摘要
"深绿"计划是美国国防部高级研究计划局支持的一项面向美国陆军、旅级的指挥控制领域的研究项目."深绿"采用基于草图交互、模型求解与态势预测和指挥系统集成等关键技术,通过信息汇聚、职能汇聚、过程汇聚,将人工智能集成到作战辅助决策中.但由于面临态势与决策的输入输出问题、对"未来"的仿真问题、数据的决定性问题、计算机与人协同问题和陆军的难题等,导致"深绿"没有达到预期效果.作战辅助决策未来发展要依靠智能技术的进步、深度学习技术和从数据到决策技术,核心在态势理解、行为模拟、决策选择和结果反馈等,要对不同类型和层次的决策辅助找到不同方法.
Deep Green is a C＆C project supported by DARPA, which is a brigade level system used for USA Army. Deep Green use AI in operation decision support systems by convergence of information, function and procedure, which is supported by some key technologies such as sketch to plan, model solving and situation forecasting, system integration technology. But Deep Green doesn’t reach the expected effect, it encounters many difficulties, for example input and output of situation and decisions, simulation of future, data problems, coordination between human and computer, Army difficulties. The future of decision support system is dependent on technology of AI, deep learning and data to decision, finding the way to solve different problems. The key point lies in awareness of battlefield situation, behavior modeling, decision selection and feedback.
出处
《指挥与控制学报》
2016年第1期22-25,共4页
Journal of Command and Control
关键词
作战
辅助决策
人工智能
战场态势
operations, decision support system, artificial intelligence, battlefield situation
