基于混沌神经网络的多目标跟踪技术研究

Chaos Neural Networks Based Multi-Target Tracking

本文应用混沌神经网络求解多目标跟踪中的数据关联问题,给出了混沌神经网络的模型,构造了数据关联的能量函数表达式,在数据关联过程中,采用退火算法。仿真结果表明,应用混沌神经网络求解数据关联比Hop fie ld网络具有更块的收敛速度和更小的关联误差。

Chaos neural networks are applied to solve data association problem of multi-target track system in this paper. Moreover, mathematic model of chaos neural networks is given, and energy function expression representing data association is constructed and anneal algorithm is proposed. Simulation result shows that chaos neural networks have faster convergence speed and smaller association error than Hopfield networks when they are used to solve data association.