Abstract:

 Target tracking becomes hard due to the complicated battlefield environment and the improved target’s maneuver. A maneuvering target tracking algorithm based on human cognition mechanism is put forward. The human “memory” mechanism is introduced into the construction of the maneuvering model. In order to make the description of the target motion reasonable, the neural network can obtain and store the information about the characteristics of the target through the off-line learning. The stored information can be used to guide the adjustment of the model parameters. Furthermore, based on the theory of the human cognitive perception-action cycle, the target state estimated information is fed back to the radar transmitter. With the minimum perceptual information entropy as the cost function, the optimal waveform is selected from the waveform library to match the target. The simulation results show that the algorithm is more accurate for the environment and target perception, and the adaptive target tracking algorithm with waveform selection is better than that with fixed waveform.

Key words:  Maneuvering targets tracking, Memory mechanism, Waveform library, Neural network