ferromagnetic abrasive particles in the oil based on the distribution density model.

[0040] The processing circuit 207 is further configured to perform the following steps: Identify the number of ferromagnetic abrasive grains corresponding to the middle area of ​​the annular magnetic field on the lower surface of the transparent protective plate 204 from the image captured by the image acquisition component 206, wherein the middle area is the area of ​​the positive projection of the imaging lens 2061 of the image acquisition component 206 on the transparent protective plate 204; Calculating the actual content of ferromagnetic abrasive particles in the oil according to the correction model, wherein the correction model includes: multiplying the number of ferromagnetic abrasive particles in the middle area by a first correction coefficient to obtain a first correction value, wherein the first correction coefficient is related to the magnetic field intensity distribution of the annular magnetic field; multiplying the first correction value by a second correction coefficient to obtain a second correction value, wherein the second correction coefficient is related to the flow rate of the oil; multiplying the second correction value by a third correction coefficient to obtain a third correction value, wherein the third correction coefficient is related to the temperature of the oil; multiplying the third correction value by a fourth correction coefficient to obtain the actual content of ferromagnetic abrasive particles in the oil, wherein the fourth correction coefficient is related to the pressure of the oil; Among them, the first correction coefficient is 1.2~1.5 when the magnetic field strength of the annular magnetic field is 0.1~0.5 Tesla, the second correction coefficient is 0.8~1.2 when the flow rate of the oil is 0.5~2 m / s, the third correction coefficient is 0.9~1.1 when the temperature of the oil is 20~80 degrees Celsius, and the fourth correction coefficient is 0.95~1.05 when the pressure of the oil is 0.1~0.5 MPa.

[0041] The feature analysis unit is used to measure the size of ferromagnetic abrasive particles through the following steps: performing edge detection on the segmented abrasive particle image to extract the contour information of the abrasive particles; calculating the maximum inscribed circle diameter and the minimum circumscribed circle diameter of the abrasive particle contour, and taking the average of the two diameters as the equivalent diameter of the abrasive particle; calculating the circumference and area of ​​the abrasive particle contour, and calculating the shape factor of the abrasive particle based on the circumference and area; the feature analysis unit is used to count the number of ferromagnetic abrasive particles through the following steps: marking the connected regions of the segmented abrasive particle image; identifying and merging adjacent abrasive particle regions; and counting the number of independent abrasive particle regions; the feature analysis unit is used to analyze the morphological characteristics of the ferromagnetic abrasive particles through the following steps: extracting the texture characteristics of the abrasive particle surface, including surface roughness, texture directionality and texture complexity; analyzing the characteristics of the abrasive particle edge, including the sharpness of the edge, the regularity of the edge and the fractal dimension of the edge; calculating the shape characteristic parameters of the abrasive particle, including roundness, aspect ratio and convexity; and establishing the morphological characteristic vector of the abrasive particle based on the above characteristic parameters.

[0042] The processing circuit 207 is also used to identify the wear type and degree of the equipment based on the size distribution, quantity change trend and morphological characteristics of the abrasive particles.

[0043] The lower surface of the transparent protective plate 204 is provided with a hydrophobic layer formed by plasma treatment, the contact angle of the hydrophobic layer is greater than 150 degrees, and a polytetrafluoroethylene anti-adhesion coating is formed on the hydrophobic layer by chemical vapor deposition, and the thickness of the polytetrafluoroethylene anti-adhesion coating is 100~200 nanometers, so that the ferromagnetic abrasive particles can smoothly detach from the lower surface of the transparent protective plate 204 when the coil is powered off; the transparent protective plate 204 is sealed and connected to the cavity through a fluororubber sealing ring, and the Shore hardness of the fluororubber sealing ring is 60~70 degrees.

[0044] A pressure sensor for detecting the oil pressure in the detection chamber 202 is provided on the side wall of the detection chamber 202. When the pressure value detected by the pressure sensor is stabilized in the range of 0.1~0.2MPa, it indicates that the oil flow rate has been reduced to a range suitable for detection of 0.1~0.2m / s; the bottom and side walls of the detection chamber 202 are coated with a light-absorbing coating, and the light-absorbing coating has an absorption rate of greater than 95% for light with a wavelength of 850nm, which is used to form a low-reflection background.

[0045] The annular light source 2062 is arranged around the imaging lens 2061 and contacts the upper surface of the transparent protective plate 204; the annular light source 2062 includes a plurality of LED light sources uniformly distributed along the circumferential direction, and the emission wavelength of each LED light source is 830 nanometers to 870 nanometers, preferably 850 nanometers, which is within the near-infrared spectrum to reduce the scattering and absorption of light by the oil; the light emitted by the annular light source 2062 is incident on the transparent protective plate 204 at an incident angle of 20 to 40 degrees, and irradiates the ferromagnetic abrasive particles fixed on the lower surface of the transparent protective plate 204 after being refracted by the transparent protective plate 204, and the light scattered by the ferromagnetic abrasive particles passes through the transparent protective plate 204 and enters the imaging lens 2061; the inner diameter of the annular light source 2062 is larger than the effective aperture of the imaging lens 2061, and the outer diameter is smaller than the inner diameter of the annular iron core to avoid affecting the distribution of the annular magnetic field; like Figure 4 As shown, the image acquisition component 206 is fixed to the upper surface of the transparent protective plate 204 through a support seat 209. The support seat 209 is an annular structure, and its inner diameter is smaller than the outer diameter of the imaging lens 2061, and its outer diameter is smaller than the inner diameter of the annular iron core and larger than the outer diameter of the imaging lens 2061; an annular groove is provided at the bottom of the support seat 209, and the annular light source 2062 is embedded in the annular groove; the support seat 209 is provided with a lens seat for mounting the imaging lens 2061; a plurality of mounting holes are provided on the outer circumferential surface of the support seat 209, which are used to fix the support seat 209 on the annular iron core to ensure that the image acquisition component 206 is coaxially arranged with the magnetic adsorption component 205; the support seat 209 is made of non-magnetic material.

[0046] Each LED light source of the annular light source 2062 is electrically connected to an LED driving circuit, and the LED driving circuit controls the luminous intensity of all LED light sources through a PWM signal; the LED driving circuit includes an LED driving unit and a control unit; the control unit adjusts the duty cycle of the PWM signal according to the requirements of image processing, and when the image is overexposed, the duty cycle of the PWM signal is reduced by 10%~20%, and when the image is underexposed, the duty cycle of the PWM signal is increased by 10%~20% until a suitable image exposure is obtained; the initial duty cycle of the PWM signal is 60% and the frequency is 1kHz; the control unit is also used to analyze the average brightness of the image in real time, and automatically adjust the duty cycle of the PWM signal when the average brightness deviates from a preset range, wherein the preset average brightness range is 128±20 (based on a grayscale value range of 0~255).

[0047] As an improvement, a slope 2021 is provided at the bottom of the detection chamber 202, and the slope 2021 is located below the transparent protective plate 204, the starting end of the slope 2021 is located between the front end of the transparent protective plate 204 and the oil inlet 201, and the terminal end of the slope 2021 is located between the rear end of the transparent protective plate 204 and the oil outlet 203; the inclination angle of the slope 2021 is 30 to 45 degrees, so that the ferromagnetic abrasive particles separated from the transparent protective plate 204 can slide quickly along the slope 2021; the detection chamber 20 The bottom surface of the oil outlet 203 of 2 is not higher than the lowest end of the slope 2021, and the diameter of the oil outlet 203 is larger than the diameter of the oil inlet 201 to ensure that the ferromagnetic abrasive particles can be discharged smoothly; the width of the slope 2021 is larger than the width of the transparent protective plate 204, the surface of the slope 2021 is a smooth plane, and the surface of the slope 2021 is made of anti-corrosion material to ensure that the ferromagnetic abrasive particles can slide smoothly and be discharged from the detection chamber 202, thereby ensuring that the electromagnet unit will not absorb the ferromagnetic abrasive particles detected in the previous detection during the next detection.

[0048] Furthermore, a flushing pipe is also provided at the bottom of the detection chamber 202, and the flushing pipe includes a main pipe and a plurality of branch pipes. The main pipe is respectively provided along the edges of both sides of the slope 2021 to form two parallel pipes, one end of the two parallel pipes is connected through a connecting pipe, and the other end is respectively connected to an external high-pressure oil pump; the plurality of branch pipes are vertically provided on the main pipe, and the plurality of branch pipes are evenly distributed along the length direction of the main pipe; one end of each branch pipe is connected to the main pipe, and the other end is provided with a nozzle; the spray direction of the nozzle is toward the slope 202 1, and the angle between the injection direction of the nozzle and the surface of the slope 2021 is 45 degrees; the nozzle is located above the surface of the slope 2021, and the vertical distance between the outlet of the nozzle and the surface of the slope 2021 is one tenth of the width of the slope 2021; a guide structure is provided at the outlet of the nozzle, and the guide structure is used to make the injected oil flow form a vortex along the surface of the slope 2021 to improve the flushing effect; a control valve is also provided on the main pipeline, and the control valve is used to control the flow rate of oil delivered by the high-pressure oil pump to the flushing pipe.

[0049] After each test is completed, the high-pressure oil pump is controlled to spray high-pressure oil flow into the test chamber 202 through the flushing pipe. The pressure of the high-pressure oil flow is 0.5~1MPa to flush away ferromagnetic abrasive particles that may be retained in the test chamber 202. A filter is also provided in the flushing pipe to prevent large particles of impurities in the external oil circuit from entering the test chamber 202.

[0050] In practical applications, the oil temperature may change, causing changes in the viscosity and refractive index of the oil. When the oil temperature rises, the oil viscosity decreases and the fluidity increases, which may lead to unstable fixation of ferromagnetic abrasive particles under the action of the magnetic field; at the same time, the change in the refractive index of the oil will affect the quality of optical imaging.

[0051] To address this technical problem, a temperature sensor is further provided on the side wall of the detection chamber 202 for real-time monitoring of the oil temperature; the control circuit 208 dynamically adjusts the current of the coil according to the temperature value detected by the temperature sensor; specifically, when the temperature is higher than 40°C, the current of the coil is increased by 10% to 20% to enhance the magnetic field strength; when the temperature is lower than 0°C, the current of the coil is reduced by 10% to 20% to avoid agglomeration of abrasive particles due to excessive magnetic field.

[0052] The control circuit 208 also includes a storage unit, which stores a corresponding relationship table between temperature and LED luminous intensity and a corresponding relationship table between temperature and image enhancement parameters; wherein the corresponding relationship table between temperature and LED luminous intensity records the LED driving current value corresponding to each temperature point of 5°C in the range of -20°C to 80°C. When the temperature rises, the LED driving current decreases accordingly, and the reduction range is 5% to 8% for every 10°C increase in the LED driving current, so as to compensate for the characteristic that the LED luminous efficiency increases with the increase in temperature; the corresponding relationship table between temperature and image enhancement parameters records the LED driving current value corresponding to each temperature point of 5°C in the range of -20°C to 80°C. The image contrast enhancement coefficient, sharpening coefficient and denoising threshold at different temperatures were recorded. Specifically, when the temperature is in the range of -20℃ to 0℃, the contrast enhancement coefficient is 1.2~1.4, the sharpening coefficient is 0.8~1.0, and the denoising threshold is 15~20; when the temperature is in the range of 0℃ to 40℃, the contrast enhancement coefficient is 1.0~1.2, the sharpening coefficient is 1.0~1.2, and the denoising threshold is 10~15; when the temperature is in the range of 40℃ to 80℃, the contrast enhancement coefficient is 0.8~1.0, the sharpening coefficient is 1.2~1.4, and the denoising threshold is 5~10.

[0053] The control circuit 208 reads the value of the temperature sensor once every predetermined time (for example, 1 second), and performs interpolation calculation in the corresponding relationship table according to the current temperature to obtain the corresponding LED driving current value and image enhancement parameters; the control circuit 208 sends the calculated LED driving current value to the LED driving circuit, and sends the calculated image enhancement parameters to the image processing unit; the LED driving circuit adjusts the luminous intensity of the LED light source according to the received driving current value, and the image processing unit processes the collected image according to the received image enhancement parameters.

[0054] In practical applications, there may be bubbles in the oil, and when these bubbles adhere to the surface of the transparent protective plate 204, they will interfere with the collection and analysis of the wear particle image, resulting in misjudgment. At the same time, the bubbles may also affect the adsorption effect of the magnetic field on the wear particles.

[0055] In order to solve this technical problem, the detection chamber 202 is provided with a plurality of tangentially arranged guide vanes arranged near the oil inlet 201. The plurality of guide vanes are evenly arranged along the circumference of the inner wall of the detection chamber 202. The angle between the mounting surface of each guide vane and the radial direction of the inner wall of the detection chamber 202 is 60 degrees, so that the extension direction of the guide vane and the flow direction of the oil form a tangential guiding effect. The frontal surface of the guide vane is an arc surface, and the back surface of the guide vane is a plane. The axial angle between the frontal surface of the guide vane and the detection chamber 202 is 30 to 35 degrees; one end of the guide vane is fixed on the inner wall of the detection chamber 202, and the other end points to the central axis of the detection chamber 202, so that the oil produces a rotational motion, and the bubbles are gathered toward the center by centrifugal force.

[0056] The top of the detection chamber 202 is provided with a plurality of strip exhaust grooves arranged along the flow direction of the oil, and the strip exhaust grooves are arranged on the top surface of the detection chamber 202 between the oil inlet 201 and the transparent protective plate 204; an exhaust hole is provided at the center position of the strip exhaust groove, and the exhaust hole is connected to a vertically arranged exhaust channel; the upper end of the exhaust channel is connected to the external atmosphere through a solenoid valve, and the control circuit 208 controls the solenoid valve to open for 3 to 5 seconds for exhaust before the detection starts, and closes the solenoid valve after the exhaust is completed.

[0057] The bottom surface of the strip exhaust groove is a V-shaped structure formed by two inclined surfaces, and the intersection of the two inclined surfaces is parallel to the flow direction of the oil; the angle between the two inclined surfaces of the V-shaped structure is 120-130 degrees, and the inclination angle of each inclined surface relative to the horizontal plane is 25-30 degrees; the cross-sectional area of ​​the strip exhaust groove gradually decreases along the flow direction of the oil; the exhaust hole is arranged at the bottom of the V-shaped groove; the bottom of the V-shaped groove is located above the opening of the V-shaped groove, so that the bubbles automatically move upward along the inclined surface and conical channel of the V-shaped groove under the action of buoyancy and gather at the top of the V-shaped groove. A plurality of fine guide grooves are also arranged on the two inclined surfaces of the V-shaped structure, and the guide grooves are used to guide small bubbles to gather to the top of the V-shaped groove.

[0058] The surface roughness Ra of the inner surface of the strip-shaped exhaust groove is less than 0.4 micrometers to reduce the adhesion of bubbles on the groove wall.

[0059] When the oil flow state suddenly changes (such as pressure fluctuation, sudden change in flow rate, etc.), the adsorbed abrasive particles may be displaced or fall off, affecting the detection accuracy. This sudden change in flow state is mainly manifested in: rapid increase or decrease in oil pressure, drastic change in oil flow rate, change in oil flow direction, etc.

[0060] In response to this technical problem, the control circuit 208 is also used to identify the stability of oil flow by real-time analysis of pressure sensor data, and to automatically adjust the coil's current when it is detected that the pressure fluctuation exceeds a preset range. Specifically, the control circuit 208 first obtains the oil pressure data at the oil inlet 201 and the oil outlet 203 through the pressure sensors arranged at the oil inlet 201 and the oil outlet 203 of the detection chamber 202, and calculates the pressure change rate; then, it is determined whether the oil flow state has suddenly changed based on the pressure change rate; when a sudden increase in pressure is detected, the coil current is increased to increase the magnetic field strength, enhance the adsorption force on the abrasive particles, and prevent the abrasive particles from shifting; when a sudden drop in pressure is detected, the coil current is kept unchanged to maintain the original magnetic field strength.

[0061] In actual working conditions, there may be mechanical vibrations in the operation of the equipment, which are transmitted to the detection device through the support structure, causing slight vibrations in the detection chamber 202. When the vibration frequency is close to a certain natural frequency of the detection system, resonance may occur, causing the wear particle image to be blurred.

[0062] In response to this technical problem, the detection device also includes a vibration detection component, a vibration reduction support component and a vibration control component; the vibration detection component includes an acceleration sensor arranged at the bottom of the detection chamber 202, which is used to detect the vibration acceleration of the detection chamber 202 in real time; the vibration reduction support component includes piezoelectric ceramic actuators arranged at the four corners of the detection chamber 202, one end of each piezoelectric ceramic actuator is fixedly connected to the detection chamber 202, and the other end is fixedly connected to the base; the vibration control component is electrically connected to the vibration detection component and the vibration reduction support component, and is used to receive the vibration signal of the vibration detection component and control the piezoelectric ceramic actuator of the vibration reduction support component to produce displacement.

[0063] The vibration control component first obtains the vibration signal detected by the acceleration sensor, and obtains the frequency and amplitude of the vibration by performing Fourier transform on the vibration signal; then calculates the compensating displacement that each piezoelectric ceramic actuator needs to generate according to the frequency and amplitude of the vibration; finally, controls each piezoelectric ceramic actuator to generate a displacement opposite to the vibration direction to offset the vibration of the detection chamber 202.

[0064] Furthermore, the processing circuit 207 is used to perform real-time analysis on the captured image, extract feature points in the image, track the motion trajectory of the feature points, and obtain the direction and amplitude of the image jitter; the vibration control component fuses the vibration data obtained by the vibration detection unit with the image jitter data obtained by the processing circuit 207 to generate a more accurate compensation control signal.

[0065] The processing circuit 207 is also used to screen stable images through the following steps: first, continuously capture multiple frames of images and extract feature points in each frame of the image; then calculate the displacement of the feature points between adjacent frames, including translation displacement and rotation displacement; then calculate the stability index of each frame of the image based on the displacement, and the stability index is inversely proportional to the average displacement of the feature points; finally, select the image with the highest stability index as the target image for subsequent wear particle feature extraction.

[0066] In practical applications, non-magnetic impurities (such as dust, fibers, etc.) and bubbles in the oil may adhere to the surface of the transparent protective plate 204, affecting the accuracy of wear particle detection. These non-magnetic impurities are not affected by the magnetic field and are difficult to remove by magnetic field control.

[0067] To address this technical problem, the detection chamber 202 is also provided with an ultrasonic generator arranged on the side wall of the cavity; the ultrasonic generator is made of piezoelectric ceramic material, with an oscillation frequency of 40~45 kHz, a power density of 0.5~1W / cm², and an adjustable amplitude range of 5~20 microns; the control circuit 208 first controls the ultrasonic generator to run for 5~10 seconds before each detection, and through the high-frequency oscillation and microbubble collapse effect generated by ultrasound in the oil, a microscopic-scale local high-pressure area and jet are formed on the surface of the transparent protective plate 204, thereby peeling off non-magnetic impurities and bubbles from the surface of the transparent protective plate 204.

[0068] The image processing unit is used to distinguish ferromagnetic abrasive particles, non-magnetic impurities and bubbles through the following steps: first, preliminary classification is performed based on the morphological characteristics of the target, wherein ferromagnetic abrasive particles are usually irregular polygonal or strip-shaped, non-magnetic impurities are fibrous or flocculent, and bubbles are round or elliptical; second, further identification is performed based on the optical characteristics of the target, including: extracting the gray value distribution characteristics of the target, the gray value distribution of the ferromagnetic abrasive particles presents a characteristic of darker center and brighter edge, the gray value distribution of non-magnetic impurities is relatively uniform, and the gray value distribution of bubbles presents a characteristic of brighter center and darker edge; calculating the edge gradient characteristics of the target, the edge gradient value of the ferromagnetic abrasive particles is large and unevenly distributed, the edge gradient value of the non-magnetic impurities is small, and the edge gradient value of the bubbles is moderate and evenly distributed; analyzing the reflection characteristics of the target at different illumination angles, the ferromagnetic abrasive particles show obvious directional reflection characteristics, the non-magnetic impurities show diffuse reflection characteristics, and the bubbles show refraction and total reflection characteristics.

[0069] The image processing unit counts the number of non-magnetic impurities and bubbles after processing each frame of the image; when the number of non-magnetic impurities detected exceeds a first predetermined number (e.g., 50 / cm2) or the number of bubbles exceeds a second predetermined number (e.g., 30 / cm2), the ultrasonic generator is automatically triggered to start.

[0070] The control circuit 208 is also used to control the ultrasonic generator to operate at a first power density (e.g., 0.5 W / cm²) and a first amplitude (e.g., 5 microns) in the first stage of cleaning, and to start the image acquisition component 206 to monitor the cleaning effect; if the image acquisition component 206 detects that the number of non-magnetic impurities and bubbles still exceeds the preset threshold after the first stage of cleaning, the second stage of cleaning is entered, the power density of the ultrasonic generator is increased to a second power density (e.g., 0.8 W / cm²) and the amplitude is increased to a second amplitude (e.g., 10 microns), and the cleaning effect is monitored by the image acquisition component 206; if the number of non-magnetic impurities and bubbles still exceeds the preset threshold after the second stage of cleaning, the third stage of cleaning is entered, the power density of the ultrasonic generator is increased to a third power density (e.g., 1 W / cm²) and the amplitude is increased to a third amplitude (e.g., 20 microns).

[0071] The image sensor 20 for online detection of oil quality provided in the present application, by setting a transparent protective plate 204 on the top of the detection chamber 202, and setting a magnetic adsorption component 205 including a ring-shaped electromagnet above the transparent protective plate 204, can use the magnetic field force to adsorb and fix ferromagnetic abrasive particles in the oil on the lower surface of the transparent protective plate 204, so that the abrasive particles remain in a stable state; by setting an image acquisition component 206 including an imaging lens 2061, a CMOS sensor and a ring-shaped light source 2062 on the upper surface of the transparent protective plate 204, and making the optical axis of the imaging lens 2061 perpendicular to the upper surface of the transparent protective plate 204, the clear image of the adsorbed and fixed abrasive particles can be obtained. The optical image is clear, avoiding the technical defect that the traditional detection method cannot obtain the morphological characteristics of the abrasive particles; the power-on time of the coil is controlled by the control circuit 208, so that the abrasive particles are adsorbed and fixed within the preset time, and the image is collected after the preset time, ensuring the timing and controllability of the detection process and improving the accuracy of the detection; the collected optical image is preprocessed by the image processing unit in the processing circuit 207, and the preprocessed abrasive particle image is analyzed by the feature analysis unit, which can not only obtain the number and size information of the abrasive particles, but also obtain the morphological characteristics of the abrasive particles, overcoming the limitation that the traditional electromagnetic induction detection method can only obtain the number and size information of the abrasive particles. In addition, the image sensor 20 adopts an online detection method, without manual sampling and microscope observation, avoiding the problem that the traditional offline iron spectrum analysis method is inefficient and susceptible to human factors, and can achieve real-time and accurate detection of abrasive particles in oil.

[0072] The embodiments of the present application are described in detail above, and the contents of this specification should not be construed as limiting the scope of protection of the present application.

Claims

1. An image sensor for online detection of oil quality, characterized in that: include: A sensor body, the sensor body comprising an oil inlet, an oil outlet and a detection chamber, the detection chamber being arranged between the oil inlet and the oil outlet, the detection chamber comprising a cavity and a transparent protective plate arranged on the top of the cavity; A magnetic adsorption component, wherein the magnetic adsorption component comprises an annular electromagnet disposed above the transparent protective plate, wherein the annular electromagnet comprises an annular iron core and a coil wound around the annular iron core; An image acquisition component, which is arranged on the upper surface of the transparent protective plate and includes an imaging lens, a CMOS sensor and a ring light source, and the optical axis of the imaging lens is perpendicular to the upper surface of the transparent protective plate; A processing circuit, the processing circuit is electrically connected to the image acquisition component, the processing circuit includes an image processing unit and a feature analysis unit, the image processing unit is used to pre-process the acquired optical image, and the feature analysis unit is used to analyze the pre-processed wear particle image; as well as A control circuit is electrically connected to the magnetic adsorption component and the image acquisition component, and is used to control the coil to be energized for a preset time to generate a magnetic field to adsorb ferromagnetic abrasive particles in the oil, and to control the image acquisition component to capture an image after the preset time.

2. The image sensor according to claim 1, characterized in that The transparent protection plate is made of one of the following materials: quartz glass, sapphire, polymethyl methacrylate, epoxy resin, and polycarbonate.

3. The image sensor according to claim 1, characterized in that The bottom of the detection chamber is provided with a slope, and the bottom surface of the oil outlet of the detection chamber is not higher than the lowest end of the slope.

4. The image sensor according to claim 1, characterized in that The inner diameter of the annular iron core is larger than the outer diameter of the imaging lens, and the outer diameter of the annular iron core is larger than the outer diameter of the annular light source.

5. The image sensor according to claim 1, characterized in that An optical coupling agent is provided between the imaging lens and the upper surface of the transparent protective plate, and the refractive index of the optical coupling agent matches the refractive index of the transparent protective plate.

6. The image sensor according to claim 1, characterized in that: A flushing pipe is also provided at the bottom of the detection chamber. The flushing pipe includes a plurality of nozzles evenly distributed along the circumference of the detection chamber. The flushing pipe is connected to an external high-pressure oil pump.

7. The image sensor according to claim 1, characterized in that: A hydrophobic layer is arranged on the lower surface of the transparent protective plate, and a polytetrafluoroethylene anti-adhesion coating is formed on the hydrophobic layer.

8. The image sensor according to claim 1, characterized in that: The image acquisition component is fixed to the upper surface of the transparent protective plate through a support seat. The support seat is an annular structure, and its inner diameter is larger than the outer diameter of the imaging lens and its outer diameter is smaller than the inner diameter of the annular iron core. An annular groove is provided at the bottom of the support seat, and the annular light source is embedded in the annular groove. The support seat is provided with a lens seat for mounting the imaging lens.

9. The image sensor according to claim 8, characterized in that: The annular light source includes a plurality of LED light sources uniformly distributed along a circumferential direction, the light emission wavelength of the LED light source is within the near-infrared spectrum, and the light emitted by the annular light source is incident on the transparent protective plate at an incident angle of 20 to 40 degrees.

10. The image sensor according to claim 9, characterized in that: The image processing unit is used to pre-process the optical image collected by the image collection component, including: performing image enhancement on the optical image to improve the clarity of the image, performing noise removal on the optical image to eliminate interference information, and performing image segmentation on the optical image to separate the image of the wear particles; The feature analysis unit is used to analyze the preprocessed abrasive image, including: identifying the size, quantity and morphological characteristics of the ferromagnetic abrasive particles fixed on the lower surface of the transparent protective plate, and identifying the size, quantity and morphological characteristics of the non-ferromagnetic abrasive particles located on the lower surface of the transparent protective plate, wherein the morphological characteristics include the shape, surface texture and edge characteristics of the abrasive particles.

Images