<title>NEKF IMM tracking algorithm</title>

Abstract

Highly maneuvering threats are a major concern for the Navy and the DoD and the technology discussed in this paper is intended to help address this issue. A neural extended Kalman filter algorithm has been embedded in an interacting multiple model architecture for target tracking. The neural extended Kalman filter algorithm is used to improve motion model prediction during maneuvers. With a better target motion mode, noise reduction can be achieved through a maneuver. Unlike the interacting multiple model architecture which uses a high process noise model to hold a target through a maneuver with poor velocity and acceleration estimates, a neural extended Kalman filter is used to predict corrections to the velocity and acceleration states of a target through a maneuver. The neural extended Kalman filter estimates the weights of a neural network, which in turn are used to modify the state estimate predictions of the filter as measurements are processed. The neural network training is performed on-line as data is processed. In this paper, the simulation results of a tracking problem using a neural extended Kalman filter embedded in an interacting multiple model tracking architecture are shown. Preliminary results on the 2 nd Benchmark Problem are also given.

Analysis of the NEKF performance in a multisensor environment

“…The Kalman filter results thus provide baseline mean error values for comparing the standard deviations of the position error. Overall performance comparisons between the Kalman filter and the NEKF in various formats can be found in a number of previous works including [1], [2], [3], [5], [6], and [7]. Figure 3 shows that the relative variation is less for the NEKF, but varies according to the relative offset in time of the measurements with the minimum variation occurring when the measurements are exactly aligned.…”

“…One approach that has been developed for tracking a target through a maneuver is that of the neural extended Kalman filter (NEKF) [1], [2], and [3]. The neural extended Kalman filter is a coupled system of a standard extended Kalman filter (EKF) that provides state estimates and an EKF neural network training parameter, as developed in [4].…”

Kinematic Prediction for Intercept Using a Neural Kalman Filter

“…As discussed in (Owen and Stubberud, 2003), this is a result of the Stone-Weiestrauss Theorem. A neural network fits the criteria of this theorem if it uses a multi-layer perceptron.…”

Supervised learning for adaptive interactive multiple model (SLAIMM) tracking

“…[12] Different approaches seek to classify the target based on its maneuver [13,14] such as determining the intent of a target [15]. In addition to the above methods, structure [16] and learning techniques have been applied to complement the IMM: including the neural-net extended Kalman Filter (NEKF) IMM [17] and the Genetic Algorithm [18]. In this paper, we use supervised reinforcement learning.…”