Simulation Modeling Technology Of Laser Jamming Effect On Space-based Early Warning System

In order to better evaluate the laser jamming effect on space-based early warning system,three key issues are studied in this paper: firstly,the simulation of the detection image of the space-based early warning system;secondly,the detection of infrared dim targets under complex background;thirdly,modeling of laser jamming effect on space-based early warning system.For the first issue,the focus of the research is to use limited data as much as possible to achieve the simulation of the detection image of the space-based early warning system.Based on the commercial high-resolution satellite infrared image,this paper studies the space-based early warning platform observation model,sensor imaging model,target trajectory movement and radiation model,laser interference spot simulation model and so on,thereby generating simulation images detected by space-based early warning systems.For the second issue,a new weak target detection algorithm based on improved anisotropic partial differential equation is proposed to suppress complex background clutter.Experimental results show that the algorithm has excellent performance in weak source target extraction,and can be used to simulate the Blue's target detection performance boundary.For the third issue,the paper puts forward the modeling method of laser jamming effect on space-based early warning system based on uniform design and XGBoost.Firstly,in order to effectively reduce the number of experiments,the simulation test method is optimized,and the uniform design is introduced into the simulation test of laser interference on space-based early warning system;secondly,the detection probability gain of target is calculated by using the improved anisotropic partial differential equation algorithm proposed in this paper,and it is used as the evaluation index and the test output result;finally,the XGBoost algorithm is used to analyze the test results,and the prediction models of each influencing factor and target detection probability gain are obtained.Simulation test results show that the method can construct a prediction model with higher accuracy under the condition of less test data.