【摘要】 床椅一体化机器人研究中，传统的多传感器融合定位算法无法有效处理时变噪声而导致定位精度较低。据此基于QR二维码和IMU惯导融合定位进行研究，提出一种自适应模糊推理无迹卡尔曼滤波算法，可实时调节无迹卡尔曼滤波的噪声，以实现床椅模块的精准定位。通过模拟传感器数据仿真，并进行机器人小车实验验证。结果表明：在传感器定位误差较大的情况下，该算法仍能得到最优的定位结果，相较与传统无迹卡尔曼滤波算法，均方误差减少了38.94%,最大误差减少了21.6%。更多还原

【Abstract】 In the research of bed and chair integrated robots, the traditional multi-sensor fusion localization algorithm cannot effectively deal with time-varying noise, resulting in low localization accuracy. Based on the research of QR code and IMU inertial navigation fusion positioning, an adaptive fuzzy inference unscented Kalman filter algorithm is proposed, which could adjust the noise of unscented Kalman filter in real time to realize accurate positioning of the bed and chair module. The results from sensor data simulation and robot car experiment show that the proposed algorithm can still obtain the optimal positioning results when the sensor positioning error is large. Compared with the traditional unscented Kalman filter algorithm, the mean square error is reduced by 38.94%, and the maximum error is reduced by 21.6%.更多还原