CN206216738U - A kind of six-DOF robot end load dynamic parameters identification device - Google Patents

Abstract

The utility model discloses a dynamic parameter identification device for a six-degree-of-freedom robot terminal load, which comprises a six-axis industrial robot, several loads of different masses replaceably arranged at the end of the six-axis industrial robot, and a real-time control system. The real-time control system is used to implement millisecond-level real-time motion data collection for the robot; the collected real-time motion data includes encoder values and torque values. The sensor system of the utility model adopts a high-efficiency real-time control system, which can realize millisecond-level real-time data collection of the robot, including displacement, motor output, etc., and the amount of collected data is scalable and has the characteristics of variability, which meets the requirements for the end load power of industrial robots. identification of the parameters.

Description

A six-degree-of-freedom robot terminal load dynamic parameter identification device

technical field

The utility model relates to a method for identifying dynamic parameters of end loads and a six-degree-of-freedom industrial robot device, in particular to an identification device for end-load dynamic parameters of a six-degree-of-freedom robot.

Background technique

Today, there are many testing devices for robot dynamic parameter identification. Including four-degree-of-freedom and six-degree-of-freedom industrial robots. Both the simulation and control of industrial robots require an accurate dynamic model of the robot. At present, many light-weight and high-performance robots widely use relatively flexible harmonic drives to drive joint motion, and the joint flexibility cannot be ignored. Therefore, a better parameter identification method is needed to establish an accurate flexible joint robot dynamic model.

For industrial robots, usually only the motor position and motor torque data can be measured. Therefore, the external sensors used to identify the dynamic parameters of flexible joint robots mainly include acceleration sensors, connecting rod position and speed sensors, and torque sensors. From a practical point of view, because for robots, adding additional internal sensors is not only expensive, but sometimes impossible to achieve. Therefore, many scholars have proposed identification methods that only need information about motor torque and motor position. In these methods, elastic deflection is not measured with additional sensors, but is obtained by solving the equations of motion involving unknown parameters.

At present, there are different methods for the identification of end load dynamic parameters. For different types of industrial robots, the established dynamic models and optimal excitation trajectories are different. The sensing system used is also different, so the way of data collection is also different.

However, the platform identification effect of many six-degree-of-freedom industrial robot terminal load dynamic parameter identification is not very ideal, because the compatibility of the software and hardware used and the improper selection of the algorithm lead to large errors in the identification results.

Utility model content

In order to solve the problem of identifying the dynamic parameters of the end load of the industrial robot, the utility model provides a device for identifying the dynamic parameters of the end load of the six-degree-of-freedom robot, which uses a series of different loads on the mechanical structure to design an experimental program for identifying the load. The optimal excitation trajectory is designed to collect data with a real-time control system, which provides external conditions for identifying the results of load dynamic parameters.

The purpose of this utility model is achieved through the following technical solutions:

A six-degree-of-freedom robot terminal load dynamic parameter identification device, including a six-axis industrial robot, a number of loads of different masses that can be replaced at the end of the six-axis industrial robot, and a real-time control system. The real-time control system uses It is used to implement millisecond-level real-time motion data collection for robots.

Further, the collected real-time motion data includes encoder values and torque values.

Further, the material of the load is steel.

Compared with the prior art, the utility model has the following advantages:

(1) The utility model is equipped with a series of different loads, and can identify the dynamic parameters of the loads, and the verification parameters are provided by the three-dimensional design drawing software.

(2) The utility model checks the accuracy of the identified load mass through the weighing experiment method, which is also evidence for verifying the accuracy of the model.

(3) The utility model realizes millisecond-level real-time data collection by adopting a high-efficiency real-time control system for data sampling, and ensures the accuracy of sample collection. The collected data includes displacement and motor output, etc., and the amount of data can be expanded, thus providing a theoretical basis for identification.

Description of drawings

Fig. 1 is a schematic diagram of a load dynamic parameter identification device for a six-degree-of-freedom industrial robot of the present invention.

Figure 2 is the first load of the identification object.

Fig. 3 is the second load of the identification object.

Figure 4. Parameter identification flow chart.

The picture shows: 1-load; 2-six-axis industrial robot.

detailed description

In order to further understand the utility model, the utility model will be further described below in conjunction with the accompanying drawings and embodiments, but it should be noted that the protection scope of the utility model is not limited to the range described in the embodiments.

Embodiment one

As shown in Figures 1 to 3, a device for identifying dynamic parameters of a six-axis robot terminal load includes a six-axis industrial robot 2, several loads 1 of different masses that can be alternatively arranged at the end of the six-axis industrial robot, A real-time control system, the real-time control system is used to implement millisecond-level real-time movement data collection for the robot.

Specifically, the collected real-time motion data includes encoder values and torque values.

Specifically, the material of the load is steel.

In this embodiment, a series of loads 1 with different configurations are designed by using three-dimensional design and drawing software, and the mass can be designed by oneself. In this experiment, the first load and the second load are designed; steel materials are used for self-designed loads, and they are processed.

Embodiment two

As shown in Figure 4, a method for identifying dynamic parameters of a six-degree-of-freedom robot end load based on the device includes steps:

(1) Establish a kinetic model according to the Lagrangian equation;

(2) Design the incentive trajectory;

(3) Implement millisecond-level real-time data collection for the robot through the real-time control system, and average multiple sampling data and improve the signal-to-noise ratio with the central difference method; the amount of collected data can be expanded, which not only ensures sufficient sampling data, but also provides The identification of the kinetic parameters of the end load provides the data basis.

(4) Identification of load parameters, substituting the parameters into the built dynamic model, and then estimating the dynamic parameters to be identified;

(5) Model verification. Using the identified load dynamic parameters to calculate the theoretical value of the moment and compare it with the actual moment value read to verify the accuracy of the model.

Specifically, described step (1) specifically includes:

(11), select the generalized coordinates, set up the finite-dimensional model, select the modal coordinates of the joint variable and the flexible link as the generalized coordinates;

(12) Establish kinetic energy, potential energy and virtual work expressions;

(13) Deriving and sorting out the Lagrangian equation, deriving the necessary inertial force term, Coriolis force and centripetal force term, gravity term and friction force term;

(14). Integrate the Lagrangian dynamic equation into a linear equation form for calculating torque.

Specifically, described step (14) specifically comprises:

(141) According to:

After derivation, the complete kinetic formula is obtained:

Among them, the calculated inertia parameter item is:

T _p is the transformation matrix of the coordinate system P relative to the base coordinate system 0; q _i is the angle value of each joint;

J _P is the pseudo-inertia matrix:

The calculated centripetal force and Coriolis force coefficient terms are:

The calculated gravitational term is

Calculated friction term:

F _vj , F _sj are viscous friction, Coulomb friction coefficient;

(142) Finally, the Lagrangian dynamic equation is simplified, and the basic dynamic parameter χ to be identified is extracted:

The linear equation for calculating moments can be reduced to:

τ=Wχ+ρ.

Specifically, the step (2) specifically includes:

The design adopts a periodic trajectory, and the excitation trajectory of each joint is the algebraic sum of sine and cosine functions, that is, a finite Fourier series function, then the joint position q _i and velocity of each joint of the robot and acceleration The plan is as follows

Specifically, the step (4) specifically includes:

(41) Combining three different load identification methods to identify the load dynamic parameters, the first load identification method uses the coupling items of the simplified basic dynamic parameters to subtract the difference, and the difference obtained is then compared with the three-dimensional drawing The load inertia data obtained by the software is identified by error analysis; the second load identification method uses the read torque value to identify whether there is a load or not; the third load identification method uses the global identification of robot dynamic parameters and load parameters.

(42) Select the optimal identification result from the three identification results.

Specifically, in the step (41),

The first load identification method includes the steps of:

(401) According to the equation τ=Wχ+ρ, calculate the estimated value χ with the method of least squares;

(402) Select the optimal track to carry out the experiment, first carry out parameter identification without loading, and obtain the basic dynamic parameters χ0 when a group of _zero loads are obtained;

(403) Then carry out the experiment with the same track again, carry out parameter identification with the first group of loads, obtain the first group of load basic dynamic parameters χ ₁ ;

(404) switch to the second group of loads for parameter identification, and obtain the second group of load basic dynamic parameters χ ₂ ;

(405) Calculating the basic dynamic parameters of the first group of loads: Δχ ₁ =χ ₁ -χ ₀ ;

(406) Calculating the basic dynamic parameters of the second group of loads: Δχ ₂ =χ ₂ −χ ₀ ;

The second load identification method includes the steps of:

(411) Read the torque value under the condition of no load, denoted as Y _WL , and then read the torque value under the condition of loading, denoted as Y _T ;

(412) The identification equation obtained is

Where, W ⁺ ＝(W ^T W) ^-1 W ^T

Y _T -Y _WL is a (6×ne)×1 matrix;

The third load identification method includes the steps of:

(421) The kinetic equation under the zero load condition and the dynamic equation when the load is put on are transformed into the form of matrix multiplication:

in:

Y _a =W _a χ

Y _b ＝W _b χ+W _L χ

(422) The above matrix equations are then solved using the weighted least squares method WLS.

In addition, described step (5) also includes the steps of:

Using the weighing experiment method to test and analyze the quality data of the load can provide a theoretical basis for the accuracy of identification. Use 3D drawing software to design 3D entities for different loads. The inertia parameters of each entity load can be obtained through 3D design drawing software, and the error analysis can be carried out with the identified theoretical values of the actual load dynamic parameters. The load dynamic parameters identified by the above-mentioned three load dynamic parameter identification methods are used for the calculation of the linear equation τ=Wχ+ρ, and the calculated torque theoretical value is compared with the read actual torque value. The accuracy of the model and the accuracy and feasibility of the identification parameter results are analyzed by using the torque error percentage.

In this embodiment, when designing the excitation trajectory, a series of constraint conditions of the robot movement are taken into consideration, such as joint angle range, speed range and so on. During excitation, the data are sampled at different speeds respectively.

The sensing system adopted in this embodiment is a real-time control system, which realizes millisecond-level real-time data collection for the movement of the robot, including displacement and motor output. The amount of collected data in the designed software structure can be expanded, which not only ensures sufficient sampling data, but also provides a data basis for identifying the dynamic parameters of the end load.

As shown in Figure 1 and Figure 2: install the first load on the end of the robot, and then use the designed excitation trajectory for excitation; in the project of data acquisition and processing, firstly, the collected joint position signals are processed by a good Tuning band-pass filtering is used to calculate the required angular velocity and angular acceleration. The second is to read the torque value, and then convert it into the required calculation parameters after data processing.

As shown in Figure 1 and Figure 3: install the second load on the end of the robot, and then use the designed optimal excitation trajectory to excite. In the data acquisition and processing project, one is to pass the collected joint position signals through A well-tuned bandpass filter is used to calculate the required angular velocity and angular acceleration. The second acquisition is to read the torque value, and then after data processing, it is converted into the required calculation parameters.

After a series of identification preparation steps, three different load identification methods are combined to identify the load dynamic parameters, and the optimal identification result is selected.

The above-mentioned embodiments of the present utility model are only examples for clearly illustrating the present utility model, and are not intended to limit the implementation of the present utility model. For those of ordinary skill in the art, other changes or changes in different forms can be made on the basis of the above description. It is not necessary and impossible to exhaustively list all the implementation manners here. All modifications, equivalent replacements and improvements made within the spirit and principles of the utility model shall be included in the protection scope of the claims of the utility model.

Claims (3)

1. A six-degree-of-freedom robot terminal load dynamic parameter identification device is characterized in that: it includes a six-axis industrial robot, a load of several different qualities that can be arranged on the end of the six-axis industrial robot, and a real-time control system, so that The real-time control system described above is used to implement millisecond-level real-time motion data collection for the robot. 2. The device for identifying dynamic parameters of a six-degree-of-freedom robot terminal load according to claim 1, wherein the collected real-time motion data includes encoder values and torque values. 3. The device for identifying dynamic parameters of the load at the end of a six-degree-of-freedom robot according to claim 1, wherein the material of the load is steel.