Study On The Method Of Robust Estimation Of Tactical Air Target Intention Based On Trajectory Information

Target intention estimation plays an important role in assisting commanders of battlefield command and also plays a pivotal role in the entire war. Current intention estimation methods is still insufficient in terms of stability, especially air target.Because of its flexible power driven, there exists jump phenomenon in the intention estimation results, which would affect the judgment of war committed by the commander. In this paper, to achieve a stable air target intention estimation result,we do some work respectively from the target steady course, target maneuvering type recognition and fusion algorithm of target intention.The main work are as follows:As aerial target has the characteristics of flexible and the time course changing rapidly, study an aerial target directional stability estimation method based on SOM network. Based on the characteristics of SOM network clustering and feature extraction of target trajectory, get a steady target estimation result. It is proved by the simulation that the method can exclude the interference of noise and fluctuations and get a correct and stable estimation of air targets.For short track segment identification maneuver is susceptible to noise, research the method of air target trajectory maneuvering type recognition and give a feature extraction strategy against airborne target trajectory characteristics.Take the extracted strategy as the input of BP neural network and get the recognition of target machine movement type. The simulation results show that this method can get correct recognition of air targets motor type identification under the impact of noise,To solve the influence of subjective and objective mixed input towards stable target intention result, propose a robust goal intention estimation method based on fuzzy stochastic Bayesian network. By defining fuzzy random variables, construct the target steady fuzzy stochastic Bayesian network model and reasoning robust intention estimation. Simulation results verify the advantages of fuzzy stochastic Bayesian networks in the robust estimation of target intention. The robust estimation of target intention based on primary and objective mixed input information is realized.A comprehensive set of the three algorithms discussed before are developed to estimate the target intention of the system. Through specific simulation, verify the the intention estimation system feasibility, accuracy, and advantages of intention estimation result in terms of stability.