The Research On Equipment Deployment Optimization Of Electro-optical System Countermeasures

Electro-optical system countermeasure is an important countermeasure method in future war,which relies on electro-optical weapon to intercept incoming targets.Electro-optical system countermeasure is a vital part of system countermeasure.The commander reasonably allocates our available electro-optical weapon according to the situation information of the attacking target when our crucial protection target is attacked by air attacking target.In Electro-optical countermeasure the Interference objects are strict and allocated electro-optical weapon under various optical constraints.Electrooptical guided weapon and unmanned aerial vehicles which have advantages such as fast response speed and autonomous control,so they are used in future wars increasingly.The electro-optical weapon used for defense becomes complicated and diversified with the development of optical technology.It is more and more difficult to give the deployment plan by analyzing the situation information of the incoming target and the electro-optical weapon equipment,so it is of great significance for the equipment employment research of the electro-optical system countermeasures.The jamming efficiency and weapon value consumption are considered in order to optimize the allocation and deployment of static electro-optical system countermeasure weapons.Based on optical spectrum matching,jamming field correlation,jamming distance effectiveness and jamming quantity,a multi-objective optimization model is established to maximize jamming efficiency against incoming targets and minimize the cost of our weapons.According to the current state of our photoelectric weapon and the interference of air attack target defined the reinforcement learning state,and the heuristic rules are used to design the reinforcement learning action,and the reward is defined by the objective function.The deep reinforcement learning algorithm based on envelope Q updating was used to solve the multi-objective optimization model,and the experimental parameters were determined by orthogonal experiment.Through the experiment,the deployment scheme of static optical system countermeasure weapon equipment was given.In order to optimize the distribution and deployment of weapons for dynamic electro-optic system countermeasures,the incoming targets are divided into multiple stages to strike our protection targets.Considering the constraints such as the exposure time of attack target,the availability of our optical weapons,optical spectrum matching,interfering distance effectiveness and correlation of interfering field of view,a dynamic multi-objective optimization model was established with the objectives of maximum interfering efficiency against incoming targets,minimum value consumption of our photoelectric weapons,and shortest response time.According to the state of our photoelectric weapon and the air attack target situation information define the state of the reinforcement learning at different stages.The heuristic rules were taken as the action and the transformed objective function was used to define the reinforcement learning reward.A deep reinforcement learning algorithm based on envelope Q updating has been used to solve the problem.According to the countermeasure process of electro-optical system,the evaluation index system of countermeasure effect is established,and TOPSIS algorithm is used to evaluate the distribution scheme.The experimental result shows that the algorithm is effective in solving the optimization problem of the dynamic electrooptical system countermeasures weapon allocation and the feasibility of the obtained allocation result.