中文摘要：

针对未知环境中,机器人同步定位与地图构建（SLAM）时,系统的统计特性发生突变问题,提出了一种基于非线性交互式多模型（IMM）的SLAM算法。该算法的主要思想是：用多个非线性高斯模型近似非线性非高斯模型;每个模型都采用扩展卡尔曼滤波（EKF）对非线性系统线性化;在每一步采用IMM方法获得融合估计值;从而演化机器人的SLAM.Monte Carlo仿真结果表明,在过程噪声均方根误差、量测噪声均方根误差和两者噪声均方根误差都发生变化的情况下,与EKF-SLAM算法和快速SLAM算法相比,该算法具有更好的估计精度。

英文摘要：

To deal with the problem concerning the statistical property mutation of a system in the unknown environment, a simultaneous localization and mapping （SLAM） algorithm based On non-linear interacting multiple model （IMM） was proposed. It is main idea of the algorithm that non-linear Gaussian model is used to approximate non-linear and non-Gaussian model ; the extended Kalman filter （EKF） algorithm is employed to linearize the non-linear system for each model; the non-linear IMM algorithm is used to get fusion estimated value in each step; SLAM of the robot can be achieved. The Monte Carlo simulation results show that at changing RMSEs of noises of process, observation and both mentioned, the proposed algorithm has better estimate precision compared with EKF-SLAM algorithm and Fast-SLAM algorithm.