Abstract:

To solve the control problem of radiation risk in the multi-sensor multi-target tracking, a scheduling algorithm of active sensors for tracking task requirement is proposed. Firstly, combined with unscented Kalman filter (UKF), the sensor scheduling policy which only considers the tracking task requirement is given. Then the radiation model based on partially observable Markov decision process (POMDP) is formulated, and the hidden Markov model (HMM) filter is used to dynamic update sensor radiation. Finally, considering the sensor constraint, the sensor scheduling problem of radiation risk control is transformed to a nonlinear optimization problem. Simulation results prove the effectiveness of the proposed algorithm.