Multi-sensor fusion detection method and system for tail end of mechanical arm

Abstract

The invention provides a multi-sensor fusion detection method and system for the tail end of a mechanical arm, and belongs to the technical field of robots, and the method comprises the steps: carrying out the environment monitoring of a region to form a three-dimensional grid space, and obtaining the coordinate information of the mechanical arm, a to-be-detected target and an obstacle; planning a mechanical arm path by referring to the shapes and spatial position parameters of a to-be-detected target and an obstacle; the multiple sensors are all moved outwards from the interior of the mechanical arm, and the positions and angles of the multiple sensors are independently controlled, so that the multiple sensors act on the to-be-detected target for detection; adjusting the angle and position of the tail end of the mechanical arm to compensate the detection deviation according to the characteristics of the to-be-detected target in the grid space; and the tail end of the mechanical arm is controlled to move according to the shape of the to-be-detected target, so that the plurality of sensors sequentially act on each detection position of the to-be-detected target. According to the multi-sensor fusion detection method for the tail end of the mechanical arm, the dimensionality and accuracy of detection data are improved, and the detection difficulty is reduced.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of robots, and more particularly relates to a mechanical arm end multi-sensor fusion detection method and system. BACKGROUND

[0002] The mechanical arm is widely used in precise detection, complex environment operation and the like scenes. In industrial quality inspection and equipment maintenance, the mechanical arm needs to acquire multi-dimensional data (such as distance, temperature, deformation, electromagnetic characteristics and the like) in real time through sensors, so as to realize accurate positioning, defect identification or environment interaction. However, the existing mechanical arm end sensor integration scheme has significant limitations. First, the detection dimension is single. The mechanical arm end is mostly provided with a single type of sensor, which leads to insufficient detection data dimension. For example, in power equipment detection, only relying on single sensor detection of special high frequency or ultrasonic cannot accurately locate the partial discharge signal, and the mechanical arm or manual intervention needs to be replaced multiple times, which seriously restricts the detection efficiency and comprehensiveness. Moreover, when the mechanical arm enters a complex environment for detection, due to different shapes of the equipment to be detected and existence of obstacles, the detection operation of the mechanical arm is difficult and the accuracy is low. SUMMARY

[0003] The application aims to provide a mechanical arm end multi-sensor fusion detection method and system, so as to solve the technical problems of insufficient detection data dimension of the mechanical arm and the sensor, and large detection difficulty and low accuracy in a complex environment in the prior art.

[0004] To achieve the above-mentioned purpose, the technical scheme adopted by the application is as follows: a mechanical arm end multi-sensor fusion detection method is provided, comprising: S1: performing environment monitoring on a region where a target to be detected is located, and forming a three-dimensional grid space; acquiring coordinate information of the mechanical arm, the target to be detected and obstacles in the grid space; referring to shape and spatial position parameters of the target to be detected and the obstacles, planning a path for the mechanical arm to approach the target to be detected; S2: moving multiple sensors to the inside of the mechanical arm, and driving the end of the mechanical arm to correspond to the target to be detected; moving the multiple sensors from the inside of the mechanical arm to the outside, and independently controlling positions and angles of the multiple sensors, so that the multiple sensors act on the target to be detected to perform detection; S3: adjusting the angle and position of the end of the mechanical arm to compensate for detection deviation according to the characteristics of the target to be detected in the grid space; S4: moving the end of the mechanical arm according to the shape of the target to be detected, so that the multiple sensors act on each detection position of the target to be detected in turn, and each sensor compensates for detection deviation when detecting.

[0005] In a possible implementation, the mechanical arm end is provided with a ring-shaped environment detector near the sensor, which is used to capture dynamic environment information around the mechanical arm end and the sensor in real time; the dynamic environment information generated by the ring-shaped environment detector is used to trigger the grid space data in S2, when an obstacle is detected in the area around the sensor, the grid accuracy of the area around the sensor is automatically increased, and a signal is sent to the mechanical arm control system to make the path planning in S1 increase the local detour sub-path.

[0006] In a possible implementation, when the plurality of sensors in S2 move outward from the inside of the mechanical arm, the ring-shaped environment detector monitors the positions of the plurality of sensors, when the sensors extend by 0-50%, the ring-shaped environment detector scans whether there is a jam in the channel inside the mechanical arm in a low-frequency mode; when extending by 50%-100%, it is switched to high-frequency scanning of the external environment, and the detection data is used to correct the extension trajectory of the sensor in real time to avoid rigid contact with the target surface or sudden obstacles.

[0007] In a possible implementation, in S2, a plurality of sensors are integrated in the concentric sleeve at the end of the mechanical arm, and a umbrella linkage mechanism is installed in the concentric sleeve; when extended, the umbrella linkage mechanism is used to realize radial diffusion to distribute the sensor array; each sensor can rotate around its own axis independently.

[0008] In a possible implementation, in S3, the deviation compensation is a double closed-loop correction, wherein the inner closed loop is used for sub-millimeter level position correction through the micro-displacement platform at the end of the mechanical arm, and the outer closed loop is used for cumulative error compensation by adjusting the offset of the spatial coordinate system of the base joint of the mechanical arm; the inner closed loop and the outer closed loop adopt a fuzzy PID algorithm to realize dynamic switching.

[0009] In a possible implementation, in S2, the plurality of sensors include a transient ground voltage sensor, an ultra-high frequency sensor, and an ultrasonic sensor; the transient ground voltage sensor is used to detect the transient ground voltage change of the target to be detected; the ultra-high frequency sensor is used to collect partial discharge signals, and the ultra-high frequency sensor and the ultrasonic sensor work alternately; the mechanical arm is provided with a swing rod for applying coupling agent to the probe of the ultrasonic sensor, and the mechanical arm is provided with a containing groove for accommodating the swing rod.

[0010] In a possible implementation, in S1, the environment monitoring uses a laser radar and a vision sensor to work cooperatively, the laser radar acquires distance information of objects in the region, the vision sensor acquires texture features of the objects, and a three-dimensional grid space containing object surface details is generated after fusion of the laser radar and the vision sensor.

[0011] In a possible implementation, in S1, the environmental monitoring adds electromagnetic compatibility detection, real-time monitoring of electromagnetic interference is performed through a spectrum analyzer, and the electromagnetic shield is started when the interference intensity exceeds a threshold.

[0012] In a possible implementation, the method further comprises: S5: after detection is completed, calibration is performed on the plurality of sensors, the plurality of sensors are moved to a preset position of a standard calibration block through the mechanical arm, a deviation of an actual detection value from a theoretical value is compared, a correction coefficient of each sensor is automatically generated, and the correction coefficient is used for initial parameter setting in next detection.

[0013] The mechanical arm end multi-sensor fusion detection method provided by the application has the following advantages: compared with the prior art, the mechanical arm end multi-sensor fusion detection method provided by the application first performs environmental monitoring and path planning, performs three-dimensional grid space modeling on a target region to be detected, accurately obtains coordinate information of a mechanical arm, a target and an obstacle, plans a path that avoids the obstacle and efficiently approaches the target in combination with a shape and a spatial position parameter of the target and the obstacle, and solves the problem that the mechanical arm is difficult to approach the target due to the obstacle in a complex environment, thereby laying a foundation for subsequent detection. Then, multi-sensor deployment and detection are performed, the plurality of sensors are moved into the mechanical arm, the mechanical arm end is driven to be aligned with the target to be detected, then the sensors are moved out of the mechanical arm, and the position and the angle of each sensor are independently controlled, so that the sensors accurately act on the target for detection. This step breaks through the limitation of a single sensor, and through the cooperative work of the plurality of sensors, multi-dimensional data such as distance, temperature and deformation can be obtained at the same time. For example, in power equipment detection, a plurality of sensing modes such as very high frequency and ultrasonic can be used at the same time, without the need for multiple replacement of the mechanical arm or manual intervention, so that the detection efficiency and comprehensiveness are greatly improved. Then, deviation compensation is performed, the angle and the position of the mechanical arm end are adjusted in real time according to the characteristics of the target to be detected in the grid space, so as to offset the deviation that may occur in the detection process, and ensure the accuracy of the detection. Finally, full-range detection is performed, the mechanical arm end is controlled to follow the shape of the target to be detected, so that the plurality of sensors act on each detection position of the target in turn, and each sensor performs deviation compensation when detecting. This step ensures omnidirectional detection of a special-shaped target, and even in the face of equipment with different shapes, the moving track that adapts to the target shape and continuous deviation compensation can be used, so that the operation difficulty is reduced and the detection accuracy is improved.

[0014] Another object of the application is to provide a mechanical arm end multi-sensor fusion detection system.

[0015] The overhead line external environment interference processing system provided by the application adopts a mechanical arm end multi-sensor fusion detection method, and effectively solves the limitations of the existing scheme by means of environment modeling and path planning, multi-sensor collaborative detection, real-time deviation compensation, and full-range detection of the target shape, and significantly improves the efficiency, comprehensiveness and accuracy of the mechanical arm in precise detection and complex environment operation. BRIEF DESCRIPTION OF DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the application, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are only some embodiments of the application, and other drawings can also be obtained by those skilled in the art without creative labor.

[0017] Figure 1 The flowchart of the mechanical arm end multi-sensor fusion detection method provided by the embodiments of the application. DETAILED DESCRIPTION

[0018] In order to make the technical problems to be solved by the application, the technical solutions and the beneficial effects more clearly, the following will further describe the application in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the application, and are not used to limit the application.

[0019] It should be noted that when an element is referred to as being "fixed to" or "disposed on" another element, it can be directly on the other element or indirectly on the other element. When an element is referred to as being "connected to" another element, it can be directly connected to the other element or indirectly connected to the other element.

[0020] It should be understood that the terms "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only used to facilitate the description of the application and simplify the description, and therefore cannot be understood as indicating or implying that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as limiting the application.

[0021] In addition, the terms "first", "second" are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the technical features indicated. Therefore, the features defined with "first", "second" can explicitly or implicitly include one or more of the features. In the description of the application, the meaning of "multiple" is two or more, unless otherwise specifically limited.

[0022] With reference to Figure 1 , the mechanical arm end multi-sensor fusion detection method provided by the present application will be described. A mechanical arm end multi-sensor fusion detection method, comprising: S1: monitoring the environment of the region where the to-be-detected target is located, and forming a three-dimensional grid space; obtaining the coordinate information of the mechanical arm, the to-be-detected target and the obstacles in the grid space; referring to the shape, spatial position parameters of the to-be-detected target and the obstacles, planning the path of the mechanical arm approaching the to-be-detected target; S2: controlling multiple sensors to move to the inside of the mechanical arm, and driving the end of the mechanical arm to correspond to the to-be-detected target; moving the multiple sensors from the inside of the mechanical arm to the outside, and independently controlling the position and angle of the multiple sensors, so that the multiple sensors act on the to-be-detected target to perform detection; S3: adjusting the angle and position of the end of the mechanical arm to compensate for detection deviation according to the characteristics of the to-be-detected target in the grid space; S4: controlling the end of the mechanical arm to move according to the shape of the to-be-detected target, so that the multiple sensors act on each detection position of the to-be-detected target in turn, and each sensor performs compensation for detection deviation when detecting.

[0023] The mechanical arm end multi-sensor fusion detection method provided by the application, compared with the prior art, first carries out environment monitoring and path planning, carries out three-dimensional grid space modeling on the target area to be detected, accurately obtains the coordinate information of the mechanical arm, the target and the obstacle, plans a path that avoids obstacles and efficiently approaches the target in combination with the shape, spatial position parameters and the like of the target and the obstacle, which solves the problem that the mechanical arm is difficult to approach the target due to obstacles in a complex environment, and lays a foundation for subsequent detection. Then multi-sensor deployment and detection, first move multiple sensors into the mechanical arm, then drive the mechanical arm end to align with the target to be detected, then move the sensors out of the mechanical arm, and independently control the position and angle of each sensor to accurately act on the target for detection. This step breaks through the limitation of a single sensor, and through the cooperative work of multiple sensors, multi-dimensional data such as distance, temperature and deformation can be obtained at the same time, for example, in power equipment detection, multiple sensing methods such as very high frequency and ultrasonic can be used at the same time, without the need for multiple changes of the mechanical arm or manual intervention, which greatly improves the detection efficiency and comprehensiveness. Then carry out deviation compensation, adjust the angle and position of the mechanical arm end in real time according to the characteristics of the target to be detected in the grid space, offset the deviation that may occur during detection, and ensure the accuracy of detection. Finally, full-range detection, control the mechanical arm end to follow the shape of the target to be detected, so that multiple sensors act on each detection position of the target in turn, and each sensor carries out deviation compensation when detecting. This step ensures the omnidirectional detection of the target to be detected, even for devices with different shapes, through the adaptive target shape moving track and continuous deviation compensation, the operation difficulty is reduced, and the detection accuracy is improved.

[0024] Through this method, with the help of environment modeling and path planning, multi-sensor cooperative detection, real-time deviation compensation and adaptive target shape full-range detection, the dimension of the detection data is improved, and the efficiency, comprehensiveness and accuracy of the mechanical arm in precision detection and complex environment operation are significantly improved.

[0025] Please refer to Figure 1, as a specific embodiment of the mechanical arm end multi-sensor fusion detection method provided by the application, the area close to the sensor of the mechanical arm end is provided with a ring-shaped environment detector, and the ring-shaped environment detector is used to capture dynamic environment information around the mechanical arm end and the sensor in real time; the dynamic environment information generated by the ring-shaped environment detector in S2 is used to trigger grid space data, when obstacles appear in the area around the sensor, the grid accuracy of the area around the sensor is automatically increased, and a signal is sent to the mechanical arm control system to increase the local bypass sub-path in the path planning in S1; by setting the ring-shaped environment detector in the area close to the sensor of the mechanical arm end, a dynamic environment monitoring and grid space linkage mechanism is formed, which has the characteristics of real-time response and accurate adaptation. The ring-shaped environment detector can continuously capture dynamic environment information around the mechanical arm end and the sensor, and in S2, the information generated thereby is directly used to trigger grid space data update: when obstacles appear around the sensor, the system will automatically increase the grid accuracy of the area, and send a signal to the control system to prompt the path planning in S1 to increase the local bypass sub-path.

[0026] This way is to dynamically adapt to complex environmental changes. Although the original path planning is based on the initial grid space, sudden obstacles may occur in actual detection. The real-time monitoring of the ring-shaped detector can quickly identify such situations, improve the local grid accuracy, make the space data of the obstacle area more detailed, provide accurate basis for the local bypass sub-path, and avoid collision between the sensor and the obstacle due to environmental mutation. At the same time, it is not necessary to interrupt the whole detection process to re-plan the global path, but only to adjust locally to deal with unexpected situations, which not only ensures the continuity of detection, but also improves the adaptability and safety of the mechanical arm in complex dynamic environment, further optimizes the flexibility of path planning and the stability of detection process, so that the multi-sensor fusion detection can still be efficiently promoted under dynamic interference.

[0027] Please refer to Figure 1As a specific embodiment of the multi-sensor fusion detection method provided by the application, when the plurality of sensors in S2 move from the inside of the mechanical arm to the outside, the annular environment detector monitors the positions of the plurality of sensors, when the sensors extend 0-50% of the stroke, the annular environment detector scans whether there is a jam in the internal channel of the mechanical arm in a low-frequency mode; when extending 50%-100% of the stroke, it is switched to high-frequency scanning of the external environment, and the detection data is used to correct the extension trajectory of the sensor in real time to avoid rigid contact with the target surface or sudden obstacles. Based on the segmented regulation of the extension stroke of the sensor, the scanning mode of the annular environment detector is regulated to realize dynamic adaptation monitoring. When the sensor extends 0-50% of the stroke, the annular environment detector scans the internal channel of the mechanical arm at a low frequency, and focuses on checking the jam hidden danger; when extending 50%-100% of the stroke, it is switched to high-frequency scanning of the external environment, and the extension trajectory is corrected in real time. By scanning at a low frequency inside the mechanical arm, the invalid energy consumption can be reduced, the jam risk of the internal channel of the mechanical arm can be accurately observed, and the smoothness of the sensor extension can be ensured; by scanning at a high frequency outside the mechanical arm, the target surface and sudden obstacles can be sharply captured, the rigid contact can be avoided through real-time trajectory correction, and the sensor and the target to be detected can be protected. This segmented monitoring strategy not only improves the safety of the sensor extension process, but also optimizes the energy consumption and detection efficiency, and enhances the reliability of multi-sensor fusion detection.

[0028] Please refer to Figure 1As a specific embodiment of the multi-sensor fusion detection method provided by the application, in S2, a mechanical arm end set is arranged with a concentric sleeve and a umbrella linkage mechanism installed on the concentric sleeve; a plurality of sensors are integrated in the concentric sleeve of the mechanical arm end; radial diffusion is realized by the umbrella linkage mechanism when the sensors are extended, so that the sensors are distributed in an array; and each sensor is independently rotated around its own axis. With the combined design of the concentric sleeve and the umbrella linkage mechanism, the multi-sensor is realized orderly stored and flexible deployed. The concentric sleeve of the mechanical arm end provides integrated storage space for the sensors, ensuring that the sensors are compactly stored in the arm when not working; when extended, the umbrella linkage mechanism drives the sensors to diffuse radially, forming an array distribution, and each sensor can independently rotate around its own axis. The concentric sleeve realizes the centralized storage of the sensors, reduces the volume of the mechanical arm end, and facilitates movement in narrow spaces; the radial diffusion of the umbrella linkage mechanism makes the sensor array cover a larger detection range, adapting to different sizes of the target to be detected; the independent rotation function allows each sensor to adjust the angle as needed, accurately aligning the target detection point, meeting the multi-dimensional data acquisition demand. For example, in complex curved surface detection, the array distribution expands the coverage range, and the independent rotation ensures that each sensor adheres to the curvature of the surface, without the need to frequently adjust the overall posture of the mechanical arm, which not only improves the detection efficiency, but also reduces the operation complexity, while avoiding mutual interference between the sensors, ensuring the accuracy and comprehensiveness of data acquisition, and further enhancing the adaptability and accuracy of multi-sensor fusion detection.

[0029] The concentric sleeve adopts a multi-layer nested structure, including a main sleeve, a middle layer adjusting cylinder and an outer layer protection cylinder from inside to outside, the wall of each layer is made of high-strength lightweight alloy material, and the inner wall is provided with a wear-resistant coating. The main sleeve is fixed to the core shaft of the mechanical arm end, the outer wall is uniformly distributed with 3-6 groups of axial guide grooves, and the ball sliding blocks are embedded in the grooves; the inner side of the middle layer adjusting cylinder is correspondingly provided with a matching sliding rail, the axial extension and contraction are realized through the sliding cooperation of the sliding block and the guide groove, and the wall is provided with radial through holes for the sensor connecting wire harness to pass through; the outer layer protection cylinder is a hollow grid structure, which has the functions of lightweight and anti-collision, and the end edge is provided with an elastic buffer ring.

[0030] The umbrella linkage mechanism takes the rotating drive disc at the end of the main sleeve as the core, including linkage rods, sensor mounting seats and elastic reset components. The drive disc is coaxial with the main sleeve and can be controlled to rotate by a servo motor, the edge is hinged with a plurality of main linkage rods, the middle part of each main linkage rod is connected with a secondary linkage rod through a hinge shaft, the other end of the secondary linkage rod is hinged with a fixed lug on the outer wall of the middle layer adjusting cylinder; the sensor mounting seat is in the shape of an arc plate, one end is hinged with the end of the main linkage rod, the other end is connected with the sensor through a ball hinge, and an angle adjusting motor is integrated in the seat body; the elastic reset component is a torsion spring sleeved on the hinge shaft of the main linkage rod and the drive disc, which ensures automatic reset when the mechanism is retracted.

[0031] When the driving disc rotates clockwise, the main linkage rod swings with it and is stretched outwards under the restraint of the secondary linkage rod, while the middle layer adjusting cylinder is synchronously extended axially along the main sleeve, driving the sensor mounting seat to diffuse radially; on the contrary, when the driving disc rotates counterclockwise, the mechanism contracts, and the sensor is stored in the concentric sleeve. This structure realizes the synchronous extension and array distribution of the sensor through mechanical linkage, and can flexibly adapt to different detection scenes by cooperating with the independent angle adjustment function.

[0032] Please refer to Figure 1 As a specific embodiment of the multi-sensor fusion detection method provided by the application, the deviation compensation in S3 is double closed loop correction, wherein the inner closed loop corrects the position at sub-millimeter level through the micro-displacement platform at the end of the mechanical arm, and the outer closed loop realizes cumulative error compensation by adjusting the offset of the spatial coordinate system of the base joint of the mechanical arm. The inner closed loop and the outer closed loop realize dynamic switching through the fuzzy PID algorithm. The inner closed loop takes the micro-displacement platform at the end of the mechanical arm as the core executive component. The platform integrates a piezoelectric ceramic driving module and can realize sub-millimeter micro-adjustment of X, Y and Z axes. The displacement accuracy is fed back in real time by the end sensor. When the sensor detects the local position deviation from the target to be detected, the signal is directly transmitted to the inner closed loop controller. The controller quickly calculates the adjustment amount through the fuzzy PID algorithm, drives the micro-displacement platform to complete the correction within 10 ms, and ensures that the sensor detection point is always aligned with the target feature area. The outer closed loop is associated with the base joint of the mechanical arm. The error sources include joint gap accumulation and positioning drift after long distance movement. The outer closed loop performs comprehensive calibration through the laser tracker installed on the base of the mechanical arm and the global environmental features collected by the sensor: when the cumulative deviation detected by the sensor exceeds 1 mm, the outer closed loop controller is started, and the overall error is controlled within 0.5 mm by adjusting the spatial coordinate system offset of the base joint.

[0033] When the inner closed loop and the outer closed loop work cooperatively, the position data output by the sensor in real time is used as the common input of the double closed loop: when the deviation is small (<1 mm), the fuzzy PID algorithm preferentially activates the inner closed loop, and uses the micro-displacement platform for high-frequency micro-adjustment to avoid the response delay caused by frequent movement of the base joint; when the deviation is too large (≥1 mm), the algorithm automatically switches to the outer closed loop dominated, and eliminates systematic errors by adjusting the base joint, while the inner closed loop assists in correcting the local residual deviation. This mechanism not only takes advantage of the fast response of the micro-displacement platform, but also solves the problem of cumulative error in long-term operation by adjusting the base joint, so that the mechanical arm can maintain high-precision detection state in complex environment.

[0034] Please refer to Figure 1As a specific embodiment of the mechanical arm end multi-sensor fusion detection method provided by the application, in S2, the plurality of sensors include a transient ground voltage sensor, an ultra-high frequency sensor, and an ultrasonic sensor. The transient ground voltage sensor is used to detect transient ground voltage changes of a target to be detected. The ultra-high frequency sensor collects partial discharge signals, and the ultra-high frequency sensor and the ultrasonic sensor work alternately. A swing rod for smearing coupling agent on a probe of the ultrasonic sensor is arranged on the mechanical arm, and a containing groove for containing the swing rod is arranged on the mechanical arm. The method is aimed at partial discharge detection of power equipment and other targets, and three types of sensors of transient ground voltage, ultra-high frequency, and ultrasonic are configured, and an ultra-high frequency and ultrasonic sensor alternation working mechanism is designed, and a coupling agent smearing swing rod with a containing groove is provided. The transient ground voltage sensor can capture transient ground voltage changes of the target, the ultra-high frequency sensor can accurately collect partial discharge signals, and the two work with the ultrasonic sensor to realize electric and acoustic multi-physical field signal fusion detection, greatly improving the defect recognition accuracy. The ultra-high frequency and ultrasonic sensor alternation working mechanism avoids signal interference and ensures data independence. The swing rod can automatically smear the coupling agent before the ultrasonic probe works, reduces acoustic energy loss, improves detection sensitivity, and the containing groove can contain the swing rod to avoid interference with the movement of the mechanical arm or the operation of the sensor. This configuration solves the problems of complex signals and multiple interferences in the partial discharge detection of power equipment, balances the detection comprehensiveness and operation convenience, and enhances the adaptability of the method to power detection scenarios.

[0035] Please refer to Figure 1, as a specific embodiment of the multi-sensor fusion detection method provided by the mechanical arm end of the application, in S1, the environment monitoring cooperates with laser radar and visual sensor, the laser radar obtains the distance information of the object in the region, the visual sensor obtains the texture characteristics of the object, and the laser radar and the visual sensor are fused to generate a three-dimensional grid space containing the surface details of the object. Through the cooperation of laser radar and visual sensor, the environment monitoring is realized, the laser radar focuses on obtaining the distance information of the object in the region, constructing the space skeleton, the visual sensor captures the texture characteristics of the object, and supplements the surface details. The three-dimensional grid space with spatial scale and surface texture is generated after fusion. With the help of laser radar, the distance and three-dimensional coordinates between objects can be accurately measured, providing a reliable spatial structure basis for the three-dimensional grid. Even in an environment with insufficient light, it can work stably and ensure the geometric accuracy of the grid space; The visual sensor can capture information such as color, texture, contour details, etc. of the object, so that the grid space not only contains position information, but also presents the surface characteristics of the target and obstacles to be detected, such as equipment nameplate, defects, obstacle material texture, etc. This fusion method avoids the limitations of a single sensor. Laser radar lacks detailed description ability, and visual sensor is easily affected by light and lacks distance measurement accuracy. The three-dimensional grid space after fusion provides more abundant environmental information for subsequent path planning, and the mechanical arm can more accurately identify the target shape and obstacle characteristics, plan a path that is more consistent with the actual environment, especially in complex scenes, which can greatly reduce the risk of collision between the mechanical arm and the obstacle, and improve the accuracy of approaching the target.

[0036] Please refer to Figure 1 , as a specific embodiment of the multi-sensor fusion detection method provided by the mechanical arm end of the application, in S1, the environment monitoring increases electromagnetic compatibility detection, which monitors electromagnetic interference in real time through a spectrum analyzer, and starts an electromagnetic shield when the interference intensity exceeds a threshold. That is, an electromagnetic compatibility detection link is added in the environment monitoring of S1, and an electromagnetic shield is automatically triggered when the interference intensity exceeds a preset threshold. The spectrum analyzer can accurately identify the frequency and intensity of electromagnetic interference, providing data support for interference evaluation, avoiding distortion of sensor detection data due to complex electromagnetic environment, for example, strong electromagnetic interference may affect the capture accuracy of partial discharge signals by a very high frequency sensor in power equipment detection. The timely start of the electromagnetic shield can form an isolation barrier, effectively attenuate external electromagnetic interference, and protect the signal acquisition quality of transient ground voltage, very high frequency and other sensors. This design makes up for the limitations of traditional environment monitoring, which only focuses on physical space information, optimizes the detection environment from the electromagnetic level, and ensures the reliability of multi-sensor fusion data.

[0037] Please refer to Figure 1As a specific embodiment of the mechanical arm end multi-sensor fusion detection method provided by the application, it further includes a sensor calibration link after detection is completed, multiple sensors are moved to a preset position of a standard calibration block by the mechanical arm, the deviation of the actual detection value of each sensor from the theoretical standard value is compared, a correction coefficient is automatically calculated and generated as the initial parameter for the next detection. First, through the reference data of the standard calibration block, the deviation degree of each sensor can be quantified, ensuring the long-term stability of the detection data. The transient voltage sensor has signal attenuation due to probe aging, and after calibration, such errors can be compensated by the correction coefficient. Second, the automatic calibration process does not require manual intervention, and the precise positioning of the mechanical arm ensures the consistency of the calibration position, avoiding the randomness of manual operation and improving the calibration efficiency and accuracy. Third, the correction coefficient is directly applied to the initial parameter setting of the next detection, forming a closed-loop mechanism of "detection-calibration-correction", so that the sensor is always in an optimal working state. Especially in the multi-sensor fusion scene, it can reduce the system error between different sensors and improve the reliability of data fusion. In addition, the preset position design of the standard calibration block adapts to multiple sensor types, meets the multi-dimensional calibration demand, and further enhances the universality of the scheme. This design fundamentally guarantees the persistence of sensor detection accuracy and provides a stable data foundation for long-term and high-frequency detection tasks.

[0038] It is not shown in the figure, and the embodiment of the application further provides a mechanical arm end multi-sensor fusion detection system. The mechanical arm end multi-sensor fusion detection system comprises the mechanical arm end multi-sensor fusion detection method of any one of the above.

[0039] The mechanical arm end multi-sensor fusion detection system comprises a control module and a computer connected with the control module. The control module is connected with a mechanical arm, a sensor, a ring environment detector, a umbrella bone linkage mechanism, a micro displacement platform, a laser radar, a vision sensor, etc. The data of the sensor, the ring environment detector, the umbrella bone linkage mechanism, the micro displacement platform, the laser radar, and the vision sensor are uploaded to the control module and then uploaded to the computer by the control module.

[0040] The mechanical arm end multi-sensor fusion detection system provided by the embodiment of the application adopts the mechanical arm end multi-sensor fusion detection method described above, and effectively solves the limitations of the existing scheme by means of environment modeling and path planning, multi-sensor collaborative detection, real-time deviation compensation, and full-range detection adapting to target shapes, significantly improving the efficiency, comprehensiveness and accuracy of the mechanical arm in precision detection and complex environment operation.

[0041] The above is only a preferred embodiment of the application and does not limit the application. Any modification, equivalent replacement and improvement made within the spirit and principle of the application shall be included in the protection scope of the application.

Claims

1. A multi-sensor fusion detection method for the end effector of a robotic arm, characterized in that, S1 includes: performing environmental monitoring on the area where the target to be detected is located and forming a three-dimensional grid space; acquiring the coordinate information of the robotic arm, the target to be detected, and obstacles in the grid space; and planning the path for the robotic arm to approach the target to be detected based on the shape and spatial position parameters of the target to be detected and obstacles. S2: Move multiple sensors into the robotic arm and align the end of the robotic arm with the target to be detected; move multiple sensors outward from the inside of the robotic arm and independently control the position and angle of each sensor so that they act on the target to be detected for detection; S3: Adjust the angle and position of the end of the robotic arm to compensate for detection deviations based on the characteristics of the target to be detected in the grid space. S4: Control the end effector of the robotic arm to move according to the shape of the target to be detected, so that multiple sensors act sequentially on each detection position of the target to be detected, and each sensor compensates for detection deviation during detection.

2. The multi-sensor fusion detection method for the end effector of a robotic arm as described in claim 1, characterized in that, A ring-shaped environmental detector is installed in the area near the sensor at the end of the robotic arm. The ring-shaped environmental detector is used to capture dynamic environmental information of the end of the robotic arm and the area around the sensor in real time. In S2, the dynamic environmental information generated by the ring-shaped environmental detector is used to trigger grid space data. When an obstacle is detected in the area around the sensor, the grid precision of the area around the sensor is automatically increased, and a signal is sent to the robotic arm control system so that the path planning in S1 adds a local detour sub-path.

3. The multi-sensor fusion detection method for the end effector of a robotic arm as described in claim 2, characterized in that, When multiple sensors in S2 move outward from inside the robotic arm, the ring-shaped environmental detector monitors the position of the multiple sensors. When the sensors extend 0-50% of their stroke, the ring-shaped environmental detector scans the internal channel of the robotic arm at a low frequency to check for any obstruction. When they extend 50%-100% of their stroke, the detector switches to a high-frequency scan of the external environment. The detection data is used to correct the extension trajectory of the sensors in real time to avoid rigid contact with the target surface or sudden obstacles.

4. The multi-sensor fusion detection method for the end effector of a robotic arm as described in claim 1, characterized in that, In S2, the end effector of the robotic arm is equipped with a concentric sleeve and an umbrella-shaped linkage mechanism installed on the concentric sleeve; multiple sensors are integrated in the concentric sleeve at the end effector of the robotic arm; when extended, radial diffusion is achieved through the umbrella-shaped linkage mechanism to distribute the sensor array; each sensor rotates independently around its own axis.

5. The multi-sensor fusion detection method for the end effector of a robotic arm as described in claim 1, characterized in that, In S3, the deviation compensation is a dual closed-loop correction. The inner closed loop performs sub-millimeter-level position correction through a micro-displacement platform at the end of the robotic arm, while the outer closed loop achieves cumulative error compensation by adjusting the spatial coordinate system offset of the robotic arm's base joints. The inner and outer closed loops are dynamically switched using a fuzzy PID algorithm.

6. The multi-sensor fusion detection method for the end effector of a robotic arm as described in claim 1, characterized in that, In S2, multiple sensors include a transient ground voltage sensor, an ultra-high frequency sensor, and an ultrasonic sensor. The transient ground voltage sensor is used to detect transient changes in the ground voltage of the target to be detected. The ultra-high frequency sensor collects partial discharge signals, and the ultra-high frequency sensor and the ultrasonic sensor operate alternately. The robotic arm is provided with a pendulum for applying coupling agent to the probe of the ultrasonic sensor, and the robotic arm is provided with a receiving groove for accommodating the pendulum.

7. The multi-sensor fusion detection method for the end effector of a robotic arm as described in claim 1, characterized in that, In S1, environmental monitoring uses a combination of lidar and a visual sensor. The lidar acquires distance information of objects within the area, and the visual sensor acquires the texture features of the objects. The lidar and the visual sensor are fused to generate a three-dimensional grid space containing details of the object's surface.

8. The multi-sensor fusion detection method for the end effector of a robotic arm as described in claim 7, characterized in that, In S1, electromagnetic compatibility testing is added to environmental monitoring. Electromagnetic interference is monitored in real time using a spectrum analyzer, and the electromagnetic shielding is activated when the interference intensity exceeds the threshold.

9. The multi-sensor fusion detection method for the end effector of a robotic arm as described in claim 1, characterized in that, Also includes: S5: After the test is completed, multiple sensors are calibrated. The robotic arm moves the multiple sensors to the preset position of the standard calibration block, compares the deviation between the actual test value and the theoretical value, and automatically generates the correction coefficient of each sensor for the initial parameter setting of the next test.

10. A multi-sensor fusion detection system for the end effector of a robotic arm, characterized in that, This includes the multi-sensor fusion detection method for the end effector of a robotic arm as described in any one of claims 1-9.

Images