Model Set Adaptation Research Articles

A variable structure multi‐model maneuvering target tracking algorithm based on Monte Carlo learning

AbstractA well‐liked maneuvering target tracking algorithm is a variable structure multi‐model (VSMM). One of the crucial elements determining the tracking effect is the successful model set adaptation (MSA). The ability to further enhance tracking accuracy for the conventional VSMM method is constrained by the absence of a mechanism to thoroughly utilise observation and tracking data to optimise MSA. We incorporate the Reinforcement learning (RL) approach into the MSA procedure to address this issue and provide a VSMM algorithm based on Monte Carlo (MC) learning. To formulate the challenge of optimising the number of effective models as a RL problem, we first used the prediction error, the number of effective model sets, and tracking accuracy to build the models of the appropriate state space, decision space, and reward. The number of efficient models was optimised using MC learning, and the entire VSMM algorithm was then created. The proposed approach was compared to the simulated experiment's five maneuvering target tracking algorithms. The outcomes demonstrate that the suggested algorithm has a lower computation scale and accurate tracking accuracy.

A novel variable structure multi-model approach based on error-ambiguity decomposition

Model Set Adaptation (MSA) plays a key role in the Variable Structure Multi-Model tracking approach (VSMM). In this paper, the Error-Ambiguity Decomposition (EAD) principle is adopted to derive the EAD-MSA criterion that is optimal in the sense of minimizing the square error between the estimate and the truth. Consequently, the EAD Variable Structure first-order General Pseudo Bayesian (EAD-VSGPB1) algorithm and the EAD Variable Structure Interacting Multiple Model (EAD-VSIMM) algorithm are constructed. The proposed algorithms are tested in two groups of maneuvering target tracking scenarios under different modes and observation error conditions. The simulation results demonstrate the effectiveness of the EAD-VSMM approach and show that, compared to some existing multi-model algorithms, the proposed EAD-VSMM algorithms achieve more robust and accurate tracking results.

Maneuvering multi-target tracking based on variable structure multiple model GMCPHD filter

The traditional multiple model cardinalized probability hypothesis density (MMCPHD) filter uses a fixed model set for all targets. It is clearly inefficient and may cause the decrease of performance in the situations where the size of the model set is large and the scene is complicated. This paper presents a variable structure multiple model (VSMM) version of the Gaussian mixture cardinalized probability hypothesis density (GMCPHD) filter based on the expected mode augmentation (EMA) for the multiple maneuvering target tracking. The GMCPHD filter which is suitable for the VSMM is derived, thus different model sets can be used for different targets. Then EMA algorithm which is an efficient model set adaptation method is introduced and applied to the VSMM version of the GMCPHD (VSMM-GMCPHD) filter. This method costs less computational time and has better estimate precision than that of the traditional multiple model version of GMCPHD (MM-GMCPHD) filter. The gating technique is utilized to further improve the computational efficiency. Simulation results verify the performance of the proposed methods. The estimate precision of VSMM-GMCPHD filters with and without the gate are almost the same, however the former one is more cost-effective. Both of the two VSMM-GMCPHD filters outperform the MM-GMCPHD filter.

An Adaptive Interacting Multiple Model for Vehicle Target Tracking Method

In the field of traffic safety vehicle target tracking prediction as the background, this paper proposes an adaptive interacting multiple model tracking algorithm. According to the field of transportation vehicle movement state characteristics, based on the uniform (CV) and uniformly accelerated motion (CA) model, based on new information structure model of motion of the likelihood function, online adaptive adjustment model of the noise variance and the Markov matrix, realization of maneuvering target movement model and model set adaptation, not only improved IMM algorithm for tracking accuracy, and enhances the real-time performance of system, the simulation results show that, the algorithm for tracking precision compared to the traditional IMM method has bigger improvement.

Best Model Augmentation for Variable-Structure Multiple-Model Estimation

A new approach, referred to as best model augmentation (BMA), for variable-structure multiple-model (VSMM) estimation is presented. Here the original set of models is augmented by a variable set of models intended to best match the unknown true mode. Based on the Kullback-Leiber (KL) information, two versions of the criterion serving as a metric of the closeness between candidate models and the true mode are derived in the space of states and measurements, respectively. The model set adaptation (MSA) in BMA turns out to be an online optimization problem based on the KL criterion, which can be solved easily. The performance of the proposed BMA approach is evaluated via several scenarios for maneuvering target tracking. Simulation results demonstrated the effectiveness of BMA compared with the interacting multiple-model (IMM) algorithm and the expected-mode augmentation (EMA) algorithm.

A novel interacting multiple model algorithm

For maneuvering target tracking, the interacting multiple model (IMM) algorithm employs a fixed model set. The performance of this algorithm depends on the model set adopted. The result of using too many models is as bad as the case of too few models. Therefore, a variable structure IMM (VSIMM) was presented and applied to ground target tracking. This algorithm improves performance and reduces computational load with using auxiliary information. But it is difficult to extend the VSIMM to other scenario (for example, aerial target), where there is not auxiliary information such as a map. A novel interacting multiple model (Novel-IMM) algorithm was presented to solve the problem of model set adaptation without auxiliary information. The Novel-IMM algorithm consists of N independent IMM filters operating in parallel, and each independent IMM filter also consists of multiple sub-filters, which operate interactively. In every time index, only one IMM output of a certain model set is used; but for a long time, the algorithm will alternatively choose an output of the model set to be the optimum final output. The Novel-IMM approach was illustrated in detail with an aerial complex maneuvering target tracking example.

Model-set adaptation using a fuzzy Kalman filter

In this paper, a fuzzy Kalman filter (KF) is proposed to combat the model-set adaptation problem of multiple model estimation. The fuzzy KF is found to be able to more exactly extract dynamic information of target maneuvers. It uses a set of fuzzy rules to adaptively control the process noise covariance of the KF and that makes it more suitable for real radar tracking. The proposed fuzzy Kalman filter is then incorporated into an interacting multiple model (IMM) algorithm, hence, a fuzzy IMM (FIMM) algorithm is obtained. The performance of the FIMM algorithm is compared with that of an adaptive IMM (AIMM) algorithm using real radar data. Simulation result shows that the FIMM algorithm greatly outperforms the AIMM algorithm in terms of both the root mean square prediction error and the number of track loss.

Multiple-model estimation with variable structure. II. Model-set adaptation

For pt.I see ibid., vol.41, p.487-93 (1996). An important, natural, and practical approach to variable-structure multiple-model (VSMM) estimation is the recursive adaptive model-set (RAMS) approach. The key to this approach is model-set adaptation (MSA), which is both theoretically and practically challenging. This paper makes theoretical contributions to MSA. Various representative problems of MSA are formulated in terms of testing hypotheses that are in general composite, N-ary, multivariate, and, worst of all, not necessarily disjoint. A number of sequential solutions are presented, which are computationally highly efficient, are easy to implement, and have some desirable optimality properties. These results form a theoretical foundation for developing good, general and practical MSA algorithms. Simulation results are provided to illustrate the usefulness and effectiveness of the solutions. The theoretical results presented herein have been applied to several RAMS algorithms in the subsequent parts of this series that are generally applicable, easily implementable, and significantly superior to the best fixed-structure MM estimators available. They are also important for model-set comparison, choice, and design for variable-structure as well as fixed-structure MM estimation.

Model-set choice for multiple-model estimation

Abstract This paper deals with the choice of model set in the multiple-model (MM) approach to adaptive estimation. Most representative problems of model-set choice commonly encountered in model-set design for MM estimation and in model-set adaptation for variable-structure MM estimation are considered. Their solutions based on sequential hypothesis tests are presented. Some of these solutions are optimal in some practical, meaningful sense. They are also extremely efficient in computation and easy to implement. Their effectiveness is verified via simulation. How these results can be used for model-set design is also demonstrated.

Multiple-model estimation with variable structure

Existing multiple-model (MM) estimation algorithms have a fixed structure, i.e. they use a fixed set of models. An important fact that has been overlooked for a long time is how the performance of these algorithms depends on the set of models used. Limitations of the fixed structure algorithms are addressed first. In particular, it is shown theoretically that the use of too many models is performance-wise as bad as that of too few models, apart from the increase in computation. This paper then presents theoretical results pertaining to the two ways of overcoming these limitations: select/construct a better set of models and/or use a variable set of models. This is in contrast to the existing efforts of developing better implementable fixed structure estimators. Both the optimal MM estimator and practical suboptimal algorithms with variable structure are presented. A graph-theoretic formulation of multiple-model estimation is also given which leads to a systematic treatment of model-set adaptation and opens up new avenues for the study and design of the MM estimation algorithms. The new approach is illustrated in an example of a nonstationary noise identification problem.