In practical application, master and slave strapdown inertial navigation system (SINS) are not set in same location. This cause that the two specific forces measured respectively by the master and slave SINS are different. Existing polar Transfer Alignment (TA) algorithms neglect the lever-arm effect caused by the different measured specific forces. Therefore, the lever-arm effect will decrease the accuracy of these algorithms. In this paper, a TA algorithm considering lever-arm effect compensation is proposed. Based on “velocity plus attitude” matching method and grid frames, a Kalman Filter (KF) is used to estimate the misalignment angles in polar region. The lever-arm effect is also compensated to improve the TA accuracy. Simulation results have demonstrated that the TA algorithm proposed in this paper is effective in achieving the precise TA with lever-arm effect in the polar region.
Network intellectualization enables networks to adapt to the dynamic environment by perceiving, thinking, and interpreting the overall environmental changes like a human. However, the thinking capability of existing intelligent networks is generally limited or even absent. To further strengthen the thinking capability, we propose a novel network-node-embedded network cognition model based on a knowledge module to provide superior network intelligence. The proposed model is designed based on the OODA cycle, which is named from its components of “Observe”, “Orient”, “Decide”, and “Act” modules, and is widely applied to improve the network intellectualization. We firstly reorganize the OODA cycle by extending the perceptive range of the “Observe” module and classifying the other three modules as “Learning Analysis” and “Optimization Decision-making” modules. Furthermore, we innovatively introduce a “Network Knowledge” module using the knowledge graph, which constructs the correlation between the network and its overall environmental changes to generate effective decisions. Besides, the proposed model is designed to be directly embedded in network nodes to verify and execute decisions simultaneously, eliminating the time delay between them. The use cases that illustrate the applicability of the proposed model are presented, and simulation results can demonstrate that the proposed model is feasible and effective.