Research On Combat Intention Identification Method Of Carrier-based Aircraft Based On Weapon Utilization Knowledge

Intention identification is widely studied and applied in many fields like natural language processing,automatic driving,human body assistance and military.In the military field,the problem is mainly referred to identifying the combat intention.Traditional combat intention identification models utilize machine learning methods to recognize the intention with the partial motion information of target objects as the input.With the development of modern combat theory,the main combat nowadays has already been turned into the combat based on So S(Systems of Systems),and the battlefield situation is becoming more and more diverse.In the maritime battlefield,which is one of the main battlefields,the Aircraft Carrier Battle Group plays a vital role,and its attack operations are mainly triggered by carrier-based aircrafts.The key problem in combat intention identification is how to fully exploit the battlefield information and identify the combat intention of carrier-based aircrafts,meanwhile improve the interpretability and accuracy of identification.The main work of this dissertation is as follows:(1)The application and characteristics of carrier-based aircraft in systematic operation are analyzedOn the background of modern war,the composition and application of Aircraft Carrier Battle Group is analyzed.Then,in view of carrier-based aircraft,which is the main practitioner of Aircraft Carrier Battle Group in operational mission,the combat application is analyzed from three aspects: main combat style,basic combat process and main characteristics.(2)A weaponry utilization knowledge structure framework is builtGuided by recognition model on entities and concepts in the systematic combat battlefield from Do DAF 2.0 concept data model and JC3 IEDM probability data model,The content of weaponry utilization knowledge and list the elements of weaponry utilization knowledge is analyzed.The elements are then utilized to represent the knowledge.From the perspective of weaponry system,the weaponry utilization knowledge structure framework is proposed that takes different types of users and different application stages into consideration.The model contains 6 perspectives and32 types of weapon utilization knowledge in total.According to the architecture of weaponry utilization knowledge structure framework,then the deployment knowledge for carrier-based aircrafts is defined,and the corresponding domain ontology model is designed.With the carrier-based aircrafts deployment domain ontology model,all types of carrier-based aircrafts utilization knowledge can be represented.(3)A combat intention identification model for Carrier-based Fighter is proposedThe procedure of identifying carrier-based aircrafts combat intention is defined.In order to better identify the combat intention,the utilization knowledge of carrierbased aircrafts is transformed into corresponding rules,and introduce the method that constructs logical architecture and defines probability distribution for utilization rules based on utilization knowledge.The carrier-based aircraft utilization knowledge domain ontology model is then extended to the probability-based carrier-based aircraft utilization knowledge domain ontology model.Further,a transformation model is proposed that generates multi-entity Bayesian frags for intention identification from domain ontology model.In detail,the bottom-up construction algorithm is employed to generate the network from the fragments,and accurate inference algorithm is used for network reasoning.(4)Case studyFirst,the settings of our case is introduced,and the probability-based carrier-based aircraft utilization knowledge is designed domain ontology model.Then the multientity Bayesian frags library for carrier-based aircrafts combat intention identification is constructed.Further,the elements of weaponry utilization knowledge are utilized to standardize the available battlefield information,the processed data is then fed into the multi-entity Bayesian network for intention identification.The experimental results show that the proposed combat intention identification model can fully use the battlefield information and accurately recognize the combat intention of carrier-based aircrafts in systematic operation.