The present technology, based on laser and visual SLAM navigation and positioning, does not apply to the event of a fire in a room where there is a large amount of smoke, for its increasingly obvious defects. In addition, traditional track deduction technology based on photoelectric encoder has accumulated error, and noise disturbance exists in the INS inertial navigation measurement technology, and the UWB positioning technology is vulnerable to NLOS disturbance caused by site occlusion. To solve the problem of accurate positioning of Autonomous Mobile fire-fighting robot in smoke scenes, the system design is optimized by adopting the following methods: using IMU-assisted residual chi-square criterion to detect whether there is NLOS in UWB, introducing IMU instantaneous compensation positioning data and adopting Chan algorithm fitting of the second multiplication to ensure the stability and accuracy of UWB data; meanwhile, a tight combination model of navigation and positioning is designed: via the improved Kalman filter algorithm, fused with the magnetic encoder track estimation pose, UWB absolute and IMU heading angle pose to realize the accurate positioning of the fire robot in the smoke scene. Finally, the fusion simulation model and algorithm are verified by MATLAB, as it shows, the method has an average positioning accuracy of 98.63% in the X-axis direction, 99.52% in the Y-axis direction, and 97.24% in the heading angle, which solves the inherent physical defects of a single positioning sensor and serves as a reliable and accurate solution for indoor robot positioning.

基于激光和视觉SLAM导航和定位的本技术由于其日益明显的缺陷而不适用于存在大量烟雾的房间中发生的火灾。另外，传统的基于光电编码器的航迹推导技术存在误差累积，惯性惯性导航测量技术中存在噪声干扰，超宽带定位技术易受场地遮挡引起的非视距干扰的影响。为解决自主移动消防机器人在烟雾场景中的准确定位问题，采用以下方法对系统设计进行了优化：利用IMU辅助的残差卡方判据检测UWB中是否存在NLOS；引入IMU瞬时补偿定位数据，并采用二次乘法的Chan算法拟合，以保证UWB数据的稳定性和准确性；同时，设计了导航与定位的紧密结合模型：通过改进的卡尔曼滤波算法，结合磁编码器轨迹估计姿态，UWB绝对值和IMU航向角姿态，实现了消防机器人在烟雾场景中的精确定位。最后，通过MATLAB对融合仿真模型和算法进行了验证，表明该方法在X轴方向的平均定位精度为98.63％，在Y轴方向的平均定位精度为99.52％，航向角的平均定位精度为97.24％。解决了单个定位传感器固有的物理缺陷，并为室内机器人定位提供了可靠而准确的解决方案。