Multiple Model Adaptive Estimation Algorithm Research Articles

Low-Complexity Architecture for Cyber-Physical Systems Model Identification

We propose a low complexity architecture for cyber-physical system (CPS) model identification based on multiple-model adaptive estimation (MMAE) algorithms. The complexity reduction is achieved by reducing the number of multiplications in the filter banks of the MMAE algorithm present in the cyber component of the CPS. The architecture has been implemented using FPGA for 16, 32, 64 filter banks as part of position and velocity estimations of autonomous auto-mobile application. It has been found up to 78% reduction in multiplications is possible, which translates to the reduction of 39% lookup tables, 13% FFs, 27% DSPs, and 43% power reduction when compared with the conventional architecture (without multiplications reduction) at 100MHz operating frequency. Furthermore, the proposed architecture is able to identify accurate model of auto-mobile application just within 510 ns, in the presence of external disturbances and abrupt changes.

Electrochemical Model Based Fault Diagnosis of Lithium Ion Battery

A Multiple Model Adaptive Estimation (MMAE) based approach of fault diagnosis for Li-Ion battery is illustrated in this paper. Electrochemical modelling approach is integrated with MMAE for fault diagnosis. This real physics based model of Li-ion battery (with Li-Co-O2 cathode chemistry) with nominal model parameters is considered as the healthy battery model. Battery fault conditions such as aging, overcharge and over discharge causes significant variations of parameters from nominal values and can be considered as separate models. Output error injection based Partial Differential Algebraic Equation (PDAE) observers are used to generate the residual voltage signals. These residuals are then used in MMAE algorithm to detect the ongoing fault conditions of the battery. Simulation results show that the fault conditions can be detected and identified accurately which indicates the effectiveness of the proposed battery fault detection method.

Multiple-model adaptive estimation for space surveillance with measurement uncertainty

Summary An efficient multiple-model adaptive estimation (MMAE) algorithm is presented for time-variant system with both system and measurement uncertainties, whose statistics are supposed to be unknown. The model uncertainties are described by a set of noise covariance matrices, such that a small model set is sufficient to achieve good estimation performance. To demonstrate the feasibility of the presented MMAE for the considered time-variant uncertain system, a proof is provided that shows the filtering convergence. The performance of the algorithm is evaluated via different operating scenarios of double line-of-sight measuring space surveillance. Simulation results demonstrate that the MMAE algorithm outperforms the robust filtering algorithms in the presence of the uncertainty and yields positioning accuracy similar to the extended Kalman filter in the absence of the uncertainty. Copyright © 2015 John Wiley & Sons, Ltd.

Improper Complex-Valued Multiple-Model Adaptive Estimation

Motivated by the problem of estimating the discrete and continuous states of an improper complex-valued stochastic hybrid system, the paper proposes a class of widely linear (augmented) multiple model adaptive estimation algorithms, referred to as the C/MMAE. We show that for an improper complex-valued signal, pseudo-covariance of the innovation sequence is not zero and, therefore, carries useful statistical information regarding the unknown behaviour mode of the hybrid system. A new Bayesian law is, therefore, derived as a function of the pseudo-covariance of the innovation sequence and used to compute the probability that a hypothesized model is in effect at a certain time. We show that the C/MMAE, which uses the new Bayesian law and utilizes the complete second-order statistical characterization of the complex-valued innovation sequence, convergencies faster than its counterpart, which only uses the conventional covariance of the innovation sequence. In order to simplify the computational complexity, we develop two circularized versions of the C/MMAE using a preprocessing step, referred to as the circularizing filter (CF). The CF is incorporated to convert the improper observations/innovations into the proper ones in order to reduce the computational complexity of the hypothesis testing step. Finally, an interacting version of the C/MMAE, referred to as C/IMM, is developed for improper complex-valued systems with Markovian switching coefficients. Simulation results indicate that the proposed hybrid estimators provide improved performance and convergence properties over their traditional counterparts.

Application of MMAE to the Fault Detection of Lithium-Ion Battery

With the advantage of high energy density, long cycle life and environmental friendliness, Lithium-ion battery has become a promising power source for hybrid and electric vehicles, which are liable to two kinds of failure, overcharge and overdischarge. Because of the capability of detecting multiple faults, Multiple Model Adaptive Estimation (MMAE) method was applied to a model-based fault detection of a lithium-ion battery with a two-order linear electrical model. Parameters that represent normal-mode and faulty-mode of the battery were obtained by a series of experiments, and three Kalman filters were designed thereafter. Finally, simulation verified the performance of the MMAE algorithm on fault detection of these two kinds of fault and it is shown that this technique is able to discern these faults rapidly and accurately.

Innovative Mars entry integrated navigation using modified multiple model adaptive estimation

A variety of anticipated missions to Mars, such as Mars base and sample return, will have to utilize precision navigation and guidance for their hypersonic entry to achieve the required landing accuracy. In order to effectively reduce the adverse impact of initial state errors and parameter uncertainties during Mars atmospheric entry and then improve entry navigation accuracy, an innovative high-precision integrated navigation algorithm for Mars entry is developed based on a modified multiple model adaptive estimation (MMAE) in this paper. First, a six degree-of-freedom (DOF) Mars entry dynamics model is derived based on the angular velocity outputs of a gyro, which can be more complete and accurate description of the state variables of an entry vehicle than the traditional three DOF dynamics models. Second, both the accelerometer outputs and radiometric measurements are adopted as the navigation observations embedded to Kalman filter bank to perform state estimation and suppress the measurement noise. Finally, a new modified multiple model adaptive estimation algorithm with exponential decay terms is proposed to overcome the inherent drawback of classical MMAE and then further improve navigation accuracy. The numerical simulation results show that the integrated navigation algorithm developed in this paper is able to accurately estimate the state variables of an entry vehicle, even in the case that the initial entry state errors and larger parameter uncertainties are taken into account.

P 256. Augmented set-shifting in Parkinson’s disease with noisy Galvanic Vestibular Stimulation

Introduction Cognitive impairment and executive dysfunction, in addition to overt motor deficits, are common in Parkinson’s disease (PD). Perseveration of behavioural paradigms in PD reflects cognitive inflexibility ( Cools, 2001 ) and is related to the inability to stop and switch motor paradigms ( Stoffers, 2001 ). Noisy Galvanic Vestibular Stimulation (GVS) is a technique that has improved motor metrics in PD ( Samoudi et al., xxxx , Yamamoto, xxxx ). Widespread vestibular areas-both cortical and subcortical-are activated by GVS ( Lopez and Blanke, 2011 ), perhaps explaining the positive effects of GVS on PD motor symptoms and also on cognitive effects, such as enhancing visual memory recall ( Wilkinson, 2008 ). Objectives We investigated whether noisy GVS has effects on PD switching abilities in a set-shifting motor task. Unlike other switching paradigms (e.g., Wisconsin Card Sorting Task), our task does not additionally depend on other cognitive components such as working memory and rule learning, thereby predicting a more sensitive measure of switching abilities per se. Materials and methods We tested 4 patients with mild to moderate PD (Hoehn and Yahr stages: 1–2), and 4 healthy age-matched control subjects. All PD patients were asked to stop taking dopaminergic medication at least 12 h prior to the study. Subjects performed 8 trials of a sinusoidal visual tracking task which alternated between 2 modes depending upon whether the displayed cursor position underestimated the actual error by 30% (’Better’) or overestimated by 200% (’Worse’). The switching was covert, unbeknownst to the subject. Either a subthreshold, noisy GVS (0.1–10 Hz, 1/f-type power spectrum) or sham current was applied during each trial in a pseudorandom order. Using Multiple Model Adaptive Estimation (MMAE) ( Oishi, 2010 ), we were able to predict which type of performance mode (‘Better’ or ‘Worse’) the subject was applying to the task. The MMAE algorithm essentially determines the likelihood a subject is behaving in a given ‘Better’ or ‘Worse’ performance mode at specific time stamps throughout a given trial. Results Noisy GVS had no significant effect on the switching detection frequency in control subjects ( Fig. 1 ). PD patients off medication demonstrated notably less switching than control subjects. Patients also responded to GVS delivery with significantly greater frequencies of switching (p = 0.02, Fig. 1 ). Conclusion Noisy GVS increased overall switching behaviour in PD patients without any effect on healthy control subjects, suggesting that Vestibular Stimulation may be able to influence cognitive flexibility deficits in PD. However, further work is needed to see if GVS can, in fact, also result in excessive switching between motor programs.

Fault detection and identification based on combining logic and model in a wall-climbing robot

A combined logic- and model-based approach to fault detection and identification (FDI) in a suction foot control system of a wall-climbing robot is presented in this paper. For the control system, some fault models are derived by kinematics analysis. Moreover, the logic relations of the system states are known in advance. First, a fault tree is used to analyze the system by evaluating the basic events (elementary causes), which can lead to a root event (a particular fault). Then, a multiple-model adaptive estimation algorithm is used to detect and identify the model-known faults. Finally, based on the system states of the robot and the results of the estimation, the model-unknown faults are also identified using logical reasoning. Experiments show that the proposed approach based on the combination of logical reasoning and model estimating is efficient in the FDI of the robot.

State estimation for hybrid systems: applications to aircraft tracking

The problem of estimating the discrete and continuous state of a stochastic linear hybrid system, given only the continuous system output data, is studied. Well established techniques for hybrid estimation, known as the multiple model adaptive estimation algorithm, and the interacting multiple model algorithm, are first reviewed. Conditions that must be satisfied to guarantee the con- vergence of these hybrid estimation algorithms are then presented. These conditions also provide a means to predict, as a function of the system parameters, which transitions in a hybrid system are relatively easy to detect. A new variant of hybrid estimation algorithms, called the residual-mean interacting multiple model (RMIMM) algorithm, is then proposed and analysed. The performance of RMIMM is demonstrated through multi-modal aircraft trajectory tracking examples.

Multiple model estimation with inter-residual distance feedback

Abstract This paper presents a modification of the Multiple Model Adaptive Estimation concept. A trade-off problem between tracking the process and distinguishing the models is pointed out and an adaption of the elemental filters is proposed. The adaption scheme modifies the filters such that the predicted measurements not become too close in some sense. This has considerable influence on the distinguishability of the filters and thereby the properties of the Multiple Model Adaptive Estimation algorithm. Stability of the method is considered, and a simulated example demonstrates the method.

Multiple Model Estimation with Inter-Residual Distance Feedback

This paper presents a modification of the Multiple Model Adaptive Estimation concept. A trade-off problem between tracking the process and distinguishing the models is pointed out and an adaption of the elemental filters is proposed. The adaption scheme modifies the filters such that the predicted measurements not become too close in some sense. This has considerable influence on the distinguishability of the filters and thereby the properties of the Multiple Model Adaptive Estimation algorithm. Stability of the method is considered, and a simulated example demonstrates the method.