Motor output torque measuring method

Abstract

The invention relates to the technical field of torque measurement, in particular to a motor output torque measurement method, which comprises the following steps of: obtaining a motor torque vector, a motor rotating speed vector, a motor power vector and a motor temperature vector; determining the local stability of the torque reference value; calculating a waveform confusion coefficient; obtaining a signal confusion coefficient; determining a motor torque reference comparison confusion degree; obtaining the interference intensity of the temperature to the torque; obtaining an anti-interference prediction quantity of the torque of the motor; determining a gain coefficient of a Kalman filtering method; and through a Kalman filtering method, a motor temperature correction torque is obtained, and a final output torque is determined. According to the invention, the influence of the motor temperature on the measured torque can be eliminated, and the more accurate measured torque can be obtained under the condition that the motor temperature changes continuously.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of torque measurement, and particularly relates to a motor output torque measurement method. BACKGROUND

[0002] In industrial automation, energy, and many fields that rely on motor driving, electrical transmission equipment plays a crucial role. They are widely used in various mechanical equipment to convert electrical energy into mechanical energy, achieving precise motion and torque control. The rack and pinion type control rod drive mechanism is a key application of electrical transmission equipment in the field of nuclear energy, mainly used to realize the functions of lifting, inserting, maintaining, and accident drop rod stopping of control rods in the reactor core. The mechanism is composed of multiple important components, including a sealed shell assembly, a motor assembly, a gear transmission assembly, an electromagnetic clutch, and a rod position detection device. Its transmission system is composed of two main transmission systems: a drive transmission system and a grabbing transmission system. The drive transmission system is composed of a drive motor, a reducer, and an electromagnetic clutch, responsible for providing the main driving torque; while the grabbing transmission system is composed of a motor, a control rod connecting device, etc., responsible for direct interaction with the control rod.

[0003] In the drive transmission system, the motor as the core electrical transmission equipment, the accurate control of its output torque is the key to ensure the safe and reliable operation of the entire control rod drive mechanism. Monitoring the motor torque is an effective means to ensure the safe and stable output of the control rod control power, therefore, obtaining accurate torque is crucial for motor control. Currently, most torque sensors use strain electric measurement principle torque sensors. This type of sensor consists of a signal gear and a coil installed on the motor output shaft (elastic shaft), and a magneto-electric signal generator is installed at both ends of the rotor. When the elastic shaft is driven by the motor, the elastic shaft will deform due to the torque, causing the signal gears and coils at both ends to produce different induced electromotive forces. By measuring the phase difference of the induced electromotive forces at both ends, the motor electromotive force can be derived, and the torque can be calculated.

[0004] However, as a typical electrical transmission equipment, the motor's temperature will continuously rise as the working environment changes, especially as the load and running time increase during operation. Changes in temperature will significantly affect the physical properties of the elastic shaft, such as the modulus of elasticity of the material, and thus affect the deformation degree and the proportional relationship of the torque. The torque sensor calculates the deformation degree by detecting the phase difference of the electrical signal, and then calculates the torque based on the deformation degree. When the physical properties of the elastic shaft change, the proportional relationship between the deformation degree and the torque will also change, resulting in measurement errors of the torque sensor, making the obtained torque measurement results inaccurate.

[0005] Therefore, developing an electric drive output torque determination method capable of overcoming the influence of temperature change on the physical properties of the electric drive equipment, thereby improving the torque measurement accuracy, has important practical significance and application value for improving the performance and reliability of the electric drive system. This not only improves the safety and reliability of the control rod drive mechanism, but also can be widely used in other industrial fields that rely on motor drive. SUMMARY

[0006] In order to solve the above technical problems, a motor output torque measurement method is provided to solve the existing problems.

[0007] The technical problem of the present application is solved by providing a motor output torque measurement method, comprising the following steps: Obtain the motor torque vector, the motor temperature vector, and the motor torque reference vector; Using a signal decomposition algorithm on the motor torque vector, the signal strength of each sub-signal at each sampling time is obtained; according to the torque of each sampling time of the motor torque reference vector and the fluctuation difference of the signal strength of each sub-signal at each sampling time, the motor torque reference contrast chaos degree of each sub-signal is determined; Based on the motor temperature vector, the temperature similar vector of each discrete class, the temperature discrete vector, the signal waveform strength value vector of each sub-signal, and the signal torque coefficient vector of each sub-signal are constructed; According to the change difference between the temperature value of each collection time in the temperature similar vector and the temperature discrete vector of each discrete class and the signal strength of the corresponding collection time in the signal waveform strength value vector of each sub-signal, and the discrete degree of the signal torque coefficient vector, the temperature-to-torque interference strength of each discrete class is determined; According to the influence degree of the motor temperature at each sampling time on the torque of each sampling time of the motor torque reference vector, the motor torque anti-interference prediction at each sampling time is determined; Based on the motor torque anti-interference prediction at each sampling time, and combined with the Kalman filtering method, the motor temperature corrected torque is obtained, and the final output torque is determined.

[0008] Preferably, the motor torque reference contrast chaos degree of each sub-signal is determined, comprising: Obtain the motor speed vector and the motor power vector, and the motor torque reference vector is a vector composed of the ratio of the motor power to the motor speed at the corresponding sampling time in the motor power vector and the motor speed vector; The vector composed of the calculation results of the first derivative of the torque at each sampling time in the motor torque reference vector is denoted as the motor torque change reference vector; analyzing local variation of the torque at each sampling time in the motor torque reference vector to determine local stability of the torque reference value at each sampling time in the motor torque variation reference vector; analyzing difference degree of fluctuation variation between the torque at all collection times in the motor torque reference vector and the signal strength at all collection times of each partial signal to determine waveform confusion coefficient of each partial signal; composing a vector of the ratio between the torque at each sampling time in the motor torque reference vector and the signal strength of each partial signal at the same sampling time as the torque coefficient vector of each partial signal; analyzing difference variation of the torque coefficient at each sampling time in the torque coefficient vector of each partial signal and the local stability of the torque reference value corresponding to the collection time to determine signal confusion coefficient of each partial signal; fusing the waveform confusion coefficient and the signal confusion coefficient to determine motor torque reference contrast confusion degree of each partial signal.

[0009] Preferably, the local stability of the torque reference value at each sampling time in the motor torque variation reference vector comprises: taking the torque at any sampling time in the motor torque variation reference vector as a target reference value, and taking the sum of the absolute value of the torque at multiple sampling times before the target reference value and the absolute value of the target reference value as the local reference coefficient of the target reference value corresponding sampling time in the motor torque variation reference vector; taking the reciprocal of the local reference coefficient as the local stability of the torque reference value at each sampling time in the motor torque variation reference vector.

[0010] Preferably, the waveform confusion coefficient of each partial signal comprises: respectively performing dimensionless processing on the torque at each sampling time of the motor torque reference vector and the signal strength of each partial signal at each sampling time to obtain a motor torque waveform reference vector and each motor torque waveform decomposition signal; analyzing the difference between the torque at each sampling time in the motor torque waveform reference vector and the signal strength at the same collection time of each motor torque waveform decomposition signal as the torque waveform difference of each partial signal at each sampling time; taking the product of the torque waveform difference of each partial signal at each sampling time and the local stability of the torque reference value at the same sampling time in the motor torque variation reference vector as the waveform contrast coefficient of each partial signal at each sampling time; fusing the waveform contrast coefficient at all sampling times to determine the waveform confusion coefficient of each partial signal.

[0011] Preferably, the signal confusion coefficient of each partial signal comprises: an absolute value of a difference between the torque coefficient at each sampling time in the torque coefficient vector of each partial signal and a mean value of the torque coefficient vector, denoted as a signal difference of each partial signal at each sampling time; a product of the signal difference of each partial signal at each sampling time and the torque reference value local stability at the corresponding sampling time, as a signal stability coefficient of each partial signal at each sampling time; a sum of the signal stability coefficients at all sampling times, as a signal confusion coefficient of each partial signal.

[0012] Preferably, the construction method of the temperature similarity vector of each discrete class, the temperature discrete vector, the signal waveform intensity value vector of each partial signal, and the signal torque coefficient vector of each partial signal is as follows: all temperature values in the motor temperature vector are divided into multiple intervals to obtain multiple discrete classes; a vector composed of all temperature values belonging to the same interval is taken as the temperature similarity vector of each discrete class; all discrete classes are numbered and sorted to obtain a serial number of each discrete class; each temperature value in the temperature similarity vector of each discrete class is replaced by the serial number of the discrete class to which it belongs to obtain the temperature discrete vector of each discrete class; each temperature value in each discrete class corresponds to a collection time in the motor temperature vector, and according to the collection time of each temperature value in the temperature similarity vector of each discrete class, the signal intensity and the torque coefficient at the corresponding collection time are extracted from each partial signal and each torque coefficient vector, respectively, to form the signal waveform intensity value vector and the signal torque coefficient vector of each partial signal of each discrete class.

[0013] Preferably, the determination of the temperature-to-torque interference intensity of each discrete class comprises: the discrete degree of the signal torque coefficient vector of each partial signal of each discrete class is taken as a torque coefficient variation scale of each partial signal of each discrete class; a ratio between each collection time temperature value in the temperature discrete vector of each discrete class and a mean value of all temperatures in the temperature similarity vector of the discrete class to which it belongs is denoted as a temperature interference coefficient of each collection time of each discrete class; a ratio between the signal intensity at each collection time in the signal waveform intensity value vector of each partial signal of each discrete class and a mean value of the signal waveform intensity value vector is denoted as a signal interference coefficient of each collection time in each partial signal of each discrete class; a difference between the temperature interference coefficient and the signal interference coefficient is determined as a temperature waveform difference of each collection time in each partial signal of each discrete class; The reciprocal of the sum of the temperature waveform differences of all sampling moments is taken as the temperature interference degree of each sub-signal of each discrete class; The temperature interference degree, the torque coefficient variation scale and the motor torque reference contrast chaos degree of all sub-signals of each discrete class are fused as the temperature-to-torque interference strength of each discrete class.

[0014] Preferably, the determination of the motor torque anti-interference prediction quantity at each sampling moment comprises: Based on the temperature-to-torque interference strength of each discrete class, the prediction quantity distribution weight at each sampling moment is calculated; The ratio between the average torque of all sampling moments in the motor torque reference vector and the average torque of all sampling moments in the motor torque vector is taken as the torque relative difference; The torque at each sampling moment in the motor torque reference vector and the torque relative difference are fused to determine the interference prediction coefficient at each sampling moment; The product of the interference prediction coefficient at each sampling moment and the prediction quantity distribution weight is taken as the temperature interference prediction degree at each sampling moment; The difference between the numerical value 1 and the prediction quantity distribution weight at each sampling moment is taken as the temperature correction weight; The product of the motor temperature correction torque at the previous moment at each sampling moment and the temperature correction weight at the corresponding sampling moment is taken as the temperature correction prediction degree; The temperature interference prediction degree and the temperature correction prediction degree are fused to determine the motor torque anti-interference prediction quantity at each sampling moment.

[0015] Preferably, the calculation method of the prediction quantity distribution weight at each sampling moment is: The median temperature in the corresponding interval of each discrete class is obtained, and the median temperature and the temperature-to-torque interference strength of each discrete class form a two-dimensional discrete array; A rectangular coordinate system is constructed for the two-dimensional discrete arrays of all discrete classes, and linear fitting is performed on all coordinate points to obtain a fitting function, which is taken as the temperature interference fitting function, to obtain the temperature interference fitting value at each sampling moment; The normalized processing result of the ratio between the torque reference value local stability at each sampling moment and the temperature interference fitting value at the corresponding sampling moment is taken as the prediction quantity distribution weight at each sampling moment.

[0016] Preferably, the obtaining of the motor temperature correction torque and the determination of the final output torque comprise: The maximum value of the temperature-to-torque interference strength of all discrete classes is taken as the maximum temperature-to-torque interference strength; The ratio of the temperature interference fitting value of each sampling time to the maximum temperature to the torque interference strength is taken as the dynamic Kalman gain of each sampling time; The motor torque anti-interference prediction value of each sampling time is taken as the prediction value of each sampling time; The torque of each sampling time in the motor torque vector is taken as the measurement value of each sampling time; The dynamic Kalman gain is taken as the gain coefficient in the Kalman filtering method, and the motor temperature corrected torque of each sampling time is obtained by combining the prediction value and the measurement value of each sampling time; The motor temperature corrected torque is taken as the measured torque of the torque sensor, and the final output torque is obtained.

[0017] The present application has at least the following beneficial effects: The present application decomposes the motor torque vector by using the signal decomposition algorithm, obtains the motor torque change reference vector by the power and the speed of the motor, obtains the motor torque reference contrast chaos degree, analyzes the difference degree between the motor torque measured by the torque sensor and the torque calculated by the motor torque formula, and considers the interference degree of the motor temperature to the torque sensor; the torque coefficient variation scale is calculated according to the motor torque change reference vector and the motor torque vector under different temperatures, and the temperature interference strength to the torque is calculated by combining the motor torque reference contrast chaos degree, which has the beneficial effect that the influence of the temperature on the measurement error of the torque sensor under different temperatures is reflected; the prediction model is constructed by combining the motor torque reference value at the current time, the motor torque reference value and the motor torque measurement value, and the motor temperature corrected torque at the last time, the motor torque anti-interference prediction value is obtained, and the dynamic Kalman gain is calculated by the temperature interference fitting function, and the measured torque is corrected by using the Kalman filtering method, which has the beneficial effect that the influence of the motor temperature on the measured torque is eliminated, which is conducive to obtaining more accurate torque size under the condition that the motor temperature is constantly changing, and improving the control effect of the driving motor. BRIEF DESCRIPTION OF DRAWINGS

[0018] The motor output torque measurement method of the present application will be further described in detail below in combination with the drawings.

[0019] Figure 1 A step flowchart of the motor output torque measurement method provided by the present application; Figure 2 A flowchart of the temperature interference strength acquisition method of each discrete class provided by the present application; Figure 3 A flowchart of the Kalman filtering provided by the present application. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical scheme and advantages of the present application more clear, the motor output torque measurement method provided by the present application is further described in detail below in combination with the drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application and not to limit the present application.

[0021] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0022] Please refer to Figure 1 which shows the step flow chart of a motor output torque measurement method provided by an embodiment of the present application, the method comprising the following steps: Step 1, obtaining motor torque vector, motor temperature vector, and obtaining motor torque reference vector.

[0023] Connect the torque sensor, speed sensor, power sensor and temperature sensor with the measured drive motor, collect the torque, speed, power and heating temperature of the electric drive motor during the motor rotation, wherein the sampling time interval t seconds, the sampling time T, according to the sampling time, the electric drive motor data collected by each sensor is respectively composed of motor torque vector, motor speed vector, motor power vector and motor temperature vector.

[0024] The ratio of the motor power and the motor speed at the corresponding sampling time in the motor power vector and the motor speed vector is composed of the vector as the motor torque reference vector.

[0025] Preferably, in the present embodiment, the sampling time interval is 0.1s, and the sampling time is 5min. The present embodiment only provides a collection scheme, and the implementer can set it according to the actual situation.

[0026] Step 2, using signal decomposition algorithm for motor torque vector, obtaining signal intensity of each partial signal at each sampling time; according to the torque of each sampling time in the motor torque reference vector and the fluctuation difference of the signal intensity of each partial signal at each sampling time, determine the motor torque reference contrast confusion degree of each partial signal.

[0027] According to the motor torque formula, the power of the motor is proportional to the product of the torque speed, and the motor torque can be calculated through the power and speed of the motor. Therefore, the influence of motor temperature on torque can be reflected by comparing and analyzing the motor torque, motor speed and motor power.

[0028] First step: according to the local change of torque of each sampling time in the motor torque reference vector, calculate the local stability of torque reference value of each sampling time in the motor torque change reference vector.

[0029] A vector composed of the first derivative of the torque at each sampling time in the motor torque reference vector is denoted as a motor torque change reference vector; Taking the torque at any sampling time in the motor torque change reference vector as a target reference value, the sum of the absolute values of the torque at a plurality of sampling times before the target reference value and the absolute value of the target reference value is taken as a local reference coefficient of the sampling time corresponding to the target reference value in the motor torque change reference vector; The reciprocal of the local reference coefficient is taken as the local stability of the torque reference value at each sampling time in the motor torque change reference vector.

[0030] Preferably, in the present embodiment, the torque at 50 sampling times before each target reference value is used to calculate the local reference coefficient.

[0031] Second step: decompose the motor torque vector by using a signal decomposition algorithm to obtain the signal intensity of each sub-signal at each sampling time and the torque coefficient vector of each sub-signal; determine the waveform confusion coefficient of each sub-signal according to the difference degree of fluctuation change between the torque at all collection times in the motor torque reference vector and the signal intensity at all collection times in each sub-signal.

[0032] The motor torque vector is decomposed by using a signal decomposition algorithm to obtain a plurality of decomposition signals, and each decomposition signal is denoted as each sub-signal. The torque of each sub-signal at each sampling time corresponds to a signal intensity, which is denoted as the signal intensity of each sub-signal at each sampling time. Preferably, in the present embodiment, the motor torque vector is decomposed by using an EMD decomposition algorithm, and the decomposition number is set to 5. The EMD decomposition algorithm is a known technology and will not be described here.

[0033] It can be understood that the present embodiment only provides one signal decomposition algorithm, i.e. the EMD decomposition algorithm, as other embodiments, the implementer can use other signal decomposition algorithms in the prior art for decomposition, such as the LMD decomposition algorithm, the EWT signal decomposition algorithm, the SSA signal decomposition algorithm, etc., and the present embodiment does not specially limit this.

[0034] Further, in order to analyze the difference degree of fluctuation change between the torque reference value of all sampling times in the motor torque reference vector and the signal intensity of all sampling times in the motor torque decomposition signal, the waveform shapes of the two are compared and analyzed, but there is a difference in the numerical value range of the two signals, so it is necessary to eliminate the dimension of the motor torque reference vector and the motor torque decomposition signal to avoid the difference in the numerical value affecting the analysis effect.

[0035] The torque at each sampling time of the motor torque reference vector and the signal intensity of each sub-signal at each sampling time are respectively dimensionalized to obtain the motor torque waveform reference vector and the decomposed signal of each motor torque waveform. Preferably, in this embodiment, the torque at each sampling time in the motor torque reference vector and the signal strength of each sub-signal at each sampling time are averaged to obtain the motor torque waveform reference vector and each motor torque waveform decomposition signal.

[0036] The difference between the torque at each sampling moment in the motor torque waveform reference vector and the signal strength at the same sampling moment of each motor torque waveform decomposition signal is analyzed as the torque waveform difference of each sub-signal at each sampling moment. The product of the difference in torque waveform of each sub-signal at each sampling time and the local stability of the torque reference value at the corresponding sampling time is used as the waveform comparison coefficient of each sub-signal at each sampling time. The waveform comparison coefficients at all sampling times are fused to determine the waveform disorder coefficient of each sub-signal; It is understandable that fusion can be represented as a multiplicative relationship, an additive relationship, etc., which is determined by the actual application, and this application does not impose any special restrictions on it.

[0037] Preferably, in this embodiment, the method for determining the waveform disorder coefficient of each sub-signal is as follows: ,in, For the first The waveform disorder coefficient of each sub-signal, The first in the motor torque waveform reference vector Torque at the sampling time, The mean value of the reference vector for the motor torque waveform. For the first The decomposed signal of the motor torque waveform of the first motor Signal strength at the sampling time, For the first The mean of the decomposed signal of the individual motor torque waveforms The first in the reference vector for motor torque change Local stability of the torque reference value at the sampling time. For all data collection times, For the first The sub-signal in the first... Differences in torque waveform at sampling time.

[0038] Step 3: Based on the difference in torque coefficient at each sampling time in the torque coefficient vector of each sub-signal and the local stability of the torque reference value at the corresponding sampling time of the motor torque change reference vector, determine the signal disorder coefficient of each sub-signal and obtain the motor torque reference comparison disorder of each sub-signal.

[0039] The vector composed of the ratios between the torque at each sampling time in the motor torque reference vector and the signal strength of each sub-signal at the same sampling time is used as the torque coefficient vector of each sub-signal. Analyze the absolute value of the difference between the torque coefficient at each sampling time and the mean value of the torque coefficient vector in the torque coefficient vector of each sub-signal, and denote it as the signal difference of each sub-signal at each sampling time; The product of the signal difference of each sub-signal at each sampling time and the local stability of the torque reference value at the same sampling time in the motor torque change reference vector is used as the signal stability coefficient of each sub-signal at each sampling time. The sum of the signal stability coefficients at all sampling times is taken as the signal disorder coefficient of each sub-signal; The waveform disorder coefficient and the signal disorder coefficient are fused to determine the motor torque reference comparison disorder degree for each sub-signal; Preferably, in this embodiment, the method for calculating the signal disorder coefficient of each sub-signal is as follows: ,in, For the first The signal disorder coefficients of each sub-signal, where... For the first In the torque coefficient vector of the nth sub-signal Torque coefficient at the sampling time, For the first The mean of the torque coefficient vector of each component signal. The first in the reference vector for motor torque change Local stability of the torque reference value at the sampling time. For all data collection times, For the first The sub-signal in the first... Signal differences at sampling times.

[0040] It should be noted that the greater the local stability of the torque reference value of each sampling time in the motor torque change reference vector, the more stable the motor parameter corresponding to the sampling time; the greater the waveform confusion coefficient, the greater the difference between the signal intensity of all sampling times of the signal and the torque waveform of all sampling times of the motor torque change reference vector; the greater the signal confusion coefficient, the less stable the ratio of the signal intensity of all sampling times of the signal to the torque of all sampling times of the motor torque change reference vector, representing the smaller the reasoning accuracy of the motor torque formula to the motor torque, and the more unstable the running state of the motor at this time; secondly, the greater the motor torque reference comparison confusion degree value, the more inconsistent the comparison results of the signal and the motor torque change reference vector, representing that the signal change in the signal is not determined by the motor torque change reference vector, and then the torque information represented by the signal is more affected by temperature change.

[0041] Step 3, based on the motor temperature vector, constructing a temperature similar vector, a temperature discrete vector, a signal waveform intensity value vector of each signal and a signal torque coefficient vector of each signal of each discrete class.

[0042] Further, at different temperatures, the influence of temperature on torque sensor error is different, and therefore, according to the difference of torque change at different temperatures, the motor temperature is divided, and the torque is analyzed step by step, and the specific method is: all temperature values in the motor temperature vector are divided into multiple intervals to obtain multiple discrete classes; Preferably, in the embodiment, the motor temperature vector is divided into 50 discrete classes.

[0043] According to the interval range to which the temperature value of each collection time in the motor temperature vector belongs, the vector composed of all temperature values belonging to the same interval is taken as the temperature similar vector of each discrete class; numbering and sorting all discrete classes to obtain the serial number of each discrete class; each temperature value in the temperature similar vector of each discrete class is replaced by the serial number value of the corresponding discrete class to obtain the temperature discrete vector of each discrete class; each temperature value in each discrete class corresponds to a collection time in the motor temperature vector, according to the collection time of each temperature value in the temperature similar vector of each discrete class, the signal intensity and torque coefficient corresponding to the collection time are extracted from each signal and each torque coefficient vector respectively to form the signal waveform intensity value vector and the signal torque coefficient vector of each discrete class.

[0044] Step 4, determining the temperature-to-torque interference strength of each discrete class according to the temperature similarity vector of each discrete class, the change difference between the temperature value of each sampling time in the temperature discrete vector and the signal strength of the corresponding sampling time in the signal waveform intensity value vector of each partial signal, the discrete degree of the signal torque coefficient vector of each partial signal, and the motor torque reference comparison confusion degree.

[0045] Based on the above analysis, the measurement result of the torque sensor can be directly taken as the torque when the temperature influence is small, but the result of the torque sensor should be corrected when the temperature influence is large, so it is necessary to analyze the temperature-to-torque interference strength to represent the interference ability of the temperature to the torque.

[0046] The discrete degree of the signal torque coefficient vector of each partial signal of each discrete class is taken as the torque coefficient variation scale of each partial signal of each discrete class. Preferably, in the embodiment, the discrete degree can be taken as the torque coefficient variation scale by calculating the standard deviation of the signal torque coefficient vector of each partial signal of each discrete class.

[0047] The ratio between the temperature value of each sampling time in the temperature discrete vector of each discrete class and the mean of all temperatures in the temperature similarity vector of the corresponding discrete class is analyzed, and is recorded as the temperature interference coefficient of each sampling time of each discrete class. The ratio between the signal strength of each sampling time in the signal waveform intensity value vector of each partial signal of each discrete class and the mean of the signal waveform intensity value vector is recorded as the signal interference coefficient of each sampling time in each partial signal of each discrete class. The difference between the temperature interference coefficient and the signal interference coefficient is determined as the temperature waveform difference of each sampling time in each partial signal of each discrete class. Preferably, in the embodiment, the absolute value of the difference between the temperature interference coefficient of each sampling time of each discrete class and the signal interference coefficient of the corresponding sampling time in each partial signal is taken as the temperature waveform difference of each sampling time in each partial signal of each discrete class.

[0048] The reciprocal of the sum of the temperature waveform differences of all sampling times is taken as the temperature interference degree of each partial signal of each discrete class. The temperature-to-torque interference strength of each discrete class is obtained by fusing the temperature interference degree, the torque coefficient variation scale, and the motor torque reference comparison confusion degree of all partial signals of each discrete class. Preferably, in the embodiment, the cumulative sum of the product of the temperature interference degree, the torque coefficient variation scale, and the motor torque reference comparison confusion degree of all partial signals is taken as the temperature-to-torque interference strength of each discrete class, and the specific calculation formula of the temperature-to-torque interference strength is: wherein, For the first The temperature-induced torque disturbance intensity of each discrete class For the first The disorder of motor torque reference comparison for individual signals For the first The first discrete class The torque coefficient variation scale of the individual signals For the first The temperature discrete vector of the discrete class is the first Temperature value at the time of sampling For the first The mean of the discrete temperature vectors of each discrete class. For the first The first discrete class The signal waveform intensity value vector of the nth sub-signal Signal strength at the sampling time, For the first The first discrete class The mean of the signal waveform intensity value vector of each sub-signal. For all sampling times in the discrete temperature vector. This is the temperature interference coefficient. The signal interference coefficient. For the first The first discrete class Within the first sub-signal Differences in temperature waveforms at sampling times For the first The first discrete class The degree of temperature interference in each signal.

[0049] Furthermore, the flowchart of the method for obtaining the temperature-torque disturbance intensity for each discrete class provided in this application is as follows: Figure 2 As shown.

[0050] It should be noted that the greater the temperature interference of each sub-signal, the smaller the difference between the signal strength change and the temperature change in each sub-signal, and the more similar their parameter changes are. In this case, the temperature interference on torque is stronger. The larger the torque coefficient variation scale of each sub-signal, the more the torque can be derived from the speed and power of each sub-signal under the discrete temperature class, and the weaker the temperature interference on torque. The greater the confusion of the motor torque reference comparison of each sub-signal, the less the torque can be derived from the speed and power of each sub-signal. In this case, the sub-signal is more important in judging the degree of temperature interference on torque. Secondly, the greater the temperature interference on torque for each discrete class, the greater the influence of temperature on the torque sensor error under the corresponding discrete temperature class. In this case, the measurement results of the torque sensor should be corrected.

[0051] Step 5, according to the influence degree of the motor temperature at each sampling time on the torque of the motor torque reference vector at each sampling time, the motor torque anti-interference prediction quantity at each sampling time is determined.

[0052] In order to correct the measured torque of the torque sensor, the measured torque is corrected by using Kalman filtering method according to the motor torque formula and historical torque data as the prediction quantity. By temperature, power, speed, and the motor temperature at the last time, the prediction model is constructed by correcting the torque, and combined with the measured torque of the torque sensor, the Kalman filtering with dynamic Kalman gain is used for torque correction, so as to eliminate the error influence of temperature on the torque sensor. The specific method is: The median temperature in the corresponding interval of each discrete class is obtained, and the median temperature of each discrete class and the torque interference intensity of temperature form a two-dimensional discrete array; A rectangular coordinate system is constructed for all two-dimensional discrete arrays of discrete classes, and linear fitting is performed on all coordinate points to obtain a fitting function, which is referred to as a temperature interference fitting function, wherein the independent variable is temperature and the dependent variable is the torque interference intensity of temperature; The motor temperature at each sampling time is obtained, and the temperature interference fitting value at each sampling time is obtained by the temperature interference fitting function; Preferably, in the embodiment, the linear interpolation algorithm is used to obtain the fitting function, wherein the linear interpolation algorithm is a known technology and will not be described here.

[0053] It can be understood that the present embodiment only provides a linear fitting method, i.e. linear interpolation algorithm, as other embodiments, the implementer can use other methods in the prior art to obtain the fitting function, which is not specially limited in the present embodiment.

[0054] The normalized processing result of the ratio of the torque reference value at each sampling time to the temperature interference fitting value at the corresponding sampling time is used as the prediction quantity distribution weight at each sampling time; The ratio between the average torque of all sampling times in the motor torque reference vector and the average torque of all sampling times in the motor torque vector is referred to as torque relative difference. The torque at each sampling time in the motor torque reference vector and the torque relative difference are fused to determine the interference prediction coefficient at each sampling time. Preferably, in the embodiment, the product of the torque at each sampling time in the motor torque reference vector and the torque relative difference is used as the interference prediction coefficient at each sampling time.

[0055] The product of the interference prediction coefficient at each sampling time and the prediction quantity distribution weight is referred to as the temperature interference prediction degree at each sampling time. The difference between the value 1 and the weight of the prediction value at each sampling time is recorded as a temperature correction weight; The product of the motor temperature correction torque at the previous time of each sampling time and the temperature correction weight corresponding to the sampling time is recorded as a temperature correction prediction degree; The temperature interference prediction degree and the temperature correction prediction degree are fused to determine the motor torque anti-interference prediction value at each sampling time.

[0056] Preferably, in the embodiment, the sum of the temperature interference prediction degree and the temperature correction prediction degree is taken as the motor torque anti-interference prediction value at each sampling time.

[0057] It should be noted that the greater the torque reference value local stability or the smaller the motor temperature interference fitting value, the better the torque derivation value prediction effect, and the greater the prediction value weight. Secondly, the interference prediction coefficient is the torque size under the stable motor parameters. Since the motor torque formula has better derivation effect in the case of motor steady state, when the motor motion parameters change, the actual torque changes slowly due to the inertia of the elastic shaft, so the motor temperature correction torque at the previous time is used to calculate the temperature correction prediction degree.

[0058] Step 6, based on the motor torque anti-interference prediction value at each sampling time, and combined with the Kalman filtering method, the motor temperature correction torque is obtained to determine the final output torque.

[0059] Based on the above analysis, through the analysis of the motor signal system, the measured torque can be corrected when the temperature causes a large error influence to determine the final output torque, and the specific method is: The maximum value of the temperature interference strength on the torque of all discrete classes is recorded as the maximum temperature interference strength on the torque; The ratio of the temperature interference fitting value at each sampling time to the maximum temperature interference strength on the torque is taken as the dynamic Kalman gain at each sampling time; The motor torque anti-interference prediction value at each sampling time is taken as the prediction value at each sampling time; The torque at each sampling time in the motor torque vector is taken as the measurement value at each sampling time; The dynamic Kalman gain is taken as the gain coefficient in the Kalman filtering method, and combined with the prediction value and the measurement value at each sampling time, the motor temperature correction torque at each sampling time is obtained; The motor temperature correction torque is taken as the measured torque of the torque sensor to obtain the final output torque.

[0060] It should be noted that when using the Kalman filter method to compensate for the temperature of the motor torque, the larger the dynamic Kalman gain, the greater the influence of temperature on the torque error of the sensor, the less reliable the measured torque value, and the stronger the correction strength of the predicted value to the measured value.

[0061] Furthermore, the flowchart of Kalman filtering provided in this application is as follows: Figure 3 As shown.

[0062] It should be understood that, although Figure 1 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 1 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0063] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0064] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application. Therefore, any simple modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of this application, without departing from the content of the technical solution of this application, shall fall within the protection scope of the technical solution of this application.

Claims

1. A method of measuring an output torque of an electric motor, characterized by, The method comprises the following steps: obtaining a motor torque vector, a motor temperature vector, and obtaining a motor torque reference vector; for the motor torque vector, a signal decomposition algorithm is used to obtain the signal strength of each sub-signal at each sampling time; according to the torque of each sampling time of the motor torque reference vector and the fluctuation difference of the signal strength of each sub-signal at each sampling time, the motor torque reference contrast confusion degree of each sub-signal is determined; based on the motor temperature vector, a temperature similar vector of each discrete class, a temperature discrete vector, a signal waveform strength value vector of each sub-signal, and a signal torque coefficient vector of each sub-signal are constructed; according to the temperature value of each collection time in the temperature similar vector and the temperature discrete vector of each discrete class, the signal strength of the corresponding collection time in the signal waveform strength value vector of each sub-signal, and the discrete degree of the signal torque coefficient vector, and the motor torque reference contrast confusion degree, the temperature-to-torque interference strength of each discrete class is determined; according to the influence degree of the motor temperature at each sampling time on the torque of each sampling time of the motor torque reference vector, the motor torque anti-interference prediction quantity at each sampling time is determined; based on the motor torque anti-interference prediction quantity at each sampling time, and combined with the Kalman filtering method, the motor temperature correction torque is obtained, and the final output torque is determined.

2. A method of measuring the output torque of an electric motor as claimed in claim 1, characterized in that, The determination of the motor torque reference contrast confusion degree of each sub-signal comprises: obtaining a motor speed vector and a motor power vector, and the motor torque reference vector is a vector composed of the ratio of the motor power and the motor speed at the corresponding sampling time in the motor power vector and the motor speed vector; a vector composed of the calculation results of the first derivative of the torque at each sampling time in the motor torque reference vector is denoted as a motor torque change reference vector; the local change of the torque at each sampling time in the motor torque reference vector is analyzed, and the local stability degree of the torque reference value at each sampling time in the motor torque change reference vector is determined; the difference degree of the fluctuation change between the torque of all collection times in the motor torque reference vector and the signal strength of all collection times of each sub-signal is analyzed, and the waveform confusion coefficient of each sub-signal is determined; a vector composed of the ratio between the torque at each sampling time in the motor torque reference vector and the signal strength of each sub-signal at the same sampling time is taken as the torque coefficient vector of each sub-signal; the difference change amount of the torque coefficient at each sampling time in the torque coefficient vector of each sub-signal and the torque reference value local stability degree of the corresponding collection time are analyzed, and the signal confusion coefficient of each sub-signal is determined; the waveform confusion coefficient and the signal confusion coefficient are fused to determine the motor torque reference contrast confusion degree of each sub-signal.

3. A method of measuring the output torque of an electric motor as claimed in claim 2, wherein, The determination of the local stability degree of the torque reference value at each sampling time in the motor torque change reference vector comprises: taking the torque at any sampling time in the motor torque change reference vector as a target reference value, the sum of the absolute values of the torque at multiple sampling times before the target reference value and the absolute value of the target reference value is taken as the local reference coefficient of the target reference value corresponding sampling time in the motor torque change reference vector. The reciprocal of the local reference coefficient is taken as the torque reference value local stability of each sampling time in the motor torque change reference vector.

4. A method of measuring the output torque of an electric motor as claimed in claim 2, wherein, The method for determining the waveform confusion coefficient of each partial signal comprises: The torque of each sampling time in the motor torque reference vector and the signal intensity of each partial signal at each sampling time are respectively subjected to dimensionless processing, so as to obtain a motor torque waveform reference vector and each motor torque waveform decomposition signal. The difference between the torque of each sampling time in the motor torque waveform reference vector and the signal intensity of each motor torque waveform decomposition signal at the same sampling time is analyzed as the torque waveform difference of each partial signal at each sampling time. The product of the torque waveform difference of each partial signal at each sampling time and the torque reference value local stability of the motor torque change reference vector at the same sampling time is taken as the waveform comparison coefficient of each partial signal at each sampling time. The waveform comparison coefficients of all sampling times are fused to determine the waveform confusion coefficient of each partial signal.

5. A method of measuring the output torque of an electric motor as defined in claim 2, wherein, The method for determining the signal confusion coefficient of each partial signal comprises: The absolute value of the difference between the torque coefficient of each sampling time in the torque coefficient vector of each partial signal and the average value of the torque coefficient vector is analyzed as the signal difference of each partial signal at each sampling time. The product of the signal difference of each partial signal at each sampling time and the torque reference value local stability of the corresponding sampling time is taken as the signal stability coefficient of each partial signal at each sampling time. The sum of the signal stability coefficients of all sampling times is taken as the signal confusion coefficient of each partial signal.

6. A method of measuring the output torque of an electric motor as defined in claim 2, wherein, The construction method of the temperature similar vector of each discrete class, the temperature discrete vector, the signal waveform intensity value vector of each partial signal and the signal torque coefficient vector of each partial signal comprises: All temperature values in the motor temperature vector are divided into multiple intervals to obtain multiple discrete classes. The vector composed of all temperature values belonging to the same interval is taken as the temperature similar vector of each discrete class. All discrete classes are numbered and sorted to obtain the serial number of each discrete class. Each temperature value in the temperature similar vector of each discrete class is replaced by the serial number value of the discrete class to obtain the temperature discrete vector of each discrete class. Each temperature value in each discrete class corresponds to a sampling time in the motor temperature vector. According to the sampling time of each temperature value in the temperature similar vector of each discrete class, the signal intensity and the torque coefficient at the corresponding sampling time are extracted from each partial signal and each torque coefficient vector respectively to form the signal waveform intensity value vector and the signal torque coefficient vector of each partial signal of each discrete class.

7. A method of measuring the torque output of an electric motor as claimed in claim 6, wherein, The method for determining the temperature-to-torque interference intensity of each discrete class comprises: The discrete degree of the signal torque coefficient vector of each partial signal of each discrete class is taken as the torque coefficient variation scale of each partial signal of each discrete class. The ratio between the temperature value of each sampling time in the temperature discrete vector of each discrete class and the average value of all temperatures in the temperature similar vector of the corresponding discrete class is analyzed as the temperature interference coefficient of each discrete class at each sampling time. The ratio between the signal intensity of each acquisition time in the signal waveform intensity value vector of each partial signal of each discrete class and the mean value of the signal waveform intensity value vector is recorded as the signal interference coefficient of each acquisition time in each partial signal of each discrete class; The difference between the temperature interference coefficient and the signal interference coefficient is determined as the temperature waveform difference of each acquisition time in each partial signal of each discrete class; The reciprocal of the sum of the temperature waveform differences of all sampling times is taken as the temperature interference degree of each partial signal of each discrete class; The temperature interference degree, the torque coefficient variation scale and the motor torque reference contrast confusion degree of all partial signals of each discrete class are fused as the temperature-to-torque interference strength of each discrete class.

8. A method of measuring the output torque of an electric motor as defined in claim 2, wherein, The determination of the motor torque anti-interference prediction quantity of each sampling time comprises: Based on the temperature-to-torque interference strength of each discrete class, the prediction quantity distribution weight of each sampling time is calculated; The ratio between the mean value of the torque of all sampling times in the motor torque reference vector and the mean value of the torque of all sampling times in the motor torque vector is recorded as the torque relative difference; The torque of each sampling time in the motor torque reference vector and the torque relative difference are fused to determine the interference prediction coefficient of each sampling time; The product of the interference prediction coefficient of each sampling time and the prediction quantity distribution weight is recorded as the temperature interference prediction degree of each sampling time; The difference between the numerical value 1 and the prediction quantity distribution weight of each sampling time is recorded as the temperature correction weight; The product of the motor temperature correction torque of the previous time of each sampling time and the temperature correction weight of the corresponding sampling time is recorded as the temperature correction prediction degree; The temperature interference prediction degree and the temperature correction prediction degree are fused to determine the motor torque anti-interference prediction quantity of each sampling time.

9. A method of measuring the torque output of an electric motor as claimed in claim 8, wherein, The calculation method of the prediction quantity distribution weight of each sampling time is: The median temperature in the corresponding interval of each discrete class is obtained, and the median temperature and the temperature-to-torque interference strength of each discrete class form a two-dimensional discrete array; A rectangular coordinate system is constructed for the two-dimensional discrete arrays of all discrete classes, and linear fitting is performed on all coordinate points to obtain a fitting function, which is recorded as the temperature interference fitting function, and the temperature interference fitting value of each sampling time is obtained; The normalized processing result of the ratio between the torque reference value local stability of each sampling time and the temperature interference fitting value of the corresponding sampling time is taken as the prediction quantity distribution weight of each sampling time.

10. A method of measuring the torque output of an electric motor as claimed in claim 9, wherein, The motor temperature correction torque is obtained, and the final output torque is determined, comprising: The maximum value of the temperature-to-torque interference strength of all discrete classes is recorded as the maximum temperature-to-torque interference strength; The ratio between the temperature interference fitting value of each sampling time and the maximum temperature-to-torque interference strength is taken as the dynamic Kalman gain of each sampling time; The motor torque anti-interference prediction quantity of each sampling time is taken as the prediction value of each sampling time; The torque of each sampling time in the motor torque vector is taken as the measurement value of each sampling time; The dynamic Kalman gain is taken as the gain coefficient in the Kalman filtering method, and the prediction value and the measurement value of each sampling time are combined to obtain the motor temperature correction torque of each sampling time. The motor temperature correction torque is used as the measured torque of the torque sensor to obtain a final output torque.

Images