技术领域Technical field
本发明属于机器人技术领域，具体涉及一种基于机理数据混合驱动的机器人动态误差预测方法及系统。The invention belongs to the field of robot technology, and specifically relates to a robot dynamic error prediction method and system based on mechanism data hybrid driving.
背景技术Background technique
工业机器人是广泛用于工业领域的多关节机械手或多自由度的机器装置，具有一定的自动性，可依靠自身的动力能源和控制能力实现各种工业加工制造功能。工业机器人被广泛应用于电子、物流、化工等各个工业领域之中。工业机器人具有较高的灵活性和较低的成本，因此被广泛应用于制造业，然而跟踪精度较低限制了其在高精度制造领域的应用。Industrial robots are multi-joint manipulators or multi-degree-of-freedom machine devices that are widely used in the industrial field. They have a certain degree of automation and can rely on their own power energy and control capabilities to achieve various industrial processing and manufacturing functions. Industrial robots are widely used in various industrial fields such as electronics, logistics, and chemical industry. Industrial robots have high flexibility and low cost, so they are widely used in manufacturing. However, low tracking accuracy limits their application in high-precision manufacturing.
目前提高工业机器人跟踪精度的方法是通过机器人动力学模型的正向应用，同时借助Simscape进行动力学仿真，该控制模式实现了对机器人终端执行机构在工作空间的连续控制。它要求在一定精度范围内严格按照预定的轨迹和速度运动。并且速度可控，运动轨迹流畅，完成任务。这种方式能够预测出机器人终端执行机构的动态误差。The current method to improve the tracking accuracy of industrial robots is through the forward application of the robot dynamics model and the use of Simscape for dynamics simulation. This control mode achieves continuous control of the robot's terminal actuator in the work space. It requires movement in strict accordance with the predetermined trajectory and speed within a certain accuracy range. And the speed is controllable, the movement trajectory is smooth, and the task is completed. This method can predict the dynamic error of the robot terminal actuator.
然而现有的动力学模型存在复杂物理过程，难以用解析形式的数学模型描述，且存在强非线性拟合能力不强等问题。However, existing dynamic models have complex physical processes, are difficult to describe with analytical mathematical models, and have problems such as weak strong nonlinear fitting capabilities.
发明内容Contents of the invention
针对现有技术存在的难以解析以及强非线性拟合能力不强导致的动态误差预测不及时的问题，本发明提供一种基于机理数据混合驱动的机器人动态误差预测方法及系统，以解决上述技术问题。In order to solve the problems of untimely dynamic error prediction caused by the difficulty of analysis and weak strong nonlinear fitting ability in the existing technology, the present invention provides a robot dynamic error prediction method and system based on mechanism data hybrid drive to solve the above-mentioned technology. question.
第一方面，本发明提供一种基于机理数据混合驱动的机器人动态误差预测方法，包括：In the first aspect, the present invention provides a robot dynamic error prediction method based on mechanism data hybrid driving, including:
获取控制指令指示的目标位置；Obtain the target position indicated by the control instruction;
基于目标位置生成运动控制参数和运动参数，所述运动参数包括位置数据；Generate motion control parameters and motion parameters based on the target position, the motion parameters including position data;
利用预先训练好的LSTM模型，基于运动控制参数和运动参数及历史运动控制参数和运动参数，预测关节空间位置残差；Use the pre-trained LSTM model to predict joint space position residuals based on motion control parameters and motion parameters and historical motion control parameters and motion parameters;
基于运动参数与关节空间位置残差生成预测位置；Generate predicted positions based on motion parameters and joint space position residuals;
将所述预测位置与所述目标位置之间的差值输出为动态误差；Output the difference between the predicted position and the target position as a dynamic error;
LSTM模型的训练方法包括：Training methods for LSTM models include:
使用蒙特卡洛法让机器人末端跑随机轮廓，使得机器人运动轨迹充满工作空间；Use the Monte Carlo method to let the end of the robot run a random contour so that the robot's motion trajectory fills the work space;
在机器人运动过程中，收集电机位置、速度、关节空间下的位置和速度及笛卡尔空间下的位置和速度，并从机器人控制系统获取相应的目标位置和实际位置，将所有数据保存至数据集；During the movement of the robot, the motor position, speed, position and speed in joint space and position and speed in Cartesian space are collected, and the corresponding target position and actual position are obtained from the robot control system, and all data are saved to the data set. ;
利用所述数据集对预先构建的LSTM模型进行训练。The pre-built LSTM model is trained using the dataset.
在一个可选的实施方式中，基于目标位置生成运动控制参数和运动参数，所述运动参数包括位置数据，包括：In an optional implementation, motion control parameters and motion parameters are generated based on the target position, and the motion parameters include position data, including:
利用PID模型基于目标位置和机器人当前位置生成运动控制参数，所述运动控制参数包括电机位置、速度和力矩；Use the PID model to generate motion control parameters based on the target position and the current position of the robot. The motion control parameters include motor position, speed and torque;
将力矩输入柔性动力学模型得到运动参数，所述运动参数包括关节空间下的位置和速度及笛卡尔空间下的位置和速度；Input torque into the flexible dynamics model to obtain motion parameters, which include position and velocity in joint space and position and velocity in Cartesian space;
将电机位置、速度和运动参数标记生成时间后保存至输入数据集；Save the motor position, speed and motion parameters to the input data set after marking the generation time;
监控所述输入数据集中的数据的生成时间，若生成时间距当前时刻的时长超过设定的时间阈值则清除相应数据。Monitor the generation time of the data in the input data set, and clear the corresponding data if the time between the generation time and the current time exceeds a set time threshold.
在一个可选的实施方式中，利用预先训练好的LSTM模型基于运动控制参数和运动参数及历史运动控制参数和运动参数，预测关节空间位置残差，包括：In an optional implementation, a pre-trained LSTM model is used to predict joint space position residuals based on motion control parameters and motion parameters and historical motion control parameters and motion parameters, including:
将所述输入数据集作为输入参数导入预先训练好的LSTM模型，得到LSTM模型预测的关节空间位置残差。The input data set is imported into the pre-trained LSTM model as input parameters to obtain the joint space position residuals predicted by the LSTM model.
在一个可选的实施方式中，基于运动参数与关节空间位置残差生成预测位置，包括：In an optional implementation, generating a predicted position based on motion parameters and joint space position residuals includes:
从所述运动参数中提取关节空间位置；Extract joint space positions from the motion parameters;
将关节空间位置与关节空间位置残差的和输出为预测位置。The sum of the joint space position and the joint space position residual is output as the predicted position.
第二方面，本发明提供一种基于机理数据混合驱动的机器人动态误差预测系统，包括：In a second aspect, the present invention provides a robot dynamic error prediction system based on mechanism data hybrid drive, including:
目标获取模块，用于获取控制指令指示的目标位置；Target acquisition module, used to acquire the target position indicated by the control instruction;
参数获取模块，用于基于目标位置生成运动控制参数和运动参数，所述运动参数包括位置数据；A parameter acquisition module for generating motion control parameters and motion parameters based on the target position, where the motion parameters include position data;
残差预测模块，用于利用预先训练好的LSTM模型，基于运动控制参数和运动参数及历史运动控制参数和运动参数，预测关节空间位置残差；The residual prediction module is used to use the pre-trained LSTM model to predict joint space position residuals based on motion control parameters and motion parameters and historical motion control parameters and motion parameters;
位置预测模块，用于基于运动参数与关节空间位置残差生成预测位置；Position prediction module, used to generate predicted positions based on motion parameters and joint space position residuals;
误差预测模块，用于将所述预测位置与所述目标位置之间的差值输出为动态误差；An error prediction module, configured to output the difference between the predicted position and the target position as a dynamic error;
LSTM模型的训练方法包括：Training methods for LSTM models include:
使用蒙特卡洛法让机器人末端跑随机轮廓，使得机器人运动轨迹充满工作空间；Use the Monte Carlo method to let the end of the robot run a random contour so that the robot's motion trajectory fills the work space;
在机器人运动过程中，收集电机位置、速度、关节空间下的位置和速度及笛卡尔空间下的位置和速度，并从机器人控制系统获取相应的目标位置和实际位置，将所有数据保存至数据集；During the movement of the robot, the motor position, speed, position and speed in joint space and position and speed in Cartesian space are collected, and the corresponding target position and actual position are obtained from the robot control system, and all data are saved to the data set. ;
利用所述数据集对预先构建的LSTM模型进行训练。The pre-built LSTM model is trained using the dataset.
在一个可选的实施方式中，所述参数获取模块包括：In an optional implementation, the parameter acquisition module includes:
第一生成单元，用于利用PID模型基于目标位置和机器人当前位置生成运动控制参数，所述运动控制参数包括电机位置、速度和力矩；A first generation unit configured to use the PID model to generate motion control parameters based on the target position and the current position of the robot, where the motion control parameters include motor position, speed and torque;
第二生成单元，用于将力矩输入柔性动力学模型得到运动参数，所述运动参数包括关节空间下的位置和速度及笛卡尔空间下的位置和速度；The second generation unit is used to input torque into the flexible dynamics model to obtain motion parameters. The motion parameters include position and velocity in joint space and position and velocity in Cartesian space;
参数保存单元，用于将电机位置、速度和运动参数标记生成时间后保存至输入数据集；The parameter saving unit is used to save the motor position, speed and motion parameters to the input data set after marking the generation time;
定期清理单元，用于监控所述输入数据集中的数据的生成时间，若生成时间距当前时刻的时长超过设定的时间阈值则清除相应数据。A regular cleaning unit is used to monitor the generation time of the data in the input data set, and clear the corresponding data if the time between the generation time and the current time exceeds a set time threshold.
在一个可选的实施方式中，所述残差预测模块包括：In an optional implementation, the residual prediction module includes:
数据输入单元，用于将所述输入数据集作为输入参数导入预先训练好的LSTM模型，得到LSTM模型预测的关节空间位置残差。A data input unit is used to import the input data set as an input parameter into a pre-trained LSTM model to obtain the joint space position residual predicted by the LSTM model.
在一个可选的实施方式中，位置预测模块包括：In an optional implementation, the location prediction module includes:
参数提取单元，用于从所述运动参数中提取关节空间位置；A parameter extraction unit, used to extract joint space positions from the motion parameters;
残差补偿单元，用于将关节空间位置与关节空间位置残差的和输出为预测位置。The residual compensation unit is used to output the sum of the joint space position and the joint space position residual as the predicted position.
第三方面，提供一种终端，包括：In the third aspect, a terminal is provided, including:
处理器、存储器，其中，processor, memory, where,
该存储器用于存储计算机程序，This memory is used to store computer programs,
该处理器用于从存储器中调用并运行该计算机程序，使得终端执行上述的终端的方法。The processor is used to call and run the computer program from the memory, so that the terminal executes the above terminal method.
第四方面，提供了一种计算机存储介质，所述计算机可读存储介质中存储有指令，当其在计算机上运行时，使得计算机执行上述各方面所述的方法。In a fourth aspect, a computer storage medium is provided. The computer-readable storage medium stores instructions, which when run on a computer, cause the computer to perform the methods described in the above aspects.
本发明的有益效果在于，本发明提供的基于机理数据混合驱动的机器人动态误差预测方法及系统，通过对PID模型对伺服电机的运动控制参数进行跟踪预测得到一个关节空间位置，再利用LSTM模型对关节空间位置进行误差预测，进而得到一个预测的实际位置，将预测的实际位置与目标位置进行比对即可得到动态误差，本发明大大降低了动态误差预测的计算量，提升了动态误差预测的效率。The beneficial effect of the present invention is that the robot dynamic error prediction method and system based on mechanism data hybrid drive provided by the present invention can obtain a joint space position by tracking and predicting the motion control parameters of the servo motor using the PID model, and then using the LSTM model to predict the motion control parameters of the servo motor. Error prediction is performed on the joint space position, and then a predicted actual position is obtained. The dynamic error can be obtained by comparing the predicted actual position with the target position. The present invention greatly reduces the calculation amount of dynamic error prediction and improves the efficiency of dynamic error prediction. efficiency.
此外，本发明设计原理可靠，结构简单，具有非常广泛的应用前景。In addition, the design principle of the invention is reliable, the structure is simple, and it has very broad application prospects.
附图说明Description of the drawings
为了更清楚地说明本发明实施例或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍，显而易见地，对于本领域普通技术人员而言，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those of ordinary skill in the art, It is said that other drawings can also be obtained based on these drawings without exerting creative work.
图1是本发明一个实施例的方法的示意性流程图。Figure 1 is a schematic flow chart of a method according to an embodiment of the present invention.
图2是本发明一个实施例的方法的另一示意性流程图。Figure 2 is another schematic flow chart of a method according to an embodiment of the present invention.
图3是本发明一个实施例的系统的示意性框图。Figure 3 is a schematic block diagram of a system according to an embodiment of the present invention.
图4为本发明实施例提供的一种终端的结构示意图。Figure 4 is a schematic structural diagram of a terminal provided by an embodiment of the present invention.
具体实施方式Detailed ways
为了使本技术领域的人员更好地理解本发明中的技术方案，下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都应当属于本发明保护的范围。In order to enable those skilled in the art to better understand the technical solutions in the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described The embodiments are only some of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts should fall within the scope of protection of the present invention.
除非另有定义，本文所使用的所有的技术和科学术语与属于本发明的技术领域的技术人员通常理解的含义相同。本文中在本发明的说明书中所使用的术语只是为了描述具体的实施例的目的，不是旨在于限制本发明。Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the technical field to which the invention belongs. The terminology used herein in the description of the invention is for the purpose of describing specific embodiments only and is not intended to limit the invention.
下面对本发明中出现的关键术语进行解释。Key terms appearing in the present invention are explained below.
PID即：Proportional（比例）、Integral（积分）、Differential（微分）的缩写。PID是经典的闭环控制算法，具有原理简单，易于实现，适用面广，控制参数相互独立，参数的选定比较简单等优点。PID算法可分为位置式PID与增量式PID两大类。在实际的编程应用中，需要使用离散化的PID算法，以适用计算机的使用环境。PID算法可以根据当前位置和目标位置对电机转速、位置和力矩进行控制。PID is the abbreviation of Proportional, Integral and Differential. PID is a classic closed-loop control algorithm. It has the advantages of simple principle, easy implementation, wide application, independent control parameters, and relatively simple parameter selection. PID algorithms can be divided into two categories: positional PID and incremental PID. In actual programming applications, discretized PID algorithms need to be used to adapt to the computer usage environment. The PID algorithm can control the motor speed, position and torque according to the current position and target position.
长短期记忆网络（LSTM，Long Short-Term Memory）是一种时间循环神经网络，是为了解决一般的RNN（循环神经网络）存在的长期依赖问题而专门设计出来的，所有的RNN都具有一种重复神经网络模块的链式形式。在标准RNN中，这个重复的结构模块只有一个非常简单的结构，例如一个tanh层。Long Short-Term Memory Network (LSTM, Long Short-Term Memory) is a kind of temporal recurrent neural network, which is specially designed to solve the long-term dependency problem of general RNN (recurrent neural network). All RNNs have a A chained form of repeating neural network modules. In standard RNN, this repeated structural module has only a very simple structure, such as a tanh layer.
本发明实施例提供的基于机理数据混合驱动的机器人动态误差预测方法由计算机设备执行，相应地，基于机理数据混合驱动的机器人动态误差预测系统运行于计算机设备中。The robot dynamic error prediction method based on mechanism data hybrid drive provided by the embodiment of the present invention is executed by a computer device. Correspondingly, the robot dynamic error prediction system based on mechanism data hybrid drive runs in the computer device.
图1是本发明一个实施例的方法的示意性流程图。其中，图1执行主体可以为一种基于机理数据混合驱动的机器人动态误差预测系统。根据不同的需求，该流程图中步骤的顺序可以改变，某些可以省略。Figure 1 is a schematic flow chart of a method according to an embodiment of the present invention. Among them, the execution subject in Figure 1 can be a robot dynamic error prediction system based on mechanism data hybrid drive. Depending on different needs, the order of steps in this flowchart can be changed, and some can be omitted.
如图1所示，该方法包括：As shown in Figure 1, the method includes:
步骤110，获取控制指令指示的目标位置；Step 110, obtain the target position indicated by the control instruction;
步骤120，基于目标位置生成运动控制参数和运动参数，所述运动参数包括位置数据；Step 120: Generate motion control parameters and motion parameters based on the target position, where the motion parameters include position data;
步骤130，利用预先训练好的LSTM模型基于运动控制参数和运动参数及历史运动控制参数和运动参数，预测关节空间位置残差；Step 130, use the pre-trained LSTM model to predict joint space position residuals based on motion control parameters and motion parameters and historical motion control parameters and motion parameters;
步骤140，基于运动参数与关节空间位置残差生成预测位置；Step 140: Generate predicted positions based on motion parameters and joint space position residuals;
步骤150，将所述预测位置与所述目标位置之间的差值输出为动态误差。Step 150: Output the difference between the predicted position and the target position as a dynamic error.
其中，LSTM模型的训练方法包括：Among them, the training methods of LSTM model include:
使用蒙特卡洛法让机器人末端跑随机轮廓，使得机器人运动轨迹充满工作空间；Use the Monte Carlo method to let the end of the robot run a random contour so that the robot's motion trajectory fills the work space;
在机器人运动过程中，收集电机位置、速度、关节空间下的位置和速度及笛卡尔空间下的位置和速度，并从机器人控制系统获取相应的目标位置和实际位置，将所有数据保存至数据集；During the movement of the robot, the motor position, speed, position and speed in joint space and position and speed in Cartesian space are collected, and the corresponding target position and actual position are obtained from the robot control system, and all data are saved to the data set. ;
利用所述数据集对预先构建的LSTM模型进行训练。The pre-built LSTM model is trained using the dataset.
为了便于对本发明的理解，下面以本发明基于机理数据混合驱动的机器人动态误差预测方法的原理，结合实施例中对机器人动态误差进行预测的过程，对本发明提供的基于机理数据混合驱动的机器人动态误差预测方法做进一步的描述。In order to facilitate the understanding of the present invention, below, based on the principle of the robot dynamic error prediction method based on mechanism data hybrid drive of the present invention, combined with the process of predicting the robot dynamic error in the embodiment, the robot dynamics based on mechanism data hybrid drive provided by the present invention will be described below. The error prediction method is further described.
具体的，请参考图2，所述基于机理数据混合驱动的机器人动态误差预测方法包括：Specifically, please refer to Figure 2. The robot dynamic error prediction method based on mechanism data hybrid drive includes:
S1、获取控制指令指示的目标位置。S1. Obtain the target position indicated by the control instruction.
从机器人控制系统提取待执行指令指示的目标位置（坐标）。Extract the target position (coordinates) indicated by the instruction to be executed from the robot control system.
S2、基于目标位置生成运动控制参数和运动参数，所述运动参数包括位置数据。S2. Generate motion control parameters and motion parameters based on the target position, where the motion parameters include position data.
利用PID模型基于目标位置和机器人当前位置生成运动控制参数，所述运动控制参数包括电机位置、速度和力矩；将力矩输入柔性动力学模型得到运动参数，所述运动参数包括关节空间下的位置和速度及笛卡尔空间下的位置和速度；将电机位置、速度和运动参数标记生成时间后保存至输入数据集；监控所述输入数据集中的数据的生成时间，若生成时间距当前时刻的时长超过设定的时间阈值则清除相应数据。The PID model is used to generate motion control parameters based on the target position and the current position of the robot. The motion control parameters include the motor position, speed and torque; the torque is input into the flexible dynamics model to obtain the motion parameters, which include the position in joint space and Speed and position and speed in Cartesian space; save the motor position, speed and motion parameters to the input data set after marking the generation time; monitor the generation time of the data in the input data set, if the generation time from the current moment exceeds The corresponding data will be cleared according to the set time threshold.
具体的，构建由PID模型和动力学模型组成的机理模型。PID模型基于目标位置和当前位置生成伺服电机控制参数，即运动控制参数，包括电机位置、速度和力矩。将力矩输入至动力学模型进行分析，可以得到运动参数，即关节空间下的位置和速度及笛卡尔空间下的位置和速度。为了提升模型精度，本实施方式中动力学模型采用柔性动力学模型。Specifically, a mechanism model consisting of a PID model and a dynamic model is constructed. The PID model generates servo motor control parameters based on the target position and current position, that is, motion control parameters, including motor position, speed and torque. By inputting the torque into the dynamic model for analysis, the motion parameters can be obtained, namely the position and velocity in joint space and the position and velocity in Cartesian space. In order to improve the accuracy of the model, the dynamic model in this embodiment adopts a flexible dynamic model.
其中构建柔性动力学模型，当考虑关节柔性时，各个连杆的递推关系为：连杆角速度、连杆角加速度、连杆线速度、连杆线加速度、连杆质心的速度。当考虑关节柔性时，各个连杆的偏速度可表示为：连杆偏角速度、连杆偏速度、连杆质心偏速度。利用Kane方法得到的各个自由度的方程针对于连杆在弹簧变性后的转角，针对电机输出的自由度：广义惯性力、广义主动力。利用Kane方程可建立考虑关节柔性系统的动力学模型：F+F*=0A flexible dynamics model is constructed. When considering joint flexibility, the recursive relationship of each link is: link angular velocity, link angular acceleration, link linear velocity, link linear acceleration, and link center of mass velocity. When joint flexibility is taken into account, the deflection velocity of each link can be expressed as: link deflection angular velocity, link deflection velocity, and link center mass deflection velocity. The equations of each degree of freedom obtained by using Kane's method are for the rotation angle of the connecting rod after the spring deformation, and for the degrees of freedom of the motor output: generalized inertial force and generalized active force. The Kane equation can be used to establish a dynamic model considering joint flexibility: F+F*=0
所导出的方程为12个2阶微分方程，当考虑关节柔性的影响时，将微分方程组为时变微分方程，需要进行数值求解。由该方程的形式可知，该方程为时变非线性微分方程，其系数矩阵中，质量矩阵、阻尼矩阵以及刚度矩阵均为坐标的函数，使用的Newmark-β方法，进行求解，具有二阶精度。The derived equations are 12 second-order differential equations. When considering the influence of joint flexibility, the differential equations are converted into time-varying differential equations, which need to be solved numerically. It can be seen from the form of this equation that it is a time-varying nonlinear differential equation. In its coefficient matrix, the mass matrix, damping matrix and stiffness matrix are all functions of coordinates. The Newmark-β method is used to solve it with second-order accuracy. .
在本实施方式中，构建的柔性动力学模型可以描述如下：In this implementation, the constructed flexible dynamic model can be described as follows:
(1) (1)
(2) (2)
(3) (3)
是关节位置、速度和加速度的矢量，/>是关节力矩矢量，/>是惯性矩阵，是离心力和科里奥利力，/>是减速器摩擦力矩，/>是重力，/>是电机位置、速度和加速度，/>是电机转子惯量，/>是电机摩擦力矩，/>是电机驱动力矩的矢量，/>是减速比，/>为扭转刚度。 is the vector of joint position, velocity and acceleration, /> is the joint moment vector,/> is the inertia matrix, are centrifugal force and Coriolis force,/> is the friction torque of the reducer,/> It’s gravity,/> are the motor position, speed and acceleration,/> is the motor rotor inertia,/> is the motor friction torque,/> is the vector of motor driving torque,/> is the reduction ratio,/> is the torsional stiffness.
动力学模型是系统的本质，描述整个机械系统物理过程，但是现有的动力学模型仍然存在对物理过程的简化和未建模部分等问题。所以仅使用动力学模型去描述机器人动态误差是不行的，并且大部分研究者将动力学模型应用于力矩预测，鲜有研究者应用于动态误差预测。The dynamic model is the essence of the system and describes the physical process of the entire mechanical system. However, existing dynamic models still have problems such as simplification of the physical process and unmodeled parts. Therefore, it is not enough to only use the dynamic model to describe the dynamic error of the robot, and most researchers apply the dynamic model to torque prediction, but few researchers apply it to dynamic error prediction.
将30min内的电机位置、速度和运动参数作为预测所需的时序数据。Use the motor position, speed and motion parameters within 30 minutes as the time series data required for prediction.
S3、利用预先训练好的LSTM模型基于运动控制参数和运动参数及历史运动控制参数和运动参数，预测关节空间位置残差。S3. Use the pre-trained LSTM model to predict joint space position residuals based on motion control parameters and motion parameters and historical motion control parameters and motion parameters.
使用蒙特卡洛法让机器人末端跑随机轮廓，使得机器人运动轨迹充满工作空间；在机器人运动过程中，收集电机位置、速度、关节空间下的位置和速度及笛卡尔空间下的位置和速度，并从机器人控制系统获取相应的目标位置和实际位置，将所有数据保存至数据集；利用所述数据集对预先构建的LSTM模型进行训练。通过这种方法产生的训练数据更加丰富和全面，有利于提高模型的预测精度和泛化性。通过引入机理模型的计算数据作为神经网络模型的训练数据，这种非端到端的数据集形式，能够将机理模型充分的融入到神经网络模型中，让神经网络模型提取到更多的特征元素，提高模型预测精度和稳定性。Use the Monte Carlo method to let the end of the robot run a random contour so that the robot's motion trajectory fills the work space; during the robot's movement, collect the motor position, speed, position and speed in joint space, and position and speed in Cartesian space, and Obtain the corresponding target position and actual position from the robot control system and save all data to a data set; use the data set to train the pre-built LSTM model. The training data generated by this method is richer and more comprehensive, which is beneficial to improving the prediction accuracy and generalization of the model. By introducing the computational data of the mechanism model as training data for the neural network model, this non-end-to-end data set format can fully integrate the mechanism model into the neural network model, allowing the neural network model to extract more feature elements. Improve model prediction accuracy and stability.
将步骤S2中的输入数据集作为输入参数导入预先训练好的LSTM模型，得到LSTM模型预测的关节空间位置残差。Import the input data set in step S2 as input parameters into the pre-trained LSTM model to obtain the joint space position residual predicted by the LSTM model.
S4、基于运动参数与关节空间位置残差生成预测位置。S4. Generate predicted positions based on motion parameters and joint space position residuals.
从所述运动参数中提取关节空间位置；将关节空间位置与关节空间位置残差的和输出为预测位置。The joint space position is extracted from the motion parameters; the sum of the joint space position and the joint space position residual is output as the predicted position.
具体的，预测位置=关节空间位置+关节空间位置残差。Specifically, predicted position = joint space position + joint space position residual.
即，利用LSTM模型预测出的位置残差对机理模型输出的关节空间位置进行补偿，得到预测的实际位置。That is, the position residual predicted by the LSTM model is used to compensate the joint space position output by the mechanism model to obtain the predicted actual position.
S5、将所述预测位置与所述目标位置之间的差值输出为动态误差。S5. Output the difference between the predicted position and the target position as a dynamic error.
将步骤S4中得到的预测位置与控制指令指示的目标位置作差，得到动态误差。The dynamic error is obtained by making a difference between the predicted position obtained in step S4 and the target position indicated by the control instruction.
通过上述方法可以实时获取机器人的动态误差，提升了机器人的动态误差预测效率。Through the above method, the dynamic error of the robot can be obtained in real time, which improves the efficiency of dynamic error prediction of the robot.
在一些实施例中，所述基于机理数据混合驱动的机器人动态误差预测系统300可以包括多个由计算机程序段所组成的功能模块。所述基于机理数据混合驱动的机器人动态误差预测系统300中的各个程序段的计算机程序可以存储于计算机设备的存储器中，并由至少一个处理器所执行，以执行（详见图1描述）基于机理数据混合驱动的机器人动态误差预测的功能。In some embodiments, the robot dynamic error prediction system 300 based on mechanism data hybrid drive may include multiple functional modules composed of computer program segments. The computer program of each program segment in the robot dynamic error prediction system 300 based on mechanism data hybrid drive can be stored in the memory of the computer device and executed by at least one processor to execute (see Figure 1 for details) based on Function of robot dynamic error prediction driven by mechanism data hybridization.
本实施例中，所述基于机理数据混合驱动的机器人动态误差预测系统300根据其所执行的功能，可以被划分为多个功能模块，如图3所示。所述功能模块可以包括：目标获取模块310、参数获取模块320、残差预测模块330、位置预测模块340、误差预测模块350。本发明所称的模块是指一种能够被至少一个处理器所执行并且能够完成固定功能的一系列计算机程序段，其存储在存储器中。在本实施例中，关于各模块的功能将在后续的实施例中详述。In this embodiment, the robot dynamic error prediction system 300 based on mechanism data hybrid drive can be divided into multiple functional modules according to the functions it performs, as shown in Figure 3 . The functional modules may include: target acquisition module 310, parameter acquisition module 320, residual prediction module 330, position prediction module 340, and error prediction module 350. The module referred to in the present invention refers to a series of computer program segments that can be executed by at least one processor and can complete fixed functions, which are stored in the memory. In this embodiment, the functions of each module will be described in detail in subsequent embodiments.
目标获取模块310，用于获取控制指令指示的目标位置；Target acquisition module 310, used to acquire the target position indicated by the control instruction;
参数获取模块320，用于基于目标位置生成运动控制参数和运动参数，所述运动参数包括位置数据；Parameter acquisition module 320, used to generate motion control parameters and motion parameters based on the target position, where the motion parameters include position data;
残差预测模块330，用于利用预先训练好的LSTM模型基于运动控制参数和运动参数及历史运动控制参数和运动参数，预测关节空间位置残差；The residual prediction module 330 is used to predict joint space position residuals based on the motion control parameters and motion parameters and historical motion control parameters and motion parameters using the pre-trained LSTM model;
位置预测模块340，用于基于运动参数与关节空间位置残差生成预测位置；Position prediction module 340, used to generate predicted positions based on motion parameters and joint space position residuals;
误差预测模块350，用于将所述预测位置与所述目标位置之间的差值输出为动态误差；Error prediction module 350, configured to output the difference between the predicted position and the target position as a dynamic error;
LSTM模型的训练方法包括：Training methods for LSTM models include:
使用蒙特卡洛法让机器人末端跑随机轮廓，使得机器人运动轨迹充满工作空间；Use the Monte Carlo method to let the end of the robot run a random contour so that the robot's motion trajectory fills the work space;
在机器人运动过程中，收集电机位置、速度、关节空间下的位置和速度及笛卡尔空间下的位置和速度，并从机器人控制系统获取相应的目标位置和实际位置，将所有数据保存至数据集；During the movement of the robot, the motor position, speed, position and speed in joint space and position and speed in Cartesian space are collected, and the corresponding target position and actual position are obtained from the robot control system, and all data are saved to the data set. ;
利用所述数据集对预先构建的LSTM模型进行训练。The pre-built LSTM model is trained using the dataset.
可选地，作为本发明一个实施例，参数获取模块包括：Optionally, as an embodiment of the present invention, the parameter acquisition module includes:
第一生成单元，用于利用PID模型基于目标位置和机器人当前位置生成运动控制参数，所述运动控制参数包括电机位置、速度和力矩；A first generation unit configured to use the PID model to generate motion control parameters based on the target position and the current position of the robot, where the motion control parameters include motor position, speed and torque;
第二生成单元，用于将力矩输入柔性动力学模型得到运动参数，所述运动参数包括关节空间下的位置和速度及笛卡尔空间下的位置和速度；The second generation unit is used to input torque into the flexible dynamics model to obtain motion parameters. The motion parameters include position and velocity in joint space and position and velocity in Cartesian space;
参数保存单元，用于将电机位置、速度和运动参数标记生成时间后保存至输入数据集；The parameter saving unit is used to save the motor position, speed and motion parameters to the input data set after marking the generation time;
定期清理单元，用于监控所述输入数据集中的数据的生成时间，若生成时间距当前时刻的时长超过设定的时间阈值则清除相应数据。A regular cleaning unit is used to monitor the generation time of the data in the input data set, and clear the corresponding data if the time between the generation time and the current time exceeds a set time threshold.
可选地，作为本发明一个实施例，残差预测模块包括：Optionally, as an embodiment of the present invention, the residual prediction module includes:
数据输入单元，用于将所述输入数据集作为输入参数导入预先训练好的LSTM模型，得到LSTM模型预测的关节空间位置残差。A data input unit is used to import the input data set as an input parameter into a pre-trained LSTM model to obtain the joint space position residual predicted by the LSTM model.
可选地，作为本发明一个实施例，位置预测模块包括：Optionally, as an embodiment of the present invention, the location prediction module includes:
参数提取单元，用于从所述运动参数中提取关节空间位置；A parameter extraction unit, used to extract joint space positions from the motion parameters;
残差补偿单元，用于将关节空间位置与关节空间位置残差的和输出为预测位置。The residual compensation unit is used to output the sum of the joint space position and the joint space position residual as the predicted position.
图4为本发明实施例提供的一种终端400的结构示意图，该终端400可以用于执行本发明实施例提供的基于机理数据混合驱动的机器人动态误差预测方法。Figure 4 is a schematic structural diagram of a terminal 400 provided by an embodiment of the present invention. The terminal 400 can be used to execute the robot dynamic error prediction method based on mechanism data hybrid driving provided by the embodiment of the present invention.
其中，该终端400可以包括：处理器410、存储器420及通信模块430。这些组件通过一条或多条总线进行通信，本领域技术人员可以理解，图中示出的服务器的结构并不构成对本发明的限定，它既可以是总线形结构，也可以是星型结构，还可以包括比图示更多或更少的部件，或者组合某些部件，或者不同的部件布置。The terminal 400 may include: a processor 410, a memory 420, and a communication module 430. These components communicate through one or more buses. Those skilled in the art can understand that the structure of the server shown in the figure does not limit the invention. It can be a bus structure, a star structure, or More or fewer components may be included than shown, or certain components may be combined, or may be arranged differently.
其中，该存储器420可以用于存储处理器410的执行指令，存储器420可以由任何类型的易失性或非易失性存储终端或者它们的组合实现，如静态随机存取存储器（SRAM），电可擦除可编程只读存储器（EEPROM），可擦除可编程只读存储器（EPROM），可编程只读存储器（PROM），只读存储器（ROM），磁存储器，快闪存储器，磁盘或光盘。当存储器420中的执行指令由处理器410执行时，使得终端400能够执行以下上述方法实施例中的部分或全部步骤。Among them, the memory 420 can be used to store execution instructions of the processor 410. The memory 420 can be implemented by any type of volatile or non-volatile storage terminal or their combination, such as static random access memory (SRAM), electronic Erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic or optical disk . When the execution instructions in the memory 420 are executed by the processor 410, the terminal 400 is enabled to perform some or all of the steps in the following method embodiments.
处理器410为存储终端的控制中心，利用各种接口和线路连接整个电子终端的各个部分，通过运行或执行存储在存储器420内的软件程序和/或模块，以及调用存储在存储器内的数据，以执行电子终端的各种功能和/或处理数据。所述处理器可以由集成电路(Integrated Circuit，简称IC) 组成，例如可以由单颗封装的IC 所组成，也可以由连接多颗相同功能或不同功能的封装IC而组成。举例来说，处理器410可以仅包括中央处理器(Central Processing Unit，简称CPU)。在本发明实施方式中，CPU可以是单运算核心，也可以包括多运算核心。The processor 410 is the control center of the storage terminal, using various interfaces and lines to connect various parts of the entire electronic terminal, by running or executing software programs and/or modules stored in the memory 420, and calling data stored in the memory, To perform various functions of the electronic terminal and/or process data. The processor may be composed of an integrated circuit (IC for short), for example, it may be composed of a single packaged IC, or it may be composed of multiple packaged ICs connected with the same function or different functions. For example, the processor 410 may only include a central processing unit (Central Processing Unit, CPU for short). In the embodiment of the present invention, the CPU may be a single computing core or may include multiple computing cores.
通信模块430，用于建立通信信道，从而使所述存储终端可以与其它终端进行通信。接收其他终端发送的用户数据或者向其他终端发送用户数据。The communication module 430 is used to establish a communication channel so that the storage terminal can communicate with other terminals. Receive user data sent by other terminals or send user data to other terminals.
本发明还提供一种计算机存储介质，其中，该计算机存储介质可存储有程序，该程序执行时可包括本发明提供的各实施例中的部分或全部步骤。所述的存储介质可为磁碟、光盘、只读存储记忆体（英文：read-only memory，简称：ROM）或随机存储记忆体（英文：random access memory，简称：RAM）等。The present invention also provides a computer storage medium, wherein the computer storage medium can store a program, and when executed, the program can include some or all of the steps in the embodiments provided by the present invention. The storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.
因此，本发明通过对PID模型对伺服电机的运动控制参数进行跟踪预测得到一个关节空间位置，再利用LSTM模型对关节空间位置进行误差预测，进而得到一个预测的实际位置，将预测的实际位置与目标位置进行比对即可得到动态误差，本发明大大降低了动态误差预测的计算量，提升了动态误差预测的效率，本实施例所能达到的技术效果可以参见上文中的描述，此处不再赘述。Therefore, the present invention obtains a joint space position by tracking and predicting the motion control parameters of the servo motor using the PID model, and then uses the LSTM model to perform error prediction on the joint space position, thereby obtaining a predicted actual position, and comparing the predicted actual position with The dynamic error can be obtained by comparing the target position. The present invention greatly reduces the calculation amount of dynamic error prediction and improves the efficiency of dynamic error prediction. The technical effects achieved by this embodiment can be found in the above description, which will not be discussed here. Again.
本领域的技术人员可以清楚地了解到本发明实施例中的技术可借助软件加必需的通用硬件平台的方式来实现。基于这样的理解，本发明实施例中的技术方案本质上或者说对现有技术做出贡献的部分可以以软件产品的形式体现出来，该计算机软件产品存储在一个存储介质中如U盘、移动硬盘、只读存储器（ROM，Read-Only Memory）、随机存取存储器（RAM，Random Access Memory）、磁碟或者光盘等各种可以存储程序代码的介质，包括若干指令用以使得一台计算机终端（可以是个人计算机，服务器，或者第二终端、网络终端等）执行本发明各个实施例所述方法的全部或部分步骤。Those skilled in the art can clearly understand that the technology in the embodiments of the present invention can be implemented by means of software plus the necessary general hardware platform. Based on this understanding, the technical solutions in the embodiments of the present invention can be embodied in the form of software products in essence or in part that contribute to the existing technology. The computer software products are stored in a storage medium such as a USB flash drive or mobile phone. Hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disk and other media that can store program code, including a number of instructions to make a computer terminal (It can be a personal computer, a server, or a second terminal, a network terminal, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention.
本说明书中各个实施例之间相同相似的部分互相参见即可。尤其，对于终端实施例而言，由于其基本相似于方法实施例，所以描述的比较简单，相关之处参见方法实施例中的说明即可。The same and similar parts among the various embodiments in this specification can be referred to each other. In particular, for the terminal embodiment, since it is basically similar to the method embodiment, the description is relatively simple. For relevant details, please refer to the description in the method embodiment.
在本发明所提供的几个实施例中，应该理解到，所揭露的系统和方法，可以通过其它的方式实现。例如，以上所描述的系统实施例仅仅是示意性的，例如，所述模块的划分，仅仅为一种逻辑功能划分，实际实现时可以有另外的划分方式，例如多个模块或组件可以结合或者可以集成到另一个系统，或一些特征可以忽略，或不执行。另一点，所显示或讨论的相互之间的耦合或直接耦合或通信连接可以是通过一些接口，系统或模块的间接耦合或通信连接，可以是电性，机械或其它的形式。In the several embodiments provided by the present invention, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are only illustrative. For example, the division of modules is only a logical function division. In actual implementation, there may be other division methods. For example, multiple modules or components may be combined or can be integrated into another system, or some features can be ignored, or not implemented. On the other hand, the coupling or direct coupling or communication connection between each other shown or discussed may be through some interfaces, indirect coupling or communication connection of the system or module, which may be in electrical, mechanical or other forms.
所述作为分离部件说明的模块可以是或者也可以不是物理上分开的，作为模块显示的部件可以是或者也可以不是物理模块，即可以位于一个地方，或者也可以分布到多个网络模块上。可以根据实际的需要选择其中的部分或者全部模块来实现本实施例方案的目的。The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical modules, that is, they may be located in one place, or they may be distributed to multiple network modules. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment.
另外，在本发明各个实施例中的各功能模块可以集成在一个处理模块中，也可以是各个模块单独物理存在，也可以两个或两个以上模块集成在一个模块中。In addition, each functional module in various embodiments of the present invention can be integrated into one processing module, or each module can exist physically alone, or two or more modules can be integrated into one module.
尽管通过参考附图并结合优选实施例的方式对本发明进行了详细描述，但本发明并不限于此。在不脱离本发明的精神和实质的前提下，本领域普通技术人员可以对本发明的实施例进行各种等效的修改或替换，而这些修改或替换都应在本发明的涵盖范围内/任何熟悉本技术领域的技术人员在本发明揭露的技术范围内，可轻易想到变化或替换，都应涵盖在本发明的保护范围之内。因此，本发明的保护范围应所述以权利要求的保护范围为准。Although the present invention has been described in detail with reference to the accompanying drawings in conjunction with preferred embodiments, the present invention is not limited thereto. Without departing from the spirit and essence of the invention, those of ordinary skill in the art can make various equivalent modifications or substitutions to the embodiments of the invention, and these modifications or substitutions should be within the scope of the invention/any Those skilled in the art can easily think of changes or substitutions within the technical scope disclosed in the present invention, and they should all be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention should be determined by the protection scope of the claims.
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311346486.XACN117086886B (en) | 2023-10-18 | 2023-10-18 | Robot dynamic error prediction method and system based on mechanism data hybrid driving |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311346486.XACN117086886B (en) | 2023-10-18 | 2023-10-18 | Robot dynamic error prediction method and system based on mechanism data hybrid driving |
| Publication Number | Publication Date |
|---|---|
| CN117086886A CN117086886A (en) | 2023-11-21 |
| CN117086886Btrue CN117086886B (en) | 2023-12-22 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311346486.XAActiveCN117086886B (en) | 2023-10-18 | 2023-10-18 | Robot dynamic error prediction method and system based on mechanism data hybrid driving |
| Country | Link |
|---|---|
| CN (1) | CN117086886B (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117272846A (en)* | 2023-11-22 | 2023-12-22 | 山东建筑大学 | A dynamic response prediction algorithm for a two-degree-of-freedom rotation and pitching motion mechanism |
| CN118404592B (en)* | 2024-07-04 | 2024-08-23 | 北京航空航天大学 | Robot control method and device and robot |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2021164276A1 (en)* | 2020-07-31 | 2021-08-26 | 平安科技（深圳）有限公司 | Target tracking method and apparatus, computer device, and storage medium |
| CN114131611A (en)* | 2021-12-16 | 2022-03-04 | 华中科技大学 | Joint error offline compensation method, system and terminal for robot gravity pose decomposition |
| CN114789454A (en)* | 2022-06-24 | 2022-07-26 | 浙江大学 | Robot digital twin trajectory completion method based on LSTM and inverse kinematics |
| CN114995132A (en)* | 2022-05-26 | 2022-09-02 | 哈尔滨工业大学（深圳） | Multi-arm spacecraft model prediction control method, equipment and medium based on Gaussian mixture process |
| CN115284276A (en)* | 2022-03-28 | 2022-11-04 | 江南大学 | Robot joint torque control method based on long short-term memory network |
| CN115577755A (en)* | 2022-11-28 | 2023-01-06 | 中环服(成都)科技有限公司 | Robot posture correction method, apparatus, computer device, and storage medium |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2021164276A1 (en)* | 2020-07-31 | 2021-08-26 | 平安科技（深圳）有限公司 | Target tracking method and apparatus, computer device, and storage medium |
| CN114131611A (en)* | 2021-12-16 | 2022-03-04 | 华中科技大学 | Joint error offline compensation method, system and terminal for robot gravity pose decomposition |
| CN115284276A (en)* | 2022-03-28 | 2022-11-04 | 江南大学 | Robot joint torque control method based on long short-term memory network |
| CN114995132A (en)* | 2022-05-26 | 2022-09-02 | 哈尔滨工业大学（深圳） | Multi-arm spacecraft model prediction control method, equipment and medium based on Gaussian mixture process |
| CN114789454A (en)* | 2022-06-24 | 2022-07-26 | 浙江大学 | Robot digital twin trajectory completion method based on LSTM and inverse kinematics |
| CN115577755A (en)* | 2022-11-28 | 2023-01-06 | 中环服(成都)科技有限公司 | Robot posture correction method, apparatus, computer device, and storage medium |
| Title |
|---|
| 周恒旭 ; 何志敏 ; 周燕 ; .基于长短期记忆网络的机械臂逆运动学解.机械与电子.2020,(06),全文.* |
| Publication number | Publication date |
|---|---|
| CN117086886A (en) | 2023-11-21 |
| Publication | Publication Date | Title |
|---|---|---|
| CN117086886B (en) | Robot dynamic error prediction method and system based on mechanism data hybrid driving | |
| CN108015763B (en) | An anti-noise interference redundant manipulator path planning method | |
| US8140189B2 (en) | Apparatus and method for computing operational-space physical quantity | |
| EP2188685B1 (en) | Inverse kinematics | |
| Urrea et al. | Design and validation of a dynamic parameter identification model for industrial manipulator robots | |
| CN111185909B (en) | Robot operation condition acquisition method and device, robot and storage medium | |
| Huang et al. | A FAT-based adaptive controller for robot manipulators without regressor matrix: theory and experiments | |
| CN116079747A (en) | Robot cross-body control method, system, computer equipment and storage medium | |
| CN114967465B (en) | Trajectory planning method, device, electronic device and storage medium | |
| Wang et al. | A UKF-based predictable SVR learning controller for biped walking | |
| CN111168682A (en) | Limited time convergence sliding mode control method of parallel robot combined with robust precise differentiator for conveying | |
| Yang et al. | Robot kinematics and dynamics modeling | |
| CN118605135B (en) | Robot pose control method for nut feeding sleeve | |
| CN117245640A (en) | Robot control methods, devices, equipment and storage media | |
| CN114967472A (en) | A UAV Trajectory Tracking State Compensation Depth Deterministic Policy Gradient Control Method | |
| CN116141314B (en) | Robot dynamic parameter identification method and system based on projective geometry algebra | |
| CN113341724A (en) | Asynchronous sliding mode control method of multi-mode mechanical arm | |
| CN114603552B (en) | Robot simulation method, electronic equipment and storage medium | |
| Axelsson | On Sensor Fusion Applied to Industrial Manipulators | |
| CN119369424B (en) | Parameter self-learning robot compliance control method, system and device | |
| CN112632803A (en) | Tracking control method and device, electronic equipment and storage medium | |
| Yaren et al. | Advanced SOA-NMPC controller design minimising real-time computational burden for dynamic obstacle avoidance in robotic manipulators | |
| CN114952825B (en) | Method, equipment and storage medium for realizing Lagrange system group consensus | |
| Ayten et al. | Optimum Trajectory Planning for Industrial Robots through Inverse Dynamics. | |
| CN113681551B (en) | Method and device for identifying dynamic parameters of SCARA robot |
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |