Research And Application System Development Of Intelligent Countermeasure Decision Model Based On Incomplete Information

Information is full of people`s daily life,people often need to make decisions on information,this process can be understood as information game.According to the completeness of the information in the scenario,the information game is divided into complete game and incomplete game.The scenario of incomplete information game is closer to people`s real life,and the study of the game decision behavior under incomplete information has a border application field,such as market auction,financial regulation,military deduction and so on.The paper based on the scenario of intelligent confrontation decision-making task under incomplete information,the agent model is designed to complete information situation assessment and optimal decision.Taking the intelligent confrontation decision model as the core content,the development of related application system is completed.The specific work of this paper includes: the study of the intelligent confrontation model with incomplete information,and the realization of the model`s situation information prediction module,heuristic resource decision module and agent confrontation deduction module.Based on the actual business needs of the intelligent countermeasure decision-making model,the requirement analysis of the relevant application system is completed.Based on the requirement analysis,the overall design and detailed design of the system are carried out,and the functional modules of the system,such as user information management,system scene operator information management,scene participant situation information prediction and resource decision-making,system participant countermeasure deduction,are realized.The SSM(Spring+Spring MVC+My Batis)framework of MVC mode is used to develop the system,the My SQL database is used to realized the system`s information management function,and the DCGAN(Deep Convolutional Generative Adversarial Networks)model and Res Net(Residual Network)model based on Python are used to complete the situation information prediction function of scene participants,the heuristic search are used to complete the optimal decision function under the huge search space,the MCTS(Monte Carlo Tree Search)are used to build a game tree to realize the confrontation between scene participants.Finally,the system function test is completed,and after the system goes online,the business login of the system is optimized and improved according to the feedback.At present,the system has been formally put into practice,each module of the system runs stably,and meets the business needs of system users.Decision-making auxiliary information is provided for participants in incomplete information scenarios,which greatly improves the success rate and decision-making accuracy rate of scenario agent confrontation deduction.The solution of the system has a good reference value for other similar problem of incomplete information scenarios.