论文部分内容阅读
本文建立了一种在杂波环境下跟踪多目标的算法。本算法具有起始目标、判决虚警以及跟踪较大机动目标的能力。本算法用自适应一阶差分预测模型代替常用的Kalman模型;将新目标起始问题化为优化决策问题;将观测数据与目标的最佳亲缘缔合问题化为多指标优化问题。本算法还使用了聚类方法技巧;使得大量目标的同时跟踪成为可能。模拟计算结果表明该算法是可以实现的。
In this paper, an algorithm to track multi-target in clutter is established. The algorithm has the ability of starting the target, judging the false alarm and tracking the larger maneuvering target. This algorithm replaces the commonly used Kalman model with the adaptive first-order difference prediction model, and turns the initial problem of the new target into the optimal decision-making problem. The best relative association problem between the observed data and the target is transformed into the multi-indicator optimization problem. The algorithm also uses clustering techniques; making it possible to track a large number of targets simultaneously. Simulation results show that the algorithm is feasible.