Abstract The strength of Generalized Pseudo Bayesian (GPB) algorithms is exploited in the presented study to enhance the target tracking precision, effective model approximation and rapid convergence of multimodel maneuvering object tracking. The GPB methods are considered to be suitable for approximating systems whose dynamics follow discrete-time and fixed state Markov process. Underwater maneuvering target tracking problems are usually solved with nonlinear Bayesian algorithms, in which kinetics of object are associated with passive bearings using state-space modeling. Here accuracy and convergence of GPB methods based on Interacting Multiple Model Extended Kalman Filter (IMMEKF), Interacting Multiple Model Extended Kalman Smoother (IMMEKS), Interacting Multiple Model Unscented Kalman Filter (IMMUKF) and Interacting Multiple Model Unscented Kalman Smoother (IMMUKS) are efficiently analyzed for tracking of multimodel maneuvering target in complex ocean environment. Application of these algorithms is systematically presented for estimating the real-time state of a maneuvering object that follows a coordinated turn trajectory. Performance analysis of IMM Kalman filters and smoothers is done with variations in the standard deviation of white Gaussian measurement noise by following Bearings Only Tracking (BOT) phenomena. Least Mean Square Error (MSE) between approximated and the real position of maneuvering target in rectangular coordinates is calculated for analyzing the performance of filtering and smoothing techniques. Simulation results of the Monte Carlo runs validate the effectiveness of IMMEKS and IMMUKS over IMMEKF and IMMUKF for scenario of given framework.

中文翻译：

---

多模型水下机动目标跟踪的广义伪贝叶斯算法

摘要 本研究利用广义伪贝叶斯（GPB）算法的优势来提高多模型机动目标跟踪的目标跟踪精度、有效模型逼近和快速收敛。GPB 方法被认为适用于逼近其动力学遵循离散时间和固定状态马尔可夫过程的系统。水下机动目标跟踪问题通常用非线性贝叶斯算法解决，其中对象的动力学与使用状态空间建模的被动轴承相关联。这里基于交互多模型扩展卡尔曼滤波器 (IMMEKF)、交互多模型扩展卡尔曼平滑器 (IMMEKS) 的 GPB 方法的准确性和收敛性，有效地分析了交互多模型无迹卡尔曼滤波器 (IMMUKF) 和交互多模型无迹卡尔曼平滑器 (IMMUKS)，以在复杂海洋环境中跟踪多模型机动目标。系统地介绍了这些算法的应用，用于估计遵循协调转弯轨迹的机动物体的实时状态。IMM 卡尔曼滤波器和平滑器的性能分析是通过遵循仅轴承跟踪 (BOT) 现象，在高斯白测量噪声的标准偏差变化的情况下完成的。为了分析滤波和平滑技术的性能，计算了机动目标在直角坐标中的近似位置和实际位置之间的最小均方误差（MSE）。