Fusion of Gaussian mixture models for possible mismatches of sensor model
Introduction
Distributed estimation fusion is an important research topic in the area of data fusion. The potential advantage of fusing the redundant and diverse sensor information contributes to increasing the reliability and flexibility of the whole multi-sensor system. Lots of valuable research results have been achieved [1], [2], [3], [4], [5], [6]. Among them, Ref. [1] summarizes practical algorithms for multi-sensor multi-target tracking in surveillance systems. In [2], some innovative multi-sensor multi-target applications and their solutions are discussed. Ref. [3] covers the basic theory and the most advanced developments in multi-sensor data fusion research area. In [4], the fusion architectures and rules are given concerning about optimal linear estimation fusion, and performance comparisons of different fusion rules are made via simulation experiments in [5]. Considering the cross-correlation of local estimation errors, Ref. [6] develops easily computable formulas for cases with linear and nonlinear observations.
However, in some practical tracking applications, sensor reports may be disturbed by some unknown and unobservable factors. If the filter is designed based on the ideal measurement model (regardless of possible model mismatch), it is hard to obtain a good filter result. Consequently, there may be a great difference among estimates from different local trackers. At this time, the covariance information is convincing no longer, which cannot reflect the accuracy of local estimates well. In this case, the direct fusion of local estimates will not produce a satisfactory result.
In [7], [8], a method combining the CI (Covariance Intersection) and CU (Covariance Union) (CI/CU) was developed, in which the Mahalanobis distance was used to identify the dissimilarity among local estimates. In [9], a probability framework based on the GMM was proposed to fuse possibly biased local estimates. The exact probability that local sensor is biased is needed to construct the Gaussian component of the mixture model. When the prior information is unknown exactly, the performance of the method may experience an evident decrease. In [10], a fast and robust approach (IT-FFGCC) was proposed according to the information-theoretic metric, which defines a parameter indicating how the local estimates differ from each other.
In this paper, we propose a novel approach (called IMM–GMM) to deal with the possible mismatch of the sensor model. The proposed approach develops a GMM representation for the system state based on the IMM estimator [11], [12], [13], and implements the fusion of two Gaussian mixtures in a probabilistic framework to achieve a better fusion performance. The IMM estimator is implemented to obtain an adaptive estimate. Note that in the IMM framework, a set of models is selected to represent the possible system patterns, and the final fused estimate is the weighted combination of estimates based on individual models. Here, the meaning of the word ‘adaptive’ lies in two aspects. One is that the weights of individual models can be adapted to reflect how the individual models match the measurement data. The second one is that the model set can be assumed variable, and be made adaptively based on measurements to capture the characteristics of online models. In our problem, we use the word ‘adaptive’ just meaning the former, and have no issue involved concerning about the MSA (model set adaptation) [13].
The contributions of the paper mainly lie in two aspects. One is to improve the filter performance of the single sensor. The other one is to adopt a suitable fusion scheme to combine local estimates. To cope with the situation where the sensor model may be possibly inaccurate, we model the measurement process for sensors in the framework of multiple models firstly; next, we describe the local estimate by a Gaussian mixture rather than a single Gaussian density, where the parameters for each Gaussian component is obtained based on the IMM recursion. As we known, the Gaussian mixture model [14], [15] severs as an important representation of the system state in many applications such as data fusion [16], [17], pattern recognition [18], [19], and supervised learning of multimedia data [20], [21]. Here, such a GMM representation of system state allows us to keep the detailed information about the local tracker, which contributes to the further fusion if treated properly. Finally, two Gaussian mixtures are fused according to the normalized product in a probabilistic framework [22], [23]. Moreover, the fusion of two Gaussian mixtures is of a closed form without complex numerical operations.
The rest of the paper is organized in the following. In Section 2, we formulate the estimation fusion in the presence of possible mismatches of sensor model, and review some main related work. Section 3 shows how we obtain the GMM representation of the local estimate, and implement the fusion of Gaussian mixture models in the probabilistic framework. Simulation results are given in Section 4 to show the performance of the proposed approach. We conclude in Section 5.
Section snippets
Problem formulation and related work
A simple distributed architecture for track fusion is considered with two sensors and one target. Each sensor produces its own local estimate , and then sends it to the fusion center where local estimates from different sensors are fused to improve the estimation accuracy. In some cases, the local sensor may be disturbed by some unknown and unobservable factors, which may result in the mismatch of measurement model. As mentioned in Section 1, some valuable research work has
Proposed IMM–GMM in the presence of model mismatch
In [9], a probability framework was developed based on the GMM to fuse possibly biased local estimates. In order to construct the Gaussian component of the mixture model, this method requires exact probabilities that local sensors are biased, which are not always available exactly in practical applications. In what follows, two main problems are addressed. One is concerned about how to produce a better local estimate based on the IMM estimator in the presence of model mismatch, without
Simulation results
Simulation results are provided to illustrate the effectiveness of the proposed approach. The motion of the target is governed by the following dynamic equationwhere represent the position component and velocity component, respectively. The covariance of the process noise wk is set to . The system matrix and weighted matrix for the process noise are given by
The revisit interval T is set to 1s. In all the
Conclusions
The IMM–GMM approach is proposed to address estimation fusion in the presence of possible mismatches of sensor model. To achieve better fusion performance, the efficient local tracker and fusion rule are suggested in this work. A GMM representation of the system state is developed based on the IMM estimator which accounts for possible mismatches of sensor model. Further, a closed form of fusing two Gaussian mixtures is derived in a probabilistic framework. Simulation results demonstrate that
Acknowledgements
The work is supported by the National Natural Science Foundation (No.61203220) and the State Key Program for Basic Research of China (2013CB329405).
References (29)
Fusion of possibly biased location estimates using Gaussian mixture models
Inform. Fus.
(2012)- et al.
Multitarget-Multisensor Tracking: Principles and Techniques
(1995) Multitarget-Multisensor Tracking: Advanced Applications
(1990)- et al.
Handbook of Multisensor Data Fusion: Theory and Practice
(2008) - X.R. Li, Y.M. Zhu, C.Z. Han, Unified optimal linear estimation fusion—part I: unified models and fusion rules, in:...
- X.R. Li, J. Wang, Unified optimal linear estimation fusion—part II: discussions and examples, in: Proceedings of the...
- X.R. Li, P. Zhang, Optimal linear estimation fusion—part III: cross-correlation of local estimation errors, in:...
Covariance consistency methods for fault-tolerant distributed data fusion
Inform. Fus.
(2002)- O. Bochardt, R. Calhoun, J. K. Uhlmann, S.J. Julier, Generalized information representation and compression using...
- et al.
Distributed estimation fusion with unavailable cross correlation
IEEE Trans. Aerosp. Electr. Syst.
(2012)
Best model augmentation for variable-structure multiple-model estimation
IEEE Trans. Aerosp. Electr. Syst.
Multiple-model estimation with variable structure
IEEE Trans. Auto. Control
Multiple-model estimation with variable structure, part II: model-set adaptation
IEEE Trans. Auto. Control
Multiresolution Gaussian Mixture Models: Theory and Application, Research Report RR404
Cited by (6)
Optimizing multi-sensor deployment via ensemble pruning for wearable activity recognition
2018, Information FusionCitation Excerpt :Sensor-based human activity recognition (HAR) system provides a low-power, high portability and high privacy environment to benefit ubiquitous computing [1–6].
Estimation of Robot Motion State Based on Improved Gaussian Mixture Model
2022, Zidonghua Xuebao/Acta Automatica SinicaMulti-robot active SLAM for observation fusion and attractor
2021, CAAI Transactions on Intelligent SystemsMachine learning techniques for monitoring the sludge profile in a secondary settler tank
2019, Applied Water ScienceMulti-platform maneuvering target tracking based on BLUE - AIMM - CI algorithm
2015, Chinese Control Conference, CCCNondestructive measurement in food and agro-products
2015, Nondestructive Measurement in Food and Agro-Products