CAJ | 학술논문

传统粒子滤波算法在实际应用中要求快速实时，而现实往往需要大量的粒子数参与运算，使得算法运算非常耗时。提出一种改进的粒子滤波新算法，根据实际应用场景，计算所需要的粒子数目，在不降低性能的情况下，动态减少粒子数，提高了运算速度。仿真实验表明，新算法显著地降低了算法运算的时间，达到了快速实时的效果。
전통입자려파산법재실제응용중요구쾌속실시，이현실왕왕수요대량적입자수삼여운산，사득산법운산비상모시。제출일충개진적입자려파신산법，근거실제응용장경，계산소수요적입자수목，재불강저성능적정황하，동태감소입자수，제고료운산속도。방진실험표명，신산법현저지강저료산법운산적시간，체도료쾌속실시적효과。
Traditional particle filter algorithms , in practical applications , require fast and real-time effects.But in reality there are often a large amount of particles needed to involve in computing , which makes the algorithmic operation very time-consuming .We propose an improved particle filter algorithm , it computes the number of particles needed according to actual applied scene and dynamically reduces the number of particles in the case of no performance degradation , which accelerates the speed of operation .Simulation experiments demonstrate that the new algorithm significantly reduces the computation time and achieves fast real -time effect.