WeiYa's Work Yard

A dog, who fell into the ocean of statistics, tries to write down his ideas and notes to save himself.

Multi-estimate extraction for SMC-PHD

Posted on
Tags: SMC-PHD

This post is based on Li, T., Corchado, J. M., Sun, S., & Fan, H. (2017). Multi-EAP: Extended EAP for multi-estimate extraction for SMC-PHD filter. Chinese Journal of Aeronautics, 30(1), 368–379.

A RFS variable is a r.v. that takes values as unordered finite sets. The cardinality of a RFS variable $X$ is random and modelled by a discrete distribution $\rho(n)=\Pr{\vert X\vert=n}$. The RFS $X$ is specified entirely by its cardinality distribution $\rho(n)$ and a family of symmetric joint dist. $p_n(x_1,\ldots,x_n)$ that characterize the distribution of its elements over the state space, conditioned on the set cardinality $n$.

The PDF of an RFS variable $X$ is denoted $f(X)$ and defined as

Code Reading

target state: $x_k=[p_{x,k},\dot p_{x,k},p_{y,k},\dot p_{y,k}, w_k]^T$

coordinated-turn state transition model:

F = @(a)[1 sin(a(5,:)*T)/a(5,:)      0    -((1-cos(a(5,:)*T))/a(5,:)) 0
         0 cos(a(5,:)*T)             0    -sin(a(5,:)*T)              0
         0 (1-cos(a(5,:)*T))/a(5,:)  1    sin(a(5,:)*T)/a(5,:)        0
         0 sin(a(5,:)*T)             0    cos(a(5,:)*T)               0
         0       0                   0    0                           1]*a;
Noise.vel = [15; 15];
Noise.turn= pi/180;

Published in categories Note