The Research On Cooperative Detection And Situation Awareness In Manned-unmanned Collaborative System

In recent years,the manned-unmanned collaborative system is widely used because of its advantages in effectiveness and safety.Among the studies in manned-unmanned collaborative system,the cooperative detection and the situation awareness(SA)are both important parts in application.The cooperative detection refers to the process to detect the environment by several reconnaissance platforms while the situation awareness refers to the process of cognition and perception of the environment based on the cooperative detection.Specifically speaking,the capacity of cooperative detection and situation awareness reflects the comprehension ability of the agent and acts as the basis for the fast reaction to emergencies.Compared to the cooperative detection and SA problems in traditional systems,there are several new factors in manned-unmanned collaborative system as follows:(1)During the process of cooperative detection,the detection performance and network lifetime are closely related to the topology of the detection network.However,current networking technology merely takes performance and lifetime in consideration at the same time,which results in a reduction in the reliability of the system.(2)During the process of situation awareness,the input of the system performs hierarchically because each agent differs in intelligence level and task complexity,which is difficult for current SA system to deal with.(3)In the mannedunmanned collaborative system,comprehensions and decisions of the commander mostly rely on the results of cooperative detection and situation awareness.However,the process of both are conducted dynamically and swiftly,which prejudices the rapid information processing and the efficient decision-making.In other words,cooperative detection and situation awareness algorithms in the manned-unmanned collaborative system need to be effectively improved.Specifically,we take the clustering algorithm,the SA algorithm and the messagepush system into consideration and propose several corresponding innovative schemes.In particular,the contributions of the thesis are listed as follows:(1)Considering the problems caused by the limited system resources and multi-targets in the environment detection,we propose an energy efficient clustering algorithm that prolongs the lifetime and retains the detection performance as well.Specifically speaking,the similarity criterion based on the covariance ellipse is proposed to build the initial cluster.Furthermore,the energy criterion is proposed to optimize the topology after the initialization of the network,which takes both the communication consumption and the computational consumption of the node into consideration.Finally,we conduct several experiments to verify the performance of the clustering algorithm.(2)Considering the differences in intelligence level,task complexity and time span of each agent,we design the hierarchy SA system and apply it to the actual combat scene,which ensures the effectiveness of the manned-unmanned collaborative system.In the hierarchy SA system,the inputs of the environment are categorized as the emergency level,regular situation level and human intervention level.Accordingly,we implement the inference procedure of each level based on the characteristics of the inputs and the effectiveness of the SA system is verified by numerous simulation.(3)Considering the man-machine interaction in manned-unmanned collaborative system,we design the message-push system because of the mass and dynamic information in application,which relieves the commander effectively.In specific,the message-push system is categorized as the threat-push module based on information entropy and the decision-push module based on rule set.Furthermore,several experiments are conducted to verify the performances of the message-push system.