Harmonic reducer performance calibration and verification method and device for humanoid robot

Abstract

The invention discloses a harmonic reducer performance calibration and verification method and device for a humanoid robot, and relates to the technical field of robotics.The method comprises the steps that firstly, when the robot is in a calibration posture, multi-source signals of a harmonic reducer are collected, and an initial efficiency-torque curve and a reverse transmission gap are established according to the multi-source signals to serve as reference performance parameters; then, according to a preset verification period, the calibration posture is reproduced, real-time multi-source signals are collected, and a real-time efficiency-torque curve and a reverse transmission gap are calculated; and comparing the real-time performance parameters with the reference parameters to obtain an efficiency attenuation rate curve and a gap increase amount, and further comprehensively evaluating a performance health degree index to form a final verification result. According to the method, multi-dimensional and periodic quantitative monitoring of performance degradation of the harmonic reducer is realized, and the problems of performance degradation and gap increase caused by long-term operation are effectively solved, so that the motion precision and the service life prediction accuracy of the humanoid robot are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of robots, in particular to a harmonic reducer performance calibration and verification method and device for humanoid robots. BACKGROUND

[0002] As the core transmission component of the joints of humanoid robots, the performance of the harmonic reducer directly affects the motion accuracy, load capacity and service life of the robot. During long-term high-load operation, the harmonic reducer is prone to problems such as efficiency decay and increased reverse transmission gap. Traditional methods usually rely on periodic disassembly detection or simple single-parameter monitoring, which cannot achieve online, multi-dimensional performance evaluation. The existing technology lacks fusion analysis of the efficiency-torque characteristic curve and the reverse gap, making it difficult to accurately reflect the comprehensive degradation state of the reducer, resulting in untimely maintenance or excessive maintenance. In addition, the dynamic tasks of humanoid robots are complex, and a fixed verification period cannot adapt to the actual working conditions, so an adaptive verification method that takes into account the running time is urgently needed to achieve performance degradation warning and life prediction.

[0003] The above content is only used to assist in understanding the technical solutions of the present application and does not represent the acknowledgement of the above content as prior art. SUMMARY

[0004] The main purpose of the present application is to provide a harmonic reducer performance calibration and verification method and device for humanoid robots, aiming to solve the technical problems of long-term running performance decay and gap increase of the harmonic reducer of humanoid robots in the prior art, which leads to decreased robot accuracy and inaccurate life prediction.

[0005] To achieve the above-mentioned purpose, the present application provides a harmonic reducer performance calibration and verification method for humanoid robots, the method comprising: When the humanoid robot is in a calibration posture, a plurality of source signals in the running state of the harmonic reducer are obtained, and an initial efficiency-torque curve and an initial reverse transmission gap are determined based on the plurality of source signals, and a reference performance parameter is obtained according to the initial efficiency-torque curve and the initial reverse transmission gap; Based on a preset verification period, the calibration posture is reproduced to obtain real-time multi-source signals, and a real-time efficiency-torque curve and a real-time reverse transmission gap are determined based on the real-time multi-source signals, and a real-time performance parameter is obtained according to the real-time efficiency-torque curve and the real-time reverse transmission gap; The reference performance parameter and the real-time performance parameter are compared to obtain an efficiency decay rate curve and a gap growth amount; A performance health index is obtained according to the efficiency decay rate curve and the gap growth amount, and a performance verification result is obtained based on the performance health index.

[0006] In an embodiment, the step of reproducing the calibration posture based on the preset checking period to obtain real-time multi-source signals, determining real-time efficiency-torque curves and real-time reverse transmission gaps based on the real-time multi-source signals, and obtaining real-time performance parameters according to the real-time efficiency-torque curves and the real-time reverse transmission gaps comprises: According to a dynamic checking period based on the actual working time of the robot, the robot is controlled to reproduce the calibration posture in an idle time window after completing a specific task; The harmonic reducer is driven at multiple constant speeds, and real-time current signals, torque signals and operating temperatures are synchronously collected, and the current signals, torque signals and operating temperatures are taken as real-time multi-source signals; Based on the real-time multi-source signals, real-time efficiency-torque curves and real-time reverse transmission gaps are calculated through parameter identification; The real-time efficiency-torque curves and the real-time reverse transmission gaps are fused to obtain real-time performance parameters.

[0007] In an embodiment, the step of calculating real-time efficiency-torque curves and real-time reverse transmission gaps based on the real-time multi-source signals through parameter identification comprises: Recursive least squares are used on the real-time multi-source signals to fit multiple groups of collected current, torque and temperature data to obtain the real-time efficiency-torque curves; The joint motor is controlled to be forward and reverse rotated in a preset interval at zero point, the reading difference of the encoder when the torque is zero is recorded, and the real-time reverse transmission gap is obtained according to the reading difference.

[0008] In an embodiment, the step of fusing the real-time efficiency-torque curves and the real-time reverse transmission gaps to obtain real-time performance parameters comprises: Efficiency values at preset torque sampling points are extracted from the real-time efficiency-torque curves to form an efficiency feature sequence; The efficiency feature sequence and the real-time reverse transmission gap are spliced in a predetermined order to form a multi-dimensional performance parameter vector; The multi-dimensional performance parameter vector is normalized to obtain real-time performance parameters.

[0009] In an embodiment, the step of comparing the real-time efficiency-torque curves with the initial efficiency-torque curves point by point, calculating an efficiency attenuation rate, and generating an efficiency attenuation rate curve comprises: Based on the sampling points on the real-time efficiency-torque curves, the relative change percentages of real-time efficiency values and initial efficiency values are respectively calculated; According to the efficiency attenuation percentages of the sampling points, an efficiency attenuation rate curve is obtained.

[0010] In an embodiment, the step of comparing the baseline performance parameter with the real-time performance parameter to obtain an efficiency decay rate curve and a gap growth amount comprises: point-by-point comparison of the real-time efficiency-torque curve with the initial efficiency-torque curve to calculate an efficiency decay rate, and generation of an efficiency decay rate curve based on the efficiency decay rate; subtraction of the real-time reverse transmission gap from the initial reverse transmission gap to obtain a gap growth amount.

[0011] In an embodiment, the step of obtaining a performance health degree index according to the efficiency decay rate curve and the gap growth amount, and obtaining a performance verification result based on the performance health degree index comprises: inputting the efficiency decay rate curve and the gap growth amount into a pre-trained performance degradation evaluation model to obtain a performance health degree index; predicting a remaining service life of the harmonic reducer according to the performance health degree index; obtaining a performance verification result according to the performance health degree index and the remaining service life.

[0012] In an embodiment, the step of inputting the efficiency decay rate curve and the gap growth amount into a pre-trained performance degradation evaluation model to obtain a performance health degree index comprises: data fusion of the efficiency decay rate curve, the gap growth amount, and a real-time collected operating temperature to form a fusion feature vector; inputting the fusion feature vector into a pre-trained performance degradation evaluation model to obtain a performance health degree index.

[0013] In an embodiment, after the step of obtaining a performance health degree index according to the efficiency decay rate curve and the gap growth amount, and obtaining a performance verification result based on the performance health degree index, the method further comprises: detecting the performance verification result to obtain a detection result; when the detection result is abnormal, generating an alarm information according to the performance verification result, and controlling the humanoid robot to stop and wait for maintenance.

[0014] In addition, to achieve the above-mentioned purpose, the present application also proposes a harmonic reducer performance calibration and verification device for a humanoid robot, which comprises: a baseline calibration module, configured to acquire multi-source signals in a running state of a harmonic reducer when a humanoid robot is in a calibration posture, and determine an initial efficiency-torque curve and an initial reverse transmission gap based on the multi-source signals, and obtain a baseline performance parameter according to the initial efficiency-torque curve and the initial reverse transmission gap. a real-time monitoring module configured to reproduce the calibration posture based on a preset checking period to obtain real-time multi-source signals, determine a real-time efficiency-torque curve and a real-time reverse transmission gap based on the real-time multi-source signals, and obtain real-time performance parameters according to the real-time efficiency-torque curve and the real-time reverse transmission gap; a parameter comparison module configured to compare the reference performance parameters with the real-time performance parameters to obtain an efficiency decay rate curve and a gap growth amount; a health assessment module configured to obtain a performance health degree index according to the efficiency decay rate curve and the gap growth amount, and obtain a performance checking result based on the performance health degree index.

[0015] In addition, to achieve the above-mentioned purpose, the present application also provides a harmonic reducer performance calibration and checking device for a humanoid robot, which comprises a memory, a processor and a computer program stored in the memory and executable on the processor, and the computer program is configured to implement the steps of the harmonic reducer performance calibration and checking method for a humanoid robot as described above.

[0016] In addition, to achieve the above-mentioned purpose, the present application also provides a storage medium, which is a computer readable storage medium, and the storage medium stores a computer program, and the computer program is executed by a processor to implement the steps of the harmonic reducer performance calibration and checking method for a humanoid robot as described above.

[0017] In addition, to achieve the above-mentioned purpose, the present application also provides a computer program product, which comprises a computer program, and the computer program is executed by a processor to implement the steps of the harmonic reducer performance calibration and checking method for a humanoid robot as described above.

[0018] The present application provides a harmonic reducer performance calibration and checking method for a humanoid robot, which comprises collecting multi-source signals of the harmonic reducer when the robot is in a calibration posture, and establishing an initial efficiency-torque curve and a reverse transmission gap as reference performance parameters. Then, the calibration posture is reproduced according to a preset checking period, and real-time multi-source signals are collected to calculate real-time efficiency-torque curve and reverse transmission gap. By comparing the real-time performance parameters with the reference parameters, an efficiency decay rate curve and a gap growth amount are obtained, and then a performance health degree index is comprehensively evaluated to form a final checking result. This method realizes multi-dimensional and periodic quantitative monitoring of performance degradation of the harmonic reducer, effectively solves the problems of performance decay and gap increase caused by long-term operation, and improves the motion accuracy and life prediction accuracy of the humanoid robot. BRIEF DESCRIPTION OF DRAWINGS

[0019] The accompanying drawings, which are incorporated herein and constitute part of the specification, illustrate embodiments consistent with the application and, together with the description, function to explain the principles of the application.

[0020] In order to more clearly illustrate the technical solutions of the embodiments of the present application or the prior art, the accompanying drawings required to be used in the embodiments or prior art description will be briefly introduced. Obviously, for those skilled in the field, based on these drawings, other drawings can also be obtained without creative labor.

[0021] Figure 1 The flowchart of the embodiment one of the performance calibration and verification method for the harmonic reducer of the humanoid robot of the present application; Figure 2 The efficiency attenuation rate-torque curve diagram of the embodiment one of the performance calibration and verification method for the harmonic reducer of the humanoid robot of the present application; Figure 3 The module structure diagram of the performance calibration and verification device for the harmonic reducer of the humanoid robot of the embodiment of the present application; Figure 4 The device structure diagram of the hardware running environment involved in the performance calibration and verification method for the harmonic reducer of the humanoid robot of the embodiment of the present application.

[0022] The implementation of the purpose of the present application, functional features and advantages will be further described with reference to the accompanying drawings. DETAILED DESCRIPTION

[0023] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application, and are not used to limit the present application.

[0024] In order to better understand the technical solutions of the present application, the following will be described in detail with reference to the accompanying drawings and specific embodiments.

[0025] The main solution of the embodiment of the present application is: when the humanoid robot is in a calibration posture, a plurality of source signals in the running state of the harmonic reducer are obtained, and the initial efficiency-torque curve and the initial reverse transmission gap are determined based on the plurality of source signals, and the reference performance parameters are obtained according to the initial efficiency-torque curve and the initial reverse transmission gap; Based on a preset verification period, the calibration posture is reproduced to obtain real-time multi-source signals, and real-time efficiency-torque curve and real-time reverse transmission gap are determined based on the real-time multi-source signals, and real-time performance parameters are obtained according to the real-time efficiency-torque curve and the real-time reverse transmission gap; The reference performance parameters and the real-time performance parameters are compared to obtain the efficiency attenuation rate curve and the gap growth amount; According to the efficiency attenuation rate curve and the gap growth amount, a performance health degree index is obtained, and a performance verification result is obtained based on the performance health degree index.

[0026] At present, as the core transmission component of the humanoid robot joint, the performance of the harmonic reducer directly affects the motion accuracy, load capacity and service life of the robot. In the long-term high-load operation process, the harmonic reducer is prone to problems such as efficiency attenuation and reverse transmission gap increase. The traditional method usually relies on regular disassembly detection or simple single-parameter monitoring, and cannot realize online and multi-dimensional performance evaluation. The existing technology lacks fusion analysis of the efficiency-torque characteristic curve and the reverse gap, and it is difficult to accurately reflect the comprehensive degradation state of the reducer, resulting in untimely maintenance or excessive maintenance. In addition, the dynamic task of the humanoid robot is complex, and it is difficult to adapt to the actual working condition with a fixed verification period. An adaptive verification method combining the running time is urgently needed to realize performance degradation warning and life prediction.

[0027] The application provides a solution. When the robot is in a calibration posture, multi-source signals of the harmonic reducer are collected, and an initial efficiency-torque curve and a reverse transmission gap are established as reference performance parameters. Then, according to a preset verification period, the calibration posture is reproduced and real-time multi-source signals are collected, and real-time efficiency-torque curves and reverse transmission gaps are calculated. By comparing the real-time performance parameters with the reference parameters, an efficiency attenuation rate curve and a gap growth amount are obtained, and a performance health degree index is comprehensively evaluated to form a final verification result. This method realizes multi-dimensional and periodic quantitative monitoring of the performance degradation of the harmonic reducer, effectively solves the problems of performance attenuation and gap increase caused by long-term operation, and improves the motion accuracy and life prediction accuracy of the humanoid robot.

[0028] It should be noted that the execution subject of the present embodiment can be a computing service device with data processing, network communication and program running functions, such as a tablet computer, a personal computer, a mobile phone, etc., or an electronic device capable of realizing the above functions, a harmonic reducer performance calibration and verification device for a humanoid robot, etc. The present embodiment does not make specific limitations on this. The harmonic reducer performance calibration and verification device for a humanoid robot will be taken as an example to describe the present embodiment and the following embodiments.

[0029] All actions of obtaining signals, information or data in the present application are carried out in compliance with the corresponding data protection regulations and policies of the country where the device is located, and with the authorization given by the corresponding device owner.

[0030] The present embodiment provides a harmonic reducer performance calibration and verification method for a humanoid robot, which is described with reference to Figure 1 , Figure 1The flowchart of the first embodiment of the method for calibrating and verifying the performance of a harmonic reducer for a humanoid robot.

[0031] In this embodiment, the method for calibrating and verifying the performance of a harmonic reducer for a humanoid robot comprises steps S10-S40: Step S10, when the humanoid robot is in a calibration posture, a plurality of source signals in the running state of the harmonic reducer are obtained, and the initial efficiency-torque curve and the initial reverse transmission gap are determined based on the plurality of source signals, and the reference performance parameters are obtained according to the initial efficiency-torque curve and the initial reverse transmission gap.

[0032] It should be noted that the humanoid robot refers to a robot with a humanoid structure, and the joint thereof usually adopts a harmonic reducer as a precision transmission component. The calibration posture refers to a specific and reproducible overall configuration of the robot set to reduce the interference of gravity, connecting rod inertia and other factors on the test results. The harmonic reducer is a high-precision reduction device that transmits motion and power through elastic deformation. The plurality of source signals are physical signals such as current, output torque, speed and temperature synchronously collected from the driver and sensor in this posture. The efficiency-torque curve describes the transmission efficiency variation law of the harmonic reducer under different load torques. The reverse transmission gap is a key indicator for measuring the reverse backlash error of the transmission chain. Finally, the reference performance parameters generated based on the data fusion provide an initial and healthy reference benchmark for subsequent performance degradation evaluation.

[0033] In a specific implementation, first, the joints of the humanoid robot need to be controlled to move to a pre-calculated calibration posture. This posture needs to be optimized dynamically to minimize the coupling effect of the gravity moment of other joints except the joint to be tested, and to ensure that the harmonic reducer to be tested is in the best stress and sensing measurement state. Subsequently, while maintaining this posture, the target harmonic reducer is driven at a plurality of different levels of constant speed, and the current at the motor end, the speed and the torque sensor data at the output end are synchronously and high-frequency collected. At the same time, the temperature sensor installed on the reducer housing monitors the operating temperature in real time, and these signals together constitute the plurality of source signals required for calibration.

[0034] After obtaining the plurality of source signals, the data processing stage is entered. By calculating the ratio of the output power to the input power under different torque load points, the efficiency-torque curve of the reducer in the initial healthy state can be drawn. For the measurement of the initial reverse transmission gap, the joint motor is usually controlled to perform a small-amplitude forward and reverse motion near zero, and the angle difference value when the torque value passes through zero is recorded by using a high-precision encoder. Finally, the initial efficiency curve and the gap value are extracted and standardized and packaged to form a complete set of reference performance parameter data set which can be used as a long-term comparison benchmark.

[0035] Step S20, based on a preset verification period, the calibration posture is reproduced to obtain real-time multi-source signals, based on the real-time multi-source signals, real-time efficiency-torque curve and real-time reverse transmission gap are determined, and real-time performance parameters are obtained according to the real-time efficiency-torque curve and the real-time reverse transmission gap.

[0036] It should be noted that the preset verification period refers to a time interval rule preset for triggering and executing performance verification tasks. The period can be based on a fixed time (such as every week or every month), or based on the cumulative working time or running mileage of the key components of the robot, and the purpose is to realize periodic and active monitoring of the performance state of the harmonic reducer, and provide regular data support for predictive maintenance.

[0037] In a specific implementation, the control system or health management platform will automatically initiate the verification task according to the preset verification period. After the task is started, the robot will first accurately reproduce the calibration posture to ensure that the boundary conditions of the two measurements are consistent, thereby ensuring data comparability. Subsequently, the system will drive the harmonic reducer and collect multi-source signals such as current, torque, speed, temperature, etc. under its real-time running state according to the same procedure as the baseline calibration stage.

[0038] After obtaining the real-time data, the same algorithm process as processing the baseline data is used to calculate and generate the current real-time efficiency-torque curve and real-time reverse transmission gap. This process is essentially a complete snapshot measurement of the key performance indicators of the harmonic reducer at a new time point. Finally, these calculated real-time performance data are packaged into real-time performance parameters, which will serve as the current state basis for directly comparing and evaluating the performance degradation degree with the baseline performance parameters.

[0039] In one possible implementation, the step of reproducing the calibration posture based on a preset verification period to obtain real-time multi-source signals, determining real-time efficiency-torque curve and real-time reverse transmission gap based on the real-time multi-source signals, and obtaining real-time performance parameters according to the real-time efficiency-torque curve and the real-time reverse transmission gap comprises: According to a dynamic verification period based on the actual working time of the robot, the robot is controlled to reproduce the calibration posture during the idle time window after completing a specific task. The harmonic reducer is driven at multiple constant speeds, and real-time current signals, torque signals and operating temperatures are synchronously collected, and the current signals, torque signals and operating temperatures are taken as real-time multi-source signals. Based on the real-time multi-source signals, real-time efficiency-torque curve and real-time reverse transmission gap are calculated through parameter identification. The real-time efficiency-torque curve and the real-time reverse transmission gap are fused to obtain real-time performance parameters: In its implementation, the timing of the verification is first determined by a dynamic verification cycle based on the robot's actual working time (rather than natural time). For example, when the cumulative running time of the harmonic reducer reaches a threshold of 100 hours, the system does not immediately interrupt the task. Instead, it waits for the robot to complete its current task and enter a preset idle time window (such as charging or standby mode) before automatically triggering the verification process. This process is implemented through a task scheduler, which monitors the robot's state machine to ensure that performance verification does not interfere with normal operations, thereby decoupling maintenance behavior from task execution.

[0040] Once the robot enters the calibration posture, the actuator controls the motor to run at a set of pre-set constant speeds (e.g., 0.1 rad / s, 0.5 rad / s, 1.0 rad / s) at the harmonic reducer. At each speed point, high-precision current and torque sensors simultaneously sample the data to obtain the real-time motor current value. and reducer output torque value Meanwhile, the temperature sensor records the temperature of the reducer housing. These data constitute real-time multi-source signals. Efficiency is calculated based on the power ratio; at each steady-state speed point, the transmission efficiency is... Calculated by the formula:

[0041] in, It is the output power. It is the input power. It is the output angular velocity of the reducer. That's the motor's angular velocity. Taking the reduction ratio into account... Given ( And motor torque With current Proportional ( (where is the motor torque constant), the formula can be simplified to:

[0042] By changing the load (i.e., changing the output torque) This allows us to obtain a series of efficiency-torque data points, which can then be used to fit a real-time efficiency-torque curve.

[0043] For real-time reverse transmission clearance Measurements are typically performed at low speeds and low loads. The motor drives the joint to slowly rotate in both directions past zero, while a high-resolution encoder records the angular position at the input. When the torque sensor at the output detects the torque value... When the motor angle changes from positive to negative (or from negative to positive), record the corresponding motor angle. and The formula for calculating the reverse gap is:

[0044] That is, the difference in the angle of the free travel of the motor side is converted to the value of the output side. Finally, this real-time efficiency-torque curve (including its fitting coefficients) and the calculated real-time reverse gap value, together with the average temperature and other environmental information at the time of acquisition, are packaged into a structured data packet, i.e., a real-time performance parameter, for subsequent comparison and analysis.

[0045] In a feasible implementation, the step of calculating the real-time efficiency-torque curve and the real-time reverse transmission gap based on the real-time multi-source signals includes: Using recursive least squares method on the real-time multi-source signals, fitting the collected multiple sets of current, torque and temperature data to obtain the real-time efficiency-torque curve; Controlling the joint motor to perform forward and reverse rotation within a preset interval at zero point, recording the reading difference of the encoder when the torque is zero, and obtaining the real-time reverse transmission gap according to the reading difference.

[0046] In a specific implementation, after obtaining the real-time multi-source signals (current , output torque , temperature ), in order to establish an accurate real-time efficiency-torque curve, system identification needs to be performed on the data. Recursive least squares (RLS) is adopted due to its excellent online real-time processing capability and tracking characteristics for time-varying parameters. The efficiency calculation model:

[0047] is converted into a parameter estimation problem. Considering the influence of temperature on the torque constant and friction characteristics, the efficiency model can be extended to a linear or nonlinear model containing temperature compensation, for example, a simplified model can be represented as where the parameter vector to be identified includes the gain and friction coefficient related to efficiency. The RLS algorithm updates the parameter estimation value online through the following formula:

[0048] where, is the current sampling time, is the regression vector, including , is the observation value (i.e. ), is the gain matrix, is the covariance matrix, is a forgetting factor (usually taken as 0.95~0.999) to give new data higher weight, thus tracking the slowly changing parameters. By processing multiple sets of steady-state data, the algorithm eventually identifies the parameters that best reflect the relationship between input current and output torque under the current working condition, and thus can draw a high-efficiency real-time efficiency-torque curve.

[0049] The measurement of real-time back-drive gap is a high-precision positioning process. The system controls the joint motor to perform low-speed and smooth reciprocating motion in a small angle preset interval (e.g. ±0.1 radian) around the theoretical zero position. During this process, the high-resolution absolute encoder continuously monitors the actual angle of the motor shaft , while the torque sensor precisely monitors the torque at the output end . The core of the measurement is to capture the motor position when the torque crosses zero. When the joint approaches the zero point from one direction, the output end remains stationary until the gear contacts due to the existence of the gap; the system accurately records the torque value when the non-zero value (positive or negative) first crosses the zero value (or a threshold close to zero), denoted as . Then, when the motor moves in the opposite direction and the torque crosses zero again, the angle at that moment is recorded .

[0050] After obtaining the key angle readings, the calculation formula of real-time back-drive gap is:

[0051] Here, is the nominal reduction ratio of the harmonic reducer. This formula converts the measured angle difference in free space on the motor side to the output side of the reducer, thus obtaining the transmission gap value in radians or degrees, which has actual physical meaning. Finally, the calculated is encapsulated with the real-time efficiency-torque curve model parameters identified by the RLS algorithm, together forming a set of real-time performance parameters that completely describe the current transmission performance of the harmonic reducer. This set of parameters provides an accurate data basis for subsequent comparison with reference parameters and health status evaluation.

[0052] In a possible implementation, the step of fusing the real-time efficiency-torque curve and the real-time back-drive gap to obtain real-time performance parameters includes: extracting the efficiency values at preset torque sampling points from the real-time efficiency-torque curve to form an efficiency feature sequence; concatenating the efficiency feature sequence and the real-time back-drive gap in a predetermined order to form a multi-dimensional performance parameter vector; ​The multi-dimensional performance parameter vector is normalized to obtain real-time performance parameters.

[0053] In a specific implementation, since the real-time efficiency-torque curve is a continuous curve, in order to be digitized and quantitatively compared, it needs to be converted into a discrete characteristic sequence. The system will preset a set of torque sampling points covering the normal working range of the reducer, for example , which are usually equally spaced or selected according to key working points. Subsequently, the real-time efficiency-torque curve model obtained through parameter identification in advance (whose functional relationship is ) is evaluated at these preset torque points , and the corresponding efficiency values are calculated. This process generates an efficiency characteristic sequence :

[0054] This sequence quantitatively captures the transmission efficiency performance of the harmonic reducer at different load levels under the current state, condensing the continuous curve information into a characteristic vector containing elements.

[0055] After obtaining the efficiency characteristic sequence and the real-time reverse transmission clearance value , the next step is to integrate these two indicators of different dimensions and physical meanings into a unified mathematical expression that can fully represent the current performance state. This is done by vector splicing in a predetermined order. Assuming that the dimension of the efficiency characteristic sequence is , then the dimension of the multi-dimensional performance parameter vector after splicing is :

[0056] This vector contains the “face” information of transmission efficiency at different loads and the key “point” information of reverse clearance, which together constitute a compact digital description of the mechanical and transmission comprehensive performance of the harmonic reducer.

[0057] The original vectors spliced directly have dimensional differences (efficiency is dimensionless, clearance is angle) and numerical scale differences (efficiency values are close to 1, clearance values may be a few degrees or a few millimeters), which are not conducive to subsequent comparative analysis algorithms (such as distance calculation, similarity). Therefore, normalization must be performed to map all components to a unified, dimensionless numerical interval (such as [0, 1]). Here, the Min-Max normalization method based on reference values is usually used. Let the reference performance parameter vector be , then the normalized real-time performance parameter vector is Each component Calculated by the following formula:

[0058] The first method (ratio method) is more commonly used, as it directly reflects the relative rate of change of real-time parameters with respect to baseline parameters. After this step, the final real-time performance parameters are obtained. This is the normalized multidimensional vector:

[0059] this The dimensionless value directly and intuitively reflects the degree of performance deviation from the baseline state (e.g., a value greater than 1 indicates higher efficiency or larger gaps, while a value less than 1 indicates performance degradation), providing a standardized input for subsequent health status assessment and life prediction.

[0060] In one feasible implementation, the step of comparing the real-time efficiency-torque curve with the initial efficiency-torque curve point by point, calculating the efficiency decay rate, and generating an efficiency decay rate curve includes: Based on the sampling points on the real-time efficiency-torque curve, the relative percentage change between the real-time efficiency value and the initial efficiency value is calculated respectively. The efficiency decay rate curve is obtained based on the efficiency decay percentage of each sampling point.

[0061] In its implementation, the system performs point-by-point comparisons and calculates the attenuation percentage. The core technology lies in precisely aligning the real-time performance curve with the initial baseline curve on the same torque coordinate axis. Specifically, the system selects a predefined sequence of torque sampling points covering the operating range. For each sampling point The algorithm starts from the function of the initial efficiency-torque curve. and the function of the real-time efficiency-torque curve In the middle, query or calculate the corresponding efficiency value. and Then, the percentage of efficiency degradation at that torque point was calculated. Its calculation formula is usually as follows:

[0062] This value directly reflects the performance under a specific load. Under these conditions, the relative rate of change of transmission efficiency with respect to the initial state. When When the value is negative, it indicates that the efficiency has decreased, and the magnitude of the absolute value indicates the severity of the decrease.

[0063] The system generates a complete efficiency degradation rate curve based on the calculation results of all sampling points. After completing the calculation of the percentage of attenuation of all preset torque sampling points , the system obtains a series of discrete data point pairs . These pairs of points clearly show the distribution of the efficiency degradation rate with the change of the output torque. Finally, the algorithm (for example, using linear interpolation or curve fitting method) connects these discrete points to form a continuous efficiency degradation rate-torque curve , referred to as the efficiency degradation rate-torque curve Figure 2 , Figure 2 . This curve is the key basis for performance degradation evaluation, as it not only reflects the overall efficiency decline level, but also reveals the dependence between attenuation and load, which helps to locate the most severe wear condition interval.

[0064] In step S30, the reference performance parameter is compared with the real-time performance parameter to obtain an efficiency degradation rate curve and a gap growth amount.

[0065] It should be noted that the efficiency degradation rate curve refers to a curve generated by comparing the efficiency-torque curve obtained by real-time testing with the initial reference curve at the same torque sampling point, calculating the percentage decline of each point efficiency value relative to the initial value, and describing the degree of transmission efficiency degradation with the change of load torque. It quantifies the performance degradation of harmonic reducer under different working conditions; while the gap growth amount is a scalar value obtained by calculating the absolute difference between the real-time measured reverse transmission gap and the initial gap value, which directly reflects the increasing degree of mechanical idle travel error caused by wear. In actual calculation, the system automatically completes the comparison of a large number of data points between curves and the scalar difference operation through an efficient numerical comparison algorithm, thereby quickly and accurately outputting the two key performance degradation indicators.

[0066] In a possible implementation, the step of comparing the reference performance parameter with the real-time performance parameter to obtain an efficiency degradation rate curve and a gap growth amount includes: point-by-point comparison of the real-time efficiency-torque curve and the initial efficiency-torque curve to calculate the efficiency degradation rate, and generation of an efficiency degradation rate curve based on the efficiency degradation rate; subtraction of the real-time reverse transmission gap from the initial reverse transmission gap to obtain the gap growth amount.

[0067] In a specific implementation, the generation of the efficiency decay rate curve is a comparative process for continuous performance curves. The system synchronously queries the corresponding efficiency values on the real-time efficiency-torque curve and the initial baseline curve at a plurality of torque sampling points preset to cover the entire working range. For each sampling point, the algorithm calculates the relative change rate (i.e., decay rate) of the real-time efficiency relative to the initial efficiency. When the calculation is completed for all sampling points, a series of discrete "torque-decay rate" data points are obtained. Finally, these points are connected by a data fitting or interpolation algorithm to form a smooth continuous curve that intuitively shows how the efficiency decay degree changes with the load torque. This curve is the core basis for evaluating the degradation of transmission performance.

[0068] The calculation of the gap growth amount is a simple scalar difference operation. The reverse transmission gap itself is a single numerical indicator representing a specific mechanical performance. The system directly reads the real-time monitored gap value and the initial baseline gap value, and then performs arithmetic subtraction. The calculation result is an absolute difference, i.e., the gap growth amount, which directly quantifies the actual increase in mechanical idle travel error caused by factors such as wear. This scalar value is often used as a key threshold parameter to trigger warnings or determine whether maintenance is needed due to its simplicity and clarity.

[0069] Step S40, obtaining a performance health degree index according to the efficiency decay rate curve and the gap growth amount, and obtaining a performance verification result based on the performance health degree index.

[0070] It should be noted that the performance health degree index is a comprehensive quantitative value that combines the efficiency decay rate curve and the gap growth amount, which reflect different aspects of transmission performance, into a single, easy-to-understand score or grade through a specific algorithm model (such as weighted scoring or fuzzy logic), to evaluate the current health status of the harmonic reducer as a whole; and the performance verification result is a final qualitative judgment made by comparing the health degree index with the preset threshold standard (such as "excellent", "qualified", "need attention", "unqualified"), which provides a direct basis for equipment predictive maintenance or replacement decision. From a technical implementation point of view, the evaluation logic built into the system automatically processes the overall efficiency decay trend represented by the curve and the absolute gap growth amount, obtains the health degree score through comprehensive operation, and instantly completes the comparison with the standard threshold, thereby outputting a clear verification conclusion.

[0071] In a feasible implementation, the step of obtaining a performance health degree index according to the efficiency decay rate curve and the gap growth amount, and obtaining a performance verification result based on the performance health degree index includes: inputting the efficiency decay rate curve and the gap growth amount into a pre-trained performance degradation evaluation model to obtain a performance health degree index; predicting a remaining useful life of the harmonic reducer according to the performance health index; obtaining a performance verification result according to the performance health index and the remaining useful life.

[0072] In a specific implementation, first, the system inputs the obtained efficiency decay rate curve and gap growth amount, which are two key features, as a set of input vectors into a performance degradation evaluation model trained using historical data in advance. The model can learn the nonlinear mapping relationship between these complex features and the overall health state, and output a quantitative performance health index. Then, the health index is input into another life prediction model, which calculates the remaining useful life prediction value of the harmonic reducer according to the correlation between the health index and historical failure data.

[0073] Finally, the generation of the performance verification result is a decision-making process. The system compares the calculated real-time performance health index and the predicted remaining useful life with the preset multi-level alarm thresholds. Through the rule engine, the system can output a comprehensive qualitative conclusion, such as “state is good, can continue to use” or “performance is seriously degraded, suggest planned maintenance”, thereby completing the closed loop from data to decision.

[0074] In a feasible implementation, the step of inputting the efficiency decay rate curve and the gap growth amount into the pre-trained performance degradation evaluation model to obtain the performance health index comprises: performing data fusion on the efficiency decay rate curve, the gap growth amount, and the real-time collected operating temperature to form a fusion feature vector; inputting the fusion feature vector into the pre-trained performance degradation evaluation model to obtain the performance health index.

[0075] In a specific implementation, the system first performs a data fusion step to build a more comprehensive health feature profile. It will align and integrate the efficiency decay rate curve reflecting long-term performance trends, the gap growth amount directly representing the scalar value of mechanical wear, and the real-time collected operating temperature data. Through feature engineering, these different types and dimensions of data are standardized and combined into a unified, multi-dimensional fusion feature vector, which can more fully describe the comprehensive health status of the harmonic reducer in a specific working state.

[0076] Then, the fusion feature vector is input into the pre-trained performance degradation evaluation model. In its internal structure, it has mastered the mapping rule from complex features to health status through learning a large amount of historical data. The model processes and reasons the input vector, and finally outputs a normalized performance health index. This index can more accurately reflect the performance under the current actual working condition by fusing real-time temperature information, and improves the reliability of the evaluation result.

[0077] In a feasible implementation, the step of obtaining the performance health index according to the efficiency decay rate curve and the gap growth amount, and obtaining the performance check result based on the performance health index further includes: detecting the performance check result to obtain a detection result; when the detection result is abnormal, generating an alarm information according to the performance check result, and controlling the humanoid robot to stop and wait for maintenance.

[0078] In specific implementation, the logic control unit built-in the system will continuously detect and judge the automatically generated performance check result, such as severe degradation or predicted life is about to run out. This detection process is actually to compare the check result in the form of text or code with the preset abnormal threshold list. Once it matches the state defined as abnormal (such as the result level exceeds the critical value), a Boolean type detection result flag will be triggered, and the system state will be switched to the alarm ready mode.

[0079] Subsequently, the system enters the safety disposal stage. The alarm management module will automatically generate structured alarm information containing fault type, severity and time based on specific abnormal check results, and send it to the operation and maintenance personnel through the man-machine interface or network. Almost at the same time, the safety link of the system is activated, and the control unit will send a high-priority stop command to the motion controller of the humanoid robot. The robot will execute the pre-defined safety stop program (such as slow folding action or immediate brake), and enter a locked state, thereby forcibly interrupting the current task and waiting for on-site maintenance by technical personnel, so as to ensure the safety of equipment and personnel.

[0080] The embodiment provides a harmonic reducer performance calibration and verification method for a humanoid robot, when the robot is in a calibration posture, multi-source signals of the harmonic reducer are collected, and an initial efficiency-torque curve and a reverse transmission gap are established as benchmark performance parameters according to the multi-source signals. Then, according to a preset verification period, the calibration posture is reproduced and real-time multi-source signals are collected, and real-time efficiency-torque curves and reverse transmission gaps are calculated. By comparing the real-time performance parameters with the benchmark parameters, an efficiency attenuation rate curve and a gap growth amount are obtained, and then a performance health degree index is comprehensively evaluated to form a final verification result. The method realizes multi-dimensional and periodic quantitative monitoring of performance degradation of the harmonic reducer, effectively solves the problems of performance attenuation and gap increase caused by long-term operation, and improves the motion accuracy and life prediction accuracy of the humanoid robot.

[0081] It should be noted that the above examples are only used for understanding the present application and do not constitute a limitation on the harmonic reducer performance calibration and verification method for a humanoid robot of the present application. More forms of simple transformation based on the technical concept are within the protection scope of the present application.

[0082] The present application also provides a harmonic reducer performance calibration and verification device for a humanoid robot, please refer to Figure 3 The harmonic reducer performance calibration and verification device for a humanoid robot comprises: A benchmark calibration module 10 is configured to acquire multi-source signals of a harmonic reducer in a running state when a humanoid robot is in a calibration posture, determine an initial efficiency-torque curve and an initial reverse transmission gap based on the multi-source signals, and obtain benchmark performance parameters according to the initial efficiency-torque curve and the initial reverse transmission gap. A real-time monitoring module 20 is configured to reproduce the calibration posture to acquire real-time multi-source signals based on a preset verification period, determine real-time efficiency-torque curves and real-time reverse transmission gaps based on the real-time multi-source signals, and obtain real-time performance parameters according to the real-time efficiency-torque curves and the real-time reverse transmission gaps. A parameter comparison module 30 is configured to compare the benchmark performance parameters with the real-time performance parameters to obtain an efficiency attenuation rate curve and a gap growth amount. A health evaluation module 40 is configured to obtain a performance health degree index according to the efficiency attenuation rate curve and the gap growth amount, and obtain a performance verification result based on the performance health degree index.

[0083] In a feasible implementation, the real-time monitoring module 20 is further configured to control the humanoid robot to reproduce the calibration posture in an idle time window after the humanoid robot completes a specific task according to a dynamic verification period based on an actual working time length of the robot. drive the harmonic reducer at multiple constant speeds, and synchronously collect real-time current signals, torque signals and operating temperatures, taking the current signals, torque signals and operating temperatures as real-time multi-source signals; Based on the real-time multi-source signals, the real-time efficiency-torque curve and the real-time reverse transmission gap are calculated through parameter identification; The real-time efficiency-torque curve and the real-time reverse transmission gap are fused to obtain real-time performance parameters.

[0084] In a feasible implementation, the real-time monitoring module 20 is further configured to use recursive least squares on the real-time multi-source signals, fit the collected multiple groups of current, torque and temperature data to obtain the real-time efficiency-torque curve; The joint motor is controlled to perform forward and reverse rotation within a preset interval at zero point, and the reading difference of the encoder at torque zero crossing is recorded to obtain the real-time reverse transmission gap according to the reading difference.

[0085] In a feasible implementation, the real-time monitoring module 20 is further configured to extract efficiency values at preset torque sampling points from the real-time efficiency-torque curve to form an efficiency feature sequence; The efficiency feature sequence and the real-time reverse transmission gap are spliced in a predetermined order to form a multi-dimensional performance parameter vector; The multi-dimensional performance parameter vector is normalized to obtain real-time performance parameters.

[0086] In a feasible implementation, the real-time monitoring module 20 is further configured to calculate the relative change percentages of real-time efficiency values and initial efficiency values based on sampling points on the real-time efficiency-torque curve; An efficiency attenuation rate curve is obtained according to the efficiency attenuation percentages of the sampling points.

[0087] In a feasible implementation, the parameter comparison module 30 is further configured to compare the real-time efficiency-torque curve and the initial efficiency-torque curve point by point to calculate an efficiency attenuation rate, and generate an efficiency attenuation rate curve based on the efficiency attenuation rate; The real-time reverse transmission gap is subtracted from the initial reverse transmission gap to obtain a gap growth amount.

[0088] In a feasible implementation, the health assessment module 40 is further configured to input the efficiency attenuation rate curve and the gap growth amount into a pre-trained performance degradation assessment model to obtain a performance health degree index; The remaining useful life of the harmonic reducer is predicted according to the performance health degree index; A performance verification result is obtained according to the performance health degree index and the remaining useful life.

[0089] In an implementable embodiment, the health assessment module 40 is further configured to perform data fusion on the efficiency decay rate curve, the gap growth amount, and the real-time collected operating temperature to form a fusion feature vector. The fusion feature vector is input into a pre-trained performance degradation assessment model to obtain a performance health index.

[0090] In an implementable embodiment, the health assessment module 40 is further configured to detect the performance calibration result to obtain a detection result. When the detection result is abnormal, an alarm information is generated according to the performance calibration result, and the humanoid robot is controlled to stop and wait for maintenance.

[0091] The harmonic reducer performance calibration and verification device for a humanoid robot provided in the present application adopts the harmonic reducer performance calibration and verification method for a humanoid robot in the above embodiments, and can solve the technical problems of long-term operation performance decay and gap increase of the harmonic reducer of the humanoid robot, which leads to robot precision decline and inaccurate life prediction. Compared with the prior art, the harmonic reducer performance calibration and verification device for a humanoid robot provided in the present application has the same beneficial effects as the harmonic reducer performance calibration and verification method for a humanoid robot provided in the above embodiments, and other technical features in the harmonic reducer performance calibration and verification device for a humanoid robot are the same as the features disclosed in the above method embodiments, which will not be repeated here.

[0092] The present application provides a harmonic reducer performance calibration and verification device for a humanoid robot, which comprises at least one processor and a memory in communication connection with the at least one processor. The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the harmonic reducer performance calibration and verification method for a humanoid robot in the above embodiment one.

[0093] Reference will be made to the following Figure 4This document illustrates a structural schematic diagram of a harmonic reducer performance calibration and verification device suitable for implementing embodiments of this application for humanoid robots. The harmonic reducer performance calibration and verification device for humanoid robots in embodiments of this application may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital radio receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), and in-vehicle terminals (e.g., in-vehicle navigation terminals), as well as fixed terminals such as digital TVs and desktop computers. Figure 4 The harmonic reducer performance calibration and verification device for humanoid robots shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0094] like Figure 4 As shown, the harmonic reducer performance calibration and verification device for humanoid robots may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in ROM (Read Only Memory) 1002 or a program loaded from storage device 1003 into RAM (Random Access Memory) 1004. RAM 1004 also stores various programs and data required for the operation of the harmonic reducer performance calibration and verification device for humanoid robots. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via bus 1005. Input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to I / O interface 1006: input devices 1007 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, LCDs (Liquid Crystal Displays), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. Communication device 1009 allows the harmonic reducer performance calibration and verification equipment for humanoid robots to exchange data wirelessly or via wired communication with other devices. Although the figure shows a harmonic reducer performance calibration and verification equipment for humanoid robots with various systems, it should be understood that it is not required to implement or possess all the systems shown. More or fewer systems can be implemented alternatively.

[0095] In particular, according to the embodiments disclosed in the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present application include a computer program product comprising a computer program carried on a computer readable medium, the computer program containing program code for executing the method shown in the flowchart. In such embodiments, the computer program can be downloaded and installed from a network through a communication device, or installed from the storage device 1003, or installed from the ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiments disclosed in the present application are executed.

[0096] The harmonic reducer performance calibration and verification device for humanoid robots provided by the present application adopts the harmonic reducer performance calibration and verification method for humanoid robots in the above-mentioned embodiments, and can solve the technical problem of harmonic reducer performance calibration and verification for humanoid robots. Compared with the prior art, the harmonic reducer performance calibration and verification device for humanoid robots provided by the present application has the same beneficial effects as the harmonic reducer performance calibration and verification method for humanoid robots provided by the above-mentioned embodiments, and other technical features in the harmonic reducer performance calibration and verification device for humanoid robots are the same as the features disclosed in the previous embodiment method, which will not be repeated here.

[0097] It should be understood that various parts of the present application can be realized by hardware, software, firmware or a combination thereof. In the description of the above-mentioned embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0098] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical scope disclosed in the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0099] The present application provides a computer readable storage medium having stored thereon computer readable program instructions (i.e. computer program) for executing the harmonic reducer performance calibration and verification method for humanoid robots in the above-mentioned embodiments.

[0100] The computer readable storage medium provided in the application may be, for example, a U disk, but is not limited to an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, system or device, or any combination thereof. More specific examples of the computer readable storage medium may include, but are not limited to, an electrical connection with one or more conductive wires, a portable computer disk, a hard disk, a RAM (Random Access Memory), a ROM (Read Only Memory), an EPROM (Erasable Programmable Read Only Memory or flash memory), an optical fiber, a CD-ROM (CD-Read Only Memory), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the embodiment, the computer readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, system or device. The program code contained on the computer readable storage medium can be transmitted by any suitable medium, including but not limited to an electrical wire, an optical cable, an RF (Radio Frequency), etc., or any suitable combination thereof.

[0101] The computer readable storage medium described above may be contained in the harmonic reducer performance calibration and verification device for a humanoid robot, or may exist separately without being assembled into the harmonic reducer performance calibration and verification device for a humanoid robot.

[0102] The computer readable storage medium described above carries one or more programs, which, when executed by the harmonic reducer performance calibration and verification device for a humanoid robot, cause the harmonic reducer performance calibration and verification device for a humanoid robot to: when the humanoid robot is in a calibration posture, acquire multi-source signals in a running state of the harmonic reducer, and determine an initial efficiency-torque curve and an initial reverse transmission gap based on the multi-source signals, and obtain a reference performance parameter according to the initial efficiency-torque curve and the initial reverse transmission gap. Based on a preset verification period, the calibration posture is reproduced to acquire real-time multi-source signals, and real-time efficiency-torque curve and real-time reverse transmission gap are determined based on the real-time multi-source signals, and real-time performance parameters are obtained according to the real-time efficiency-torque curve and the real-time reverse transmission gap. The reference performance parameters are compared with the real-time performance parameters to obtain an efficiency decay rate curve and a gap growth amount. A performance health degree index is obtained according to the efficiency decay rate curve and the gap growth amount, and a performance verification result is obtained based on the performance health degree index.

[0103] Computer program code for carrying out operations of the present application can be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C++ or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider).

[0104] The flow diagrams and the block diagrams in the drawings are meant as methodological and functional description of implementations of systems, methods, and computer program products according to various embodiments of the present application. In this regard, each block in the flow diagrams or block diagrams can represent a module, a segment, or a portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that in some alternative implementations, the functions noted in the blocks can occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently or the blocks may

[0105] The modules involved in the embodiments of the present application can be implemented in software or hardware. In some cases, the names of the modules do not constitute a limitation on the modules themselves.

[0106] The readable storage medium provided by the present application is a computer readable storage medium, which stores computer readable program instructions (i.e. computer programs) for executing the above-mentioned harmonic reducer performance calibration and verification method for humanoid robots, and can solve the technical problem of harmonic reducer performance calibration and verification for humanoid robots. Compared with the prior art, the computer readable storage medium provided by the present application has the same beneficial effects as the harmonic reducer performance calibration and verification method for humanoid robots provided by the above-mentioned embodiments, and will not be described here.

[0107] The application also provides a computer program product comprising a computer program which, when executed by a processor, implements the steps of the method for calibrating and checking the performance of a harmonic reducer of a humanoid robot as described above.

[0108] The computer program product provided by the application can solve the technical problem of calibrating and checking the performance of a harmonic reducer of a humanoid robot. Compared with the prior art, the beneficial effects of the computer program product provided by the application are the same as those of the method for calibrating and checking the performance of a harmonic reducer of a humanoid robot provided by the above-mentioned embodiments, and are not described here.

[0109] The above is only some embodiments of the application, and does not limit the patent scope of the application. Any equivalent structural transformation made by using the content of the specification and drawings, or direct / indirect application in other related technical fields within the technical concept of the application is included in the patent protection scope of the application.

Claims

1. A method for performance calibration and verification of a harmonic reducer for a humanoid robot, characterized by, The harmonic reducer performance calibration and verification method for the humanoid robot comprises: When the humanoid robot is in a calibration posture, a plurality of source signals in a running state of the harmonic reducer are acquired, and initial efficiency-torque curves and initial reverse transmission clearances are determined based on the plurality of source signals, and reference performance parameters are obtained according to the initial efficiency-torque curves and the initial reverse transmission clearances; Based on a preset verification period, the calibration posture is reproduced to acquire real-time source signals, real-time efficiency-torque curves and real-time reverse transmission clearances are determined based on the real-time source signals, and real-time performance parameters are obtained according to the real-time efficiency-torque curves and the real-time reverse transmission clearances; The reference performance parameters are compared with the real-time performance parameters to obtain an efficiency attenuation rate curve and a clearance growth amount; Performance health degree indexes are obtained according to the efficiency attenuation rate curve and the clearance growth amount, and performance verification results are obtained based on the performance health degree indexes.

2. The method of claim 1, wherein, The step of reproducing the calibration posture based on a preset verification period to acquire real-time source signals, determining real-time efficiency-torque curves and real-time reverse transmission clearances based on the real-time source signals, and obtaining real-time performance parameters according to the real-time efficiency-torque curves and the real-time reverse transmission clearances comprises: According to a dynamic verification period based on an actual working time of the robot, the humanoid robot is controlled to reproduce the calibration posture in an idle time window after completing a specific task; The harmonic reducer is driven at a plurality of constant speeds, and real-time current signals, torque signals and running temperatures are synchronously collected, and the current signals, the torque signals and the running temperatures are taken as real-time source signals; Real-time efficiency-torque curves and real-time reverse transmission clearances are calculated based on the real-time source signals through parameter identification; The real-time efficiency-torque curves and the real-time reverse transmission clearances are fused to obtain real-time performance parameters.

3. The method of claim 2, wherein, The step of calculating real-time efficiency-torque curves and real-time reverse transmission clearances based on the real-time source signals through parameter identification comprises: Recursive least squares are used on the real-time source signals to fit a plurality of groups of collected current, torque and temperature data to obtain the real-time efficiency-torque curves; The joint motor is controlled to perform forward and reverse rotation in a preset interval at zero point, a reading difference value of the encoder at torque zero is recorded, and the real-time reverse transmission clearance is obtained according to the reading difference value.

4. The method of claim 2, wherein, The step of fusing the real-time efficiency-torque curves and the real-time reverse transmission clearances to obtain real-time performance parameters comprises: Efficiency values at preset torque sampling points are extracted from the real-time efficiency-torque curves to form an efficiency feature sequence; The efficiency feature sequence and the real-time reverse transmission clearance are spliced in a predetermined order to form a multi-dimensional performance parameter vector; The multi-dimensional performance parameter vector is normalized to obtain real-time performance parameters.

5. The method of claim 4, wherein, The step of comparing the real-time efficiency-torque curves with the initial efficiency-torque curves point by point, calculating an efficiency attenuation rate, and generating an efficiency attenuation rate curve comprises: Calculate the relative change percentage of the real-time efficiency value and the initial efficiency value based on the sampling points on the real-time efficiency-torque curve, respectively; Obtain an efficiency attenuation rate curve according to the efficiency attenuation percentages of the sampling points.

6. The method of claim 1, wherein, The step of comparing the baseline performance parameter with the real-time performance parameter to obtain an efficiency attenuation rate curve and a gap growth amount includes: Point-by-point compare the real-time efficiency-torque curve with the initial efficiency-torque curve to calculate the efficiency attenuation rate, and generate an efficiency attenuation rate curve based on the efficiency attenuation rate; Subtract the real-time reverse transmission gap from the initial reverse transmission gap to obtain the gap growth amount.

7. The method of claim 1, wherein, The step of obtaining a performance health degree index according to the efficiency attenuation rate curve and the gap growth amount, and obtaining a performance verification result based on the performance health degree index includes: Input the efficiency attenuation rate curve and the gap growth amount into a pre-trained performance degradation evaluation model to obtain a performance health degree index; Predict the remaining service life of the harmonic reducer according to the performance health degree index; Obtain a performance verification result according to the performance health degree index and the remaining service life.

8. The method of claim 7, wherein, The step of inputting the efficiency attenuation rate curve and the gap growth amount into a pre-trained performance degradation evaluation model to obtain a performance health degree index includes: Data fusion of the efficiency attenuation rate curve, the gap growth amount, and the real-time collected operating temperature to form a fusion feature vector; Input the fusion feature vector into a pre-trained performance degradation evaluation model to obtain a performance health degree index.

9. The method of claim 1, wherein, After the step of obtaining a performance health degree index according to the efficiency attenuation rate curve and the gap growth amount, and obtaining a performance verification result based on the performance health degree index, further includes: Detect the performance verification result to obtain a detection result; When the detection result is abnormal, generate an alarm information according to the performance verification result, and control the humanoid robot to stop working and wait for maintenance.

10. A harmonic reducer performance calibration and verification device for a humanoid robot, characterized by, The harmonic reducer performance calibration and verification device for a humanoid robot includes: A baseline calibration module configured to acquire multi-source signals in a running state of a harmonic reducer when a humanoid robot is in a calibration posture, determine an initial efficiency-torque curve and an initial reverse transmission gap based on the multi-source signals, and obtain a baseline performance parameter according to the initial efficiency-torque curve and the initial reverse transmission gap; A real-time monitoring module configured to reproduce the calibration posture based on a preset verification period to acquire real-time multi-source signals, determine a real-time efficiency-torque curve and a real-time reverse transmission gap based on the real-time multi-source signals, and obtain a real-time performance parameter according to the real-time efficiency-torque curve and the real-time reverse transmission gap; A parameter comparison module configured to compare the baseline performance parameter with the real-time performance parameter to obtain an efficiency attenuation rate curve and a gap growth amount; A health evaluation module configured to obtain a performance health degree index according to the efficiency attenuation rate curve and the gap growth amount, and obtain a performance verification result based on the performance health degree index.

Images