摘要
精密位姿测量是多自由度电机实现精密运动控制的前提,现有的多自由度电机位姿测量方案多采用编码器、惯性测量单元等外部传感器,存在结构复杂、可靠性低等不足.该文提出一种基于冗余磁场信息的三自由度位姿测量方法,通过布置冗余磁传感器阵列检测电机自身转子永磁体所产生的磁场,并设计由所测磁感应强度信息快速解算电机多自由度位姿的测量算法和永磁体磁场模型参数的标定算法,实现电机位姿的实时精密测量.该方法基于电机自身磁场进行位姿测量,不易受外部环境如振动噪声、灰尘污染等影响,具备高精密性、高可靠性等特点,且永磁体磁场既为电机动力来源又为位姿测量基准,因此能够有效提高电机系统集成度与结构紧凑性.仿真和实验结果表明,该方法能够快速准确测量三自由度旋转电机位姿,三旋转轴位姿测量的均方根误差分别为0.031、0.025和0.056 rad.
Abstract
[Objective]Accurate pose measurement is crucial for precise motion control in multi-degree-of-freedom motors.Traditional methods for pose measurement often rely on external sensors such as encoders,inertial measurement units,and optical sensors,which can be complex and less reliable.This paper introduces a three-degree-of-freedom pose measurement technique that uses redundant magnetic field information.By utilizing the motor's intrinsic magnetic field,this approach aims to simplify the system design and improve its robustness.[Methods]The proposed method employs an array of redundant magnetic sensors to detect the magnetic field generated by the motor's rotor permanent magnets.A measurement algorithm is developed to quickly calculate the motor's multi-degree-of-freedom pose from the detected magnetic flux density data.In addition,a calibration algorithm is introduced to ensure real-time,precise measurement of the motor's pose by accurately aligning the magnetic field model parameters with the actual magnetic field generated by the motor.This calibration process leverages simulation data and a linear magnetic field model to achieve accurate alignment.To further enhance accuracy,spatial transformation matrices are used to construct a mathematical model of the magnetic sensor signals.This allows the system to effectively map the detected magnetic field information to the motor's pose.The Gauss-Newton method is then employed to solve the overdetermined equations arising from the redundant information provided by multiple sensors,and multi-degree-of-freedom pose measurement and position parameter calibration are realized.[Results]A significant advantage of this method is its reliance on the motor's intrinsic magnetic field for pose measurement.This intrinsic measurement approach simplifies the overall system structure and enhances its robustness.By measuring the pose based on the motor's magnetic field,the system can simultaneously determine the installation position of the sensor array and calibrate the parameter offsets of the permanent magnets.This dual functionality is achieved without any special requirements for the sensor arrangement,providing versatility and adaptability to different motor configurations.In addition,this method is less susceptible to external environmental factors such as vibrational noise and dust contamination.Traditional external sensors can be significantly affected by such factors,resulting in measurement errors and reduced reliability.In contrast,the proposed method provides a more reliable and precise measurement system that can be easily integrated into various motor configurations and withstand demanding environmental conditions.[Conclusions]To validate the proposed method,experiments were conducted on a self-built physical system that included the Real-Time eXtension(RTX)operating system,a Windows software platform,a magnetic sensor array,and an FPGA hardware platform.These experiments covered permanent magnet parameter offset calibration,stability tests,and comparative performance evaluations.The results from both simulations and experiments show that the proposed method can rapidly and accurately measure the pose of a three-degree-of-freedom rotary motor.The root-mean-square errors in orientation measurement for the three rotational axes were 0.031,0.025,and 0.056 rad,respectively,with a resolution of approximately 1.77 × 10-4 rad.These results confirm the method's capability to provide high-precision pose measurements,positioning it as a promising solution for advanced motion control applications in multi-degree-of-freedom motors.Both simulation and experimental results validate the high-precision pose measurement capability of the proposed method,highlighting its potential for such advanced motion control applications.
维普
万方数据