论文部分内容阅读
针对量测不确定条件下多传感器量测数据的有效利用问题,提出一种多传感器自适应粒子滤波算法.利用随机采样策略和量测模型转移概率实现当前时刻多传感器量测集合的采样,通过粒子滤波中重采样步骤完成估计状态和量测集合的更新,进而依据重采样后单个传感器量测数目在传感器量测集合中的比重实现当前时刻传感器量测的确认.该算法通过有效量测的合理选择,改善了扰动对滤波精度和计算量的不利影响.理论分析和仿真实验均验证了所提出算法的有效性.
Aiming at the problem of the effective use of multi-sensor measurement data under uncertain conditions, a multisensor adaptive particle filter algorithm is proposed. By using the random sampling strategy and the transition probability of the measurement model, the sampling of the multisensor measurement set at the current time is achieved. Particle filter resampling steps to complete the estimation state and the update of the measurement set, and then based on re-sampling the number of single sensor measurements in the sensor measurement set to achieve the current moment the proportion of sensor to confirm the measurement of the effective measurement of the algorithm Reasonable selection improves the adverse effect of disturbance on the filtering accuracy and computation, and the theoretical analysis and simulation experiments verify the effectiveness of the proposed algorithm.