【摘要】 为了准确判断跟随过程中的障碍物距离，提高四足机器人在复杂环境下的跟随能力，提出了一种基于卡尔曼滤波(KF)和多模态信息融合的四足机器人避障方法。首先，对传感器采集到的数据进行KF,降低数据中包含的噪声；然后，通过融合算法处理滤波后的数据；最后，将融合后的数据与实际的距离数据进行比较。仿真分析结果表明：传感器采集到的数据经过滤波和融合算法处理后，同真实距离的最大差值为0.5 cm,比融合前的数据更接近于真实距离值。所提出的四足机器人避障方法对含有噪声的数据有良好的处理作用，处理后的数据波动更小，能满足四足机器人的检测要求。更多还原

【Abstract】 In order to accurately judge the obstacle distance in following process and improve the following ability of quadruped robot in complex environment, based on Kalman filtering(KF)and multimodal information fusion, an obstacle avoidance method for quadruped robot is proposed.Firstly, KF is applied to the data collected by the sensor to reduce the noise contained in data.Then, the filtered data are processed by fusion algorithm.Finally, the fused data is compared with the actual distance data.Simulation analysis results show that the maximum difference between the data processed by the filtering and fusion algorithm and the real distance is 0.5 cm, which is closer to the real distance than the data before fusion.The proposed obstacle avoidance method of quadruped robot has good processing effect on the data with noise, and the processed data has less fluctuation, which can meet the detection requirements of quadruped robot.更多还原