WO2023029903A1 - 负载惯量辨识方法、装置、电子设备及系统 - Google Patents

负载惯量辨识方法、装置、电子设备及系统 Download PDF

Info

Publication number
WO2023029903A1
WO2023029903A1 PCT/CN2022/111093 CN2022111093W WO2023029903A1 WO 2023029903 A1 WO2023029903 A1 WO 2023029903A1 CN 2022111093 W CN2022111093 W CN 2022111093W WO 2023029903 A1 WO2023029903 A1 WO 2023029903A1
Authority
WO
WIPO (PCT)
Prior art keywords
mechanical transmission
transmission system
state
current
load
Prior art date
Application number
PCT/CN2022/111093
Other languages
English (en)
French (fr)
Inventor
李明洋
许雄
朱春晓
Original Assignee
节卡机器人股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 节卡机器人股份有限公司 filed Critical 节卡机器人股份有限公司
Publication of WO2023029903A1 publication Critical patent/WO2023029903A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02P—CONTROL OR REGULATION OF ELECTRIC MOTORS, ELECTRIC GENERATORS OR DYNAMO-ELECTRIC CONVERTERS; CONTROLLING TRANSFORMERS, REACTORS OR CHOKE COILS
    • H02P21/00—Arrangements or methods for the control of electric machines by vector control, e.g. by control of field orientation
    • H02P21/14—Estimation or adaptation of machine parameters, e.g. flux, current or voltage
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02P—CONTROL OR REGULATION OF ELECTRIC MOTORS, ELECTRIC GENERATORS OR DYNAMO-ELECTRIC CONVERTERS; CONTROLLING TRANSFORMERS, REACTORS OR CHOKE COILS
    • H02P21/00—Arrangements or methods for the control of electric machines by vector control, e.g. by control of field orientation
    • H02P21/13—Observer control, e.g. using Luenberger observers or Kalman filters
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02P—CONTROL OR REGULATION OF ELECTRIC MOTORS, ELECTRIC GENERATORS OR DYNAMO-ELECTRIC CONVERTERS; CONTROLLING TRANSFORMERS, REACTORS OR CHOKE COILS
    • H02P25/00—Arrangements or methods for the control of AC motors characterised by the kind of AC motor or by structural details
    • H02P25/02—Arrangements or methods for the control of AC motors characterised by the kind of AC motor or by structural details characterised by the kind of motor
    • H02P25/022—Synchronous motors
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02P—CONTROL OR REGULATION OF ELECTRIC MOTORS, ELECTRIC GENERATORS OR DYNAMO-ELECTRIC CONVERTERS; CONTROLLING TRANSFORMERS, REACTORS OR CHOKE COILS
    • H02P2207/00—Indexing scheme relating to controlling arrangements characterised by the type of motor
    • H02P2207/05—Synchronous machines, e.g. with permanent magnets or DC excitation

Definitions

  • the present application relates to the technical field of automatic control, for example, to a load inertia identification method, device, electronic equipment and system.
  • Servo system also known as servo system, is a feedback control system used to accurately follow or reproduce a process.
  • Servo motor refers to the engine that controls the operation of mechanical components in the servo system, and is an auxiliary motor indirect transmission device.
  • the servo system mainly relies on pulses for positioning. When the servo motor receives one pulse, it will rotate the angle corresponding to one pulse to achieve displacement.
  • the servo motor itself has the function of sending pulses, so every time the servo motor rotates an angle, it will send out a corresponding number of pulses, which echoes the pulses received by the servo motor, or is called a closed loop, so that the servo system will know how much it sent
  • the pulse is sent to the servo motor, and how many pulses are received back at the same time, the rotation of the motor can be controlled, so as to achieve positioning.
  • the system load torque and moment of inertia also change.
  • the present application provides a load inertia identification method, device, electronic equipment and system for accurately identifying the rotational inertia of the load end of the system.
  • This application provides a load inertia identification method, including:
  • the load inertia information of the mechanical transmission system in the current state is determined.
  • the application provides an inertia identification device, including:
  • the first acquisition module is configured to acquire the input control command, motor speed and load speed of the mechanical transmission system to be tested in the previous state;
  • a prediction module configured to predict and obtain the current system predicted state of the mechanical transmission system according to the input control command, the motor speed, the load speed and the state prediction equation of the mechanical transmission system;
  • the second acquisition module is configured to acquire the current motor speed measurement value and the current load speed measurement value of the mechanical transmission system
  • a calculation module configured to determine the current optimal estimated state of the mechanical transmission system according to the current predicted state of the system, the current measured value of the motor speed and the measured value of the current load speed;
  • the determination module is configured to determine the load inertia information of the mechanical transmission system in the current state according to the current optimal estimated state.
  • the present application provides an electronic device, including: a memory for storing a computer program; a processor for executing the computer program, so as to realize the above load inertia identification method.
  • the application provides a load inertia identification system, including:
  • a servo driver connected to the mechanical transmission system, configured to drive the mechanical transmission system to run;
  • An inertia recognizer connected to the mechanical transmission system and the servo driver, is configured to implement the above load inertia identification method to identify the load inertia information of the mechanical transmission system.
  • FIG. 1 is a schematic structural diagram of an electronic device provided by an embodiment of the present application.
  • FIG. 2A is a schematic diagram of a load inertia identification system model provided by an embodiment of the present application
  • Fig. 2B is a schematic diagram of a dual-inertia mechanical transmission device provided by an embodiment of the present application.
  • FIG. 3 is a schematic flowchart of a load inertia identification method provided by an embodiment of the present application
  • FIG. 4 is a schematic flowchart of another load inertia identification method provided by an embodiment of the present application.
  • FIG. 5 is a schematic waveform diagram of servo motion data of a mechanical arm provided by an embodiment of the present application.
  • FIG. 6 is a schematic structural diagram of a load inertia identification device provided by an embodiment of the present application.
  • Icons 1-electronic equipment; 10-bus; 11-processor; 12-memory; 2-load inertia identification system; 3-double inertia mechanical transmission device; 110-servo driver; 120-mechanical transmission system; 130-load mechanism ;140-inertia identifier; 150-motor; 160-transmission mechanism; 600-load inertia identification device; 601-first acquisition module; 602-prediction module; 603-second acquisition module; 604-calculation module; 605-confirmation module.
  • FIG. 1 is a schematic structural diagram of an electronic device provided by an embodiment of the present application.
  • the electronic device 1 includes: at least one processor 11 and a memory 12 .
  • One processor is taken as an example in FIG. 1 .
  • the processor 11 and the memory 12 are connected through the bus 10, and the memory 12 stores instructions that can be executed by the processor 11, and the instructions are executed by the processor 11, so that the electronic device 1 can execute all or part of the processes of the methods in the following embodiments .
  • the electronic device 1 may be a mobile phone, a tablet computer, a notebook computer, a desktop computer and other devices.
  • FIG. 2A is a model schematic diagram of a load inertia identification system provided by an embodiment of the present application.
  • the load inertia identification system 2 includes a servo drive 110 , a mechanical transmission system 120 , a load mechanism 130 , and an inertia identification device 140 .
  • a permanent magnet synchronous motor is used as the above-mentioned servo driver 110 .
  • the servo driver 110 is connected to the load mechanism 130 through a mechanical transmission system 120 .
  • An Extended Kalman Filter Extended Kalman Filter (Extended Kalman Filter, EKF) (also called an Extended Kalman Observer) is used as the inertia identifier 140 mentioned above.
  • the current sampling circuit detects the three-phase currents ia , ib and ic of the motor in the three-phase stationary coordinate system through the Hall sensor, and converts them into stationary two-phase currents through Clark transformation (3s/2s).
  • the current sampling circuit can also detect and obtain the three-phase currents ia , ib and ic of the motor in the three-phase stationary coordinate system through other sampling methods.
  • the current values i ⁇ and i ⁇ in the stationary two-phase coordinate system are transformed into two-phase feedback calculation excitation current i d_fdb and feedback calculation rotation in the rotor rotating coordinate system through Park transformation (2s/2r).
  • Moment current i q_fdb The given excitation current id_ref is compared with the feedback calculated excitation current id_fdb , and after adjustment by the current regulator, the d-axis given output voltage ud of the two-phase rotating coordinates is obtained.
  • the given output voltage u q of the q-axis of the two-phase rotating coordinates is obtained.
  • the two-phase given output voltages u d and u q in the rotating coordinate system are converted into two-phase voltages u ⁇ and u ⁇ in the stationary two-phase coordinate system after Park inverse transformation (2r/2s) inverse transformation, and then pulse width modulated (Pulse Width Modulation, PWM) generation module is adjusted to generate a PWM wave, which drives the servo driver 110 to work after passing through the three-phase power inverter circuit.
  • PWM pulse width modulated
  • the known parameters include the torsional stiffness K of the mechanical transmission system 120, the moment of inertia J 1 of the rotor of the servo drive 110, the feedback calculation torque current i q_fdb and the feedback speed of the servo drive 110 side Load mechanism 130 side feedback speed
  • the output is the moment of inertia J 2 of the load side of the inertia identifier system 2 .
  • FIG. 2B is a schematic diagram of a dual-inertia mechanical transmission device provided by an embodiment of the present application.
  • the schematic diagram of the dual-inertia mechanical transmission device is a schematic diagram of a dual-inertia mechanical transmission device 3 abstracted from the load inertia identification system 2 , wherein the dual inertia mechanical transmission device 3 includes a motor 150 , a transmission mechanism 160 and a load mechanism 130 .
  • J 1 is the moment of inertia of the rotor on the side of the motor 150, which is a basic parameter of the motor 150, which is obtained by the test of the manufacturer of the motor 150, and marked in the manual of the motor 150;
  • C 1 is the damping coefficient of the motor 150, determined by the motor 150.
  • 150 is obtained from the test of the manufacturer of the motor 150 and marked in the manual of the motor 150;
  • T e is the electromagnetic torque of the motor 150, and the q-axis current I q is obtained by calculating the three-phase current of the motor 150 through sensors.
  • the q-axis current I q is proportional to its electromagnetic torque T e , and its proportional factor is K e ;
  • ⁇ 1 is the electromagnetic rotation angle output by the position sensor on the rotor side of the motor 150;
  • Tw will be generated during torsional deformation;
  • Cw is the damping coefficient of the transmission mechanism 160, because the transmission mechanism 160 will be lubricated, so Cw is very small, provided by the manufacturer of the transmission mechanism 160, sometimes it can be set to 0;
  • K is the torsional stiffness coefficient, which is a constant value or parameter table determined by the manufacturer of the transmission mechanism 160, and is given in the user manual of the transmission mechanism 160.
  • the torsional stiffness coefficient K represents the motor 150 in the double inertia mechanical transmission device 3
  • the flexibility connected with the load mechanism 130 ⁇ 2 is the load rotation angle output by the position sensor on the side of the load mechanism 130; J2 is the equivalent moment of inertia of the load mechanism 130, which needs to be identified; C2 is the damping coefficient of the load mechanism 130, because The load mechanism 130 will be lubricated, so C 2 is very small, provided by the manufacturer of the load mechanism 130, sometimes it can be set to 0; T 1 is the torque on the load mechanism 130 side.
  • the motor 150 , the transmission mechanism 160 and the load mechanism 130 constitute a typical dual-inertia mechanical transmission device 3 .
  • the transmission mechanism 160 connects the motor 150 and the load mechanism 130, and has a certain torsional stiffness K and a damping coefficient Cw .
  • a torque T w will be generated.
  • the electromagnetic torque T e and the transmission mechanism 160 side torque T w jointly act on the rotating shaft of the motor 150 with the moment of inertia J 1 and the damping coefficient C 1 .
  • the equivalent moment of inertia on the side of the load mechanism 130 is J 2
  • the damping coefficient is C 2
  • the torque T w of the transmission mechanism 160 and the load torque T l act together on the actuator to finally determine the speed ⁇ l of the load side.
  • J 1 is the moment of inertia of the rotor on the side of the motor 150
  • ⁇ 1 is the electromagnetic rotation angle
  • T e is the electromagnetic torque of the motor 150
  • C 1 is the damping coefficient of the motor 150
  • T w is the torque of the transmission mechanism 160
  • J2 is the equivalent moment of inertia of the load mechanism 130
  • ⁇ 2 is the rotation angle of the load mechanism 130 side
  • C2 is the damping coefficient of the load mechanism 130
  • T1 is the torque on the load mechanism 130 side
  • Cw is the damping of the transmission mechanism 160 coefficient
  • K is the torsional stiffness coefficient of the transmission mechanism 160
  • ⁇ m is the rotational speed of the motor 150 side
  • ⁇ l is the rotational speed of the load mechanism 130 side.
  • the damping coefficient in the mechanical transmission device 3 is very small and can be set to 0; in order to carry out the inertia identification calculation, the Laplace change is performed on the formula (1), and the formula ( 2):
  • J 1 is the moment of inertia of the rotor on the side of the motor 150
  • ⁇ 1 is the electromagnetic rotation angle
  • s is the Laplace operator
  • T e is the electromagnetic torque of the motor 150
  • T w is the torque of the transmission mechanism 160
  • J 2 is the equivalent moment of inertia of the load mechanism 130
  • ⁇ 2 is the load rotation angle
  • T 1 is the torque on the load mechanism 130 side
  • K is the torsional stiffness coefficient
  • ⁇ m is the rotational speed of the motor 150 side
  • ⁇ 1 is the load mechanism 130 side speed.
  • FIG. 3 is a schematic flowchart of a load inertia identification method provided by an embodiment of the present application.
  • the method can be executed by the electronic device 1 shown in FIG. 1 , and can be applied to load inertia identification scenarios as shown in FIGS. 2A to 2B , so as to realize accurate identification of load inertia.
  • the method includes:
  • the input control command of the mechanical transmission system to be tested in the previous state can be a number of parameters that characterize the system in the previous state, such as the electromagnetic torque T e calculated from the three-phase current and the motor model, the motor speed in the previous state It can be the feedback speed ⁇ m of the motor 150 side obtained after conversion of the electromagnetic angle ⁇ 1 measured by the position sensor on the motor 150 side, or it can be directly or indirectly measured by other sensors; the load speed in the previous state can be obtained by The load rotation angle ⁇ 2 measured by the position sensor on the load mechanism 130 side is converted to obtain the feedback speed ⁇ l on the load mechanism 130 side, which can also be directly or indirectly measured by other sensors.
  • the input control command of the mechanical transmission system to be tested in the previous state can also be the q-axis torque current obtained by calculating the three-phase current; the three-phase current of the motor measured by the sensor and the basic specifications of the motor itself
  • the electromagnetic torque T e obtained by calculating parameters in other ways; the input control command of the mechanical transmission system to be tested in the previous state in this application can be selected by the inertia identification system according to actual needs, and is not limited thereto.
  • the current system prediction state of the mechanical transmission system can be predicted according to the obtained input control command, motor speed, load speed and state prediction equation of the mechanical transmission system.
  • the current system prediction state is calculated using the following formula:
  • k represents the cycle of the previous state of the mechanical transmission system
  • k+1 represents the cycle of the current state of the mechanical transmission system
  • T s is the sampling period of the mechanical transmission system
  • u(k) is the input control command of the mechanical transmission system in the kth period
  • u(k) is the input control command of the mechanical transmission system in the kth period
  • Q(k) is the covariance of the system error of the mechanical transmission system in the kth period
  • J 1 is the rotation of the motor rotor of the mechanical transmission system Inertia
  • M l is the reciprocal of the load side moment of inertia J
  • k represents the period of the previous state of the mechanical transmission system
  • T s is the sampling period of the mechanical transmission system
  • J 1 is the moment of inertia of the motor rotor of the mechanical transmission system
  • M l is the moment of inertia of the load side of the mechanical transmission system Reciprocal of J 2 .
  • the current measured value of the motor speed can be measured indirectly by the position sensor arranged on the motor 150 side, or can be measured by other sensors arranged on the motor 150 side; the current load speed measured value can be indirectly measured by the position sensor arranged on the load mechanism 130 side It is known that it can also be measured by other sensors provided on the load mechanism 130 side.
  • the Kalman gain of a mechanical transmission system can be calculated using the following formula:
  • the error mean square of the current optimal estimated state is calculated by the following formula:
  • k represents the cycle of the previous state of the mechanical transmission system
  • k+1 represents the cycle of the current state of the mechanical transmission system
  • K(k+1) represents the cycle of the mechanical transmission system in the (k+1)th cycle.
  • described Kalman gain is the predicted noise covariance of the mechanical transmission system at the (k+1)th period
  • R(k) is the covariance of the measurement error of the mechanical transmission system at the kth period
  • y(k+1) is the mechanical transmission system at the (k+1)th cycle determined based on the current motor speed measurement value and the current load speed measurement value output state measurements for the period
  • the load inertia information of the mechanical transmission system in the current state is determined.
  • the torsional stiffness coefficient K, the motor side speed ⁇ m , the load side speed ⁇ l , and the transmission shaft rotation speed are introduced.
  • T w and load torque T l so as to establish the state model and observation matrix of the dual inertia system; then determine the system state prediction matrix and covariance matrix; and then use the extended Kalman filter algorithm to iteratively identify the load inertia, improving the load side Accuracy of inertia identification.
  • the load inertia information of the mechanical transmission system in the current state can be determined based on the current optimal estimated state of the mechanical transmission system, which is different from the traditional motor Compared with the load as an inertia parameter, the moment of inertia at the load end of the system can be more accurately identified.
  • FIG. 4 is a schematic flowchart of another load inertia identification method provided by an embodiment of the present application.
  • the method can be executed by the electronic device 1 shown in FIG. 1 , and can be applied to load inertia identification scenarios as shown in FIGS. 2A to 2B , so as to realize accurate identification of load inertia.
  • the method includes:
  • the parameters in the dual-inertia mechanical transmission device 3 include: the moment of inertia J 1 of the rotor on the side of the motor 150 , the damping coefficient C 1 of the motor 150 , and the electromagnetic torque of the motor 150 T e , electromagnetic rotation angle ⁇ 1 , torque T w of transmission mechanism 160 , damping coefficient C w of transmission mechanism 160 , torsional stiffness coefficient K of transmission mechanism 160 , rotation angle ⁇ 2 of load mechanism 130 side, load mechanism 130 , etc.
  • matrix C in formula (12) is as shown in formula (10),
  • the extended Kalman filter can be used to directly solve the characteristics of nonlinear equations in an iterative manner, that is, the system state variables of the previous state can be predicted according to the system state equation, and the system state of the current state can be calculated. Based on formulas (11)-(17), the state prediction equation of the mechanical transmission system is obtained:
  • k represents the cycle of the previous state of the mechanical transmission system
  • k+1 represents the cycle of the current state of the mechanical transmission system
  • x(k) is the predicted value of the kth cycle of the mechanical transmission system
  • x(k+ 1) is the predicted value of the predicted state of the mechanical transmission system in the (k+1)th period
  • T s is the sampling period of the mechanical transmission system
  • u(k) is the input control of the mechanical transmission system in the kth period
  • y(k) is the predicted value of the output variable of the kth period of the mechanical transmission system.
  • w represents the impact of the system parameter error
  • v represents the noise and interference in the measurement process, including the quantization error of the position signal measured by the code wheel.
  • the noise can be stationary white Gaussian noise with an average value of zero. Therefore, the covariance matrix of noise can be defined as:
  • the identified load inertia information can be output, and the output load inertia information can be used for adjustment of the mechanical transmission system.
  • FIG. 5 is a schematic diagram of a servo motion data waveform of a robotic arm provided by an embodiment of the present application.
  • the load inertia identification method in the above embodiments is used to observe the servo motion of the manipulator.
  • the servo driver 110 is the base joint servo driver of the six-axis manipulator, and the EKF is used as the inertia identifier 140 to observe the load inertia at the rear end of the joint in real time during the operation of the manipulator.
  • the ordinate is the inertia of the load side, the unit is (Kg ⁇ m 2 ), and the abscissa is the sampling time, the unit is ms.
  • the solid line is the change curve of the inertia of the rear end of the joint of the manipulator base calculated according to the theoretical dynamic model of the manipulator
  • the dotted line around the solid line is the load inertia of the rear end of the joint observed in real time by the Kalman observer of this embodiment . It can be seen from the data graph that the Kalman online inertia observer of the dual inertia system can well follow the change of load inertia and has good real-time performance.
  • FIG. 6 is a load inertia identification device provided by an embodiment of the present application.
  • the method can be implemented by the electronic device 1 shown in FIG. 1 , and can be applied to load inertia identification scenarios as shown in FIGS. 2A to 2B , so as to identify the load inertia more accurately.
  • the device 600 includes: a first acquisition module 601, a prediction module 602, a second acquisition module 603, a calculation module 604, and a determination module 605.
  • the principle relationship of the multiple modules is as follows:
  • the first acquisition module 601 is configured to acquire the input control command, the motor speed and the load speed of the mechanical transmission system under test in the last state.
  • the prediction module 602 is configured to predict and obtain the current system predicted state of the mechanical transmission system according to the input control command, the motor speed, the load speed and the state prediction equation of the mechanical transmission system.
  • the second acquiring module 603 is configured to acquire the current motor speed measurement value and the current load speed measurement value of the mechanical transmission system.
  • the calculation module 604 is configured to determine the current optimal estimated state of the mechanical transmission system according to the current predicted state of the system, the current measured value of the motor speed and the current measured value of the load speed.
  • the determination module 605 is configured to determine the load inertia information of the mechanical transmission system in the current state according to the current optimal estimated state.
  • the load inertia identification device 600 also includes a building module, which is configured to: determine the state equation and output equation of the mechanical transmission system according to the dual inertia parameters and the mechanical motion equation of the mechanical transmission system; A state equation and an output equation determine said state prediction equation for the mechanical transmission system.
  • the establishment module uses the following formula to determine the state equation and output equation of the mechanical transmission system:
  • k represents the period of the last state of the mechanical transmission system
  • x is the state variable of the mechanical transmission system
  • y is the output of the mechanical transmission system
  • ⁇ m is the motor speed of the mechanical transmission system
  • ⁇ l is the Load speed
  • T w is the shafting torque of the mechanical transmission system
  • T l is the load torque of the mechanical transmission system
  • M l is the reciprocal of the moment of inertia J2 at the load side of the mechanical transmission system
  • K e is the proportional factor
  • I q is The torque current of the mechanical transmission system
  • f(x) represents the prediction function of the kth cycle of the mechanical transmission system
  • k is an integer.
  • the prediction module 602 calculates the current system prediction state by using the following formula:
  • k represents the cycle of the previous state of the mechanical transmission system
  • k+1 represents the cycle of the current state of the mechanical transmission system
  • T s is the sampling period of the mechanical transmission system
  • u(k) is the input control command of the mechanical transmission system in the kth period
  • u(k) is the input control command of the mechanical transmission system in the kth period
  • Q(k) is the covariance of the system error of the mechanical transmission system at the kth period
  • J 1 is the rotation of the motor rotor of the mechanical transmission system Inertia
  • M l is the reciprocal of the load side moment of inertia J
  • the calculation module 604 calculates the Kalman gain of the mechanical transmission system using the following formula:
  • the current best estimate is calculated using the following formula:
  • the error mean square of the current optimal estimate is calculated by the following formula:
  • k represents the cycle of the previous state of the mechanical transmission system
  • k+1 represents the cycle of the current state of the mechanical transmission system
  • K(k+1) represents the Karl of the mechanical transmission system at the (k+1)th cycle Mann gain
  • R(k) is the covariance of the measurement error of the mechanical transmission system at the kth period
  • y(k+1) is the mechanical transmission system at the (k+1)th cycle determined based on the current motor speed measurement value and the current load speed measurement value output state measurements for the period
  • the load inertia identification device 600 further includes: an output module configured to output load inertia information of the mechanical transmission system.

Landscapes

  • Engineering & Computer Science (AREA)
  • Power Engineering (AREA)
  • Control Of Electric Motors In General (AREA)

Abstract

本文公开一种负载惯量辨识方法、装置、电子设备及系统。该负载惯量辨识方法包括获取待测的机械传动系统在上一状态的输入控制指令、电机转速和负载转速；根据输入控制指令、电机转速、负载转速以及机械传动系统的状态预测方程，预测得到机械传动系统的当前系统预测状态；获取机械传动系统的当前电机转速测量值和当前负载转速测量值；根据当前系统预测状态、当前电机转速测量值和当前负载转速测量值，确定机械传动系统的当前最优估计状态；根据当前最优估计状态，确定机械传动系统在当前状态下的负载惯量信息。

Description

负载惯量辨识方法、装置、电子设备及系统
本申请要求在2021年09月01日提交中国专利局、申请号为202111018435.5的中国专利申请的优先权，该申请的全部内容通过引用结合在本申请中。
技术领域
本申请涉及自动控制技术领域，例如涉及一种负载惯量辨识方法、装置、电子设备及系统。
背景技术
伺服系统(servomechanism)又称随动系统，是用来精确地跟随或复现一个过程的反馈控制系统。伺服电机(servo motor)是指在伺服系统中控制机械元件运转的发动机，是一种补助马达间接变速装置。伺服系统主要靠脉冲来定位，伺服电机接收到1个脉冲，就会旋转1个脉冲对应的角度，从而实现位移。伺服电机本身具备发出脉冲的功能，所以伺服电机每旋转一个角度，都会发出对应数量的脉冲，和伺服电机接受的脉冲形成了呼应，或者叫闭环，如此一来，伺服系统就会知道发了多少脉冲给伺服电机，同时又收了多少脉冲回来，就能够控制电机的转动，从而实现定位。电机运行中，随着电机运行工况的变化，系统负载转矩和转动惯量也随之改变。为了提高伺服系统的动态抗扰性能，需要相应地调节控制参数，使伺服系统运行特性为最佳状态，因此对转动惯量的辨识显得尤为重要。
发明内容
本申请提供一种负载惯量辨识方法、装置、电子设备及系统，用以准确辨识系统的负载端的转动惯量。
本申请提供了一种负载惯量辨识方法，包括：
获取待测的机械传动系统在上一状态的输入控制指令、电机转速和负载转速；
根据所述输入控制指令、所述电机转速、所述负载转速以及所述机械传动系统的状态预测方程，预测得到所述机械传动系统的当前系统预测状态；
获取所述机械传动系统的当前电机转速测量值和当前负载转速测量值；
根据所述当前系统预测状态、所述当前电机转速测量值和所述当前负载转速测量值，确定所述机械传动系统的当前最优估计状态；
根据所述当前最优估计状态，确定所述机械传动系统在当前状态下的负载惯量信息。
本申请提供了一种惯量辨识装置，包括：
第一获取模块，设置为获取待测的机械传动系统在上一状态的输入控制指令、电机转速和负载转速；
预测模块，设置为根据所述输入控制指令、所述电机转速、所述负载转速以及所述机械传动系统的状态预测方程，预测得到所述机械传动系统的当前系统预测状态；
第二获取模块，设置为获取所述机械传动系统的当前电机转速测量值和当前负载转速测量值；
计算模块，设置为根据所述当前系统预测状态、所述当前电机转速测量值和所述当前负载转速测量值，确定所述机械传动系统的当前最优估计状态；
确定模块，设置为根据所述当前最优估计状态，确定所述机械传动系统在当前状态下的负载惯量信息。
本申请提供了一种电子设备，包括：存储器，用以存储计算机程序；处理器，用以执行所述计算机程序，以实现上述的负载惯量辨识方法。
本申请提供了一种负载惯量辨识系统，包括：
机械传动系统；
伺服驱动器，连接所述机械传动系统，设置为驱动所述机械传动系统运行；
惯量识别器，连接所述机械传动系统和所述伺服驱动器，设置为实现上述的负载惯量辨识方法以识别所述机械传动系统的负载惯量信息。
附图说明
图1为本申请一实施例提供的一种电子设备的结构示意图；
图2A为本申请一实施例提供的一种负载惯量辨识系统模型示意图；
图2B为本申请一实施例提供的一种双惯量机械传动装置示意图；
图3为本申请一实施例提供的一种负载惯量辨识方法的流程示意图；
图4为本申请一实施例提供的另一种负载惯量辨识方法的流程示意图；
图5为本申请一实施例提供的一种机械臂伺服运动数据的波形示意图；
图6为本申请一实施例提供的一种负载惯量辨识装置的结构示意图。
图标：1-电子设备；10-总线；11-处理器；12-存储器；2-负载惯量辨识系统；3-双惯量机械传动装置；110-伺服驱动器；120-机械传动系统；130-负载机构；140-惯量识别器；150-电机；160-传动机构；600-负载惯量辨识装置；601-第一获取模块；602-预测模块；603-第二获取模块；604-计算模块；605-确定模块。
具体实施方式
下面将结合本申请实施例中的附图，对本申请实施例中的技术方案进行描述。在本申请的描述中，术语“第一”、“第二”等仅用于区分描述，而不能理解为指示或暗示相对重要性。
请参阅如图1，图1其为本申请一实施例提供的一种电子设备的结构示意图，电子设备1包括：至少一个处理器11和存储器12，图1中以一个处理器为例。处理器11和存储器12通过总线10连接，存储器12存储有可被处理器11执行的指令，指令被处理器11执行，以使电子设备1可执行下述的实施例中方法的全部或部分流程。
于一实施例中，电子设备1可以是手机、平板电脑、笔记本电脑、台式计算机等设备。
如图2A所示，图2A为本申请一实施例提供的一种负载惯量辨识系统的模型示意图。负载惯量辨识系统2包括伺服驱动器110、机械传动系统120、负载机构130、以及惯量识别器140。
本实施例中，采用永磁同步电机作为上述的伺服驱动器110。伺服驱动器110与负载机构130之间通过机械传动系统120连接。采用扩展卡尔曼滤波器(Extended Kalman Filter，EKF)(也可称为拓展卡尔曼观测器)作为上述的惯量识别器140。
本实施例中，电流采样电路通过霍尔传感器检测电机在三相静止坐标系下的三相电流i a，i b以及i c，经过克拉克(Clark)变换(3s/2s)，转换为静止两相坐标系下的电流值i α、i β，再将速度外环中的给定转速与由位置传感器微分得到的反馈速度相比较的误差，经过速度外环控制器调节后，输出转子旋转坐标系下的给定转矩电流和静止两相坐标系下的电流值i α、i β以及转子侧位置传感器输出的电磁角度θ 1。
于一实施例中，电流采样电路也可由其他采样方式检测获得电机在三相静止坐标系下的三相电流i a，i b以及i c。
本实施例中，静止两相坐标系下的电流值i α、i β经过派克(Park)变换(2s/2r)转换为转子旋转坐标系下的两相反馈计算励磁电流i d_fdb和反馈计算转矩电流i q_fdb。给定励磁电流i d_ref与反馈计算励磁电流i d_fdb相比较，经过电流调节器调节之后，得到两相旋转坐标的d轴给定输出电压u d。
本实施例中，给定转矩电流i q_ref与反馈计算转矩电流i q_fdb相比较之后，经过电流调节器调节后，得到两相旋转坐标的q轴给定输出电压u q。旋转坐标系下的两相给定输出电压u d与u q经过Park逆变换(2r/2s)逆变换之后转换为静止两相坐标系下的两相电压u α与u β，经过脉冲宽度调制(Pulse Width Modulation，PWM)发生模块的调节，产生PWM波，经过三相功率逆变电路之后，驱动伺服驱动器110工作。已知的参数包括机械传动系统120的扭转刚度K、伺服驱动器110转子的转动惯量J 1，反馈计算转矩电流i q_fdb与伺服驱动器110侧反馈速度
Figure PCTCN2022111093-appb-000001
负载机构130侧反馈速度
Figure PCTCN2022111093-appb-000002
作为惯量识别器140的输入，输出即为惯量辨识系统2的负载侧转动惯量J 2。
请参阅图2B，图2B为本申请一实施例提供的一种双惯量机械传动装置示意图，该双惯量机械传动装置示意图是由负载惯量辨识系统2中抽象出的双惯量机械传动装置3的示意图，其中，双惯量机械传动装置3包括电机150、传动机构160和负载机构130。图中，J 1为电机150侧转子的转动惯量，是电机150的基本参数，由电机150的生产厂家测试获得，标注在电机150的使用手册中；C 1为电机150的阻尼系数，由电机150的生产厂家测试获得，标注在电机150的使用手册中；T e为电机150的电磁转矩，通过传感器采集电机150的三相电流经计算获得q轴电流I q，近似的，电机150的q轴电流I q与其电磁转矩T e成正比，其正比因子为K e；θ 1为电机150转子侧位置传感器输出的电磁转动角度；T w为传动机构160的转矩，传动机构160发生扭转变形时将生成T w；C w为传动机构160的阻尼系数，由于会对传动机构160进行润滑，因此C w非常小，由传动机构160的生产制造厂商提供，有时可设定为0；K为抗扭刚度系数，是由传动机构160的生产制造厂商测定的常数值或参数表，在传动机构160的使用手册中给定，抗扭刚度系数K表征双惯量机械传动装置3中电机150与负载机构130连接的柔性；θ 2为负载机构130侧位置传感器输出的负载转动角度；J 2为负载机构130等效转动惯量，是需要辨识的；C 2为负载机构130的阻尼系数，由于会对负载机构130进行润滑，因此C 2非常小，由负载机构130的生产制造厂商提供，有时可设定为0；T l为负载机构130侧的转矩。
本实施例中，电机150、传动机构160和负载机构130组成的典型的双惯量机械传动装置3。其中传动机构160连接电机150和负载机构130，具有一定的抗扭刚度K和阻尼系数C w。当传动机构160发生扭转形变时将产生转矩T w。电磁转矩T e和传动机构160侧转矩T w共同作用于转动惯量为J 1和阻尼系数为C 1的电机150的转轴。负载机构130侧等效转动惯量为J 2、阻尼系数为C 2，传动机构160转矩T w与负载转矩T l共同作用于执行机构最终决定了负载侧转速ω l。其中，电机150侧转速
Figure PCTCN2022111093-appb-000003
负载机构130侧转速
Figure PCTCN2022111093-appb-000004
综合上述多个参数变量及其含义建立微分方程组，如公式(1)所示。
Figure PCTCN2022111093-appb-000005
式中，J 1为电机150侧转子的转动惯量、θ 1为电磁转动角度、T e为电机150的电磁转矩、C 1为电机150的阻尼系数、T w为传动机构160的转矩、J 2为负载机构130等效转动惯量、θ 2为负载机构130侧转动角度、C 2为负载机构130的阻尼系数、T l为负载机构130侧的转矩、C w为传动机构160的阻尼系数、K为传动机构160抗扭刚度系数、ω m为电机150侧转速、ω l为负载机构130侧转速。
由于会对双惯量机械传动装置3进行润滑，机械传动装置3中的阻尼系数很小，可设定为0；为进行惯量辨识计算，对式(1)进行拉普拉斯变化，得式(2)：
Figure PCTCN2022111093-appb-000006
式中，J 1为电机150侧转子的转动惯量、θ 1为电磁转动角度、s为拉普拉斯算子、T e为电机150的电磁转矩、T w为传动机构160的转矩、J 2为负载机构130等效转动惯量、θ 2为负载转动角度、T l为负载机构130侧的转矩、K为抗扭刚度系数、ω m为电机150侧转速、ω l为负载机构130侧转速。
请参阅图3，图3为本申请一实施例提供的一种负载惯量辨识方法的流程示意图。该方法可由图1所示的电子设备1来执行，并可应用于如图2A至2B所示的负载惯量辨识场景中，以实现负载惯量的准确辨识。该方法包括：
301：获取待测的机械传动系统在上一状态的输入控制指令、电机转速和负载转速。
待测的机械传动系统在上一状态的输入控制指令可以是表征系统在上一状态多项参数，例如由根据三相电流及电机模型计算得到的电磁转矩T e，上一状态的电机转速可以是由电机150侧的位置传感器所测数据电磁角度θ 1经转换后得到的电机150侧反馈速度ω m，也可以是由其他传感器直接或间接测量得到；上一状态的负载转速可以是由负载机构130侧的位置传感器所测数据负载转动角度θ 2经转换后得到的负载机构130侧反馈速度ω l，也可以是由其他传感器直接或间接测量得到。
于一实施例中，待测的机械传动系统在上一状态的输入控制指令还可以是三相电流经计算得到的q轴转矩电流；根据传感器测得的电机三相电流和电机自身基本规格参数经其他方式计算得到的电磁转矩T e；本申请中的待测的机械传动系统在上一状态的输入控制指令可由惯量辨识系统根据实际需求进行选择，并不以此为限。
302：根据输入控制指令、电机转速、负载转速以及机械传动系统的状态预测方程，预测得到机械传动系统的当前系统预测状态。
可以根据得到的输入控制指令、电机转速、负载转速以及机械传动系统的状态预测方程，预测得到机械传动系统的当前系统预测状态。
于一实施例中，采用如下公式计算得到所述当前系统预测状态：
Figure PCTCN2022111093-appb-000007
Figure PCTCN2022111093-appb-000008
Figure PCTCN2022111093-appb-000009
式中，k表示机械传动系统的上一状态所在的周期，k+1表示机械传动系统的当前状态所在的周期，
Figure PCTCN2022111093-appb-000010
为机械传动系统在第k个周期的系统状态，
Figure PCTCN2022111093-appb-000011
为机械传动系统在第k个周期的系统状态的预测函数，
Figure PCTCN2022111093-appb-000012
为机械传动系统在第(k+1)个周期的系统预测状态，T s为机械传动系统的采样周期，
Figure PCTCN2022111093-appb-000013
为机械传动系统在第(k+1)个周期的预测输出状态，u(k)为机械传动系统在第k个周期的输入控制指令，
Figure PCTCN2022111093-appb-000014
为机械传动系统在第k个周期的噪声协方差，
Figure PCTCN2022111093-appb-000015
为机械传动系统在第(k+1)个周期的预测噪声协方差，Q(k)为机械传动系统在第k个周期的系统误差的协方差，J 1为机械传动系统的电机转子的转动惯量，M l为机械传动系统的负载侧转动惯量J 2的倒数。
其中，雅可比矩阵F(x)为：
Figure PCTCN2022111093-appb-000016
式中，k表示机械传动系统的上一状态所在的周期、T s为机械传动系统的采样周期、J 1为机械传动系统的电机转子的转动惯量、M l为机械传动系统的负载侧转动惯量J 2的倒数。
303：获取机械传动系统的当前电机转速测量值和当前负载转速测量值。
当前电机转速测量值可由设置在电机150侧的位置传感器间接测量得知，也可以由设置在电机150侧的其他传感器测得；当前负载转速测量值可由设置 在负载机构130侧的位置传感器间接测量得知，也可以由设置在负载机构130侧的其他传感器测得。
304：根据当前系统预测状态、当前电机转速测量值和当前负载转速测量值，确定机械传动系统的当前最优估计状态。
可以采用如下公式计算机械传动系统的卡尔曼增益：
Figure PCTCN2022111093-appb-000017
采用如下公式计算当前最优估计状态：
Figure PCTCN2022111093-appb-000018
采用如下公式计算所述当前最优估计状态的误差均方：
Figure PCTCN2022111093-appb-000019
式中，
Figure PCTCN2022111093-appb-000020
其中，k表示机械传动系统的上一状态所在的周期，k+1表示机械传动系统的当前状态所在的周期，K(k+1)表示机械传动系统在第(k+1)个周期的所述卡尔曼增益；
Figure PCTCN2022111093-appb-000021
为机械传动系统在第(k+1)个周期的预测噪声协方差；R(k)为机械传动系统在第k个周期的测量误差的协方差；
Figure PCTCN2022111093-appb-000022
为机械传动系统在第(k+1)个周期的最优估计状态；
Figure PCTCN2022111093-appb-000023
为机械传动系统在第(k+1)个周期的系统预测状态；
Figure PCTCN2022111093-appb-000024
为机械传动系统在第(k+1)个周期的预测输出状态；y(k+1)为基于当前电机转速测量值和当前负载转速测量值确定的机械传动系统在第(k+1)个周期的输出状态测量值；
Figure PCTCN2022111093-appb-000025
为机械传动系统在第(k+1)个周期的最优估计状态的误差均方。
305：根据当前最优估计状态，确定机械传动系统在当前状态下的负载惯量信息。
可以根据得到的在第(k+1)个周期的最优估计状态
Figure PCTCN2022111093-appb-000026
从而确定机械传动系统在当前状态的负载惯量信息。
上述负载惯量辨识方法中，通过将机械传动系统中的电机与负载端之间的连接认定为柔性连接，引入抗扭刚度系数K、电机侧转速ω m、负载侧转速ω l、传动轴系转矩T w以及负载转矩T l，从而建立双惯量系统状态模型和观测矩阵；而后确定系统状态预测矩阵和协方差矩阵；进而利用拓展卡尔曼滤波算法对于负载惯量进行迭代辨识，提高了负载侧惯量辨识的准确性。本实施例由于将机械传动系统的电机转速和负载转速分别进行考虑，因此可以基于该机械传动系统的当前最优估计状态确定所述机械传动系统在当前状态下的负载惯量信息， 与传统将电机与负载作为一个惯量参数相比，可以更加准确辨识系统的负载端的转动惯量。
请参阅图4，图4为本申请一实施例提供的另一种负载惯量辨识方法的流程示意图。该方法可由图1所示的电子设备1来执行，并可应用于如图2A至2B所示的负载惯量辨识场景中，以实现负载惯量的准确辨识。该方法包括：
401：根据机械传动系统的双惯量参数和机械运动方程，确定机械传动系统的状态方程和输出方程。
以图2B所示的双惯量机械传动装置3为例，双惯量机械传动装置3中的参数有：电机150侧转子的转动惯量J 1、电机150的阻尼系数C 1、电机150的电磁转矩T e、电磁转动角度θ 1、传动机构160的转矩T w、传动机构160的阻尼系数C w、传动机构160的抗扭刚度系数K、负载机构130侧转动角度θ 2、负载机构130等效转动惯量J 2、负载机构130的阻尼系数C 2、负载机构130侧的转矩T l；基于双惯量机械传动装置3与上述参数，从而建立公式(1)，基于公式(1)可获得公式(2)，对于公式(1)与公式(2)的解释参见对于图2B的描述，此处不再重复赘述。
因此，基于公式(2)，采用如下公式确定所述机械传动系统的状态方程和输出方程：
Figure PCTCN2022111093-appb-000027
y＝Cx  (12)
其中，公式(12)中的矩阵C如公式(10)中所示，
x＝[ω m ω l T w T l M l] T  (13)
y＝[ω m ω l] T  (14)
u＝T e＝K e·I q  (15)
Figure PCTCN2022111093-appb-000028
Figure PCTCN2022111093-appb-000029
式中，k表示机械传动系统的上一状态所在的周期；x为机械传动系统的状 态变量；y为机械传动系统的输出变量；ω m为机械传动系统的电机转速；ω l为机械传动系统的负载侧转速；T w为传动轴系转矩；T l为负载转矩；J 1为电机转子的转动惯量；M l为负载侧转动惯量J 2的倒数；电磁转矩T e可通过采集电机的q轴电流I q得到；近似的，电机的q轴电流I q与电磁转矩T e成正比，K e为其正比因子；T s为机械传动系统的采样周期。由于T l和M l变化较为缓慢，可视其导数为0。
402：根据机械传动系统的状态方程和输出方程，确定机械传动系统的状态预测方程。
利用扩展Kalman滤波器可以通过迭代方式直接求解非线性方程的特征，即可由上一状态的系统状态变量根据系统状态方程做出预测，计算当前状态的系统状态。基于公式(11)-(17)得到机械传动系统的状态预测方程：
x(k+1)＝x(k)+T s·f(x(k))+T s·Bu(k)  (18)
y(k)＝Cx(k)  (19)
式中，k表示机械传动系统的上一状态所在的周期、k+1表示机械传动系统的当前状态所在的周期、x(k)为机械传动系统第k个周期的预测值、x(k+1)为机械传动系统在第(k+1)个周期的预测状态的预测值、T s为机械传动系统的采样周期、u(k)为所述机械传动系统在第k个周期的输入控制指令、y(k)为机械传动系统第k个周期的输出变量的预测值。
于一实施例中，系统中存在不确定性和可变性，如存在测量噪声，因此公式(18)-(19)可表示为：
x(k+1)＝x(k)+T s·f(k(k))+T s·Bu(k)+w(k)  (20)
y(k)＝Cx(k)+v(k)  (21)
式中，w代表系统参数误差所带来的影响，而v代表测量过程中的噪声和干扰，包括码盘测量位置信号的量化误差。噪声可以是平稳的高斯白噪声，其平均值为零。因此，噪声的协方差矩阵可定义为:
Q＝cov(w)＝E{ww T}
R＝cov(v)＝E{vv T}
403：获取待测的机械传动系统在上一状态的输入控制指令、电机转速和负载转速。详细参见上述实施例中对于301的描述。
404：根据输入控制指令、电机转速、负载转速以及机械传动系统的状态预测方程，预测得到机械传动系统的当前系统预测状态。详细参见上述实施例中对于302的描述。
405：获取机械传动系统的当前电机转速测量值和当前负载转速测量值。详 细参见上述实施例中对于303的描述。
406：根据当前系统预测状态、当前电机转速测量值和当前负载转速测量值，确定机械传动系统的当前最优估计状态。详细参见上述实施例中对于304的描述。
407：根据当前最优估计状态，确定机械传动系统在当前状态下的负载惯量信息。详细参见上述实施例中对于305的描述。
408：输出机械传动系统的负载惯量信息。
辨识到的负载惯量信息可以输出，输出的负载惯量信息可以用于机械传动系统进行调节。
请参阅图5，图5为本申请一实施例提供的一种机械臂伺服运动数据波形示意图。采用了上述实施例中的负载惯量辨识方法对于机械臂伺服运动进行观测。
本实施例中，伺服驱动器110为六轴机械臂的基座关节伺服驱动器，利用EKF作为惯量识别器140，在机械臂运行过程中对关节后端负载惯量进行实时观测。其纵坐标为负载侧惯量，单位为(Kg·m 2)，横坐标为采样时间，单位为ms。其中，实线为根据机械臂理论动力学模型计算的机械臂基座关节后端惯量的变化曲线，实线周围的虚线为通过本实施例的卡尔曼观测器实时观测到的关节后端负载惯量。从数据图形中可以看出，双惯量系统卡尔曼在线惯量观测器可以很好的跟随负载惯量变化，实时性较好。
请参阅图6，图6为本申请一实施例提供的一种负载惯量辨识装置。该方法可由图1所示的电子设备1来执行，并可应用于如图2A至2B所示的负载惯量辨识场景中，以更加准确地对于负载惯量进行辨识。该装置600包括：第一获取模块601、预测模块602、第二获取模块603、计算模块604和确定模块605，多个模块的原理关系如下：
第一获取模块601，设置为获取待测的机械传动系统在上一状态的输入控制指令、电机转速和负载转速。
预测模块602，设置为根据输入控制指令、电机转速、负载转速以及机械传动系统的状态预测方程，预测得到机械传动系统的当前系统预测状态。
第二获取模块603，设置为获取机械传动系统的当前电机转速测量值和当前负载转速测量值。
计算模块604，设置为根据当前系统预测状态、当前电机转速测量值和当前负载转速测量值，确定机械传动系统的当前最优估计状态。
确定模块605，设置为根据当前最优估计状态，确定机械传动系统在当前状 态下的负载惯量信息。
于一实施例中，负载惯量辨识装置600还包括建立模块，建立模块设置为：根据机械传动系统的双惯量参数和机械运动方程，确定机械传动系统的状态方程和输出方程；根据机械传动系统的状态方程和输出方程，确定机械传动系统的所述状态预测方程。
于一实施例中，建立模块采用如下公式确定机械传动系统的状态方程和输出方程：
Figure PCTCN2022111093-appb-000030
y＝Cx
其中，
x＝[ω m ω l T w T l M l] T
y＝[ω m ω l] T
u＝T e＝K e·I q
Figure PCTCN2022111093-appb-000031
Figure PCTCN2022111093-appb-000032
Figure PCTCN2022111093-appb-000033
其中，k表示机械传动系统的上一状态所在的周期，x为机械传动系统的状态变量，y为机械传动系统的输出量，ω m为机械传动系统的电机转速，ω l为机械传动系统的负载转速，T w为机械传动系统传动轴系转矩，T l为机械传动系统负载转矩，M l为机械传动系统的负载侧转动惯量J 2的倒数，K e为正比因子，I q为机械传动系统的转矩电流，f(x)表示机械传动系统的第k个周期的预测函数，k为整数。
于一实施例中，预测模块602采用如下公式计算得到当前系统预测状态：
Figure PCTCN2022111093-appb-000034
Figure PCTCN2022111093-appb-000035
Figure PCTCN2022111093-appb-000036
其中，
Figure PCTCN2022111093-appb-000037
其中，k表示机械传动系统的上一状态所在的周期，k+1表示机械传动系统的当前状态所在的周期，
Figure PCTCN2022111093-appb-000038
为机械传动系统在第k个周期的系统状态，
Figure PCTCN2022111093-appb-000039
为机械传动系统在第k个周期的系统状态的预测函数，
Figure PCTCN2022111093-appb-000040
为机械传动系统在第(k+1)个周期的系统预测状态，T s为机械传动系统的采样周期，
Figure PCTCN2022111093-appb-000041
为机械传动系统在第(k+1)个周期的预测输出状态，u(k)为机械传动系统在第k个周期的输入控制指令，
Figure PCTCN2022111093-appb-000042
为机械传动系统在第k个周期的噪声协方差，
Figure PCTCN2022111093-appb-000043
为机械传动系统在第(k+1)个周期的预测噪声协方差，Q(k)为机械传动系统在第k个周期的系统误差的协方差，J 1为机械传动系统的电机转子的转动惯量，M l为机械传动系统的负载侧转动惯量J 2的倒数。
于一实施例中，计算模块604采用如下公式计算机械传动系统的卡尔曼增益：
Figure PCTCN2022111093-appb-000044
采用如下公式计算所述当前最优估计：
Figure PCTCN2022111093-appb-000045
采用如下公式计算所述当前最优估计的误差均方：
Figure PCTCN2022111093-appb-000046
其中，k表示机械传动系统的上一状态所在的周期，k+1表示机械传动系统的当前状态所在的周期，K(k+1)表示机械传动系统在第(k+1)个周期的卡尔曼增益；
Figure PCTCN2022111093-appb-000047
为机械传动系统在第(k+1)个周期的预测噪声协方差；R(k)为机械传动系统在第k个周期的测量误差的协方差；
Figure PCTCN2022111093-appb-000048
为机械传动系统在第(k+1)个周期的最优估计状态；
Figure PCTCN2022111093-appb-000049
为机械传动系统在第(k+1)个周期的系统预测状态；
Figure PCTCN2022111093-appb-000050
为机械传动系统在第(k+1)个周期的预测输出状态；y(k+1)为基于当前电机转速测量值和当前负载转速测量值确定的机械传动系 统在第(k+1)个周期的输出状态测量值；
Figure PCTCN2022111093-appb-000051
为机械传动系统在第(k+1)个周期的最优估计状态的误差均方。
于一实施例中，负载惯量辨识装置600还包括：输出模块，设置为输出机械传动系统的负载惯量信息。
上述负载惯量辨识装置600的描述，请参见上述实施例中相关方法的描述。

Claims (10)

  1. 一种负载惯量辨识方法，包括：
    获取待测的机械传动系统在上一状态的输入控制指令、电机转速和负载转速；
    根据所述输入控制指令、所述电机转速、所述负载转速以及所述机械传动系统的状态预测方程，预测得到所述机械传动系统的当前系统预测状态；
    获取所述机械传动系统的当前电机转速测量值和当前负载转速测量值；
    根据所述当前系统预测状态、所述当前电机转速测量值和所述当前负载转速测量值，确定所述机械传动系统的当前估计状态；
    根据所述当前估计状态，确定所述机械传动系统在当前状态下的负载惯量信息。
  2. 根据权利要求1所述的方法，还包括：建立所述机械传动系统的所述状态预测方程；
    所述建立所述机械传动系统的所述状态预测方程，包括：
    根据所述机械传动系统的双惯量参数和机械运动方程，确定所述机械传动系统的状态方程和输出方程；
    根据所述机械传动系统的所述状态方程和所述输出方程，确定所述机械传动系统的所述状态预测方程。
  3. 根据权利要求2所述的方法，其中，所述输入控制指令为电机电磁转矩指令；所述根据所述机械传动系统的双惯量参数和机械运动方程，确定所述机械传动系统的状态方程和输出方程，包括：
    采用如下公式确定所述机械传动系统的所述状态方程和所述输出方程：
    Figure PCTCN2022111093-appb-100001
    y＝Cx；
    其中，
    x＝[ω m ω l T w T l M l] T；
    y＝[ω m ω l] T；
    u＝T e＝K e·I q；
    Figure PCTCN2022111093-appb-100002
    Figure PCTCN2022111093-appb-100003
    Figure PCTCN2022111093-appb-100004
    其中，k表示所述机械传动系统的上一状态所在的周期，x为所述机械传动系统的状态变量，y为所述机械传动系统的输出变量，ω m为所述机械传动系统的电机转速，ω l为所述机械传动系统的负载转速，T w为所述机械传动系统的传动轴系转矩，T l为所述机械传动系统的负载转矩，M l为所述机械传动系统的负载侧转动惯量J 2的倒数，u为所述输入控制指令携带的输入控制变量，T e是电机电磁转矩，K e为正比因子，I q为所述机械传动系统的转矩电流，f(x)表示所述机械传动系统的第k个周期的预测函数，k为整数，T s为所述机械传动系统的采样周期，J 1为所述机械传动系统的电机转子的转动惯量。
  4. 根据权利要求3所述的方法，其中，所述根据所述输入控制指令、所述电机转速、所述负载转速以及所述机械传动系统的状态预测方程，预测得到所述机械传动系统的当前系统预测状态，包括：
    采用如下公式计算得到所述当前系统预测状态：
    Figure PCTCN2022111093-appb-100005
    Figure PCTCN2022111093-appb-100006
    Figure PCTCN2022111093-appb-100007
    其中，
    Figure PCTCN2022111093-appb-100008
    其中，k+1表示所述机械传动系统的当前状态所在的周期，
    Figure PCTCN2022111093-appb-100009
    为所述机械传动系统在第k个周期的系统状态，
    Figure PCTCN2022111093-appb-100010
    为所述机械传动系统在第k个周期的系统状态的预测函数，
    Figure PCTCN2022111093-appb-100011
    为所述机械传动系统在第(k+1)个周期的系统预测状态，
    Figure PCTCN2022111093-appb-100012
    为所述机械传动系统在第(k+1)个周期的预测输出状 态，u(k)为所述机械传动系统在第k个周期的输入控制指令，
    Figure PCTCN2022111093-appb-100013
    为所述机械传动系统在第k个周期的噪声协方差，
    Figure PCTCN2022111093-appb-100014
    为所述机械传动系统在第(k+1)个周期的预测噪声协方差，Q(k)为所述机械传动系统在第k个周期的系统误差的协方差。
  5. 根据权利要求4所述的方法，其中，所述根据所述当前系统预测状态、所述当前电机转速测量值和所述当前负载转速测量值，确定所述机械传动系统的当前估计状态，包括：
    采用如下公式计算所述机械传动系统的卡尔曼增益：
    Figure PCTCN2022111093-appb-100015
    采用如下公式计算所述当前估计状态：
    Figure PCTCN2022111093-appb-100016
    采用如下公式计算所述当前估计状态的误差均方：
    Figure PCTCN2022111093-appb-100017
    其中，K(k+1)表示所述机械传动系统在第(k+1)个周期的所述卡尔曼增益；R(k)为所述机械传动系统在第k个周期的测量误差的协方差；
    Figure PCTCN2022111093-appb-100018
    为所述机械传动系统在第(k+1)个周期的估计状态；y(k+1)为基于所述当前电机转速测量值和所述当前负载转速测量值确定的所述机械传动系统在第(k+1)个周期的输出状态测量值；
    Figure PCTCN2022111093-appb-100019
    为所述机械传动系统在第(k+1)个周期的估计状态的误差均方。
  6. 根据权利要求1所述的方法，还包括：
    输出所述机械传动系统的所述负载惯量信息。
  7. 一种负载惯量辨识装置，包括：
    第一获取模块，设置为获取待测的机械传动系统在上一状态的输入控制指令、电机转速和负载转速；
    预测模块，设置为根据所述输入控制指令、所述电机转速、所述负载转速以及所述机械传动系统的状态预测方程，预测得到所述机械传动系统的当前系统预测状态；
    第二获取模块，设置为获取所述机械传动系统的当前电机转速测量值和当前负载转速测量值；
    计算模块，设置为根据所述当前系统预测状态、所述当前电机转速测量值和所述当前负载转速测量值，确定所述机械传动系统的当前估计状态；
    确定模块，设置为根据所述当前估计状态，确定所述机械传动系统在当前状态下的负载惯量信息。
  8. 根据权利要求7所述的装置，其中，所述装置还包括建立模块，所述建立模块设置为：
    根据所述机械传动系统的双惯量参数和机械运动方程，确定所述机械传动系统的状态方程和输出方程；
    根据所述机械传动系统的所述状态方程和所述输出方程，确定所述机械传动系统的所述状态预测方程。
  9. 一种电子设备，包括：
    存储器，设置为存储计算机程序；
    处理器，设置为执行所述计算机程序，以实现如权利要求1至6中任一项所述的负载惯量辨识方法。
  10. 一种负载惯量辨识系统，包括：
    机械传动系统；
    伺服驱动器，连接所述机械传动系统，设置为驱动所述机械传动系统运行；
    惯量识别器，连接所述机械传动系统和所述伺服驱动器，设置为实现如权利要求1至6中任一项所述的负载惯量辨识方法以识别所述机械传动系统的负载惯量信息。
PCT/CN2022/111093 2021-09-01 2022-08-09 负载惯量辨识方法、装置、电子设备及系统 WO2023029903A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202111018435.5A CN113708690A (zh) 2021-09-01 2021-09-01 负载惯量辨识方法、装置、电子设备及系统
CN202111018435.5 2021-09-01

Publications (1)

Publication Number Publication Date
WO2023029903A1 true WO2023029903A1 (zh) 2023-03-09

Family

ID=78658519

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/111093 WO2023029903A1 (zh) 2021-09-01 2022-08-09 负载惯量辨识方法、装置、电子设备及系统

Country Status (2)

Country Link
CN (1) CN113708690A (zh)
WO (1) WO2023029903A1 (zh)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113708690A (zh) * 2021-09-01 2021-11-26 上海节卡机器人科技有限公司 负载惯量辨识方法、装置、电子设备及系统
CN114603554B (zh) * 2022-02-21 2023-06-20 苏州艾利特机器人有限公司 机器人负载转动惯量的标定方法、装置及存储介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2006034003A (ja) * 2004-07-16 2006-02-02 Yaskawa Electric Corp サーボ制御装置およびその負荷イナーシャ算出方法
US20090033259A1 (en) * 2006-08-03 2009-02-05 Stmicroelectronics S.R.I. Method of estimating the state of a system and relative device for estimating position and speed of the rotor of a brushless motor
CN110572107A (zh) * 2019-08-28 2019-12-13 哈尔滨工业大学 一种伺服系统免调试控制方法及装置
CN110739893A (zh) * 2019-10-23 2020-01-31 西安理工大学 一种改进自适应无轨迹卡尔曼滤波转动惯量辨识方法
CN113708690A (zh) * 2021-09-01 2021-11-26 上海节卡机器人科技有限公司 负载惯量辨识方法、装置、电子设备及系统

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2006034003A (ja) * 2004-07-16 2006-02-02 Yaskawa Electric Corp サーボ制御装置およびその負荷イナーシャ算出方法
US20090033259A1 (en) * 2006-08-03 2009-02-05 Stmicroelectronics S.R.I. Method of estimating the state of a system and relative device for estimating position and speed of the rotor of a brushless motor
CN110572107A (zh) * 2019-08-28 2019-12-13 哈尔滨工业大学 一种伺服系统免调试控制方法及装置
CN110739893A (zh) * 2019-10-23 2020-01-31 西安理工大学 一种改进自适应无轨迹卡尔曼滤波转动惯量辨识方法
CN113708690A (zh) * 2021-09-01 2021-11-26 上海节卡机器人科技有限公司 负载惯量辨识方法、装置、电子设备及系统

Also Published As

Publication number Publication date
CN113708690A (zh) 2021-11-26

Similar Documents

Publication Publication Date Title
WO2023029903A1 (zh) 负载惯量辨识方法、装置、电子设备及系统
EP3040690B1 (en) Angle error correction device and angle error correction method for position detector
US20190339161A1 (en) Failure diagnosis device for reduction gear, method of diagnosing failure, and mechanical apparatus provided with the failure diagnosis device
CN113510693B (zh) 一种基于摩擦力的机器人控制方法、装置及设备
JP5277787B2 (ja) 同期モータ駆動制御装置
KR20030020272A (ko) 전동기 제어 장치
WO2020003738A1 (ja) 制御パラメータ調整方法および制御パラメータ調整装置
JPH09146643A (ja) サーボモータの外乱負荷推定方法
JP4196400B2 (ja) モータの制御装置及び制御方法
US11863107B2 (en) Device and method for controlling rotary electric machine
CN112271965A (zh) 基于初始相位辨识的相序控制方法、装置、电子设备
CN115566954B (zh) 嵌入式电机调速控制补偿方法及系统
CN108880369B (zh) 基于分数阶滑模控制的电机抗干扰控制方法、系统及装置
CN113131818B (zh) 一种霍尔传感器安装误差辨识方法、装置及电机控制系统
JP4449693B2 (ja) ロボット制御装置およびその制御方法
JP2021197752A (ja) トルク推定装置およびトルク推定方法
JP2798217B2 (ja) 高速位置決め制御方法
CN112234897B (zh) 永磁同步电机参数辨识方法及其辨识装置和辨识系统
US20240085867A1 (en) System and method for static auto-tuning electric motor
RU2771794C1 (ru) Способ идентификации постоянной времени ротора и взаимной индуктивности статора и ротора асинхронного двигателя в системе векторного полеориентированного управления
Huazhong PMSM parameter identification motor test system framework based on virtual instrument technology
CN114900085B (zh) 一种机器人关节伺服电机模型预测参数优化方法及装置
CN112234888B (zh) 一种用于电机的电流辨识方法、电机控制器及车辆
WO2020255584A1 (ja) 電動工具
KR20170084881A (ko) 영구자석형 동기전동기의 초기위치 판별을 위한 장치 및 방법

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22863044

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE