A Method for Judicious Fusion of Inconsistent Multiple Sensor Data

Abstract

One of the major problems in sensor fusion is that sensors frequently provide spurious observations which are difficult to predict and model. The spurious measurements from sensors must be identified and eliminated since their incorporation in the fusion pool might lead to inaccurate estimation. This paper presents a unified sensor fusion strategy based on a modified Bayesian approach that can automatically identify the inconsistency in sensor measurements so that the spurious measurements can be eliminated from the data fusion process. The proposed method adds a term to the commonly used Bayesian formulation. This term is an estimate of the probability that the data is not spurious, based upon the measured data and the unknown value of the true state. In fusing two measurements, it has the effect of increasing the variance of the posterior distribution when measurement from one of the sensors is inconsistent with respect to the other. The increase or decrease in variance can be estimated using the information theoretic measure "entropy." The proposed strategy was verified with the help of extensive computations performed on simulated data from three sensors. A comparison was made between two different fusion schemes: centralized fusion in which data obtained from all sensors were fused simultaneously, and a decentralized or sequential Bayesian scheme that proved useful for identifying and eliminating spurious data from the fusion process. The simulations verified that the proposed strategy was able to identify spurious sensor measurements and eliminate them from the fusion process, thus leading to a better overall estimate of the true state. The proposed strategy was also validated with the help of experiments performed using stereo vision cameras, one infrared proximity sensor, and one laser proximity sensor. The information from these three sensing sources was fused to obtain an occupancy profile of the robotic workspace.Index Terms-Bayesian approach, decentralized fusion, sensor fusion, sequential fusion, spurious data.

Hybrid Data Fusion Method Using Bayesian Estimation and Fuzzy Cluster Analysis for WSN

“…Dealing with the spurious measurement, there are several methods reported in the literature. A Bayesian method is proposed in [8], it is effective in sensor validation and identification of inconsistent data. However, the spurious data is not eliminated totally and still has some impacts on the fusion result.…”

“…The conclusion of [9] is used in this paper. And the methods in [7,8] are used as comparisons during simulation.…”

“…Since the denominator depends only on the measurement (the summation is carried out over all possible values of state), an intuitive estimation can be made by maximizing this posterior distribution, by maximizing the numerator. This is called maximum a posteriori (or MAP) estimate [8], and the fusion formula is given by…”

“…Data fusion is the process of combining data from multiple sensors in order to minimize the amount of data [1]. Many algorithms about data fusion have been proposed in literatures and books include arithmetic mean, Kalman filter, Bayesian estimation, Entropy theory, DempsterShafer theory, fuzzy logic and neural network [2][3][4][5][6][7][8]. As the energy of the network is limited, the data fusion algorithm for WSN is required to be easy.…”

Fusing actigraphy signals for outpatient monitoring

“…In the case of imperfect data the main approaches followed were the probabilistic [7,8], the evidential [9,10,11], fuzzy reasoning [12,13,14], possibilistic [15,16], rough set theoretic [17,18,19], hybridization [20,14] and random set theoretic [21,22]. In the case of outliers and missing data, the most common approaches are based on sensor validation techniques [23,24,25] and on stochastic adaptive sensor modelling [26]. Finally the approaches followed to solve the problems related to heterogeneous sensors were highly depending on the sensors used and the desired target.…”