Servo module control method and device for equipment safety overload protection

Abstract

The invention relates to a servo module control method and device for equipment safety overload protection, and the method comprises the steps: obtaining equipment data of intelligent equipment, carrying out the basic safety analysis according to a preset equipment threshold parameter, and obtaining an equipment safety state; acquiring a user file and a real-time motion stage of the intelligent equipment, and performing intelligent fusion on the equipment data to obtain an equipment dynamic overload parameter; performing equipment grading judgment based on the equipment dynamic overload parameter and the equipment safety state to obtain an equipment safety level; according to the real-time motion stage, the equipment dynamic overload parameter and the equipment safety level, instruction control analysis is carried out, and a servo module control scheme is obtained. According to the invention, accurate safety control for different training intensities and modes can be realized, individual differences and special requirements are adapted, and the equipment safety and the user experience are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of equipment safety protection, and in particular to a servo module control method and device for equipment safety overload protection. Background Art

[0002] Existing smart device security protection mechanisms typically employ static threshold judgment methods, presetting fixed security parameter thresholds and triggering protective measures when device operating data exceeds these thresholds. This simplistic security protection mechanism fails to fully account for individual user differences, stage characteristics, and the dynamic changes in the device's real-time status, making it difficult to meet the security needs of varying usage intensities and modes. This static threshold judgment method can lead to untimely or excessive protective measures during high-intensity use or by special populations, in particular. Summary of the Invention

[0003] The main purpose of the present invention is to provide a servo module control method and device for equipment safety overload protection, which can accurately and safely control different training intensities and modes, adapt to individual differences and special needs, and improve equipment safety and user experience.

[0004] To achieve the above objectives, the present invention provides a servo module control method for equipment safety overload protection, comprising: Obtain device data from smart devices and perform basic security analysis based on preset device threshold parameters to determine the device security status; Obtaining the user profile and real-time motion stage of the smart device, intelligently integrating the device data, and obtaining a dynamic overload parameter of the device; Performing equipment classification judgment based on the dynamic overload parameter of the equipment and the safety status of the equipment to obtain the equipment safety level; A command control analysis is performed based on the real-time motion stage, the dynamic overload parameter of the device and the safety level of the device to obtain a servo module control solution.

[0005] Furthermore, the smart device data of the smart device is obtained, and basic security analysis is performed according to preset device threshold parameters to obtain the device security status, including: Acquire servo torque data and motor current data of the smart device, perform device torque identification, and obtain device torque load data; Acquire position encoder data and speed sensor data of the smart device, perform motion trajectory analysis, and obtain device motion state data; Performing multi-dimensional fusion of the equipment according to the equipment torque load data and the equipment motion state data to obtain the equipment data; The device data is compared with the preset device threshold parameters to obtain the device safety status.

[0006] Furthermore, the acquiring of the user profile and real-time motion stage of the smart device and intelligent fusion of the device data to obtain the device dynamic overload parameter includes: Acquire visual sensor data and pressure sensor data of the smart device, perform motion stage recognition, and obtain a real-time motion stage; Performing action recognition on the visual sensing data based on preset action specification information to obtain action posture data; Performing pressure recognition on the action posture data based on the pressure sensing data to obtain posture pressure parameters; Obtain the user profile through a user database connected to the smart device, perform historical training identification, and obtain user strength assessment data; Dynamic motion fusion is performed on the device data according to the real-time motion stage, the user strength assessment data and the posture pressure parameter to obtain the device dynamic overload parameter.

[0007] Furthermore, performing device classification judgment based on the device dynamic overload parameter and the device safety status to obtain the device safety level includes: Performing composite load data extraction on the dynamic overload parameters of the equipment to obtain mechanical load data and motion load data of the equipment; Performing state quantization encoding on the safety state of the device to obtain a safety state quantization value; performing abnormal attenuation calibration on the equipment mechanical load data according to the safety state quantified value to obtain calibrated mechanical load data; Performing load coefficient mapping on the motion load data according to the safety state quantification value to obtain a motion load safety factor; An equipment safety assessment is performed on the calibrated mechanical load data and the motion load safety factor based on a preset safety level conversion rule to obtain the equipment safety level.

[0008] Furthermore, the equipment safety assessment is performed on the calibration mechanical load data and the motion load safety factor based on a preset safety level conversion rule to obtain the equipment safety level, including: Performing joint torque analysis on the motion load safety factor to obtain a joint torque distribution vector; Performing mechanical equivalent transformation on the joint torque distribution vector to obtain a posture damage safety value; Performing a safety margin correction on the motion load safety factor according to the posture damage safety value to obtain a human body protection priority coefficient; performing equipment conflict optimization on the human protection priority coefficient and the calibrated mechanical load data based on a preset anti-saturation constraint rule to obtain an equipment protection strength attenuation factor; The device protection strength attenuation factor is reconstructed according to the security level conversion rule to obtain the device security level.

[0009] Furthermore, the command control analysis is performed according to the real-time motion stage, the dynamic overload parameter of the device and the safety level of the device to obtain a servo module control solution, including: Performing hierarchical rule matching on the device security level based on preset hierarchical intervention measures to obtain an intervention level identifier; Performing real-time load extraction on the equipment data to obtain real-time operating parameters of the equipment; Performing action phase segmentation on the real-time motion phase to obtain an action interval set; Performing intensity iteration on the intervention level identifier in combination with the dynamic overload parameter of the device and the real-time operating condition parameter of the device to obtain a dynamic intervention intensity calibration value; Performing servo control instruction conversion on the dynamic intervention intensity calibration value to obtain an overload control instruction set; The action interval set is coupled with the overload control instruction set to obtain the servo module control solution.

[0010] Furthermore, the device security level is matched with hierarchical rules based on the preset hierarchical intervention measures to obtain an intervention level identifier, including: Performing overload threshold stratification on the safety level of the equipment to obtain multi-level overload protection threshold intervals; Performing servo response mapping on the hierarchical intervention measures according to the multi-level overload protection threshold intervals to obtain a servo intervention measure mapping table; Perform overload parameter identification on the equipment safety level to obtain a real-time value of the overload status; Perform load interval positioning on the multi-level overload protection threshold interval according to the real-time value of the overload state to obtain a current overload interval; The servo intervention measure mapping table is dynamically matched according to the current overload interval to obtain the intervention level identifier.

[0011] Furthermore, the method further includes obtaining execution feedback information of the servo module control solution, and performing device security recording together with the device data and the device security level to obtain a security event log file: Performing multi-dimensional execution feedback collection on the servo module control solution to obtain the execution feedback information; performing abnormal pattern recognition on the device data according to the execution feedback information to obtain a dynamic device operation status indicator; Performing deep semantic fusion on the dynamic device operating status indicator and the device security level to obtain a hierarchical security status description; Perform reasoning and matching on the hierarchical security status description according to a preset security event knowledge graph to obtain an accurate event type identification; The precise event type identifier and the hierarchical security status description are written hierarchically into a preset encrypted log file to obtain the security event log file.

[0012] The present invention further provides a servo module control device for equipment safety overload protection, which is applied to any of the above-mentioned servo module control methods for equipment safety overload protection, comprising: A collection module is used to obtain device data of smart devices and perform basic security analysis based on preset device threshold parameters to obtain the device security status; An analysis module, configured to obtain a user profile and real-time motion stage of the smart device, intelligently fuse the device data, and obtain a dynamic overload parameter of the device; an association module, the association module being configured to perform device classification judgment based on the dynamic overload parameter of the device and the safety status of the device to obtain a device safety level; A processing module is used to perform command control analysis according to the real-time motion stage, the dynamic overload parameter of the device and the safety level of the device to obtain a servo module control solution.

[0013] The present invention provides a servo module control method and device for equipment safety overload protection, which has the following beneficial effects: Through real-time safety analysis based on preset threshold parameters, potential hidden dangers can be discovered in a timely manner, solving the problem of delayed response in traditional static judgments. By combining user profiles and real-time exercise stages for intelligent data fusion, dynamic overload parameters can be accurately calculated, breaking through the limitations of a single data dimension and comprehensively improving the accuracy of safety risk identification. The multi-level safety assessment system enables the system to take appropriate measures according to different risk levels, avoiding excessive or insufficient protective intervention and enhancing system adaptability. Command control analysis based on multi-dimensional parameters achieves precise safety control for different training intensities and modes, adapting to individual differences and special needs, and improving equipment safety and user experience. Taking into account user characteristics and training stage characteristics, the system can flexibly adjust protection strategies to adapt to diverse training scenarios, effectively balancing training effects and safety assurance, and providing a more scientific and safe fitness experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1This is a flow chart of a servo module control method for equipment safety overload protection provided by the present invention; Figure 2 This is a structural diagram of a servo module control device for equipment safety overload protection provided by the present invention.

[0015] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0016] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0017] The present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0018] Reference Figure 1 As shown, the present invention provides a servo module control method for equipment safety overload protection, comprising: Step S1: Obtain device data of the smart device and perform basic security analysis based on preset device threshold parameters to obtain the device security status; Step S2: Obtain the user profile and real-time motion stage of the smart device, perform intelligent fusion on the device data, and obtain the dynamic overload parameter of the device; Step S3: Perform equipment classification based on the equipment dynamic overload parameter and equipment safety status to obtain the equipment safety level; Step S4: Perform command control analysis based on the real-time motion stage, the dynamic overload parameters of the equipment, and the equipment safety level to obtain a servo module control solution.

[0019] Based on the above steps, the detailed process is as follows: Step S1: A network of multiple sensors collects operating parameters of intelligent devices in real time, including key operational data such as motor speed, torque load, current power, mechanical tension, temperature, and vibration frequency. This data is aggregated by the device's built-in data acquisition module and transmitted to the safety control unit for processing. The safety control unit performs basic safety analysis on the collected data based on a pre-defined device threshold parameter library. This threshold parameter library contains the upper and lower limits of various parameters required for normal operation of the device, such as the maximum allowable continuous operating temperature, maximum allowable current, torque limit, and vibration safety threshold. The analysis process utilizes a multi-dimensional parameter monitoring algorithm to perform real-time comparisons of various parameters to determine whether the device is within its safe operating range.

[0020] When any parameter exceeds the safety threshold, the system records the corresponding abnormal state. The basic safety analysis outputs device safety status indicators, including normal state, mild abnormal state, moderate abnormal state, and severe abnormal state. This device safety status serves as an important input parameter for subsequent steps, providing the basis for the device's dynamic protection strategy. It is worth noting that the basic safety analysis uses static threshold comparison, which focuses on security at the device hardware level and does not consider individual user differences and the dynamic characteristics of the movement process. This is precisely what needs to be supplemented in subsequent steps.

[0021] Step S2: The system extracts user profiles from the user management database, including basic physiological parameters (height, weight, age, gender, etc.), training history (training frequency, intensity, duration, etc.), health status markers (cardiopulmonary function, muscle strength level, joint mobility, etc.), and personal training goal settings. It also identifies the system's current real-time exercise phase, including warm-up, main training phase, recovery and adjustment phase, or cool-down phase. The system uses a motion curve feature analysis algorithm, combined with exercise time, intensity trends, and biofeedback signals (such as heart rate changes), to accurately identify the current exercise phase.

[0022] Intelligent fusion processing utilizes a multi-source heterogeneous data fusion algorithm to integrate and analyze acquired device data with user profiles and real-time motion stage information. The algorithm uses a weighted feature extraction method to dynamically adjust the weight coefficients of various parameters based on the characteristics of different motion stages, forming personalized device dynamic overload parameters for the current user and current motion state. These dynamic overload parameters include user-adaptive maximum load limits, stage-by-stage power adjustment coefficients, fatigue compensation factors, motion safety margins, etc. Unlike static thresholds, the generated device dynamic overload parameters are highly personalized and context-adaptive, and can be adjusted in real time based on individual user differences and dynamic changes in the motion process, providing a more accurate decision-making basis for subsequent safety level determinations. The output result of this step, the device dynamic overload parameters, will work together with the device safety status in the device safety level determination process.

[0023] Step S3: The hierarchical judgment utilizes a hybrid decision-making approach combining fuzzy logic control algorithms with neural networks, constructing a five-dimensional risk assessment space encompassing the dimensions of device status, user adaptability, movement phase, cumulative load, and safety margin. In the device status dimension, the system assesses the extent to which hardware parameters deviate from the safe range; the user adaptability dimension considers the match between the current load and the user's training capabilities; the movement phase dimension focuses on specific risk factors in different movement phases; the cumulative load dimension assesses the cumulative effects of fatigue over the continuous operation of the device; and the safety margin dimension calculates the safe distance between the current operating state and the limit state.

[0024] Based on the comprehensive scores of these five dimensions, the system divides the equipment safety level into four levels: green safety level (normal operation, no risk), blue caution level (minor risk, requires monitoring), yellow warning level (moderate risk, requires adjustment) and red danger level (serious risk, requires immediate intervention). During the judgment process, the system uses dynamic weighting to assign weights to different dimensions. For example, during high-intensity training, the weight of the user adaptability dimension increases; and after the equipment has been running for a long time, the weight of the cumulative load dimension increases. This multi-dimensional risk assessment method significantly improves the accuracy and adaptability of safety judgments and can capture potential safety hazards under complex conditions. The output result of step S3 - the equipment safety level, serves as the direct basis for formulating a protection and control plan in step S4, ensuring that the system can take corresponding levels of protection measures according to the degree of risk.

[0025] Step S4: Comprehensively analyze the three core elements of real-time motion phase, equipment dynamic overload parameters, and equipment safety level, and generate the optimal protection control plan through adaptive decision-making control algorithm. During the control plan formulation process, the system constructs a multi-objective optimization model to balance the two goals of ensuring training effect and equipment safety protection. For different safety levels, the system adopts differentiated response strategies: At the green safety level, normal control parameters are maintained and preventive monitoring is initiated. At the blue caution level, the resistance curve is fine-tuned, peak load is reduced, and intelligent power limiting is initiated, while a slight prompt message is sent to the user interface. At the yellow warning level, the load curve is significantly adjusted, the exercise intensity is reduced by 20%-50%, the transition time is extended, and a clear warning is sent to the user; At the red danger level, the emergency protection mechanism is activated, including gradual load reduction until safe shutdown, electromagnetic brake intervention, and cutting off high-power circuits.

[0026] Based on the characteristics of the real-time exercise phase, the implementation of protective controls is optimized. For example, if risks are detected during the warm-up phase, the warm-up period is extended and the starting load is reduced. During the main training phase, the load curve is replanned to ensure training effectiveness while avoiding risk points. The control scheme is executed using a servo closed-loop feedback mechanism, monitoring the execution effect in real time and making fine-tuning adjustments. This step also includes a learning and optimization module, which records the effectiveness of each protective control and user feedback, and continuously optimizes the control strategy library.

[0027] The present invention provides a servo module control method for equipment safety overload protection. Through real-time safety analysis based on preset threshold parameters, it can timely discover potential hidden dangers and solve the problem of delayed response of traditional static judgment. By combining user profiles and real-time motion stages for data intelligent fusion, dynamic overload parameters are accurately calculated, breaking through the limitations of a single data dimension and comprehensively improving the accuracy of safety risk identification. The multi-level safety assessment system enables the system to take corresponding measures according to different risk levels, avoiding excessive or insufficient protection intervention and enhancing system adaptability. Based on the command control analysis of multi-dimensional parameters, precise safety control for different training intensities and modes is achieved, adapting to individual differences and special needs, and improving equipment safety and user experience. Taking into account user characteristics and training stage characteristics, the system can flexibly adjust protection strategies, adapt to diverse scenarios, effectively balance training effects and safety guarantees, and provide a more scientific and safe fitness experience.

[0028] In one embodiment, smart device data of a smart device is obtained, and basic security analysis is performed based on preset device threshold parameters to obtain the device security status, including: The smart device's built-in torque sensor and current monitoring module collect real-time torque and motor current data from the servo motor. Servo torque data directly reflects the device's current load status. The torque sensor samples this data at a high frequency of 100 times per second to obtain precise torque values ​​in Newton-meters (N·m). Motor current data is collected from the motor drive circuit via a Hall effect current sensor at a sampling frequency of 200Hz in amperes (A). These two sets of data are amplified and filtered by the signal conditioning circuit before being transmitted to the main controller.

[0029] The device torque identification process utilizes a torque-current relationship model to compare and analyze servo torque data with motor current data, eliminating the effects of electrical interference and mechanical noise to extract the true load torque value. A temperature compensation factor is incorporated into the model calculation to eliminate the effects of temperature changes on torque measurement. This device torque identification calculation yields device torque load data, including key parameters such as instantaneous torque, average torque, torque fluctuation, maximum torque, and torque-time curve.

[0030] This torque load data comprehensively reflects the mechanical characteristics of the equipment under different motion states, providing basic data support for subsequent multi-dimensional integration of equipment. Accurate acquisition of equipment torque load data is crucial for determining whether the equipment is in an overload state.

[0031] Position encoder data comes from a high-precision photoelectric encoder mounted on the servo motor shaft, with a resolution of 1024 pulses per revolution. It records the motor shaft's rotational angle changes in real time. Velocity sensor data, obtained from a Hall-effect velocity sensor or through differential calculation of position data, records the linear and angular velocities of the device's moving parts. These two sets of data are collected synchronously to ensure timestamp consistency. Motion trajectory analysis utilizes kinematic calculations to process the raw position and velocity data and generate a complete set of motion parameters. This analysis includes position signal denoising, velocity curve smoothing, acceleration calculation, and motion direction determination.

[0032] The analysis results generate equipment motion state data, including the position coordinates of the moving parts, motion velocity vectors, acceleration values, motion direction indications, motion range limits, and complete position-time and velocity-time curves. These equipment motion state data comprehensively describe the motion characteristics of the equipment in three-dimensional space and reflect whether the equipment's motion state meets the expected design. The motion state data, together with the torque load data obtained in the previous step, constitute a complete description of the physical state of the equipment, laying the foundation for the next step of multi-dimensional equipment integration. Abnormal changes in motion state data are often early signs of equipment failure or improper use. By accurately capturing these changes, it is helpful to promptly identify potential safety hazards.

[0033] The equipment torque load data and equipment motion status data obtained in the first two steps are integrated and analyzed in multiple dimensions. The multi-dimensional fusion of equipment uses data association analysis methods to establish a three-dimensional relationship model of torque, position and speed. Through data pairing and cross-validation at corresponding time points, the logical association between various physical quantities is realized. The fusion process calculates a series of comprehensive performance indicators, including power indicators (the product of torque and angular velocity), energy consumption indicators (the integral of current and voltage), efficiency indicators (the ratio of output power to input power), stability indicators (the variance of torque fluctuations and speed fluctuations), and coordination indicators (the smoothness of the position-torque curve). These indicators are organized into structured equipment data to form a comprehensive description of the working status of the equipment.

[0034] Device data consists of three layers: raw data, characteristic parameters, and performance indicators, forming a complete digital representation of the device's operating status. This data integrates the device's mechanical, kinematic, and energy characteristics, reflecting its comprehensive performance during actual use. Multi-dimensional device integration not only improves data reliability and integrity but also reveals the device's internal state, which is difficult to visualize with single-dimensional data.

[0035] Systematically compare device data against preset threshold parameters to determine the current safe operating status of the device. Preset device threshold parameters are stored in the device control system's parameter library and include safety limits such as maximum allowable torque, maximum allowable speed, maximum allowable power, maximum allowable operating temperature, maximum allowable vibration amplitude, and maximum allowable continuous operating time. These threshold parameters are based on device design specifications, material strength, motor characteristics, and safety standards, and represent the boundary conditions for safe device operation.

[0036] Threshold comparison analysis utilizes a multi-parameter threshold determination method, comparing each metric in the device data obtained in the previous step with the corresponding threshold parameters. This comparison process calculates the safety margin for each metric—the percentage difference between the actual value and the threshold. Based on these comparison results, a comprehensive assessment is conducted to determine the device's safety status, which is categorized into four levels: safe (all parameters are below 80% of the threshold), caution (any parameter reaches 80%-90% of the threshold), warning (any parameter reaches 90%-100% of the threshold), and critical (any parameter exceeds the threshold).

[0037] The device safety status assessment includes the overall safety level, specific parameters that trigger warnings, a quantitative value for the degree of danger, and a safety margin analysis. This assessment, as the final output, directly influences the development of subsequent protection strategies.

[0038] This embodiment achieves accurate equipment torque identification by synchronously collecting servo torque data and motor current data, providing a reliable basis for the early detection of load anomalies. Combining the motion trajectory analysis of position encoder and speed sensor data, the system can fully grasp the motion state of the equipment, avoiding the limitations of single parameter monitoring. Multi-dimensional fusion technology integrates torque load data with motion state data to construct a three-dimensional monitoring model of equipment operation, revealing potential risks that are difficult to detect with traditional single-dimensional monitoring. Through refined threshold comparison and analysis, the system can achieve accurate judgment of the four-level safety status, thereby providing early warning before the equipment becomes abnormal, greatly improving the sensitivity and accuracy of safety protection.

[0039] In one embodiment, the user profile and real-time motion stage of the smart device are obtained, and the device data is intelligently integrated to obtain the device dynamic overload parameter, including: The smart device's array of high-definition visual sensors captures the user's movements during training in real time, including visual feature data such as body position, joint angles, and motion trajectory. Simultaneously, a matrix of pressure sensors at key contact points collects pressure distribution data generated by the user's interaction with the device, recording parameters such as force intensity, pressure center migration, and pressure duration.

[0040] These two types of sensor data are preprocessed by the device's built-in data processing unit to eliminate environmental noise and standardize the data format. The exercise phase recognition module uses time series feature extraction technology to segment and analyze the continuously collected data, identifying typical characteristic patterns of exercise phases. Based on a training cycle model, this module divides the overall training process into a warm-up phase, a main training phase, a transition and adjustment phase, and a cool-down and recovery phase.

[0041] The warm-up phase is characterized by a low-intensity, progressive exercise pattern, with pressure sensor data showing a gradually increasing trend; the main training phase is characterized by a high-intensity, repetitive exercise pattern, with significant fluctuations in pressure data and a high peak value; the transition adjustment phase is characterized by reduced intensity and short pauses; the relaxation recovery phase presents a continuously reduced pressure pattern and soothing movement characteristics.

[0042] The results of motion phase recognition are output as real-time motion phase identifiers, including information such as the current phase type, duration, and degree of completion. This provides a timing framework for subsequent dynamic overload parameter calculations. The real-time motion phase information output from this step directly influences the selection of strategies for subsequent dynamic action fusion, ensuring consistency between safety measures and the training process.

[0043] The motion recognition processing unit draws upon a pre-stored library of motion specifications, which contains standardized reference data for various fitness movements, including standard posture models, joint angle ranges, and motion trajectory curves. These specifications, defined by professional athletes and fitness coaches, cover the full range of training movements supported by smart devices. The motion recognition processing unit then performs deep feature extraction on the visual sensor data acquired in the previous step, locating key points of the user's body, including skeletal joints such as the head, shoulders, elbows, wrists, hips, knees, and ankles.

[0044] The 3D skeleton model of the user's current posture is reconstructed based on the positional relationships between skeletal points. Key posture parameters such as joint angles, body tilt, and movement symmetry are calculated. The action recognition module compares the reconstructed user posture with standard action information in real time, calculating posture matching, deviations, and potential risk points. This comparison process accounts for individual differences and allows for reasonable posture variations within a safe range. The action recognition results output action posture data, including multi-dimensional information such as the currently executed action type, a score for matching the standard action, identification of key deviation points, and an assessment of posture stability.

[0045] This posture data not only reflects the standardization and safety of the user's current movements but also indirectly reflects their technical proficiency and body control. As an important intermediate output, posture data will be combined with the pressure identification results in the next step to form a comprehensive assessment of the user's movement quality and potential risks, providing key input for the final calculation of the device's dynamic overload parameters.

[0046] The pressure recognition module receives the motion posture data and raw pressure sensor data output from the previous step and constructs a posture-pressure mapping relationship. A network of pressure sensors is deployed at user contact points such as the device's handles, seat cushion, backrest, and pedals, providing comprehensive pressure monitoring coverage. The pressure recognition processing unit first spatially registers the raw pressure data, accurately mapping the pressure point locations to the body parts in the motion posture data and establishing a pressure-posture spatial correlation model. This model then calculates the pressure distribution characteristics of each contact point, including parameters such as the pressure center location, pressure uniformity, and pressure change rate. For dynamic motion, the pressure recognition module tracks the temporal pattern of pressure changes, identifying the timing, duration, and release characteristics of force application, and assessing the coordination and stability of the user's force application. Combined with the joint angle information in the motion posture data, the pressure recognition module calculates the joint force state, assesses the appropriate force ratios for different body parts, and identifies dangerous postures that may lead to excessive joint stress.

[0047] The pressure recognition results output posture pressure parameters, which include multi-dimensional information such as force distribution balance, force stability, joint pressure safety index, and pressure peak and valley values. The posture pressure parameters comprehensively reflect the user's force characteristics and potential risk points during the execution of the action, revealing improper force issues that are not easily discovered from purely visual data. The output of this step forms a progressive relationship with the previous two steps, deeply analyzing the user's movement state from movement phase, movement posture to mechanical characteristics, providing comprehensive basic data for subsequent user force assessment and dynamic overload parameter calculation.

[0048] Smart devices connect to the user's database through a secure authentication channel, accessing authorized training profile data. These profiles contain basic user information (age, gender, height, weight, and health status), training history (training frequency, duration, and completion rate), historical load parameters (weight, resistance, and speed settings), and training effect feedback (fatigue assessment, soreness feedback, and progress). The Historical Training Recognition module analyzes this data over time to create a user's training development curve, identifying strength growth patterns, adaptability characteristics, and potential bottlenecks.

[0049] By comparing the quality of execution and load parameters of similar exercises over time, the module assesses the user's proficiency progression and strength gains in specific movements. Historical Training Recognition also analyzes performance characteristics during training breaks and recovery periods, analyzing adaptive recovery patterns and assessing the gap between current training status and optimal performance. The module comprehensively considers the user's recent training density, intensity changes, and subjective feedback to assess potential fatigue accumulation.

[0050] The historical training recognition results output user strength assessment data, including multi-dimensional information such as absolute strength level score, relative historical best state ratio, fatigue risk assessment, and strength difference distribution of different movement types. This user strength assessment data reflects the user's current training ability level and physical condition, providing an important basis for personalized adjustment of the device's dynamic overload parameters. This step complements the previous three steps, which focus on current state analysis, while this step provides a longitudinal comparison of historical dimensions, allowing the final dynamic overload parameters to fully consider the user's training history and individual characteristics.

[0051] The dynamic motion fusion module receives the outputs of the first four steps: real-time motion phase, motion posture data, posture pressure parameters, and user force assessment data, combined with device data (such as motor torque, running resistance, speed, and other device operating parameters). Dynamic motion fusion uses a hierarchical weighted integration method to construct a multidimensional safety assessment space.

[0052] The first level of fusion focuses on the temporal dimension, adjusting safety thresholds based on the real-time phase of movement: conservative safety parameters are used during the warm-up phase, higher loads are allowed during the main training phase, and the upper limit of load is significantly reduced during the recovery phase.

[0053] The second layer of fusion focuses on the action quality dimension, combining posture data with posture pressure parameters to assess the standardization and safety of the current action execution and identify key risk points. When irregular posture and unreasonable pressure distribution are detected simultaneously, the safety risk rating is significantly improved.

[0054] The third layer of fusion integrates individual user characteristics, using user strength assessment data as a benchmark to set personalized load safety intervals, providing more stringent protection limits for weak movements. The dynamic motion fusion module generates dynamic overload parameters for the device through these three layers of fusion calculations, including control parameters such as the maximum allowable load factor, dynamic safety margin, overload response time constant, and force distribution optimization coefficient. These parameters not only reflect the safety threshold under the current motion state, but also include preventative adjustment strategies for potential risks. The dynamic overload parameters of the device serve as the final output and will be passed to subsequent device classification judgments. Together with the device safety status, they determine the device safety level, ultimately influencing the formulation of the protection control plan.

[0055] This embodiment implements a safety overload protection control method based on multimodal data fusion on smart devices, thereby achieving comprehensive monitoring and dynamic adjustment of the user's training process. This method combines visual sensor data with pressure sensor data to identify the stage of motion, which not only accurately determines the user's training stage, but also adjusts safety parameters according to the characteristics of different stages. Posture recognition based on preset action specification information is combined with pressure recognition to comprehensively evaluate the quality of user movements and force characteristics, and can promptly discover potential risks that are not easy to detect through a single sensor. Introducing user profiles for historical training identification enables the system to formulate personalized safety thresholds based on the user's individual characteristics and training history, avoiding the problem of over-protection or under-protection caused by fixed parameters. Through multi-level dynamic action fusion, the system can generate dynamic overload parameters for the device in real time, providing a precise safety control basis for smart devices, significantly improving the device safety performance and user training experience. In one embodiment, a device classification judgment is performed based on the device dynamic overload parameter and the device safety status to obtain the device safety level, including: The process of extracting composite load data from the dynamic overload parameters of the equipment involves separating two types of key load indicators from the dynamic overload parameters of the equipment. This process uses a data decomposition algorithm to perform multi-dimensional decomposition of the dynamic overload parameters of the equipment to extract the mechanical load data and motion load data of the equipment. The mechanical load data of the equipment mainly includes load parameters related to the equipment hardware, such as motor torque, bearing stress, transmission system tension, and mechanical structure vibration frequency. These parameters directly reflect the actual stress state of the physical components of the equipment. The motion load data focuses on the load characteristics related to the user's training behavior, including parameters such as the motion power curve, speed change rate, impact load frequency, and motion duration. These parameters reflect the dynamic load characteristics generated by the user's training process on the equipment.

[0056] Composite load data extraction utilizes a combined frequency-domain and time-domain analysis method. By converting the time-domain signal to the frequency domain, load components in different frequency bands are identified. The time-domain signal is then reconstructed through an inverse transform, effectively separating mechanical and motion loads. This step also includes outlier detection and smoothing to remove data noise that could interfere with subsequent analysis.

[0057] This step clearly separates the previously mixed dynamic overload parameters of the equipment into two independent but interrelated load data sets, laying the foundation for subsequent differentiated processing. The mechanical load data will undergo abnormal attenuation correction in subsequent steps, while the motion load data will be mapped to load factors. Together, these two data sets form the core basis for determining the equipment's safety level.

[0058] The quantification and encoding of the equipment's safety status converts it into a numerical representation that can be quantified and calculated. This process utilizes a multi-level fuzzy quantification method to convert the qualitative description of the equipment's safety status into a quantitative safety status value. The quantification code uses a standardized scale of 0-100, with 0-25 indicating a severe abnormality, 25-50 indicating a moderate abnormality, 50-75 indicating a mild abnormality, and 75-100 indicating a normal state. The quantification process comprehensively considers multiple safety indicators, including the degree of parameter excursion, the duration of the abnormality, the frequency of parameter fluctuations, and the effects of multiple parameter linkages.

[0059] For different safety parameter categories, weight coefficients are used to adjust their influence on the final quantized value. For example, different weights are assigned to electrical safety parameters, mechanical safety parameters, and thermal safety parameters. The state quantization encoding process also includes a historical state fusion mechanism. This mechanism uses an exponentially weighted average to combine the current state with the historical state to form a state quantization value with temporal continuity, thus avoiding misjudgments caused by transient fluctuations. This quantization encoding not only transforms the representation of the safety state but also provides a unified quantization benchmark for subsequent attenuation calibration and load factor mapping.

[0060] The safety state quantification value serves as a key intermediate variable and will be used in the subsequent two steps to adjust the equipment's mechanical load data and motion load data, respectively, to achieve differentiated load processing based on the safety state. Lower quantification values ​​indicate a worse safety state, leading to more stringent treatment measures in the subsequent attenuation calibration and load factor mapping.

[0061] By using a nonlinear attenuation function, the mechanical load data of the equipment is calibrated and adjusted to reflect the changes in the equipment's actual bearing capacity under different safety conditions. The calibration algorithm constructs a dynamic attenuation factor matrix based on the quantified safety state value, making targeted adjustments to each dimension of the mechanical load data. Attenuation calibration follows the principle of "the worse the state, the stronger the attenuation." When the quantified safety state value is in the low value range (0-25), a high-intensity attenuation coefficient (0.4-0.6) is used to significantly reduce the mechanical load. When the quantified value is in the mid-range (25-75), a moderate attenuation coefficient (0.7-0.9) is used for moderate adjustment. When the quantified value is in the high value range (75-100), a slight attenuation coefficient (0.9-1.0) is used for fine-tuning.

[0062] The calibration process differentiates between different types of mechanical loads, employing differentiated attenuation strategies for structural stress loads, dynamic impact loads, thermal loads, and more. Furthermore, the calibration algorithm considers the cumulative effect of loads, applying an additional attenuation factor to prolonged high loads to prevent fatigue damage. The calibrated mechanical load data more accurately reflects the actual load-bearing capacity of the equipment in its current safety state, providing a revised mechanical load baseline for the final equipment safety assessment. This output, along with the motion load safety factor processed in subsequent steps, contributes to the final equipment safety level determination.

[0063] The load factor mapping process of the motion load data based on the safety state quantification value converts the motion load data into a load factor for safety assessment. This process uses an adaptive mapping model to establish a corresponding relationship between the safety state quantification value and the motion load safety factor.

[0064] The mapping model adopts a piecewise function structure and adopts different mapping strategies in different safety state intervals: when the quantified value of the safety state is lower than the critical threshold, the mapping function decreases exponentially, resulting in an extremely low exercise load safety factor and forcibly limiting exercise intensity; when the quantified value of the safety state is in the middle interval, the mapping function changes linearly, resulting in a load safety factor proportional to the safety state; when the quantified value of the safety state is close to the optimal interval, the mapping function tends to be flat, resulting in a load safety factor close to but not exceeding 1.0, to ensure maximum training effect.

[0065] The mapping process takes into account the multidimensional characteristics of the exercise load, sets independent mapping channels for different types of exercise load parameters (such as speed, power, and impact), and generates a multidimensional exercise load safety factor. The system integrates the personal adaptability factors in the user profile and makes personalized adjustments to the mapping function, so that the same safety state quantification value produces differentiated load safety factors for users with different training levels. The mapping result - the exercise load safety factor is a dimensionless value between 0 and 1 that directly represents the degree of impact of the current exercise load on equipment safety. The lower the value, the greater the threat posed by the exercise load to equipment safety. This exercise load safety factor, together with the calibrated mechanical load data, constitutes a dual input for equipment safety assessment.

[0066] The equipment safety assessment process combines the previously processed results of the calibrated mechanical load data and motion load safety factors based on pre-defined safety level conversion rules to generate a final equipment safety level. This process uses a decision tree approach combined with threshold determination to perform a multi-dimensional assessment according to pre-defined safety level conversion rules.

[0067] The safety assessment utilizes a two-dimensional decision matrix, with the horizontal axis representing calibrated mechanical load data and the vertical axis representing the exercise load safety factor. The matrix is ​​divided into four safety zones: green safety zone (low mechanical load, high safety factor), blue caution zone (medium-low mechanical load or medium-high safety factor), yellow warning zone (medium-high mechanical load or medium-low safety factor), and red danger zone (high mechanical load, low safety factor). A device type correction factor is incorporated into the assessment process to adjust the decision matrix's zone division thresholds for different types of smart devices (such as treadmills, elliptical machines, and strength trainers). Furthermore, based on the current real-time performance phase, more relaxed safety standards are applied during warm-up or cool-down phases, standard safety standards are used during main training phases, and stricter safety standards are applied during sprint or extreme training phases.

[0068] The safety assessment also considers time, analyzing the changing trends of calibrated mechanical load data to identify rapid load increases or cumulative effects, and issuing early warnings. The resulting equipment safety level directly determines the selected protection and control scheme. A green safety level allows normal equipment operation, a blue caution level triggers minor adjustments, a yellow warning level activates moderate intervention measures, and a red danger level initiates mandatory protection mechanisms. This safety assessment mechanism accurately quantifies and categorizes equipment operational risks, providing a scientific basis for subsequent control decisions.

[0069] This embodiment achieves accurate separation of mechanical load and motion load by extracting composite load data from the dynamic overload parameters of the equipment, laying the foundation for subsequent differentiated processing. A multi-level fuzzy quantization method is used to encode the safety status of the equipment, and the qualitative description is converted into a standardized scale so that the safety status can be accurately calculated. Based on the nonlinear attenuation calibration mechanism of the safety status quantification value, the mechanical load data can reflect the actual load-bearing capacity of the equipment under different safety conditions. The motion load safety factor is generated by an adaptive mapping model, which effectively quantifies the degree of influence of the motion load on the safety of the equipment. Finally, a two-dimensional decision matrix is ​​used for safety assessment, which comprehensively considers the equipment type, motion stage and time factors to achieve accurate classification of equipment operation risks. The whole set of methods establishes a complete equipment safety assessment system, which significantly improves the safety control accuracy and reliability of the servo module under overload conditions.

[0070] In one embodiment, the equipment safety assessment is performed on the calibrated mechanical load data and the motion load safety factor based on a preset safety level conversion rule to obtain the equipment safety level, including: The joint torque analysis process for motion load safety factors establishes a mapping relationship between human kinematics and equipment loads. This modeling process, based on the principles of human biomechanics, converts the motion load safety factors into torque data for each joint. The modeling process utilizes a multibody dynamics model, simplifying the human body into a joint system connected by rigid rods. This includes major motion joints such as the ankle, knee, hip, lumbar spine, shoulder, and elbow. Through forward kinematic analysis, the angular change, angular velocity, and angular acceleration of each joint during motion are calculated. The external torque acting on each joint is calculated by combining the resistance characteristics provided by the equipment and the motion load safety factors.

[0071] The model incorporates personalized parameters such as body weight distribution and limb length ratios to adapt to individual differences. The modeling outputs a joint torque distribution vector, which contains the torque values ​​experienced by all key joints during the current motion state. These torque values ​​reflect the actual effects of motion loads on various joints, providing foundational data for subsequent mechanical equivalent conversions and safety assessments.

[0072] The mechanical equivalent conversion of joint torque distribution vectors transforms local joint forces into global motion risk. This conversion applies the principle of biomechanical equivalence to integrate the torque data of each joint into a unified postural injury safety value. During the mechanical equivalent conversion process, different joint torque values ​​are assigned different weighting coefficients to reflect the differences in joint tolerance and injury sensitivity. The joint torque is then ratioed to its corresponding maximum safe load torque to determine the relative load rate of each joint.

[0073] The conversion process considers the coupling effect between joints and jointly analyzes the stress states of adjacent joints to assess the combined impact of compound movements on joints. A movement duration factor is also introduced to assess the cumulative risk of joint damage from long-term repetitive movement. The posture injury safety value (PAS) derived from the equivalent conversion is a comprehensive indicator that quantifies the overall risk of injury to the human joint system caused by the current movement posture. This safety value uses a standardized scale of 0-100, with lower values ​​indicating a higher risk of PAS injury.

[0074] The safety margin correction process for the motion load safety factor based on the posture injury safety value achieves load adjustment based on human safety. This correction process generates a human protection priority coefficient by establishing a corresponding relationship between the posture injury safety value and the safety margin. The correction process utilizes a nonlinear mapping function. When the posture injury safety value falls below the warning threshold, the safety margin is significantly increased; when the safety value is within the normal range, a moderate safety margin is maintained.

[0075] The correction algorithm incorporates a dynamic compensation mechanism to predictively adjust the safety margin based on the changing trend of the posture injury safety value during exercise. Safety margin correction takes into account the user's exercise level and physical condition, employing differentiated correction strategies for different user types. The corrected human protection priority coefficient is an adjustment factor between 0 and 1 that directly affects the strength of the device's protection control. A lower coefficient indicates a higher level of human protection, resulting in a more conservative device control strategy. Safety margin correction ensures that while maintaining training effectiveness, human safety is prioritized to prevent injuries caused by improper loading.

[0076] The conflict between human and equipment protection is resolved by optimizing the equipment's conflicting priorities and calibrating mechanical load data based on pre-set anti-saturation constraints. This optimization process balances human safety requirements with equipment operating limits through anti-saturation constraints. Upper and lower equipment load limits are set during the optimization process to prevent overly aggressive protection measures from causing equipment operating points to deviate from their normal range.

[0077] Among them, the anti-saturation constraint rules include: Load limit rules: Set upper and lower thresholds for equipment load to ensure that the actual workload of the equipment will not exceed its physical limits or fall below the minimum effective operating point regardless of how safety protection measures are activated.

[0078] Dynamic adjustment rules: Dynamically adjust the rate of change of protection strength according to the current operating status of the equipment to prevent sudden changes in protection measures from causing jumps or oscillations in equipment control parameters.

[0079] Smooth transition rule: When it is necessary to reduce or increase the protection strength, a gradual adjustment mechanism is adopted to ensure a smooth transition of the equipment operating parameters and avoid mechanical shock caused by sudden changes.

[0080] Priority balancing rule: When there is a conflict between human protection needs and equipment protection needs, a compromise solution is determined based on the preset priority matrix to maximize equipment protection while ensuring human safety.

[0081] Feedback compensation rules: By monitoring the response of the equipment in real time, closed-loop adjustments are made to the protection intensity to compensate for control deviations caused by nonlinear load characteristics.

[0082] Time constraint rules: Considering the time accumulation effect of equipment load, a higher tolerance is allowed for load peaks in a short period of time, but stricter constraints are imposed on continuous high loads.

[0083] Anti-saturation constraints ensure that the equipment's actual operating parameters remain within its safe operating range under all circumstances. The optimization algorithm simultaneously processes the human protection priority coefficient and calibrated mechanical load data to generate a device protection strength attenuation factor. This attenuation factor comprehensively considers human protection requirements and the equipment's load capacity, dynamically adjusting to find the optimal balance when these conflict. The optimization results ensure human safety while avoiding reduced training effectiveness due to excessive equipment protection. The device protection strength attenuation factor, a key control parameter, will be directly used in subsequent safety level determinations.

[0084] The final device security level determination is completed by reconstructing the device protection strength attenuation factor according to the security level conversion rules. This reconstruction process applies the preset security level conversion rules to map the device protection strength attenuation factor to a specific security level.

[0085] Rule reconstruction adopts a hierarchical judgment structure, dividing the value range of the attenuation factor into multiple intervals, each of which corresponds to a different safety level. The reconstruction process integrates the characteristics of the device type, the characteristics of the motion stage, and the user's personalized parameters to dynamically adjust the basic judgment rules. The output of the rule reconstruction is a four-level safety level: green safety level, blue caution level, yellow warning level, and red danger level. Each safety level corresponds to a specific control strategy and protection measures. Rule reconstruction ensures the accuracy and real-time nature of the safety level judgment and provides a clear execution basis for equipment protection control. This step is the final link in the entire safety assessment process, and the output of the equipment safety level directly determines the selection and implementation of the subsequent protection control plan.

[0086] Among them, the security level conversion rules include the following rules: Basic threshold rules: Green safety level: attenuation factor ≥ 0.85; blue caution level: 0.85> attenuation factor ≥ 0.70; yellow warning level: 0.70> attenuation factor ≥ 0.50; red danger level: attenuation factor < 0.50.

[0087] Equipment type characteristic rules: High-precision equipment: the threshold is increased by 5%; high-load equipment: the threshold is decreased by 3%; high-frequency motion equipment: the threshold is increased by 8%; complex equipment: the adjustment coefficient is selected based on the main functional characteristics.

[0088] Rules for the movement phase: Start-up phase: the threshold is relaxed by 10%; stable operation phase: the standard threshold is adopted; deceleration and stop phase: the threshold is tightened by 15%; emergency state: the threshold is tightened by 25%.

[0089] Cumulative effect rule: Continuous running time exceeds the preset value: the threshold is tightened by 2% for every hour exceeding the preset value; frequent start and stop: the threshold is tightened by 1% for each start and stop; load fluctuation: when the fluctuation amplitude exceeds 20%, the threshold is tightened by 5%.

[0090] Environmental factor rules: Temperature exceeds the standard: the threshold is tightened by 1% for every degree exceeding the standard; humidity exceeds the standard: the threshold is tightened by 2% for every 5% exceeding the standard; vibration exceeds the standard: the threshold is tightened by 3% for every 10% exceeding the baseline value.

[0091] User parameter rules: Novice users: the threshold is tightened by 10%; professional users: the threshold is relaxed by 5%; special groups: the threshold is adjusted according to their physical condition.

[0092] Linkage protection rules: If multiple indicators are abnormal at the same time: take the most stringent level; if the trend deteriorates rapidly: issue an early warning one level in advance; if the span of safety levels is too large: enforce a gradual transition.

[0093] Recovery mechanism rules: After the fault is eliminated: the normal threshold is restored in stages; after maintenance: the threshold is temporarily tightened by 10%; after the system upgrade: the recalibration process is executed.

[0094] Special operating conditions rules: Overload operation: the threshold is tightened by 20%; low load operation: the threshold is relaxed by 15%; intermittent operation: dynamic adjustment according to the duty cycle.

[0095] Safety redundancy rules: Critical components: add an additional 5% safety margin; vulnerable parts: add an 8% safety margin; safety protection devices: add a 10% safety margin.

[0096] This embodiment achieves accurate mapping of motion load and human joint force by modeling the motion load safety factor through human joint torque, making the safety assessment closer to the actual motion state. The torque of each joint is integrated into a unified posture damage safety value through mechanical equivalent conversion, effectively quantifying the overall motion risk. The safety margin correction mechanism based on the posture damage safety value realizes the dynamic adjustment of the human protection priority, ensuring personal safety during training. The anti-saturation constraint rule is used for equipment conflict optimization, which effectively resolves the contradiction between human protection and equipment protection, maintaining the training effect while ensuring safety. Through the reconstruction of the safety level conversion rules, a clear four-level safety level system is established, which provides a reliable execution basis for equipment protection control. The overall solution realizes intelligent safety protection of human-machine collaboration, significantly improving the safety and reliability of equipment use.

[0097] In one embodiment, a command control analysis is performed based on the real-time motion stage, the dynamic overload parameters of the device, and the device safety level to obtain a servo module control solution, including: Tiered intervention measures include four basic intervention levels: observation, warning, restriction, and emergency. During the hierarchical rule matching process, device security levels are mapped to pre-set tiered intervention measures. The green safety level is mapped to the observation level, resulting in the observation level intervention identifier O-Level; the blue caution level is mapped to the warning level, resulting in the warning level intervention identifier W-Level; the yellow warning level is mapped to the restriction level, resulting in the restriction level intervention identifier L-Level; and the red danger level is mapped to the emergency level, resulting in the emergency level intervention identifier E-Level.

[0098] Each intervention level identifier contains the corresponding basic intervention parameter set. The observation layer intervention parameter set includes regular monitoring frequency, data recording period, etc.; the warning layer intervention parameter set includes load adjustment coefficient, power limit threshold, etc.; the restriction layer intervention parameter set includes forced deceleration ratio, overload protection trigger point, etc.; the emergency layer intervention parameter set includes emergency braking parameters, safe shutdown timing, etc.

[0099] The intervention level identifier serves as a benchmark for subsequent dynamic intervention intensity calculations, providing a basis for hierarchical execution of equipment protection control. The intervention level identifier output from this step carries complete hierarchical protection information, ensuring accurate execution of subsequent control plans.

[0100] The real-time load extraction process continuously samples and analyzes equipment operating data. Sampled data includes core parameters such as motor torque, speed, current, and power. This real-time data is acquired through the data acquisition module. The load extraction algorithm filters the raw data to eliminate sampling noise and extract the effective load signal. Real-time equipment operating parameters include key indicators such as real-time load index, power utilization, equipment response time, and load fluctuation rate.

[0101] The real-time load index reflects the actual load level currently borne by the equipment; the power utilization rate represents the ratio of the equipment's output power to its rated power; the equipment's response time describes its dynamic response to load changes; and the load fluctuation rate characterizes load stability. These operating parameters constitute a real-time characterization of the equipment's operating status, providing real-time data support for subsequent dynamic intervention intensity calculations.

[0102] Movement phase segmentation, based on the principles of sports biomechanics, divides a complete movement cycle into multiple functional movement intervals. The movement interval set includes basic units such as the preparation interval, the power interval, the transition interval, and the recovery interval. The preparation interval corresponds to the preparatory phase before the movement begins; the power interval contains the main load phase; the transition interval is the movement transition process; and the recovery interval is the buffering period after the movement is completed. Each interval is marked with timing information such as the start time, end time, and duration.

[0103] The action interval set is divided based on motion trajectory characteristics, and its boundaries are identified by combining kinematic parameters such as angle, velocity, and acceleration. This action interval set provides a timing framework for subsequent scheme coupling, ensuring the coordination and consistency of the protection control scheme with the motion rhythm.

[0104] The intensity iteration process combines dynamic overload parameters with real-time operating parameters to dynamically adjust the basic intervention parameters in the intervention level identifier. The dynamic intervention intensity calibration value includes core parameters such as the intervention timing coefficient, intervention intensity coefficient, and intervention duration coefficient. The intervention timing coefficient determines when the protective measure is triggered; the intervention intensity coefficient determines the intensity of the protective measure; and the intervention duration coefficient specifies the duration of the protective measure.

[0105] During calibration, each coefficient is updated in real time based on the severity of the overload parameters and the changing trends of the operating parameters. The dynamic intervention intensity calibration value reflects the most appropriate protection intensity for the current situation, providing precise control parameters for command conversion.

[0106] Servo control command conversion converts the dynamic intervention intensity calibration value into specific execution instructions. The overload control command set includes basic command types such as speed control, torque control, and position control. Speed ​​control commands are used to adjust movement speed; torque control commands are used to limit output torque; and position control commands are used to constrain movement range. The command conversion process considers the dynamic characteristics of the servo system to ensure the executable nature of the commands. The overload control command set provides specific execution-level instructions for the final solution coupling.

[0107] Solution coupling integrates and matches the timing information of the action interval set with the control information of the overload control instruction set. The servo module control solution includes execution plans such as the interval control strategy, transition control strategy, and emergency control strategy. The interval control strategy specifies the specific control instruction sequence within each action interval; the transition control strategy defines the smooth transition method when switching between intervals; and the emergency control strategy specifies the handling process in abnormal situations.

[0108] The coupling scheme ensures the synchronization of control instructions and action rhythm, guaranteeing the precise intervention and smooth execution of protective measures. The resulting servo module control scheme has complete execution logic and timing arrangements, achieving precise control of equipment safety protection.

[0109] This embodiment achieves precise intervention identification from the observation layer to the emergency layer by matching safety level grading rules based on preset hierarchical intervention measures, ensuring the accuracy of graded execution of protective measures. Real-time load extraction is performed on device data to obtain key operating parameters, including real-time load index and power utilization, providing real-time data support for dynamic intervention intensity calculation. The real-time motion phase is divided into functional action intervals such as preparation, force generation, transition, and recovery, allowing protective measures to be accurately mapped to different stages of the motion cycle. Intensity iteration is performed by combining the device's dynamic overload parameters with real-time operating condition parameters to generate intervention timing, force, and duration coefficients, improving the accuracy and adaptability of protective measures. Servo control instruction conversion converts abstract intervention intensity into specific speed, torque, and position control instructions, ensuring reliable execution. Ultimately, by coupling the action interval set with the overload control instruction set, synchronization of protective measures with the motion rhythm is achieved, ensuring precise control and smooth execution of safety protection.

[0110] In one embodiment, a hierarchical rule matching is performed on the device security level based on the preset hierarchical intervention measures to obtain an intervention level identifier, including: The overload threshold stratification process establishes a refined overload protection threshold interval system based on the risk level of the equipment safety level. The multi-level overload protection threshold interval includes four main levels: safe operation interval (0-60%), mild overload interval (61%-75%), moderate overload interval (76%-90%) and severe overload interval (91%-100%). The safe operation interval corresponds to the normal working state of the equipment, and the various parameters of the equipment in this interval are within the rated range; the mild overload interval indicates that the equipment begins to show a slight overload phenomenon, but it is still within the controllable range; the moderate overload interval indicates that the equipment load has reached the warning level and there is a potential risk; the severe overload interval indicates that the equipment is in a dangerous load state and urgently needs intervention and protection.

[0111] Each interval is assigned specific parameter boundaries, including key indicators such as the upper limit of motor speed, torque limit, power threshold, and upper temperature limit. These threshold intervals are determined based on a combination of equipment performance parameters, safety factors, and historical operating data, providing a baseline reference for subsequent servo response mapping. This hierarchical structure enables precise quantification of overload conditions, laying the foundation for the accurate triggering of protective measures.

[0112] The servo response mapping establishes a correspondence between overload protection threshold intervals and specific intervention measures. The servo intervention measure mapping table contains four response levels: monitoring response, early warning response, limiting response, and emergency response. Monitoring response corresponds to the safe operating range and includes basic measures such as routine data collection and status monitoring; early warning response corresponds to the mild overload range and involves mild interventions such as load warnings and parameter fine-tuning; limiting response corresponds to the moderate overload range and includes significant interventions such as power limiting and speed reduction; and emergency response corresponds to the severe overload range and includes mandatory measures such as emergency braking and safe shutdown.

[0113] The mapping table details the specific execution parameters for each response level, such as response time requirements, execution priority, and intervention intensity coefficient. This mapping ensures a precise match between intervention measures and overload levels, providing a clear basis for real-time control. The establishment of this mapping table standardizes and systematizes protective measures, ensuring consistency and controllability of intervention actions.

[0114] The overload parameter identification process monitors and analyzes the equipment's operating status in real time. The real-time overload status value is composed of several key parameters: load factor index, overload duration, load change rate, and temperature rise rate. The load factor index reflects the ratio of the current load to the rated load; the overload duration records the cumulative duration of the load exceeding the normal load; the load change rate describes the dynamic characteristics of the load; and the temperature rise rate indicates the temperature trend of the equipment. These parameters are continuously collected through a sensor network and processed to obtain standardized state quantification values. The real-time overload status value provides a precise description of the equipment's current operating status, providing a basis for subsequent load range positioning. This identification process ensures the real-time and accurate overload status assessment.

[0115] Load interval positioning compares and analyzes the real-time overload status value with the preset multi-level overload protection threshold intervals. The interval positioning process calculates the specific position of the real-time overload status value within the threshold interval and determines the overload interval to which the current operating point belongs. The positioning result includes information such as the interval type identifier, interval position percentage, and trend direction. The interval type identifier indicates the current overload level; the interval position percentage shows the relative position within the current interval; and the trend direction indicates the direction of load change. The determination of the current overload interval provides accurate interval information for subsequent dynamic matching. This positioning process enables real-time quantitative assessment of the overload status.

[0116] The dynamic matching process associates the current overload interval with the response strategy in the servo intervention action mapping table. The intervention level identifier includes elements such as the response level code, execution parameter set, and timing control information. The response level code identifies the current intervention level to be executed; the execution parameter set contains the specific control parameter values; and the timing control information specifies the execution order and timing requirements.

[0117] The matching process considers the location and changing trends of the overload zone to select the most appropriate intervention measure. The intervention level identifier serves as a direct instruction for subsequent control execution, ensuring the timeliness and adaptability of protective measures. This matching process completes the transition from state identification to control instructions, providing the basis for execution of equipment protection control.

[0118] This embodiment establishes a refined four-level protection threshold interval system by stratifying the overload threshold of the equipment safety level, realizing accurate quantitative management of the equipment operating status and improving the accuracy and reliability of overload protection. By establishing a servo response mapping table, a correspondence is established between the overload protection threshold interval and the specific intervention measures, making the implementation of protection measures more standardized and systematic, and enhancing the pertinence and effectiveness of the intervention measures. Through real-time overload parameter identification and load interval positioning, dynamic monitoring and precise positioning of the equipment operating status are achieved, improving the real-time and accuracy of overload status assessment. Through the dynamic matching mechanism, the precise correspondence between intervention measures and overload levels is ensured, making protection control more timely and adaptable, effectively reducing the risk of equipment overload, extending the service life of the equipment, and improving the safety performance of smart devices.

[0119] In one embodiment, the method further includes obtaining execution feedback information of the servo module control solution, and performing device security recording together with device data and device security level to obtain a security event log file: The execution feedback collection process of the servo module control solution uses a multi-dimensional sensor network to obtain various state parameters after the execution of the control instructions in real time. This step establishes a complete closed-loop monitoring system, covering four dimensions: electrical parameter feedback, mechanical response feedback, thermodynamic feedback, and user interaction feedback. Electrical parameter feedback includes key indicators such as the servo motor's current change curve, power fluctuations, driver temperature changes, and response time delay; mechanical response feedback covers physical quantities such as resistance adjustment accuracy, mechanical motion trajectory deviation, vibration spectrum characteristics, and mechanical component stress distribution; thermodynamic feedback monitors the temperature rise rate, heat distribution pattern, and heat dissipation efficiency of each key component; and user interaction feedback records the user's response to protective measures, such as whether to adjust the training posture or follow the load reduction recommendations.

[0120] Multi-dimensional execution feedback collection utilizes a distributed data collection architecture. Each subsystem independently collects data and synchronizes them via timestamps to ensure data temporal consistency. The collection frequency is differentiated based on parameter importance: high-frequency sampling (above 100 Hz) is used for critical safety parameters, medium-frequency sampling (10-50 Hz) for general status parameters, and low-frequency sampling (1-5 Hz) for environmental parameters. After preliminary filtering and outlier detection, all collected data is converted into a structured execution feedback information data packet containing attributes such as parameter name, value, timestamp, and confidence level. This execution feedback information not only reflects the actual execution performance of the protection and control scheme but also provides a complete data foundation for subsequent anomaly pattern recognition. The quality of execution feedback information directly impacts the accuracy of subsequent anomaly pattern recognition. Therefore, a data quality assessment mechanism is introduced during the collection process to conduct real-time evaluations of data integrity, consistency, and accuracy to ensure that subsequent analysis is based on high-quality data.

[0121] The abnormal pattern recognition phase compares and analyzes execution feedback information with raw device data to identify deviation patterns and abnormal characteristics in device operation. This step establishes a multi-layered anomaly detection framework, consisting of a statistical feature analysis layer, a pattern analysis layer, and an association rule analysis layer. The statistical feature analysis layer identifies abnormal changes in parameter distribution by calculating statistical quantities such as the mean, variance, skewness, and kurtosis of the parameters. The pattern analysis layer utilizes time series decomposition techniques to decompose the parameter sequence into trend terms, cyclic terms, and random terms, detecting abnormal changes in each component. The association rule analysis layer explores the coupling relationships between multiple parameters and identifies abnormal patterns of coordinated changes between parameters. The abnormal pattern recognition process comprehensively considers both instantaneous anomalies and cumulative effects. Instantaneous anomalies focus on sudden changes in parameters and significant deviations within a short period of time, while cumulative effects track slow drift of parameters over long periods of time.

[0122] Based on the identified anomaly patterns, the system generates dynamic device operating status indicators, including device health index, stability score, anomaly severity, development trend forecast, and potential risk assessment. Dynamic device operating status indicators use a combination of quantitative and qualitative expressions. Quantitative indicators represent the degree of each state using a numerical range of 0-100, while qualitative indicators describe the type and characteristics of anomalies. Unlike static threshold monitoring, dynamic device operating status indicators reflect the performance characteristics of the device in the actual operating environment, taking into account the combined influence of environmental factors, user behavior, and the device's own characteristics. This indicator system is adaptive and can continuously adjust the judgment criteria based on the device's usage history to improve the accuracy of anomaly identification. Dynamic device operating status indicators provide structured data support for subsequent deep semantic fusion and are a core component of hierarchical security status descriptions.

[0123] The deep semantic fusion step constructs a multi-dimensional, multi-level semantic representation of the security status by integrating dynamic device operating status indicators with device safety level information. This step uses a knowledge representation learning method to transform numerical status indicators and symbolic safety levels into a unified semantic space representation. The fusion process establishes an equipment status ontology model, which defines the core concepts, attribute relationships, and semantic rules of status description to form a standardized semantic framework. Based on the ontology model, the system maps dynamic device operating status indicators into status feature descriptors, including four categories: functional status, performance status, reliability status, and safety status; at the same time, it transforms the equipment safety level into a risk level descriptor, which includes three dimensions: risk severity, scope of impact, and urgency.

[0124] Through semantic association analysis, the system establishes a mapping relationship between feature descriptors and risk descriptors, forming a unified semantic network. Based on this semantic network, the system generates a hierarchical safety status description, including a low-level description of parameter anomalies, a mid-level description of functional impacts, and a high-level description of safety risks. The low-level description focuses on specific parameter anomalies, such as "the motor temperature continuously exceeds the rated value by 15%." The mid-level description focuses on the impact of anomalies on device functions, such as "the cooling system's efficiency decreases, resulting in frequent triggering of temperature control protection." The high-level description focuses on the overall assessment of safety risks, such as "the device faces a moderate risk of overheating, which may lead to performance degradation and shortened lifespan."

[0125] The hierarchical security status description uses structured natural language, preserving data accuracy while providing a human-understandable representation, facilitating subsequent knowledge graph matching and manual review. This description provides a standardized semantic foundation for the precise classification and recording of security incidents, ensuring consistency and interpretability of security event records.

[0126] The security event knowledge graph matching phase uses semantic reasoning technology to map hierarchical security status descriptions to a predefined event type system, enabling accurate classification and identification of security events. This step is based on a pre-built security event knowledge graph, which contains structured knowledge about various types of security events that may occur on a device, including an event type classification system, event feature descriptions, causal association networks, and severity assessment criteria.

[0127] The knowledge graph utilizes a multi-level event classification system, including primary classification (e.g., hardware failure, software anomaly, improper use, environmental interference), secondary classification (e.g., electrical system failure, mechanical system failure), and tertiary classification (e.g., overcurrent protection triggering, torque limiter activation). The matching process utilizes a graph structure similarity calculation method to convert hierarchical safety status descriptions into event feature graphs, which are then structurally matched with typical event patterns in the knowledge graph. The matching algorithm considers node attribute similarity and edge relationship consistency to calculate a comprehensive similarity score.

[0128] For multiple candidate event types whose similarity reaches the threshold, the system further screens them through a causal reasoning network to analyze the causal consistency between the current state and each event type. The final matching result forms a precise event type identification, which includes the event type code, event name, severity level, and processing priority. The event type code adopts a hierarchical coding scheme, such as E01.03.02, which represents an event of the first category, third subcategory, and second specific type. This precise event type identification provides a standardized classification basis for security events, facilitating subsequent statistical analysis and knowledge accumulation. The knowledge graph has the ability to continuously learn. By recording the consistency between the matching results and the actual situation, it continuously adjusts and optimizes the matching rules to improve classification accuracy. The precise event type identification and the hierarchical security status description together constitute the core content of the security event log, providing structured knowledge support for security management and equipment optimization.

[0129] The security event log generation step writes the precise event type identifier and hierarchical security status description into an encrypted log file according to a hierarchical storage strategy, forming a complete security event record. This step implements a hierarchical storage mechanism, dividing security event information into three levels: core information, detailed information, and raw data, based on the severity and sensitivity of the event.

[0130] The core information layer contains the time of event occurrence, event type identification, impact degree and basic status description, and is open to all permission levels; the detailed information layer contains a complete hierarchical security status description, parameter anomaly details and implemented protection measures, and is only open to maintenance personnel; the original data layer contains the original parameter records and complete execution logs before and after the event, and is only open to system administrators.

[0131] The writing process utilizes a time-series database structure to ensure the chronological order and relevance of event records, supporting time-window-based event query and analysis. To ensure data security, the system implements multiple protection measures for log files, including encryption algorithms to encrypt data content, digital signature technology to ensure data integrity, access control mechanisms to restrict log reading permissions, and data backup mechanisms to prevent log loss. Log files utilize a structured format and support standard data exchange formats such as JSON and XML, facilitating integration with other systems and data analysis.

[0132] The log generation process also includes data compression and storage optimization mechanisms, employing differentiated compression strategies for different data types to reduce storage space usage. Security event log files not only record detailed information about individual events but also, through event correlation analysis, establish temporal relationships and causal links between events, forming a comprehensive security event knowledge base. This knowledge base provides data support for device performance optimization, security policy adjustments, and user behavior guidance, achieving closed-loop optimization of device security management. The standardized recording of security event log files ensures the traceability and analyzability of security events, providing a solid foundation for security management throughout the device lifecycle.

[0133] This embodiment achieves all-round monitoring of multiple dimensions such as electrical, mechanical, thermal and user interaction by collecting multi-dimensional execution feedback of the servo module control solution, significantly improving the integrity and accuracy of data collection. Through deep semantic fusion technology, dynamic equipment operating status indicators are integrated with equipment security levels to build a standardized semantic expression system, enhancing the comprehensibility and explainability of security status descriptions. Using the preset security event knowledge graph for reasoning and matching, accurate classification and identification of security events are achieved, improving the accuracy and efficiency of event classification. The use of a hierarchical encryption storage mechanism to generate security event log files not only ensures data security, but also implements differentiated access control for different authority levels, providing reliable data support and decision-making basis for equipment security management.

[0134] Reference Figure 2 As shown, the present invention also provides a servo module control device for equipment safety overload protection, which is applied to any of the above-mentioned servo module control methods for equipment safety overload protection, including: The acquisition module is used to obtain device data of smart devices and perform basic security analysis based on preset device threshold parameters to obtain the device security status; Analysis module: The analysis module is used to obtain the user profile and real-time motion stage of the smart device, intelligently integrate the device data, and obtain the dynamic overload parameters of the device; The association module is used to make equipment classification judgment based on the dynamic overload parameters of the equipment and the equipment safety status to obtain the equipment safety level; The processing module is used to perform command control analysis based on the real-time motion stage, the dynamic overload parameters of the equipment and the equipment safety level to obtain the servo module control solution.

[0135] The present invention provides a servo module control device for equipment safety overload protection. Through real-time safety analysis based on preset threshold parameters, it can timely discover potential hidden dangers and solve the problem of delayed response of traditional static judgment. By combining user profiles and real-time motion stages for data intelligent fusion, dynamic overload parameters are accurately calculated, breaking through the limitations of a single data dimension and comprehensively improving the accuracy of safety risk identification. The multi-level safety assessment system enables the system to take corresponding measures according to different risk levels, avoiding excessive or insufficient protection intervention and enhancing system adaptability. Based on the command control analysis of multi-dimensional parameters, precise safety control for different training intensities and modes is achieved, adapting to individual differences and special needs, and improving equipment safety and user experience. Taking into account user characteristics and training stage characteristics, the system can flexibly adjust protection strategies, adapt to diverse training scenarios, effectively balance training effects and safety guarantees, and provide a more scientific and safe fitness experience.

[0136] It should be noted that, those skilled in the art will clearly understand that, for the sake of convenience and brevity of description, the specific working processes of the above-described system and each module can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0137] The above description is only a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the present invention description and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A servo module control method for equipment safety overload protection, characterized in that: include: Obtain device data from smart devices and perform basic security analysis based on preset device threshold parameters to determine the device security status; Obtaining the user profile and real-time motion stage of the smart device, intelligently integrating the device data, and obtaining a dynamic overload parameter of the device; Performing equipment classification judgment based on the dynamic overload parameter of the equipment and the safety status of the equipment to obtain the equipment safety level; A command control analysis is performed based on the real-time motion stage, the dynamic overload parameter of the device and the safety level of the device to obtain a servo module control solution.

2. The servo module control method for equipment safety overload protection according to claim 1, characterized in that: The step of obtaining smart device data of the smart device and performing basic security analysis based on preset device threshold parameters to obtain the device security status includes: Acquire servo torque data and motor current data of the smart device, perform device torque identification, and obtain device torque load data; Acquire position encoder data and speed sensor data of the smart device, perform motion trajectory analysis, and obtain device motion state data; Performing multi-dimensional fusion of the equipment according to the equipment torque load data and the equipment motion state data to obtain the equipment data; The device data is compared with the preset device threshold parameters to obtain the device safety status.

3. The servo module control method for equipment safety overload protection according to claim 1, characterized in that: The acquiring of the user profile and real-time motion stage of the smart device, and intelligent fusion of the device data to obtain the dynamic overload parameter of the device includes: Acquire visual sensor data and pressure sensor data of the smart device, perform motion stage recognition, and obtain a real-time motion stage; Performing action recognition on the visual sensing data based on preset action specification information to obtain action posture data; Performing pressure recognition on the action posture data based on the pressure sensing data to obtain posture pressure parameters; Obtain the user profile through a user database connected to the smart device, perform historical training identification, and obtain user strength assessment data; Dynamic motion fusion is performed on the device data according to the real-time motion stage, the user strength assessment data and the posture pressure parameter to obtain the device dynamic overload parameter.

4. The servo module control method for equipment safety overload protection according to claim 1, characterized in that: The performing device classification judgment based on the device dynamic overload parameter and the device safety status to obtain the device safety level includes: Performing composite load data extraction on the dynamic overload parameters of the equipment to obtain mechanical load data and motion load data of the equipment; Performing state quantization encoding on the safety state of the device to obtain a safety state quantization value; performing abnormal attenuation calibration on the equipment mechanical load data according to the safety state quantified value to obtain calibrated mechanical load data; Performing load coefficient mapping on the motion load data according to the safety state quantification value to obtain a motion load safety factor; An equipment safety assessment is performed on the calibrated mechanical load data and the motion load safety factor based on a preset safety level conversion rule to obtain the equipment safety level.

5. The servo module control method for equipment safety overload protection according to claim 4, characterized in that: The performing of equipment safety assessment on the calibration mechanical load data and the motion load safety factor based on a preset safety level conversion rule to obtain the equipment safety level includes: Performing joint torque analysis on the motion load safety factor to obtain a joint torque distribution vector; Performing mechanical equivalent transformation on the joint torque distribution vector to obtain a posture damage safety value; Performing a safety margin correction on the motion load safety factor according to the posture damage safety value to obtain a human body protection priority coefficient; performing equipment conflict optimization on the human protection priority coefficient and the calibrated mechanical load data based on a preset anti-saturation constraint rule to obtain an equipment protection strength attenuation factor; The device protection strength attenuation factor is reconstructed according to the security level conversion rule to obtain the device security level.

6. The servo module control method for equipment safety overload protection according to claim 1, characterized in that: The command control analysis is performed according to the real-time motion stage, the dynamic overload parameter of the device and the safety level of the device to obtain a servo module control solution, including: Performing hierarchical rule matching on the device security level based on preset hierarchical intervention measures to obtain an intervention level identifier; Performing real-time load extraction on the equipment data to obtain real-time operating parameters of the equipment; Performing action phase segmentation on the real-time motion phase to obtain an action interval set; Performing intensity iteration on the intervention level identifier in combination with the dynamic overload parameter of the device and the real-time operating condition parameter of the device to obtain a dynamic intervention intensity calibration value; Performing servo control instruction conversion on the dynamic intervention intensity calibration value to obtain an overload control instruction set; The action interval set is coupled with the overload control instruction set to obtain the servo module control solution.

7. The servo module control method for equipment safety overload protection according to claim 6, characterized in that: The step of performing hierarchical rule matching on the device security level based on the preset hierarchical intervention measures to obtain an intervention level identifier includes: Performing overload threshold stratification on the safety level of the equipment to obtain multi-level overload protection threshold intervals; Performing servo response mapping on the hierarchical intervention measures according to the multi-level overload protection threshold intervals to obtain a servo intervention measure mapping table; Perform overload parameter identification on the equipment safety level to obtain a real-time value of the overload status; Perform load interval positioning on the multi-level overload protection threshold interval according to the real-time value of the overload state to obtain a current overload interval; The servo intervention measure mapping table is dynamically matched according to the current overload interval to obtain the intervention level identifier.

8. The servo module control method for equipment safety overload protection according to claim 1, characterized in that: The method further includes obtaining execution feedback information of the servo module control solution, and performing device security recording with the device data and the device security level to obtain a security event log file: Performing multi-dimensional execution feedback collection on the servo module control solution to obtain the execution feedback information; performing abnormal pattern recognition on the device data according to the execution feedback information to obtain a dynamic device operation status indicator; Performing deep semantic fusion on the dynamic device operating status indicator and the device security level to obtain a hierarchical security status description; Perform reasoning and matching on the hierarchical security status description according to a preset security event knowledge graph to obtain an accurate event type identification; The precise event type identifier and the hierarchical security status description are written hierarchically into a preset encrypted log file to obtain the security event log file.

9. A servo module control device for equipment safety overload protection, characterized in that: The servo module control method for equipment safety overload protection applied to any one of claims 1 to 8 above comprises: A collection module is used to obtain device data of smart devices and perform basic security analysis based on preset device threshold parameters to obtain the device security status; An analysis module, configured to obtain a user profile and real-time motion stage of the smart device, intelligently fuse the device data, and obtain a dynamic overload parameter of the device; an association module, the association module being configured to perform device classification judgment based on the dynamic overload parameter of the device and the safety status of the device to obtain a device safety level; A processing module is used to perform command control analysis according to the real-time motion stage, the dynamic overload parameter of the device and the safety level of the device to obtain a servo module control solution.

Images