Multiple Model Algorithm Research Articles

Многоальтернативный алгоритм одномаяковой навигации автономного необитаемого подводного аппарата без априорных данных о его местоположении. Часть 1. Математическое описание

Многоальтернативный алгоритм одномаяковой навигации автономного необитаемого подводного аппарата без априорных данных о его местоположении. Часть 1. Математическое описание

Target Classification Aided Variable-Structure Multiple-Model Algorithm

Because the varieties of general aviation aircraft and their performance differences become increasingly large, the traditional multiple model tracking algorithm needs to employ more motion models to describe the actual maneuver model of a moving target. This fact is easy to degrade the tracking accuracy and brings a large computational load. Thus, a target classification aided variable-structure multiple-model algorithm (TCA-VSMM) is proposed to solve the above problem. In the proposed TCA-VSMM algorithm, the target classification aided is introduced to improve the accuracy of state estimation. Then, the target classification aided in the screening of the motion model set is analyzed. Concretely, the target classification information and velocity information in automatic dependent surveillance-broadcast (ADS-B) measurements are incorporated into the variable-structure multiple-model filter. Consequently, the screening model set is more approximated to the real motion model of a moving target. Experiments show that the TCA-VSMM algorithm can obtain better performance with small computation load and high estimation accuracy than those of the model-group switching variable-structure multiple-model algorithm.

Unmanned aerial vehicle trajectory data fusion based on an active and passive feedback system

The surveillance data obtained from an unmanned aerial vehicle during real-time operation are very important for unmanned aerial vehicle supervision. In this study, in order to obtain more reliable and accurate unmanned aerial vehicle flight paths, we have proposed a trajectory fusion method of unmanned aerial vehicle data based on an active and passive feedback system. The active surveillance data are obtained from automatic dependent surveillance – broadcast and the passive surveillance data are derived from the unmanned aerial vehicle ground control station. The convex combination fusion algorithm is employed to fuse the two sets of data to obtain the accurate flight state and a continuous stable running track for the unmanned aerial vehicle. The simulation results showed that the fusion trajectory obtained using the convex combination fusion algorithm was smoother than the local trajectory obtained by the interacting multiple model algorithm.

Interacting Multiple Model-Based Human Pose Estimation Using a Distributed 3D Camera Network

Distributed camera network for human pose estimation can solve the problem of limited view and occlusion of single view, which has great potential for wide area surveillance applications. To fuse different field of view information, we propose a distributed human pose estimation method by combining the interactive multiple model (IMM) algorithm with the distributed information fusion of human skeleton joints. Compared with the state-of-the-art works which often use a single motion model to depict the motion of human skeleton joints, e.g., constant velocity, the novelty of our work is that the maneuvering property of human action is handled by the IMM, i.e., the motion model of human skeleton joints in the filter is approximated using constant velocity, constant acceleration, and Singer motion models. Based on the advantages of IMM algorithm for maneuvering target tracking, our method can not only solve the single-view occlusion problem, but also solve the problem of joint point fluctuation caused by the estimation error of each sensor node after distributed information fusion. The final human action recognition experimental results show that the proposed method can improve the action recognition rate on the datasets captured by Kinect V2. In addition, we built a distributed camera network using embedded machine learning boards, such that deep learning-based human pose estimation methods can be employed in our framework to handle the limitations of original Kinect SDK.

Research into the integrated navigation of a deep-sea towed vehicle with USBL/DVL and pressure gauge

Underwater Towed Vehicle navigation and localization in deep sea environments are particularly challenging without high-precision inertial sensors. In this paper, the multi-sensors navigation fusion method of a towed underwater vehicle based on the ultrashort baseline (USBL) system, Dopplervelocitylog (DVL) and a pressure gauge in the deep sea is investigated. Combined with GPS real-time tide detection technology, the positioning principle of the USBL based on the depth constraint is given, and a straight-line model be proposed to estimate the position of the vehicle used the average heading when underwater towed vehicle is working normally in deep sea without DVL. The interacting multiple model (IMM) algorithm based on smoothing algorithm is proposed to improve the positioning accuracy in the deep sea, and the nonlinearity of the model, precision and stability of filtering are considered simultaneously using the adaptive robust square root cubature Kalman filter. The results of the simulated and practical experiments at 4800 m verify that the new algorithm can take full use of high-precision observation information and greatly improve the positioning accuracy and robustness of a deep sea underwater towed system.

Atmospheric PM2.5Concentration Prediction Based on Time Series and Interactive Multiple Model Approach

Urbanization, industrialization, and regional economic integration have developed rapidly in China in recent years. Air pollution has attracted more and more attention. However, PM2.5is the main particulate matter in air pollution. Therefore, how to predict PM2.5accurately and effectively has become a concern of experts and scholars. For the problem, atmosphere PM2.5concentration prediction algorithm is proposed based on time series and interactive multiple model in this paper. PM2.5concentration is collected by using the monitor at different air quality levels. The time series models are established by historical PM2.5concentration data, which were given by the autoregressive model (AR). In the paper, three PM2.5time series models are established for three different air quality levels. Then, the three models are converted to state equation, respectively, by autoregressive integrated with Kalman filter (AR-Kalman) approaches. Besides, the proposed interactive multiple model (IMM) algorithm is, respectively, compared with autoregressive (AR) model algorithm and AR-Kalman prediction algorithm. It is turned out the proposed IMM algorithm is more accurate than the other two approaches for PM2.5prediction, and it is effective.

A Fixed-Lag Smoothing Interactive Multiple Model Tracking and Interception System for Maneuvering Target

In this work, an improved interactive multiple model-based smoothing algorithm is proposed for tracking and interception of a maneuvering target. The smoothed states are obtained at some past scan ($$k-N$$), but the target interception happens at the current scan. This helps to obtain the real-time benefits of the smoothing. Fixed-lag smoothing interactive multiple model algorithm is employed to follow the target more precisely. Real-time benefit of smoothing is used to improve the decision making in order to achieve better target interception before it reaches its desired destination. Three different simulation environments have been used to analyze the performance and validity of the proposed algorithm. The simulation results are further extended to compare each case in terms of hit error and root-mean-square error.

Equality‐constrained state estimation for hybrid systems

The authors investigate the problem of state estimation for stochastic hybrid dynamical systems with state equality constraints. They divided this problem into three categories depending on the linearity or non-linearity of equality constraints as well as on the dependence of constraints on the operating mode. For the mode-independent equality-constrained linear case, they present sufficient conditions on process dynamics and filter initial conditions so that the classical interacting multiple model (IMM) algorithm yields state estimates satisfying the linear equality constraint for all subsequent times. For both linear and non-linear systems, the mode-dependent equality constraints must be enforced over time by the filter. They present a modified version of the IMM filter to enforce the equality constraints in such cases. Their numerical results show that the proposed methods provide more accurate estimates than the IMM unconstrained estimates.

A Hybrid IMM Based INS/DVL Integration Solution for Underwater Vehicles

To enhance the performance of the navigation system under the complex underwater environment, a hybrid interacting multiple model (HIMM) aided inertial navigation system (INS)/Doppler velocity logs (DVL) integration solution is proposed. First, to employ the acoustic Doppler current profiler mode of DVL, a novel INS/DVL mechanism is constructed where the water current velocity is estimated in real time and both bottom-track and water-track velocity measurements of DVL are involved in the observation vector. Meanwhile, to deal with the outliers and observation noises in the DVL's measurements, a HIMM algorithm is proposed to adaptively select the proper models to describe the changing environment. Simulations and field test are conducted to evaluate the effectiveness of the proposed algorithm, where the interacting multiple model (IMM) algorithm is employed for comparison. The results indicate that the proposed HIMM-aided INS/DVL integration solution shows superiority than the traditional IMM method when the observation noises and outliers exist. Meanwhile, the proposed integration scheme can successfully bridge the DVL's bottom-track outages.

Bayesian State Estimation for Markovian Jump Systems: Employing Recursive Steps and Pseudocodes

In this article, we review several existing Bayesian state-estimation algorithms for Markovian jump systems (MJSs) and illustrate their distinctive properties. Specifically, we analyze the jump Markov linear systems (JMLSs), for which autonomous multiple-model (AMM) algorithms, first-order generalized pseudo-Bayesian (GPB1) algorithms, secondorder generalized pseudo-Bayesian (GPB2) algorithms, interacting multiple-model (IMM) algorithms, and reweighed IMM (RIMM) algorithms are introduced. For nonlinear MJSs, we consider the sampling importance resampling (SIR) particle filter (PF) as well as the IMM PF to demonstrate how to handle nonlinearity in the framework of MJSs. For convenience, the realization codes of GPB1 and IMM algorithms and SIR PFs are provided.

Multiple-Model State Estimation Based on Variational Bayesian Inference

In this paper, we propose a new approach to state estimation of multiple state-space models. Unlike the traditional methods (including the interacting multiple-model algorithm) that approximate a Gaussian mixture distribution with a single Gaussian distribution, the proposed method approximates the joint probability density functions of the state and model identity through Bayesian inference. It is shown that the proposed method reduces the approximation error considerably, and improves estimation accuracy without increasing computational cost. Analysis of its specific features as well as a potential extension is also presented. Numerical examples with a practically oriented simulation are employed to illustrate the effectiveness of the proposed method.

Estimation of the tire slip angle under various road conditions without tire–road information for vehicle stability control

This paper presents a tire slip angle estimator based on an Interacting Multiple Model (IMM) algorithm for vehicle stability control. The proposed algorithm is capable of estimating the tire slip angles under various road conditions without the prior knowledge of tire and road condition by only using on-board vehicle sensors. Instead of employing tire and road information, the proposed algorithm utilizes multiple numbers of model candidates to represent all aspects of vehicle motions under various road conditions. Each model candidate is a combination of lateral vehicle dynamics and transient/steady state tire dynamics. The proposed algorithm evaluates the fidelity of each model candidate to the current vehicle dynamics with probability. Moreover, in the proposed algorithm, multiple numbers of Kalman filters are embedded with these model candidates as process models. The final estimate of the proposed algorithm in each time step is a linear sum of the posteriori states from multiple embedded filters with the calculated probability as coefficients. The proposed estimation algorithm has been evaluated via vehicle tests. The tests have been conducted on dry asphalt and wet asphalt using a luxury passenger car equipped with a high-performance GPS for reference and data logging computer. The results have shown that the proposed estimator can successfully estimate tire slip angles with satisfactory accuracy under various road conditions

An Enhanced Particle Filter for Detection and Tracking Maneuvering Dim Target

Track-before-detect (TBD) is a powerful technique for detection and tracking dim targets. To solve the problem of target maneuvering, the basic idea of multiple model (MM) algorithm is introduced into the basic multiple model particle filter TBD, a novel and enhanced particle filter algorithm is proposed in this paper. The algorithm proposed formulates the probability of existence calculation respectively, avoiding the need for hybrid estimation. This simulation results show that the algorithm proposed is efficient in dim maneuvering target tracking.

A novel nonlinear least-squares approach to highly maneuvering target tracking

Abstract Trajectories of aerial and marine vehicles are typically made of a succession of smooth trajectories, linked by abrupt changes, i.e. maneuvers. Notably, modern highly maneuvering targets are capable of very brutal changes in the heading, with accelerations of up to 15 g . As a result, we model the target behavior using piecewise deterministic Markov models, driven by parameters that jump at unknown times. Over the past years, real-time (or incremental) optimization-based smoothing methods have become a popular alternative to nonlinear filters, such as the Extended Kalman Filter (EKF), owing to the successive relinearizations that mitigate the linearization errors that inherently affect the EKF estimates. In the present paper, we propose to combine such methods for tracking the target during non-jumping phases with a probabilistic approach to detect jumps. Our algorithm is shown to compare favorably to the state-of-the-art Interacting Multiple Model (IMM) algorithm, especially in terms of target's velocity estimation, on a set of meaningful and challenging trajectories.

Fuzzy-Logic-Based, Obstacle Information-Aided Multiple-Model Target Tracking

Incorporating obstacle information into maneuvering target-tracking algorithms may lead to a better performance when the target when the target maneuver is caused by avoiding collision with obstacles. In this paper, we propose a fuzzy-logic-based method incorporating new obstacle information into the interacting multiple-model (IMM) algorithm (FOIA-MM). We use convex polygons to describe the obstacles and then extract the distance from and the field angle of these obstacle convex polygons to the predicted target position as obstacle information. This information is fed to two fuzzy logic inference systems; one system outputs the model weights to their probabilities, the other yields the expected sojourn time of the models for the transition probability matrix assignment. Finally, simulation experiments and an Unmanned Aerial Vehicle experiment are carried out to demonstrate the efficiency and effectiveness of the proposed algorithm.

Reliable flight performance assessment of multirotor based on interacting multiple model particle filter and health degree

Multirotor has been applied to many military and civilian mission scenarios. From the perspective of reliability, it is difficult to ensure that multirotors do not generate hardware and software failures or performance anomalies during the flight process. These failures and anomalies may result in mission interruptions, crashes, and even threats to the lives and property of human beings. Thus, the study of flight reliability problems of multirotors is conductive to the development of the drone industry and has theoretical significance and engineering value. This paper proposes a reliable flight performance assessment method of multirotors based on an Interacting Multiple Model Particle Filter (IMMPF) algorithm and health degree as the performance indicator. First, the multirotor is modeled by the Stochastic Hybrid System (SHS) model, and the problem of reliable flight performance assessment is formulated. In order to solve the problem, the IMMPF algorithm is presented to estimate the real-time probability distribution of hybrid state of the established SHS-based multirotor model, since it can decrease estimation errors compared with the standard interacting multiple model algorithm based on extended Kalman filter. Then, the reliable flight performance is assessed with health degree based on the estimation result. Finally, a case study of a multirotor suffering from sensor anomalies is presented to validate the effectiveness of the proposed method.

Robust Interacting Multiple Model With Modeling Uncertainties for Maneuvering Target Tracking

This paper proposes an improved robust interacting multiple model (RIMM) algorithms with modeling uncertainties for maneuvering target tracking with changing dynamics. To mitigate the effects of the modeling uncertainty, a compensation step is introduced to adjust the degree of dependence of the filtering on the system or the measurement model based on the orthogonality principle between the state estimation error and innovation sequence of the subfilter model in the RIMM algorithm. By relying on the compensation parameter, the proposed algorithm fully utilizes the useful information in the innovation sequence and reduces the impact of system model error. The numerical simulation and car-mounted experiments using time difference of arrival (TDOA) measurements of the maneuvering target tracking with changing dynamics are conducted to verify the effectiveness of the proposed RIMM algorithm. Compared with the conventional approaches, the proposed RIMM algorithm results in a remarkable improvement in the state estimation accuracy and small bias while improving the consistency of the filter.

Qualitative study of Riccati difference equation on maneuvering target tracking and fault diagnosis of wind turbine gearbox

Various performance analyses using Interacting Multiple Model (IMM) algorithm for the stability of tracking a maneuvering target are dealt with the state error covariance. A noble way of analyzing ...

Fuzzy health degree-based dynamic performance evaluation of quadrotors in the presence of actuator and sensor faults

This paper proposes a fuzzy health degree-based dynamic performance evaluation algorithm of quadrotors in the presence of actuator and sensor faults. First, an augmented stochastic hybrid system (SHS) model for quadrotors is established. In the SHS model, the discrete modes are assigned with sensor normal mode and other senor anomalous modes. In each mode, a process equation and an observation equation are provided to describe the continuous behavior of quadrotors, where the process equation is augmented to model actuator fault by introducing effectiveness coefficients, and different observation equations are built to model different sensor faults. Then, a modified interacting multiple model algorithm is used to estimate the hybrid state of the SHS model, and a concept of fuzzy health degree is introduced to measure dynamic performance of the quadrotor based on the state estimation result. Finally, a simulation of a quadrotor suffering from successive actuator and sensor faults is presented to validate the effectiveness of the proposed algorithm.

Fully adaptive update rate for non‐linear trackers

Modern phased-array multifunction radars have the ability to change or schedule the tasks of the beam in order to accomplish all their missions in an optimised fashion. The resource manager of the radar must then control the update rate of the measurement task of target active tracking, so as to minimise the radar computer load without losing targets. Scarce measurements lead to low radar load, but they also lead to an increased number of illuminations at each measurement epoch to find the target. On the basis of this rationale, a sound procedure was proposed by Blackman and van Keuk to derive an optimal measurement rate. Their optimisation criterion is established using a linear Singer target model and a linear Kalman filter. In this study, their method is extended, and the authors propose a versatile optimal update rate algorithm that is applicable to virtually any non-linear target model combined with any non-linear filter able to output an error covariance matrix. This includes extended Kalman filter (EKF), Unscented Kalman Filter (UKF), Interacting Multiple Model (IMM) algorithm, and particle filters. For numerical experiments and validation, they consider a non-linear target model based on Frenet–Serret three-dimensional equations, and the tracking is performed by a non-linear invariant EKF.