A gear power transmission device for a speed reducer

Abstract

This invention provides a gear power transmission device for a speed reducer, relating to the field of gear reducer technology. The device includes a speed reducer body and an intelligent monitoring and self-cleaning system mounted on a base plate. This invention, through real-time monitoring of multiple parameters such as oil viscosity, abrasive particle concentration, and morphology, can accurately capture early anomalies and effectively distinguish wear types. After diagnosing anomalies, it can automatically purify the lubricating oil online, removing harmful abrasive particles without stopping the machine, effectively extending the lifespan of the oil and equipment. The highly integrated multi-functional modules within a unified detection box result in a compact structure, improving the overall reliability of the system and avoiding the looseness and unreliability issues of external structures in harsh industrial environments, thus enhancing the intelligence level of fault diagnosis. The introduction of machine learning-based automatic particle morphology recognition technology replaces offline analysis relying on professionals, lowering the technical threshold and making fault diagnosis more intuitive and efficient.

Description

Technical Field

[0001] This invention relates to the field of gear reducer technology, specifically to a gear reducer power transmission device. Background Technology

[0002] In the field of industrial transmission, gear reducers, as core power transmission devices, directly determine the operational stability and lifespan of the entire equipment system through their reliability. With the popularization of intelligent manufacturing and predictive maintenance concepts, traditional preventative maintenance based on fixed cycles is gradually being replaced by predictive maintenance strategies based on the real-time status of the equipment, due to the risks of over-maintenance or under-maintenance. This shift makes real-time, online monitoring of the health status of key components within the reducer, such as gear meshing and bearing operation, particularly important. The aim is to achieve fault warning and precise maintenance by capturing early fault characteristics. To adapt to this trend, some reducer products on the market are attempting to integrate sensor technology. For example, they use vibration sensors to capture abnormal mechanical vibrations or integrate simple oil quality sensors to monitor the basic physicochemical parameters of the oil. These explorations indicate that these products are continuously developing towards intelligence and digitalization, and have initially acquired the ability to perceive their condition.

[0003] However, existing intelligent monitoring reducers still have significant drawbacks in terms of the depth and breadth of their technological application. Firstly, their monitoring methods are often relatively simple and lack sufficient analytical dimensions: while vibration analysis is sensitive to certain mechanical faults, it is easily affected by background noise and struggles to identify early wear; existing oil monitoring systems can mostly only make rough judgments on the overall quality of the oil (such as dielectric constant) or a single wear indicator (such as the total amount of ferromagnetic particles), failing to accurately identify and analyze the morphology, size distribution, and composition of wear particles. This results in a high false alarm and false negative rate for system fault warnings, making it difficult to effectively distinguish between normal running-in wear and abnormal cutting or fatigue wear, thus lacking sufficiently accurate data support for maintenance decisions. Secondly, in terms of system integration and functional completeness, the integration of existing monitoring systems with the reducer body is generally low, often employing external or bypass designs. This not only results in a less compact structure and larger footprint but also poses challenges to reliability in complex industrial environments. More importantly, these systems are limited to the "monitoring" stage. After detecting oil contamination or abnormal wear, they generally lack online, proactive self-cleaning capabilities and still require manual shutdown for oil changes or filter replacements. This "diagnosis-only, no treatment" model disrupts the continuous operation of the equipment and fails to truly achieve intelligent closed-loop management of "monitoring-diagnosis-self-cleaning-feedback," thereby affecting the overall operating efficiency and automation level of industrial production lines.

[0004] Therefore, there is an urgent need to deeply improve and optimize the structure and system of existing gear reducer power transmission devices. Through highly integrated and modular design, advanced multi-parameter and multi-principle sensing technologies (such as viscosity monitoring reflecting oil lubrication status, inductive monitoring for early metal shavings abnormalities, and optical particle morphology analysis for accurate wear mechanism identification) are deeply integrated with online and active filtration and self-purification functions to construct a fully functional intelligent health management system. This aims to fundamentally solve the core problems of existing technologies, such as single monitoring dimensions, inaccurate early warning, and passive maintenance relying on manual interruption, thus driving the gear reducer to evolve from a single transmission component into an intelligent unit with self-sensing, self-diagnostic, and self-maintenance capabilities. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a reducer gear power transmission device, which solves the problem that existing reducer gear power transmission devices cannot truly achieve intelligent closed-loop management of "monitoring-diagnosis-self-purification-feedback", thus affecting the overall operating efficiency and automation level of industrial production lines.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a reducer gear power transmission device, comprising:

[0007] The reducer body includes a base plate, a housing consisting of a lower housing and an upper housing, an input shaft, an output shaft, and a reduction gear set disposed within the housing. An oil reservoir is formed within the housing, and an oil outlet and an oil return port are respectively provided on the front and rear side walls of the lower housing.

[0008] The intelligent monitoring and self-purification system is installed on the base plate. The intelligent monitoring and self-purification system includes a circulating oil pump fixed by a first connecting foot, a detection box fixed by a second connecting foot, a filter box fixed by a third connecting foot, and a control box.

[0009] The inlet of the circulating oil pump is connected to the outlet via an oil outlet pipe, and its outlet is provided with an oil delivery pipe.

[0010] The detection box is equipped with a horizontal detection tube inside, the front end of which serves as an inlet, penetrating the front wall of the detection box and connecting to the oil delivery pipe.

[0011] The rear end of the detection tube is located inside the detection box, and an electrically controlled diverter valve is fixedly connected to its rear end. The electrically controlled diverter valve is controlled by the central controller and is used to switch the oil circuit according to the detection result. The two outlets of the electrically controlled diverter valve are respectively connected to a first branch pipe and a second branch pipe. The rear end of the first branch pipe penetrates through the rear wall of the detection box and is connected to the filter box. An optical particle imaging module is fixedly connected to the rear end of the second branch pipe, and the outlet of the optical particle imaging module is connected to the filter box through a connecting pipe. The connecting pipe penetrates through the rear wall of the detection box.

[0012] The outlet of the filter box is connected to the oil return port on the lower housing via an oil return pipe;

[0013] The control box is fixedly connected to the side wall of the detection box. The control box contains a central controller, which is electrically connected to the circulating oil pump, the detection box, the electronically controlled diverter valve, and the optical particle imaging module.

[0014] Preferably, the outer wall of the detection tube is provided with the following from front to back:

[0015] The viscosity detection module is a MEMS viscosity-temperature sensor chip encapsulated in a stainless steel threaded sleeve, with its probe penetrating the outer wall of the detection tube and extending into the interior.

[0016] An inductive abrasive particle detection module includes a frame sleeved on the outer wall of a detection tube and a three-coil differential inductive sensor disposed between the inner wall of the frame and the outer wall of the detection tube. The frame and the detection tube are encapsulated and fixed by potting compound.

[0017] Preferably, the electrically controlled diverter valve is a two-position three-way solenoid valve, fixed at the rear end of the detection tube, with its inlet connected to the detection tube, its first outlet connected to the first branch pipe, and its second outlet connected to the second branch pipe.

[0018] Preferably, the optical particle imaging module includes a fixed frame, a miniature CMOS camera, a pulsed LED light source, and two sets of sapphire glass.

[0019] The fixed frame has a cavity, and the two sets of sapphire glass are respectively disposed on the left and right inner walls of the cavity, and are separated by a pad to form a flat observation pool. The pad has an oil passage hole.

[0020] The miniature CMOS camera and the pulsed LED light source are respectively adjustablely mounted on the left and right sides of the fixed frame through a set of adjustment sleeves, and the axes of the two coincide with the center line of the observation pool; the adjustment sleeves are locked in their relative positions to the fixed frame by locking nuts.

[0021] Preferably, the filter box is provided with a filter module inside, the filter module including a filter frame that is slidably inserted into the filter cavity of the filter box, a filter core disposed in the filter frame, and a panel fixed to the outside of the filter frame for sealing the filter cavity.

[0022] Preferably, the control box is equipped with a display screen, status indicator lights, and a buzzer on the side away from the detection box.

[0023] Preferably, the oil outlet is located on the front wall of the lower casing and at a lower position near the oil storage tank, and the oil return port is located on the rear wall of the lower casing and above the liquid surface of the oil storage tank.

[0024] Preferably, the second branch pipe is connected to the front wall of the fixed frame and communicates with the interior of the observation pool.

[0025] Preferably, the central controller is programmed to:

[0026] It receives and processes signals from the viscosity detection module and the three-coil differential inductive sensor in real time;

[0027] When the signal of the three-coil differential inductive sensor exceeds the preset threshold, the electronically controlled shunt valve is switched to the second outlet path, and the miniature CMOS camera and pulsed LED light source are activated for image acquisition and analysis.

[0028] Based on the results of multi-sensor data fusion, the device health status diagnosis and maintenance suggestions are output through the display screen, status indicator lights and buzzer.

[0029] Preferably, the image analysis program of the central controller includes a particle morphology recognition model based on machine learning algorithms, used to classify and recognize the acquired abrasive particle images.

[0030] This invention provides a gear power transmission device for a speed reducer. It has the following advantages:

[0031] Compared with existing technologies, this reducer gear power transmission device achieves a leap from single monitoring to multi-dimensional collaborative diagnosis. By integrating real-time monitoring of multiple physical quantities such as oil viscosity, abrasive particle concentration and size, and abrasive particle morphology, it can accurately capture early abnormal wear signals and effectively distinguish wear types (such as fatigue wear, cutting wear, etc.), fundamentally solving the problems of traditional monitoring methods being singular, inaccurate in early warning, and unable to identify fault mechanisms.

[0032] Compared with existing technologies, this reducer gear power transmission device constructs an intelligent closed loop of "monitoring-diagnosis-self-purification". This system can not only diagnose oil abnormalities in real time, but also automatically start the bypass precision filter unit when excessive wear particles are detected to purify the lubricating oil online. It can remove harmful wear particles without stopping the machine, effectively extending the life of oil and equipment, and solving the drawbacks of existing technologies that "only diagnose, not treat" and whose maintenance relies on manual intervention during downtime.

[0033] Compared with existing technologies, this reducer gear power transmission device achieves a highly integrated and compact design. Core functional modules such as multi-parameter sensing units, flow dividers, and optical observation cells are highly integrated into a compact detection box and managed by a unified controller. This compact structure avoids the looseness and unreliability of external designs, improving the overall integrity and stability of the system in harsh industrial environments.

[0034] Compared with existing technologies, this reducer gear power transmission device improves the intuitiveness and intelligence of fault diagnosis. By introducing machine learning-based automatic particle morphology recognition technology, it can automatically classify and diagnose wear particles, replacing the traditional offline oil analysis that relies on the experience of professional personnel. This reduces the technical dependence on operators and makes fault diagnosis more intuitive, intelligent, and efficient. Attached Figure Description

[0035] Figure 1 This is a schematic diagram of the structure of the present invention;

[0036] Figure 2 This is a schematic diagram of the circulating oil pump structure of the present invention;

[0037] Figure 3 This is a schematic diagram of the detection box and control box structure of the present invention;

[0038] Figure 4 This is a schematic diagram of the filter box structure of the present invention;

[0039] Figure 5 This is a schematic diagram of the filter frame structure of the present invention;

[0040] Figure 6 This is a top sectional view of the internal structure of the detection box of the present invention;

[0041] Figure 7 This is a cross-sectional schematic diagram of the skeleton, the three-coil differential inductive sensor, and the connection structure of the detection tube of the present invention;

[0042] Figure 8 This is a top sectional view of the internal structure of the optical particle imaging module of the present invention;

[0043] Figure 9 This is a schematic diagram of the pad structure of the present invention.

[0044] The components include: 1. Base plate; 2. Lower housing; 3. Upper housing; 4. Input shaft; 5. Output shaft; 6. Circulating oil pump; 601. First connecting pin; 602. Oil outlet pipe; 603. Oil delivery pipe; 7. Detection box; 701. Second connecting pin; 8. Filter box; 801. Third connecting pin; 9. Control box; 901. Display screen; 902. Status indicator light; 903. Buzzer; 10. Detection tube; 11. First branch pipe; 12. Connecting pipe; 3. Panel; 1301. Filter frame; 1302. Filter element; 14. Oil return pipe; 15. Viscosity detection module; 16. Frame; 17. Electrically controlled diverter valve; 18. Fixing frame; 19. Second branch pipe; 20. Miniature CMOS camera; 21. Pulse LED light source; 22. Sapphire glass; 23. Pad; 24. Oil passage hole; 25. Three-coil differential inductor sensor; 26. Encapsulating glue; 27. Adjusting sleeve; 28. Locking nut. Detailed Implementation

[0045] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0046] Example

[0047] like Figures 1 to 9 As shown, to achieve real-time, multi-dimensional monitoring of the reducer's operating status, this embodiment includes a reducer body and an intelligent monitoring and self-cleaning system. The reducer body includes a base plate 1, a housing composed of a lower housing 2 and an upper housing 3, an input shaft 4, an output shaft 5, and a reduction gear set disposed within the housing. An oil reservoir is formed within the housing. An oil outlet and an oil return port are respectively disposed on the front and rear side walls of the lower housing 2, wherein the oil outlet is located on the front wall of the lower housing 2 and is close to the lower position of the oil reservoir, and the oil return port is located on the rear wall of the lower housing 2 and is above the oil surface in the oil reservoir. The intelligent monitoring and self-cleaning system is disposed on the base plate 1 and includes a circulating oil pump 6 fixed by a first connecting foot 601, a detection box 7 fixed by a second connecting foot 701, a filter box 8 fixed by a third connecting foot 801, and a control box 9 fixedly connected to the side wall of the detection box 7. Through the above structure, a closed-loop monitoring and purification oil circuit is constructed that is independent of the reducer body but closely connected to it, providing a physical basis for realizing online, uninterrupted condition monitoring and proactive maintenance, and solving the problems of low integration and reliance on external equipment in existing monitoring systems.

[0048] To construct a stable and controllable monitoring oil circuit, the inlet of the circulating oil pump 6 is connected to the oil outlet of the lower housing 2 via an oil outlet pipe 602, and its outlet is equipped with an oil delivery pipe 603. Inside the detection box 7, there is a horizontal detection pipe 10, the front end of which serves as an inlet penetrating the front wall of the detection box 7 and connecting to the oil delivery pipe 603. When the circulating oil pump 6 starts, it continuously pumps lubricating oil from the oil storage tank into the detection pipe 10, providing a stable flow of oil sample for subsequent online analysis, ensuring the continuity and representativeness of the monitoring data.

[0049] To achieve basic monitoring of the physicochemical properties and abrasive particle concentration of lubricating oil, a viscosity detection module 15 and an inductive abrasive particle detection module are sequentially arranged on the outer wall of the detection tube 10 from front to back. The viscosity detection module 15 is a MEMS viscosity-temperature sensor chip encapsulated within a stainless steel threaded sleeve. Its probe penetrates the outer wall of the detection tube 10 and extends into the interior, directly sensing the viscosity and temperature of the oil. The inductive abrasive particle detection module includes a frame 16 fitted onto the outer wall of the detection tube 10 and a three-coil differential inductive sensor 25 disposed between the inner wall of the frame 16 and the outer wall of the detection tube 10. The frame 16 and the detection tube 10 are encapsulated and fixed together with potting compound 26. Through this structure, the viscosity detection module 15 can monitor the health status of the oil in real time (such as oxidation, dilution, or contamination), while the three-coil differential inductive sensor 25 can sensitively detect the concentration and size distribution of ferromagnetic abrasive particles in the oil, achieving a preliminary and rapid assessment of equipment wear conditions. This solves the problem of traditional monitoring methods being singular and unable to comprehensively assess the condition of the oil and equipment.

[0050] To achieve precise diagnosis and particle morphology analysis of abnormal wear, the rear end of the detection tube 10 is located inside the detection box 7, and an electrically controlled diverter valve 17 is fixedly connected to its rear end. This electrically controlled diverter valve 17 is a two-position three-way solenoid valve, with its inlet connected to the detection tube 10, its first outlet connected to the first branch pipe 11, and its second outlet connected to the second branch pipe 19. An optical particle imaging module is fixedly connected to the rear end of the second branch pipe 19. This module includes a fixing frame 18, a miniature CMOS camera 20, a pulsed LED light source 21, and two sets of sapphire glass 22. The fixing frame 18 has a cavity, and the two sets of sapphire glass 22 are respectively disposed on the left and right inner walls of the cavity, separated by a spacer 23 to form a flat observation pool. An oil passage hole 24 is provided on the spacer 23. The miniature CMOS camera 20 and the pulsed LED light source 21 are adjustablely mounted on the left and right sides of the fixed frame 18 via a set of adjusting sleeves 27, with their axes coinciding with the center line of the observation pool. The adjusting sleeves 27 are locked in their relative positions to the fixed frame 18 by locking nuts 28. The second branch pipe 19 is connected to the front wall of the fixed frame 18 and communicates with the interior of the observation pool. The outlet of the optical particle imaging module is connected to the filter box 8 via a connecting pipe 12. When the inductive abrasive particle detection module detects an abnormality, the electrically controlled diverter valve 17 switches the oil circuit, allowing the oil to flow through the observation pool. Under the illumination of the pulsed LED light source 21, the miniature CMOS camera 20 takes high-definition pictures of the flowing abrasive particles, thereby achieving accurate identification of the abrasive particle shape (such as cutting, flake, or spherical), solving the fundamental problem of only being able to monitor abrasive particle concentration but not be able to determine the wear mechanism.

[0051] To achieve online self-purification and convenient maintenance of the lubricating oil, the rear end of the first branch pipe 11 penetrates the rear wall of the detection box 7 and connects to the filter box 8. Similarly, the connecting pipe 12 penetrates the rear wall of the detection box 7 and connects to the filter box 8. The filter box 8 contains a filter module, which includes a filter frame 1301 slidably inserted into the filter chamber of the filter box 8, a filter element 1302 disposed within the filter frame 1301, and a panel 13 fixed to the outside of the filter frame 1301 to seal the filter chamber. The outlet of the filter box 8 is connected to the oil return port on the lower housing 2 via a return oil pipe 14. Through this structure, regardless of which branch the oil flows through, it must ultimately be purified by the filter element 1302 before returning to the reducer body, achieving active cleaning during the monitoring process. When the filter element needs to be replaced, simply loosen the fixing screws of the panel 13 to pull out the entire filter module. Maintenance is simple and quick, solving the drawbacks of existing technologies that lack online purification capabilities and require shutdown for maintenance.

[0052] To achieve intelligent control and human-machine interaction, the control box 9 is equipped with a central controller. A display screen 901, status indicator lights 902, and a buzzer 903 are located on the side furthest from the detection box 7. The central controller is electrically connected to the circulating oil pump 6, viscosity detection module 15, three-coil differential inductor sensor 25, electrically controlled diverter valve 17, miniature CMOS camera 20, and pulsed LED light source 21. The central controller is programmed to: receive and process sensor signals in real time; when the signal from the three-coil differential inductor sensor 25 exceeds a preset threshold, control the electrically controlled diverter valve 17 to switch to the second outlet path and activate the miniature CMOS camera 20 and pulsed LED light source 21 for image acquisition and analysis; based on the multi-sensor data fusion results, output equipment health status diagnosis and maintenance suggestions through the display screen 901, status indicator lights 902, and buzzer 903. The image analysis program of the central controller includes a particle morphology recognition model based on machine learning algorithms, used to classify and identify the acquired abrasive particle images. Through the above control logic, the system can automatically complete the entire process from basic monitoring and anomaly judgment to precise diagnosis, and transform complex monitoring results into intuitive diagnostic conclusions and alarms, effectively reducing the technical threshold of equipment maintenance and the reliance on the experience of professional personnel, and realizing true predictive intelligent maintenance.

[0053] Working principle: Under the unified management of the central controller, the intelligent monitoring and self-purification system mainly operates in two modes: normal monitoring and abnormal diagnosis.

[0054] The normal monitoring mode works as follows: The central controller (which can be an embedded industrial control board based on ARM architecture, such as Raspberry Pi Compute Module 4) first starts the circulating oil pump 6 (which can be a TACMINAS ZF series micro magnetic gear pump). Lubricating oil is drawn out from the oil outlet on the front wall of the reducer lower housing 2, flows sequentially through the oil outlet pipe 602, the circulating oil pump 6, and the oil delivery pipe 603, and enters the horizontal detection pipe 10 in the detection box 7. When the oil flows through the detection pipe 10, the viscosity detection module 15 (which can be an AMETEK Brookfield Viscosphere series MEMS chip) senses the viscosity and temperature of the oil in real time, while the three-coil differential inductive sensor 25 continuously monitors the concentration and size of ferromagnetic abrasive particles in the flowing oil. Subsequently, the oil passes through the electrically controlled diversion valve 17 (which can be an SMCVQ211U series two-position three-way solenoid valve) in the default position, and enters the filter box 8 through the first branch pipe 11. Inside the filter box 8, the oil passes through the filter element 1302 (with an accuracy of not less than 10µm, conforming to ISO2941 standard), where impurities and abrasive particles are filtered out. The clean oil is then returned to the oil storage tank via the return oil pipe 14 from the return port on the rear wall of the lower housing 2. Through this workflow, the system continuously provides clean lubricating oil to the reducer without manual intervention and establishes baseline data on the equipment's health status, realizing the transformation from "periodic maintenance" to "routine monitoring."

[0055] The working principle of its abnormal diagnosis mode is as follows: The central controller processes the signal of the three-coil differential inductive sensor 25 in real time. When the concentration of abrasive particles of a specific size (e.g., >100µm) exceeds the preset threshold, the system determines that there is a risk of abnormal wear. At this time, the controller immediately changes the passage of the electronically controlled diverter valve 17 to the trigger position and simultaneously starts the pulsed LED light source 21 and the miniature CMOS camera 20 (FLIR Blackfly S series board-level camera can be selected). The oil then changes its flow direction and enters the observation pool of the optical particle imaging module through the second branch pipe 19. The abrasive particles in the oil flowing through the flat observation pool are illuminated by the instantaneous high-brightness backlight emitted by the pulsed LED light source 21 and are captured by the miniature CMOS camera 20 at high speed. The captured particle images are sent to the central controller, which analyzes them using its built-in particle morphology recognition model based on machine learning algorithms (such as convolutional neural networks) and automatically classifies the abrasive particles into typical shapes such as cutting, flake, or spherical. Through the above diagnostic process, the system not only detected abnormalities but also accurately identified the type of wear (such as cutting wear indicating scoring and spherical particles indicating bearing fatigue), providing a decisive scientific basis for maintenance decisions and fundamentally solving the pain point of traditional monitoring that "only reports abnormalities but not causes".

[0056] The central controller operates as follows: It continuously collects viscosity, temperature, and abrasive data, and integrates this data with particle morphology information obtained in diagnostic mode. For example, if the system simultaneously detects "normal viscosity, a sudden increase in the concentration of large abrasive particles, and mostly flake-like abrasive particles," it can diagnose "fatigue pitting on the gear tooth surface." In this case, the controller drives the yellow status indicator 902 on the control box 9 to flash slowly, displays a clear diagnostic conclusion "gear fatigue wear warning" on the display screen 901, and recommends planned inspections. If the situation worsens, the controller will drive the red status indicator 902 to remain constantly lit and activate the buzzer 903 to issue a continuous alarm. Through this intelligent decision-making and human-machine interaction mechanism, the system transforms complex multi-source sensor data into highly intuitive operating instructions and maintenance suggestions, significantly reducing reliance on on-site operators' technical skills and achieving intelligent and accessible equipment health management.

[0057] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A reducer gear power transmission device characterized by: include: The reducer body includes a base plate (1), a housing consisting of a lower housing (2) and an upper housing (3), an input shaft (4), an output shaft (5), and a reduction gear set disposed in the housing. An oil storage pool is formed inside the housing, and an oil outlet and an oil return port are respectively provided on the front and rear side walls of the lower housing (2). The intelligent monitoring and self-purification system is installed on the base plate (1). The intelligent monitoring and self-purification system includes a circulating oil pump (6) fixed by the first connecting foot (601), a detection box (7) fixed by the second connecting foot (701), a filter box (8) fixed by the third connecting foot (801), and a control box (9). The inlet of the circulating oil pump (6) is connected to the oil outlet through the oil outlet pipe (602), and its outlet is provided with an oil delivery pipe (603). The detection box (7) is equipped with a horizontal detection tube (10) inside. The front end of the detection tube (10) serves as an inlet, penetrating the front wall of the detection box (7) and connecting to the oil delivery pipe (603). The outer wall of the detection tube (10) is provided with the following components from front to back: The viscosity detection module (15) is a MEMS viscosity-temperature sensor chip encapsulated in a stainless steel threaded sleeve, and its probe penetrates the outer wall of the detection tube (10) and extends into the interior. The inductive abrasive detection module includes a frame (16) sleeved on the outer wall of the detection tube (10) and a three-coil differential inductive sensor (25) disposed between the inner side wall of the frame (16) and the outer wall of the detection tube (10). The frame (16) and the detection tube (10) are encapsulated and fixed by potting compound (26). The rear end of the detection tube (10) is located inside the detection box (7), and an electrically controlled diverter valve (17) is fixedly connected to its rear end. The electrically controlled diverter valve (17) is controlled by the central controller and is used to switch the oil circuit according to the detection result of the inductive abrasive particle detection module. The two outlets of the electrically controlled diverter valve (17) are respectively connected to a first branch pipe (11) and a second branch pipe (19). The rear end of the first branch pipe (11) penetrates the rear wall of the detection box (7) and is connected to the filter box (8). The rear end of the second branch pipe (19) is fixedly connected to an optical particle imaging module, and the outlet of the optical particle imaging module is connected to the filter box (8) through a connecting pipe (12). The connecting pipe (12) penetrates the rear wall of the detection box (7). The outlet of the filter box (8) is connected to the oil return port on the lower housing (2) through the oil return pipe (14); The control box (9) is fixedly connected to the side wall of the detection box (7). The control box (9) is equipped with a central controller, which is electrically connected to the circulating oil pump (6), the detection box (7), the electronically controlled diverter valve (17), and the optical particle imaging module.

2. A speed reducer gear power transmission device according to claim 1, wherein The electrically controlled diverter valve (17) is a two-position three-way solenoid valve, fixed at the rear end of the detection tube (10). Its inlet is connected to the detection tube (10), its first outlet is connected to the first branch tube (11), and its second outlet is connected to the second branch tube (19).

3. A reduction gear power transmission device according to claim 2, characterised in that, The optical particle imaging module includes a fixed frame (18), a miniature CMOS camera (20), a pulsed LED light source (21), and two sets of sapphire glass (22). The fixed frame (18) has a cavity, and the two sets of sapphire glass (22) are respectively set on the left and right inner walls of the cavity, and are separated by a pad (23) to form a flat observation pool. The pad (23) has an oil hole (24). The miniature CMOS camera (20) and the pulsed LED light source (21) are respectively adjustablely mounted on the left and right sides of the fixed frame (18) by a set of adjustment sleeves (27), and the axes of the two coincide with the center line of the observation pool; the adjustment sleeves (27) are locked in relative position with the fixed frame (18) by locking nuts (28).

4. A speed reducer gear power transmission device according to claim 1, wherein The filter box (8) is provided with a filter module inside. The filter module includes a filter frame (1301) that is slidably inserted into the filter cavity of the filter box (8), a filter core (1302) that is disposed in the filter frame (1301), and a panel (13) that is fixed to the outside of the filter frame (1301) for sealing the filter cavity.

5. A speed reducer gear power transmission device according to claim 1, wherein The control box (9) is equipped with a display screen (901), a status indicator light (902) and a buzzer (903) on the side away from the detection box (7).

6. A gear power transmission device for a reducer according to claim 1, characterized in that, The oil outlet is located on the front wall of the lower shell (2) and near the lower position of the oil storage tank, while the oil return port is located on the rear wall of the lower shell (2) and above the liquid surface of the oil storage tank.

7. A gear power transmission device for a reducer according to claim 3, characterized in that, The second branch pipe (19) is connected to the front wall of the fixed frame (18) and is connected to the interior of the observation pool.

8. A reducer gear power transmission device according to any one of claims 2 to 7, characterized in that, The central controller is programmed to: Signals from the viscosity detection module (15) and the three-coil differential inductance sensor (25) are received and processed in real time. When the signal of the three-coil differential inductor sensor (25) exceeds the preset threshold, the electronically controlled shunt valve (17) is switched to the second outlet path, and the micro CMOS camera (20) and pulsed LED light source (21) are activated to acquire and analyze images. Based on the results of multi-sensor data fusion, the device health status diagnosis and maintenance suggestions are output through the display screen (901), status indicator (902) and buzzer (903).

9. A gear power transmission device for a reducer according to claim 8, characterized in that, The image analysis program of the central controller includes a particle morphology recognition model based on machine learning algorithms, which is used to classify and identify the acquired abrasive particle images.

Images