自动化生物操作算法理论与软件集成

摘要

在生命科学研究中,生物操作主要存在操作者训练时间长、操作成功率和效率低、操作缺乏一致性和重复性等问题,自动化生物操作可以解决这些问题。将多个商用设备集成在一起,利用视觉系统与机械手系统的映射矩阵建立误差校正方程,采用局部递归和全局优化标定的方法减小目标点的定位误差。在此基础上,利用卡尔曼滤波器建立系统控制方程,控制操作探针精确定位。最后,用程序实现了算法理论,并将该算法与其他设备驱动模块集成在一起,建立了柔性的、用户界面友好的软件系统。实验结果表明：建立的算法理论和软件系统,可以自动控制操作探针精确运动,运动精度在1个像素范围内。%For the study of life sciences,bio-manipulation faces such challenges as long time for operator training,low success rate,low efficiency,operation inconsistency and repetition.We believe that automatic bio-manipulation can solve these challenges.We integrate several commercial devices and use the matrix that maps the visual system with the robotic system to derive the error correction equation.We also use the local recurrence and global optimization calibration method to reduce the target point′s positioning error.On this basis,using the Kalman filter,we establish the systems′ control equations to control the precise positioning of the bio-manipulation probe.Finally,we implement the algorithm theory and integrate the algorithm with the drive modules of other devices,thus establishing a flexible and user-friendly software system.The experimental results show that the algorithm theory and software systems thus integrated can automatically control the probe′s accurate motion,the motion accuracy being within one pixel.