Search

In this paper, the observability in BOT problem is analyzed and defined in form of leg by leg, the basic relationship between observability and the number of particles is opened out, and a fast particle filter with variable number of particles which calculated by this relationship is proposed to adopt the observability in BOT to save the computation resource. The simulation results show that the definition of observability accords with the reality and the proposed method decreases the computation time while remaining the tracking quality.

Due to the reasons of sensors in itself and the communication delays, the research of the asynchronous multisensor data fusion problem is more practical than that of synchronous senosrs. Based on minimizing the trace of the fusion error covariance, the asynchronous multisensor data fusion algorithm is presented. The multisensor track fusion is valid for asynchronous sensors as well as synchronous sensors. The algorithm makes up for the drawback that asynchronous fusion is limited to sneor-to-sensor data fusion. Significantly, the generalized multisensor asynchronous sensor fusion model is established. The analyse of algorithm in the theoretical deduction is presented. The format of algorithm is concise and applicable. Finally, simulation results show the asynchronous mulitsensor fusion is effective and the performance of the algorithm is superior to that of single sensor.

Separation of spatial-time signals is a very meaningfully work in the real world. The spatial separability of Blind source separation (BSS) technique is studied in this paper. Two main problems about spatial separability are discussed in the paper. One is what relationship between the separability of the signals and the spatial locations of the sources, especially their azimuths with respect to the receiver. Another is how close of two source signals in field space can be separated using the BSS technique. A join approximate diagonalization eigen-matrices algorithm is proposed for this purpose. Simulations show interesting results.