在空战场协同攻击中，常涉及到多传感器协同探测及跟踪，由于目标的出现与消失具有随机性，所以在协同中既要考虑已有目标的跟踪，更要重视新生目标的及时探测和捕获。为此，建立了新生目标的探测概率模型，并阐述了不同传感器联盟对新生目标的探测能力，依据后验克拉美-罗下界（Posterior Cramer-Rao Lower Bound，PCRLB）对已跟踪目标组建传感器联盟，利用二值粒子群优化（Binary Particle Swarm Optimization，BPSO）算法及PCRLB 研究基于动态联盟的多传感器协同探测与跟踪方法。仿真表明，该方法跟踪精度较高，误差小且稳定。

In cooperative attacks, multi-sensor cooperative detection and tracking is usually used. Because targets randomly appear or vanish, we should think about tracking new appearing targets and pay attention to detect or even capture new targets. Therefore, a probabilistic model for detecting new targets was established, stating the ability of different sensor coalitions to detect new targets. A sensor coalition for tracked targets was set up according to PCRLB, and the method of multi-sensor cooperative detection and tracking based on dynamic coalition using the BPSO algorithm and PCRLB was studied. The simulation result indicates that the method possesses higher tracking precision, less error, and more stability.