| The omnidirectional penetration of combat simulation field has a profound guiding significance to combat in reality,making great changes in weapons and equipment,command system,etc.The situation of modern battlefield changes rapidly,and the real-time performance and effectiveness of combat decision-making affect the final trend of war.Therefore,it is of great significance to carry out reasonable intelligent decision-making simulation to assist combat commanders in making combat plans.This thesis focuses on tactical reasoning,battle path planning and fire distribution.Firstly,the tactical reasoning is carried out according to the combat situation of the enemy side.Secondly,the optimal path scheme is planned according to the characteristics of fire coverage.Finally,combined with the situation planning of the enemy and the optimal fire distribution scheme.Its main research contents are as follows:(1)The traditional tactical decision-making methods mainly carry out tactical reasoning through commanders and data-driven methods,which have a strong dependence on the experience and number of samples of combat commanders.To solve this problem,firstly,a hybrid tactical reasoning model based on individual fitness threshold,novelty and fitness hybrid driven differential evolution algorithm and K-nearest neighbor density peak clustering algorithm was proposed.Secondly,in order to improve the accuracy of case retrieval,a case retrieval strategy based on individual fitness threshold of novelty and fitness hybrid driven differential evolution algorithm was proposed to optimize the weight of feature attributes.Finally,the density peak clustering algorithm based on K-nearest neighbor is used to maintain the case base,and the redundancy and addition of cases are realized.(2)The previous path planning research methods usually consider the whole path,but in the process of combat path planning,it is necessary to divide the whole path into fast burst section and hidden burst section according to the characteristics of fire coverage,and carry out path planning for different path sections respectively.To solve this problem,firstly,A segmented path planning model based on the fusion of adaptive elite regeneration Mustang optimization algorithm and adaptive weight A* algorithm was proposed.Secondly,the segmented path planning model considering the characteristics of fire coverage was used to divide the path into two stages: rapid burst segment and covert burst segment.Thirdly,the Mustang optimization algorithm based on adaptive elite regeneration is proposed for the path planning of fast burst segment.Finally,an adaptive weight A* algorithm is proposed to optimize the path of covert burst segment.The final path scheme is obtained by fusing the path scheme of fast burst segment and hidden burst segment.(3)Relying on the previous fire distribution model and the experience of combat commanders in the fire distribution of infantry vehicles,although it can achieve a certain damage effect,it often causes a waste of resources.In order to solve this problem,firstly,an expert experience knowledge base for the firepower allocation of the infantry vehicle unit is constructed.Secondly,a search space optimization model is constructed based on the expert experience knowledge base.Finally,an improved social group optimization algorithm is used to output a fire distribution scheme that minimizes the ammunition consumption cost under the condition of achieving the damage expectation. |
