Track following method and system based on permanent magnet image

Abstract

The application provides a permanent magnet image-based trajectory tracking method and system, relates to the image trajectory tracking technical field, and the method comprises the following steps: collecting video information of permanent magnet operation, and extracting a continuous image sequence; according to the geometric structure characteristics and the magnetic pole distribution characteristics of the permanent magnet in an executing motor, a visual equivalent model of the permanent magnet is constructed; based on the visual equivalent model, the permanent magnet in each frame of image in the continuous image sequence is detected, and corresponding spatial position information is extracted; the spatial position information of the permanent magnet is associated and predicted in a time sequence, the motion trajectory of the permanent magnet is continuously tracked in combination with the motion constraint of the executing motor, time sequence trajectory data is generated, and the motion trajectory of the permanent magnet is obtained. Through the application, the technical problem of poor trajectory tracking precision based on the permanent magnet image in the prior art can be solved, and the technical effect of improving the trajectory tracking precision based on the permanent magnet image is achieved.

Description

Technical Field

[0001] This application relates to the field of image trajectory tracking technology, and in particular to a trajectory tracking method and system based on permanent magnet images. Background Technology

[0002] With the widespread application of actuators in industrial automation, precision manufacturing, intelligent equipment, and high-reliability electromechanical systems, higher requirements are placed on the accuracy of motor operating status perception, online monitoring capabilities, and early fault identification. This is especially true for motor types such as permanent magnet synchronous motors, which use permanent magnets as key functional components. The spatial position and rotation trajectory of the permanent magnets directly reflect the rotor operating status, assembly accuracy, and load condition changes. Therefore, how to accurately acquire and continuously track the motion status of permanent magnets without introducing additional sensors or damaging the motor structure has become an important research direction in the field of motor condition monitoring and intelligent diagnosis.

[0003] Currently, most existing methods for permanent magnet motion state detection and rotor state analysis rely on indirect observation methods such as current signals, vibration signals, or magnetic field signals, or use simple image detection methods to identify the position of permanent magnets. These methods usually lack a unified modeling mechanism for the intrinsic relationship between the geometric features of permanent magnets, the characteristics of magnetic pole distribution, and the motion law of the actuator. At the image processing level, permanent magnets are often treated as ordinary targets and detected frame by frame, failing to fully utilize the inherent periodicity, structural stability, and physical constraint characteristics of permanent magnets during rotation. This leads to easy jumps or misjudgments in the identification results under complex lighting conditions, noise interference, occlusion, or speed fluctuations. At the same time, existing technologies generally use simple time smoothing or filtering strategies in the trajectory tracking process, lacking a time-series correlation prediction and constraint correction mechanism based on the physical motion law of the motor, making it difficult to guarantee the physical consistency of the trajectory.

[0004] In summary, existing technologies suffer from several technical problems. The lack of a visual modeling method that combines the structural features of permanent magnets and the periodicity of magnetic poles with the motion laws of the actuator motor, and the lack of a tracking mechanism for temporal correlation prediction and physical consistency constraints on the motion trajectory of permanent magnets, makes it difficult to guarantee the continuity and authenticity of the motion trajectory of permanent magnets. This further affects the accuracy of rotor state monitoring, the reliability of anomaly identification, and the level of actuator motor operation status analysis based on permanent magnet motion information. Summary of the Invention

[0005] The purpose of this application is to provide a trajectory tracking method and system based on permanent magnet images, in order to solve the technical problems in the prior art. Due to the lack of a visual modeling method that combines the structural features of permanent magnets and the periodicity of magnetic poles with the motion law of the actuator, and the lack of a tracking mechanism for temporal correlation prediction and physical consistency constraints on the motion trajectory of permanent magnets, the continuity and authenticity of the motion trajectory of permanent magnets are difficult to guarantee, which further affects the accuracy of rotor state monitoring, the reliability of anomaly identification, and the level of actuator operating state analysis based on permanent magnet motion information.

[0006] In view of the above problems, this application provides a trajectory tracking method and system based on permanent magnet images.

[0007] Firstly, this application provides a trajectory tracking method based on permanent magnet images, implemented through a trajectory tracking system based on permanent magnet images. The method includes: acquiring video information of a permanent magnet in operation and extracting a continuous image sequence containing the permanent magnet; constructing a visual equivalent model of the permanent magnet based on its geometric structure features and magnetic pole distribution features in the actuator, used to constrain the position and motion state of the permanent magnet in the image; detecting the permanent magnet in each frame of the continuous image sequence based on the visual equivalent model and extracting its corresponding spatial position information; predicting the spatial position information of the permanent magnet in a time series, and continuously tracking the motion trajectory of the permanent magnet in conjunction with the motion constraints of the actuator, generating time-series trajectory data to obtain the motion trajectory of the permanent magnet.

[0008] Preferably, the trajectory tracking method based on permanent magnet images further includes: extracting image frames from the video information in chronological order to construct an original image frame sequence; performing permanent magnet identification judgment on each image frame in the original image frame sequence, retaining image frames containing identifiable features of permanent magnets, and constructing a continuous image sequence of permanent magnets; based on the continuous image sequence, performing continuous constraint judgment on the motor motion of adjacent image frames in terms of time and position, and determining the continuous image sequence of image frames that meet the constraint conditions as the continuous image sequence.

[0009] Preferably, the trajectory tracking method based on permanent magnet images further includes: extracting stable geometric features based on the structural information of the permanent magnet, including the spatial positional relationship of the permanent magnet relative to the rotation center of the actuator rotor, the equivalent rotation radius, and the range of changes in overall size or contour; mapping the periodic features of the magnetic poles during rotation to periodic angular features or periodic structural features in the image based on the magnetic pole distribution features, inserting the spatial position of the geometric features, and establishing a visual equivalent relationship; and obtaining the equivalent parameters of the permanent magnet based on the visual equivalent relationship, including at least one of the equivalent rotation center parameters, the equivalent rotation radius parameters, and the periodic angular parameters, and establishing a visual equivalent model of the permanent magnet.

[0010] Preferably, the trajectory tracking method based on permanent magnet images further includes: obtaining the number of magnetic poles of the permanent magnet; calculating the angular interval and occurrence period of the magnetic poles during rotation based on the motor speed and magnetic pole distribution; recording the periodic changes in the image based on the brightness and shape changes of the magnetic poles during rotation; and establishing a permanent magnet angle-position mapping relationship based on image features to form an equivalent model at the visual level by correspondingly associating the periodic visual changes in the image with the physical rotation angle of the magnetic poles.

[0011] Preferably, the trajectory tracking method based on permanent magnet images further includes: determining the theoretical position sequence of each magnetic pole relative to the image coordinate system during rotation based on the fixed installation relationship and rotational motion law of the permanent magnet in the motor; providing physical constraints for the position of each magnetic pole in the image using the geometric features of the permanent magnet in the motor; and combining the physical position with the visual periodic spatial position based on the physical constraints so that the position of the magnetic pole in the image is consistent with the actual physical position in the motor.

[0012] Preferably, the trajectory tracking method based on permanent magnet images further includes: establishing positional spatial constraints based on the visual equivalent model, generating a search region and feature template for the permanent magnet in the image; based on the search region and feature template, identifying the position, contour features, and rotation angle of the permanent magnet in each frame of the continuous image sequence to obtain permanent magnet identification features; and based on the permanent magnet identification features, performing position optimization and verification through geometric and periodic constraints in the visual equivalent model, and outputting the spatial position information of the permanent magnet in each frame.

[0013] Preferably, the trajectory tracking method based on permanent magnet images further includes: sorting the extracted spatial position information of the permanent magnet in time to obtain a time sequence reflecting the continuous position changes of the permanent magnet; predicting the spatial position in future frames based on the position change trend reflected in the time sequence and the motion constraints of the actuator; correcting the predicted spatial position using the motion constraints of the actuator, and performing smoothing filtering and trajectory fitting on the position data of multiple consecutive frames to obtain the continuous motion trajectory of the permanent magnet; generating time-series trajectory data based on the continuous motion trajectory, and calculating and outputting the motion state parameters of the permanent magnet, including rotor angle, speed, and acceleration, according to the time-series trajectory data.

[0014] Preferably, the trajectory tracking method based on permanent magnet images further includes: during the tracking of the permanent magnet's motion trajectory, when an abnormal change in the permanent magnet's position is detected, analyzing the performance of the abnormal change in terms of temporal continuity, spatial consistency, or periodicity, and determining whether the abnormal change meets the motion law of the motor rotor; when the abnormal change does not meet the motion law of the motor rotor, determining the abnormal change as image noise or observation error, and correcting or suppressing the corresponding trajectory data.

[0015] Preferably, the trajectory tracking method based on permanent magnet images further includes: when the abnormal change meets the rotor motion law of the motor, the abnormal change is identified as a physical abnormality caused by rotor vibration, eccentricity or torque fluctuation, and the abnormal change is retained for rotor state analysis.

[0016] Secondly, this application also provides a trajectory tracking system based on permanent magnet images, used to execute the trajectory tracking method based on permanent magnet images as described in the first aspect, comprising: a continuous image sequence extraction module, used to acquire video information of permanent magnet operation and extract a continuous image sequence containing permanent magnets; a visual equivalent model construction module, used to construct a visual equivalent model of permanent magnets based on the geometric structural features and magnetic pole distribution features of permanent magnets in the actuator motor, used to constrain the position and motion state of permanent magnets in the images; a spatial position information extraction module, used to detect permanent magnets in each frame of the continuous image sequence based on the visual equivalent model and extract corresponding spatial position information; and a motion trajectory obtaining module, used to correlate and predict the spatial position information of permanent magnets in the time series, combine with the motion constraints of the actuator motor, continuously track the motion trajectory of permanent magnets, generate time-series trajectory data, and obtain the motion trajectory of permanent magnets.

[0017] The technical solution provided in this application has at least the following technical effects or advantages: by achieving the technical goal of accurately acquiring the continuous motion trajectory of a permanent magnet based on a visual equivalent model and tracking the physical consistency constraint, it can still stably and continuously acquire the real motion trajectory of the permanent magnet under complex imaging conditions and operating conditions, thereby improving the technical effect of improving the accuracy of rotor operation status monitoring, the reliability of anomaly identification, and the overall operation status analysis and diagnosis capabilities of the actuator.

[0018] The above description is merely an overview of the technical solution of this application. To enable a clearer understanding of the technical means of this application and to facilitate its implementation according to the description, and to make the above and other objects, features, and advantages of this application more apparent, specific embodiments of this application are described below. It should be understood that the content described in this section is not intended to identify key or important features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent through the following description. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0020] Figure 1 This is a flowchart illustrating the trajectory tracking method based on permanent magnet images according to this application; Figure 2 This is a schematic diagram of the trajectory tracking system based on permanent magnet images according to this application.

[0021] Figure labeling: 1. Continuous image sequence extraction module; 2. Visual equivalent model construction module; 3. Spatial location information extraction module; 4. Motion trajectory acquisition module. Detailed Implementation

[0022] This application provides a trajectory tracking method and system based on permanent magnet images, solving the technical problems in existing technologies. These problems stem from the lack of a visual modeling method that combines the structural features and periodicity of permanent magnets with the motion laws of the actuator, and the lack of a tracking mechanism for temporal correlation prediction and physical consistency constraints on the permanent magnet trajectory. This leads to difficulties in ensuring the continuity and realism of the permanent magnet trajectory, further affecting the accuracy of rotor state monitoring, the reliability of anomaly identification, and the level of actuator operating state analysis based on permanent magnet motion information. The application achieves the technical goal of accurately acquiring and tracking the continuous motion trajectory of permanent magnets with physical consistency constraints based on a visual equivalent model. This enables stable and continuous acquisition of the true motion trajectory of permanent magnets even under complex imaging conditions and operating conditions, thereby improving the accuracy of actuator rotor operating state monitoring, the reliability of anomaly identification, and the overall operating state analysis and diagnostic capabilities.

[0023] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. It should be understood that this application is not limited to the exemplary embodiments described herein. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application. It should also be noted that, for ease of description, only the parts related to this application are shown in the accompanying drawings, not all of them.

[0024] Example 1, please refer to the appendix. Figure 1 This application provides a trajectory tracking method based on permanent magnet images, which is applied to a trajectory tracking system based on permanent magnet images, and specifically includes the following steps: The system collects video information of the permanent magnet's operation and extracts a continuous image sequence containing the permanent magnet.

[0025] Furthermore, this application also includes: extracting image frames from the video information in chronological order to construct an original image frame sequence; performing permanent magnet identification judgment on each image frame in the original image frame sequence, retaining image frames containing identifiable features of permanent magnets, and constructing a continuous image sequence of permanent magnets; based on the continuous image sequence, performing continuous constraint judgment on the motor motion of adjacent image frames in terms of time and position, and determining the continuous image sequence of image frames that meet the constraint conditions as the continuous image sequence.

[0026] Specifically, extracting image frames from video information in chronological order to construct an original image frame sequence involves analyzing the acquired continuous video data frame by frame, extracting discrete image frame data according to the chronological order of the video timeline, thereby forming a set of original image frames that reflects the changes of the permanent magnet in the actuator over time. The original image frame sequence maintains the temporal continuity and spatial consistency of the video acquisition, serving as the foundational data source for subsequent permanent magnet identification and trajectory analysis.

[0027] Furthermore, permanent magnet identification is performed on each image frame in the original image frame sequence. Image frames containing identifiable features of permanent magnets are retained to construct a continuous image sequence of permanent magnets. This means that based on the brightness distribution, contour shape, structural features, or periodic visual features caused by magnetic poles of permanent magnets in the image, it is determined whether there is a stably identifiable permanent magnet target in each image frame. This process eliminates image frames that do not contain effective information about permanent magnets or have insufficient identification confidence, and only retains image frames that can reflect the actual position and attitude changes of permanent magnets to form a continuous image sequence of permanent magnets for subsequent tracking processing.

[0028] Subsequently, based on the continuous image sequence, the motor motion continuity constraint judgment of adjacent image frames in terms of time and position is performed. The continuous image sequence that meets the constraint conditions is determined as a continuous image sequence. This means that on the basis of the screened image frames containing permanent magnets, the motion law of the actuator is introduced, and the time interval and spatial displacement of the permanent magnets in adjacent image frames are jointly constrained and analyzed. It is determined whether the position change of the permanent magnets between adjacent frames conforms to the speed, displacement and direction change characteristics of the motor rotor in the continuous rotation process. Thus, discontinuous frame sequences caused by frame loss, misidentification or abnormal interference are eliminated, and only the image frame sequence that satisfies the continuity of motor motion in both time and space is retained as the final continuous image sequence.

[0029] Based on the geometric structure and magnetic pole distribution characteristics of the permanent magnet in the actuator, a visual equivalent model of the permanent magnet is constructed to constrain the position and motion state of the permanent magnet in the image.

[0030] Furthermore, this application also includes: extracting stable geometric structural features based on the structural information of the permanent magnet, including the spatial positional relationship of the permanent magnet relative to the rotation center of the actuator rotor, the equivalent rotation radius, and the range of changes in overall size or contour; mapping the periodic features of the magnetic poles during rotation to periodic angular features or periodic structural features in the image based on the magnetic pole distribution features, inserting the spatial position of the geometric structural features, and establishing a visual equivalent relationship; and obtaining the equivalent parameters of the permanent magnet based on the visual equivalent relationship, including at least one of the equivalent rotation center parameters, the equivalent rotation radius parameters, and the periodic angular parameters, and establishing a visual equivalent model of the permanent magnet.

[0031] Furthermore, this application also includes: obtaining the number of magnetic poles of the permanent magnet; calculating the angular interval and occurrence period of the magnetic poles during rotation based on the motor speed and magnetic pole distribution; recording the periodic changes in the image based on the brightness and shape changes of the magnetic poles during rotation; and establishing a permanent magnet angle-position mapping relationship based on image features to form an equivalent model at the visual level by correspondingly associating the periodic visual changes in the image with the physical rotation angle of the magnetic poles.

[0032] Furthermore, this application also includes: determining the theoretical position sequence of each magnetic pole relative to the image coordinate system during rotation based on the fixed installation relationship and rotational motion law of the permanent magnet in the motor; providing physical constraints for the position of each magnetic pole in the image using the geometric characteristics of the permanent magnet in the motor; and combining the physical position with the visual periodic spatial position based on the physical constraints so that the position of the magnetic pole in the image is consistent with the actual physical position in the motor.

[0033] Specifically, stable geometric features are extracted based on the structural information of the permanent magnet, including the spatial positional relationship of the permanent magnet relative to the rotation center of the actuator rotor, the equivalent rotation radius, and the range of overall size or contour changes. This refers to the abstract description of the geometric shape of the permanent magnet in the image based on its installation structure, fixing method, and physical structural characteristics that remain unchanged as the rotor rotates. This determines the spatial relative positional relationship of the permanent magnet relative to the rotation center of the rotor, and combines the equivalent rotation radius formed when it moves around the rotation center, as well as the overall size range or contour change boundary that remains stable under changing imaging conditions, to provide a stable and repeatable geometric constraint basis for subsequent visual modeling.

[0034] Furthermore, based on the characteristics of magnetic pole distribution, the periodic features of the magnetic poles during rotation are mapped to periodic angular features or periodic structural features in the image. The spatial position of the geometric structural features is inserted to establish an equivalent relationship at the visual level. This means using the number and arrangement of magnetic poles on the permanent magnet and their periodic occurrence during rotation to convert the angular changes and periodic distribution of the magnetic poles in physical space into observable angular changes, brightness changes, or structural repetition features in the image. The periodic visual features are then fused with the aforementioned geometric structural features in the same image space coordinate system, so that the visual representation of the magnetic poles and the physical spatial position of the permanent magnet form a one-to-one equivalent mapping relationship, thereby establishing an equivalent relationship at the visual level that can reflect the true rotation state.

[0035] Among them, determining the theoretical position sequence of each magnetic pole relative to the image coordinate system during rotation, based on the fixed installation relationship and rotational motion law of the permanent magnet in the motor, refers to deriving and calculating the possible spatial positions of each magnetic pole at different rotation angles based on the fixed connection between the permanent magnet and the rotor of the actuator motor and the motion characteristics of the rotor rotating periodically around its rotation center, combined with the image coordinate system established by camera imaging, thereby forming a theoretical position sequence that can reflect the change of the magnetic pole over time, used to describe the ideal motion trajectory of the magnetic pole in the image space.

[0036] Furthermore, by utilizing the geometric characteristics of permanent magnets in the motor, physical constraints are provided for the position of each magnetic pole in the image. This means that the possible areas of magnetic poles in the image are limited based on the overall size, shape profile, equivalent radius of rotation of the permanent magnet, and the fixed geometric relationship between the magnetic poles and the rotor center. This ensures that the identification results of the magnetic pole positions must meet the geometric rationality at the physical structure level, thereby ensuring that the position of the permanent magnet in the image is consistent with the actual physical movement of the motor.

[0037] Subsequently, based on physical constraints, the physical position and the visual periodic spatial position are combined to make the magnetic pole position in the image consistent with the actual physical position in the motor. This means that, based on geometric constraints, the periodic visual changes in the position of the magnetic pole in the image are jointly corrected with the actual rotation position in the motor, so that the magnetic pole position detected in the image not only conforms to the visual feature change law, but also remains consistent with the actual physical position of the permanent magnet inside the motor, thereby achieving positional alignment between physical space and image space.

[0038] Simultaneously, the number of permanent magnet poles is obtained, and the angular interval and reappearance period of the poles during rotation are calculated based on the motor speed and pole distribution. This involves obtaining information about the number of poles on the permanent magnet, combining it with the motor speed parameters and the uniform or non-uniform distribution characteristics of the poles in the circumferential direction, to calculate the angular interval between adjacent poles and the time period corresponding to the reappearance of a single pole at the same observation position during rotation. This is used to characterize the periodic characteristics of the pole rotational motion. The angular interval is calculated as 360° / number of poles. Predictions are made using the known relationship between motor speed and the number of poles to ensure that the angular interval conforms to the actual physical motion of the permanent magnet within the motor.

[0039] Furthermore, based on the brightness and shape changes of the magnetic poles during rotation, the periodic changes in the image are recorded. This refers to the continuous observation and recording of visual features such as brightness fluctuations and contour changes of the magnetic poles caused by changes in magnetic field distribution, surface structure, or imaging angle in the image sequence. In this way, image change features that can reflect the periodic appearance and disappearance of the magnetic poles are extracted to characterize the visual periodicity of the magnetic poles during rotation.

[0040] Subsequently, the periodic visual changes in the image are correlated with the physical rotation angle of the magnetic poles to establish a permanent magnet angle-position mapping relationship based on image features, forming a visual equivalent model. This means that the extracted periodic change features of the image are correlated one-to-one with the angle position of the magnetic poles during the physical rotation process, thereby constructing a mapping relationship between the rotation angle and spatial position of the permanent magnet derived from the image features, and forming a visual equivalent model that can reflect the true rotation state of the permanent magnet.

[0041] Subsequently, based on the visual equivalence relationship, the equivalent parameters of the permanent magnet are obtained, including at least one of the equivalent rotation center parameter, equivalent rotation radius parameter, and periodic angle parameter. A visual equivalent model of the permanent magnet is established to constrain the position and motion state of the permanent magnet in the image. This means that, based on the equivalent mapping relationship, key parameters that can characterize the rotation behavior of the permanent magnet are inverted from the image features. By determining the position of the equivalent rotation center, the size of the equivalent rotation radius, and the angle parameter reflecting the periodic change of the magnetic poles in the image coordinate system, a visual equivalent model corresponding to the actual physical structure and motion law of the permanent magnet is constructed. The visual equivalent model is then used to uniformly constrain and verify the position distribution, attitude change, and motion continuity of the permanent magnet in each image frame.

[0042] Based on the visual equivalent model, permanent magnets in each frame of the continuous image sequence are detected, and their corresponding spatial location information is extracted.

[0043] Furthermore, this application also includes: establishing positional spatial constraints based on the visual equivalent model to generate a search region and feature template for the permanent magnet in the image; based on the search region and feature template, identifying the position, contour features, and rotation angle of the permanent magnet in each frame of the continuous image sequence to obtain permanent magnet identification features; and based on the permanent magnet identification features, performing position optimization and verification through geometric and periodic constraints in the visual equivalent model to output the spatial position information of the permanent magnet in each frame.

[0044] Specifically, based on the visual equivalent model, establishing positional spatial constraints and generating the search area and feature template of the permanent magnet in the image refers to using parameters such as the equivalent rotation center, equivalent rotation radius, and magnetic pole periodicity angle contained in the visual equivalent model to limit the spatial range in which the permanent magnet may appear in the image coordinate system and determine the search area for target detection. At the same time, based on the stable appearance characteristics of the permanent magnet at different rotation angles, corresponding feature templates are constructed to reduce the search space and improve the targeting and stability of the subsequent recognition process.

[0045] Furthermore, based on the search area and feature template, the position, contour features, and rotation angle of the permanent magnet are identified in each frame of the continuous image sequence to obtain permanent magnet identification features. This means that within the aforementioned defined search area, combined with the pre-constructed feature template, each frame of the continuous image sequence is analyzed and processed. By matching or identifying the areas in the image that match the appearance features of the permanent magnet, information such as the spatial position, shape contour, and corresponding rotation angle of the permanent magnet in the current image frame is extracted, forming a set of identification features that can characterize the current state of the permanent magnet.

[0046] Subsequently, based on the permanent magnet identification features, position optimization and verification are performed through geometric and periodic constraints in the visual equivalent model. The spatial position information of the permanent magnet in each frame is output. This means that the position, angle and contour features of the permanent magnet initially identified in each frame are substituted into the visual equivalent model. The rationality of the position is verified through geometric constraints, and the consistency verification and correction of the identification results are performed by combining the angle constraints corresponding to the periodic changes of the magnetic poles. In this way, the spatial position of the permanent magnet is optimized, and finally, the spatial position information of the permanent magnet in each image frame that conforms to the laws of physical motion and visual consistency is output.

[0047] The spatial position information of the permanent magnet is correlated and predicted in the time series. Combined with the motion constraints of the actuator, the motion trajectory of the permanent magnet is continuously tracked to generate time-series trajectory data and obtain the motion trajectory of the permanent magnet.

[0048] Furthermore, this application also includes: sorting the extracted spatial position information of the permanent magnet in time to obtain a time series reflecting the continuous position changes of the permanent magnet; predicting the spatial position in future frames based on the position change trend reflected in the time series and the motion constraints of the actuator; correcting the predicted spatial position using the motion constraints of the actuator, and smoothing and fitting the position data of multiple consecutive frames to obtain the continuous motion trajectory of the permanent magnet; generating time series trajectory data based on the continuous motion trajectory, and calculating and outputting the motion state parameters of the permanent magnet, including rotor angle, speed, and acceleration, according to the time series trajectory data.

[0049] Furthermore, this application also includes: during the tracking of the permanent magnet's motion trajectory, when an abnormal change in the permanent magnet's position is detected, the abnormal change is analyzed in terms of its temporal continuity, spatial consistency, or periodicity to determine whether the abnormal change meets the motion law of the motor rotor; when the abnormal change does not meet the motion law of the motor rotor, the abnormal change is determined to be image noise or observation error, and the corresponding trajectory data is corrected or suppressed.

[0050] Furthermore, this application also includes: when the abnormal change meets the rotor motion law of the motor, the abnormal change is identified as a physical abnormality caused by rotor vibration, eccentricity or torque fluctuation, and the abnormal change is retained for rotor state analysis.

[0051] Specifically, the spatial position information of the permanent magnet is correlated and predicted in the time series. Combined with the motion constraints of the actuator motor, the motion trajectory of the permanent magnet is continuously tracked. The extracted spatial position information of the permanent magnet is sorted in time to obtain a time series reflecting the continuous position changes of the permanent magnet. This means arranging the spatial position information of the permanent magnet acquired in each image frame according to the time sequence of image acquisition, and performing correlation analysis on adjacent position points in the time dimension. The motion constraint condition of continuous rotation of the actuator motor rotor is introduced so that the position information of the permanent magnet at different time points can form a continuous position sequence that conforms to the laws of physical motion, thereby constructing a time series to describe the changes of the permanent magnet over time.

[0052] Furthermore, based on the position change trend reflected in the time sequence and combined with the motion constraints of the actuator, spatial position prediction in future frames is performed. This means that by analyzing the trend characteristics of the permanent magnet position changing with time in the time sequence and combining the motion constraints of the actuator in terms of speed, acceleration and rotation direction, the spatial position of the permanent magnet in subsequent unprocessed image frames is predicted, so as to improve the performance of trajectory tracking in terms of temporal continuity and robustness.

[0053] Subsequently, the predicted spatial position is corrected using the motion constraints of the actuator motor. The position data of multiple consecutive frames is smoothed and fitted to obtain the continuous motion trajectory of the permanent magnet. This means that, based on the predicted position, the physical constraints followed by the actual movement of the actuator motor rotor are introduced to correct the prediction result and make it consistent with the actual motion state. The noise influence is reduced by smoothing and filtering the position data of multiple frames. Then, the discrete position points are made continuous by the trajectory fitting method, thereby forming a continuous motion trajectory describing the overall motion process of the permanent magnet.

[0054] Finally, based on the continuous motion trajectory, time-series trajectory data is generated, and based on the time-series trajectory data, the motion state parameters of the permanent magnet are calculated and output, including rotor angle, velocity, and acceleration. This means organizing and storing the continuous motion trajectory in the form of a time series, and by analyzing the changes in the trajectory in the time dimension, calculating the key parameters that can characterize the motion state of the permanent magnet and the rotor in which it is located, thereby outputting motion state parameters that reflect the changes in rotor angle, rotational speed, and acceleration characteristics.

[0055] Furthermore, during the tracking of the permanent magnet's trajectory, when an abnormal change in the permanent magnet's position is detected, the analysis of the abnormal change's temporal continuity, spatial consistency, or periodicity characteristics determines whether the abnormal change conforms to the motion law of the actuator motor rotor. This means that during the continuous tracking of the permanent magnet's spatial position over time, if a deviation from the existing trajectory trend is detected between adjacent time points, the change is not immediately judged as noise. Instead, it is treated as a candidate physical anomaly possibly caused by changes in the physical state of the actuator motor rotor. The analysis considers whether the change disrupts the continuity of the permanent magnet's motion from a temporal perspective, whether it remains within the reasonable range allowed by the motor's geometry and rotational constraints from a spatial perspective, and whether it still conforms to the periodicity of the magnetic pole distribution and rotation period from a periodicity perspective. This comprehensive assessment determines whether the abnormal change conforms to the motion law of the actuator motor rotor in actual operation. When an abnormal change conforms to the motion law of the actuator motor rotor, the abnormal change is identified as a physical anomaly caused by rotor vibration, eccentricity, or torque fluctuation, and is retained for rotor state analysis. This means that if the abnormal change still conforms to the motion characteristics that the actuator motor rotor may have during actual operation in terms of amplitude, trend, and frequency, the abnormal change is determined to be caused by a real physical state change of the rotor during operation. Rotor vibration refers to the periodic or non-periodic micro-amplitude oscillation of the rotor during rotation due to structural or operating condition changes. Rotor eccentricity refers to the trajectory deviation caused by the deviation of the rotor's rotation center from the geometric center or ideal rotation axis. Torque fluctuation refers to the unstable phenomenon of speed or angular acceleration caused by the change of driving torque over time. The corresponding abnormal trajectory changes are retained as important information reflecting the rotor's operating state for subsequent analysis and evaluation of the rotor's operating state, health status, and abnormal mechanism.

[0056] Furthermore, when abnormal changes do not conform to the motion law of the motor rotor, the abnormal changes are judged as image noise or observation error, and the corresponding trajectory data is corrected or suppressed. This means that after multi-dimensional consistency analysis, if the abnormal changes cannot match the actual motion characteristics of the motor rotor in terms of temporal continuity, spatial consistency, or periodicity, then the abnormal changes are determined not to be caused by the actual physical state of the rotor, but by factors such as image acquisition noise, lighting interference, occlusion, recognition error, or observation deviation. The trajectory data corresponding to the abnormality is then processed by position correction, abnormal point suppression, or smoothing replacement to avoid the abnormal changes from adversely affecting the continuity and accuracy of the overall motion trajectory of the permanent magnet.

[0057] In summary, the trajectory tracking method based on permanent magnet images provided in this application has the following technical effects: by achieving the technical goal of accurately acquiring and tracking the continuous motion trajectory of a permanent magnet based on a visual equivalent model, it can stably and continuously acquire the real motion trajectory of the permanent magnet under complex imaging conditions and operating conditions, thereby improving the accuracy of rotor operation status monitoring, the reliability of anomaly identification, and the overall operation status analysis and diagnosis capabilities of the actuator.

[0058] Example 2: Based on the same inventive concept as the trajectory tracking method based on permanent magnet images in the foregoing examples, this application also provides a trajectory tracking system based on permanent magnet images. Please refer to the appendix. Figure 2 The system includes: a continuous image sequence extraction module 1, used to collect video information of the permanent magnet's operation and extract a continuous image sequence containing the permanent magnet; a visual equivalent model construction module 2, used to construct a visual equivalent model of the permanent magnet based on the geometric structure features and magnetic pole distribution features of the permanent magnet in the actuator, used to constrain the position and motion state of the permanent magnet in the image; a spatial position information extraction module 3, used to detect the permanent magnet in each frame of the continuous image sequence based on the visual equivalent model and extract the corresponding spatial position information; and a motion trajectory acquisition module 4, used to correlate and predict the spatial position information of the permanent magnet in the time series, combine it with the motion constraints of the actuator, continuously track the motion trajectory of the permanent magnet, generate time-series trajectory data, and obtain the motion trajectory of the permanent magnet.

[0059] Furthermore, the trajectory tracking system based on permanent magnet images is also used to: extract image frames from the video information in chronological order to construct an original image frame sequence; perform permanent magnet identification judgment on each image frame in the original image frame sequence, retain image frames containing identifiable features of permanent magnets, and construct a continuous image sequence of permanent magnets; based on the continuous image sequence, perform continuous constraint judgment on the motor motion of adjacent image frames in terms of time and position, and determine the continuous image sequence of image frames that meet the constraint conditions as the continuous image sequence.

[0060] Furthermore, the trajectory tracking system based on permanent magnet images is also used for: extracting stable geometric features based on the structural information of the permanent magnet, including the spatial positional relationship of the permanent magnet relative to the rotation center of the actuator rotor, the equivalent rotation radius, and the range of overall size or contour changes; mapping the periodic features of the magnetic poles during rotation to periodic angular features or periodic structural features in the image based on the magnetic pole distribution features, inserting the spatial position of the geometric features, and establishing a visual equivalent relationship; and obtaining the equivalent parameters of the permanent magnet based on the visual equivalent relationship, including at least one of the equivalent rotation center parameters, the equivalent rotation radius parameters, and the periodic angular parameters, and establishing a visual equivalent model of the permanent magnet.

[0061] Furthermore, the trajectory tracking system based on permanent magnet images is also used to: obtain the number of magnetic poles of the permanent magnet; calculate the angular interval and occurrence period of the magnetic poles during rotation based on the motor speed and magnetic pole distribution; record the periodic changes in the image based on the brightness and shape changes of the magnetic poles during rotation; and correlate the periodic visual changes in the image with the physical rotation angle of the magnetic poles to establish a permanent magnet angle-position mapping relationship based on image features, forming an equivalent model at the visual level.

[0062] Furthermore, the trajectory tracking system based on permanent magnet images is also used to: determine the theoretical position sequence of each magnetic pole relative to the image coordinate system during rotation based on the fixed installation relationship and rotational motion law of the permanent magnet in the motor; provide physical constraints for the position of each magnetic pole in the image using the geometric features of the permanent magnet in the motor; and combine the physical position with the visual periodic spatial position based on the physical constraints so that the position of the magnetic pole in the image is consistent with the actual physical position in the motor.

[0063] Furthermore, the trajectory tracking system based on permanent magnet images is also used for: establishing positional spatial constraints based on the visual equivalent model, generating a search region and feature template for the permanent magnet in the image; based on the search region and feature template, identifying the position, contour features, and rotation angle of the permanent magnet in each frame of the continuous image sequence to obtain permanent magnet recognition features; and based on the permanent magnet recognition features, performing position optimization and verification through geometric and periodic constraints in the visual equivalent model, and outputting the spatial position information of the permanent magnet in each frame.

[0064] Furthermore, the trajectory tracking system based on permanent magnet images is also used for: sorting the extracted spatial position information of the permanent magnet in time to obtain a time-series sequence reflecting the continuous position changes of the permanent magnet; predicting the spatial position in future frames based on the position change trend reflected in the time-series sequence and the motion constraints of the actuator; correcting the predicted spatial position using the motion constraints of the actuator, smoothing and filtering the position data of multiple consecutive frames and fitting the trajectory to obtain the continuous motion trajectory of the permanent magnet; generating time-series trajectory data based on the continuous motion trajectory, and calculating and outputting the motion state parameters of the permanent magnet, including rotor angle, velocity, and acceleration, according to the time-series trajectory data.

[0065] Furthermore, the trajectory tracking system based on permanent magnet images is also used to: when an abnormal change in the position of the permanent magnet is detected during the tracking of the permanent magnet's motion trajectory, analyze the performance of the abnormal change in terms of temporal continuity, spatial consistency, or periodicity to determine whether the abnormal change meets the motion law of the motor rotor; when the abnormal change does not meet the motion law of the motor rotor, determine the abnormal change as image noise or observation error, and correct or suppress the corresponding trajectory data.

[0066] Furthermore, the trajectory tracking system based on permanent magnet images is also used to: when the abnormal change meets the rotor motion law of the motor, confirm the abnormal change as a physical abnormality caused by rotor vibration, eccentricity or torque fluctuation, and retain the abnormal change for rotor state analysis.

[0067] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The trajectory tracking method and specific examples based on permanent magnet images in the aforementioned embodiment one are also applicable to the trajectory tracking system based on permanent magnet images in this embodiment. Through the foregoing detailed description of the trajectory tracking method based on permanent magnet images, those skilled in the art can clearly understand the trajectory tracking system based on permanent magnet images in this embodiment. Therefore, for the sake of brevity, it will not be described in detail here.

[0068] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0069] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application also intends to include such modifications and variations.

Claims

1. A trajectory tracking method based on permanent magnet images, characterized in that, include: Collect video information of the permanent magnet's operation and extract a continuous image sequence containing the permanent magnet; Based on the geometric structure and magnetic pole distribution characteristics of the permanent magnet in the actuator, a visual equivalent model of the permanent magnet is constructed to constrain the position and motion state of the permanent magnet in the image. Based on the visual equivalent model, permanent magnets in each frame of the continuous image sequence are detected, and their corresponding spatial location information is extracted. The spatial position information of the permanent magnet is correlated and predicted in the time series. Combined with the motion constraints of the actuator, the motion trajectory of the permanent magnet is continuously tracked to generate time-series trajectory data and obtain the motion trajectory of the permanent magnet. Based on the geometric structure and magnetic pole distribution characteristics of permanent magnets in actuators, a visual equivalent model of the permanent magnet is constructed, including: Based on the structural information of the permanent magnet, stable geometric features are extracted, including the spatial positional relationship of the permanent magnet relative to the rotation center of the actuator rotor, the equivalent rotation radius, and the range of overall size or contour variation. Based on the magnetic pole distribution characteristics, the periodic features of the magnetic poles during rotation are mapped to periodic angular features or periodic structural features in the image, and the spatial position of the geometric structural features is inserted to establish an equivalent relationship at the visual level. Based on the visual equivalent relationship, the equivalent parameters of the permanent magnet are obtained, including at least one of the equivalent rotation center parameter, equivalent rotation radius parameter, and periodic angle parameter, and a visual equivalent model of the permanent magnet is established. Establish visual equivalence relationships, including: The number of magnetic poles of the permanent magnet is obtained, and the angular interval and occurrence period of the magnetic poles during rotation are calculated based on the motor speed and magnetic pole distribution. Based on the changes in brightness and shape of the magnetic poles during rotation, periodic changes in the image are recorded; By associating periodic visual changes in images with the physical rotation angles of magnetic poles, an angle-position mapping relationship for permanent magnets based on image features is established, forming an equivalent model at the visual level. The spatial location for inserting the geometric feature includes: Based on the fixed installation relationship and rotational motion law of the permanent magnet in the motor, the theoretical position sequence of each magnetic pole relative to the image coordinate system during rotation is determined; By utilizing the geometric features of permanent magnets in an electric motor, physical constraints are provided for the position of each magnetic pole in the image; Based on the physical constraints, the physical location and the visual periodic spatial location are combined to make the magnetic pole position in the image consistent with the actual physical position in the motor.

2. The trajectory tracking method based on permanent magnet images according to claim 1, characterized in that, Acquire video information of the permanent magnet's operation and extract a continuous image sequence containing the permanent magnet, including: Image frames are extracted from the video information in chronological order to construct an original image frame sequence; Permanent magnet identification is performed on each image frame in the original image frame sequence, and image frames containing identifiable features of permanent magnets are retained to construct a continuous image sequence of permanent magnets. Based on the continuous image sequence, the motor motion continuity constraint is determined for adjacent image frames in terms of time and position, and the continuous sequence of image frames that meet the constraint conditions is determined as the continuous image sequence.

3. The trajectory tracking method based on permanent magnet images according to claim 1, characterized in that, Based on the aforementioned visual equivalence model, permanent magnets in each frame of the continuous image sequence are detected, and their corresponding spatial location information is extracted, including: Based on the visual equivalent model, positional spatial constraints are established to generate the search region and feature template of the permanent magnet in the image. Based on the search region and feature template, the position, contour features and rotation angle of the permanent magnet are identified in each frame of the continuous image sequence to obtain the permanent magnet identification features. Based on the permanent magnet identification features, position optimization and verification are performed through geometric and periodic constraints in the visual equivalent model, and the spatial position information of the permanent magnet in each frame is output.

4. The trajectory tracking method based on permanent magnet images according to claim 3, characterized in that, Generate time-series trajectory data, including: The extracted spatial position information of the permanent magnet is sorted over time to obtain a time series reflecting the continuous position changes of the permanent magnet. Based on the position change trend reflected in the time sequence, combined with the motion constraints of the actuator, the spatial position in the future frame is predicted; The predicted spatial position is corrected by using the motion constraints of the actuator motor, and the position data of multiple consecutive frames are smoothed and fitted to obtain the continuous motion trajectory of the permanent magnet. Based on the continuous motion trajectory, time-series trajectory data is generated, and the motion state parameters of the permanent magnet, including rotor angle, velocity, and acceleration, are calculated and output according to the time-series trajectory data.

5. The trajectory tracking method based on permanent magnet images according to claim 1, characterized in that, Also includes: During the tracking of the permanent magnet's trajectory, when an abnormal change in the permanent magnet's position is detected, the abnormal change is analyzed in terms of its temporal continuity, spatial consistency, or periodicity to determine whether the abnormal change meets the motion law of the motor rotor. When the abnormal change does not meet the motion law of the motor rotor, the abnormal change is determined to be image noise or observation error, and the corresponding trajectory data is corrected or suppressed.

6. The trajectory tracking method based on permanent magnet images according to claim 5, characterized in that, After determining whether the abnormal change satisfies the motion law of the motor rotor, the process further includes: When the abnormal change meets the rotor motion law of the motor, the abnormal change is identified as a physical abnormality caused by rotor vibration, eccentricity or torque fluctuation, and the abnormal change is retained for rotor state analysis.

7. A trajectory tracking system based on permanent magnet images, characterized in that, The steps for implementing the trajectory tracking method based on permanent magnet images according to any one of claims 1 to 6 include: The continuous image sequence extraction module is used to acquire video information of the permanent magnet's operation and extract continuous image sequences containing the permanent magnet; The visual equivalent model construction module is used to construct a visual equivalent model of the permanent magnet based on the geometric structure features and magnetic pole distribution features of the permanent magnet in the actuator, which is used to constrain the position and motion state of the permanent magnet in the image. The spatial location information extraction module is used to detect permanent magnets in each frame of the continuous image sequence based on the visual equivalent model and extract the corresponding spatial location information. The motion trajectory acquisition module is used to correlate and predict the spatial position information of the permanent magnet in the time series, and combine it with the motion constraints of the actuator to continuously track the motion trajectory of the permanent magnet, generate time-series trajectory data, and obtain the motion trajectory of the permanent magnet.

Images