天基信息支援体系离散事件仿真与效能评估

To solve the problem that accuracy and real-time performance cannot be balanced in the effectiveness evaluation of space-based information support system, based on the system simulation, a method of constructing the performance evaluation model based on discrete event simulation is proposed. Firstly, a discrete event simulation model is established based on simulation tools to simulate the structure and operation process of the system, and an evaluation index system is constructed based on the simulation model to generate performance evaluation data samples. Then, by using the powerful expressive ability of deep learning for complex mapping, the performance evaluation proxy model is constructed based on the deep belief network, and the training and parameter optimization of the network arc realized by combining unsupervised pre-training and supervised tuning. Finally, the evaluation model verification experiment is carried out on the test set samples generated by simulation. The results show that the proposed method has a high degree of reduction to the original evaluation model while calculating quickly.