Research On The High Precision Trajectory Tracking Of Boost-glide Vehicle

In recent years,the research on boost-glide vehicles has gradually matured.The world's aerospace powers have carried out a large number of related tests,and the boost-glide vehicles of some countries have entered the state of combat duty.Different from ballistic targets with fixed motions,boost-glide vehicles have the characteristics of multiple flight stages,strong maneuverability,and variable maneuvering forms.Its strong penetration capability poses a great challenge to the existing missile defense system.The development of aircraft interception technology is of great urgency and is of great strategic significance for protecting our country's national security and the safety of people's lives and property.The boost-glide vehicle trajectory tracking technology provides target information support for the entire interception process.How to improve the tracking accuracy of the boost-glide vehicle is the primary problem in the development of the interception technology,and it is the most important part that affects the success or failure of the mission.In this paper,aiming at the two main flight stages of boost-glide vehicle boost and gliding,the ballistic characteristic analysis,motion model establishment,and high-precision trajectory tracking method are studied.The specific research contents include:1)Based on the analysis of ballistic characteristics,the construction of the high-precision motion model of the boost-glide vehicles was completed.First of all,starting from the force analysis of the boost phase of the boost-glide vehicle,the ballistic characteristics of the vehicle are summarized.Furthermore,reasonable assumptions and simplifications were made on the motion differential equation of the boost section of the aircraft,and a motion model of the boost section was established which had the advantages of both in line with the characteristics of motion and the advantages of simple form.Then,aiming at the problem of strong maneuverability and variable maneuvering forms in the glide section,the characteristics of common maneuvering forms in the glide section are analyzed and a complete model set is established.Among them,a "current" statistical model was established to describe the form of equilibrium glide and lateral maneuvering,and the sine wave autocorrelation adaptive mean value random process with exponential decay term was used to describe the maneuvering target acceleration in the jump glide phase,therefore,proposed a non-zero mean attenuation oscillating sine wave model,which is more in line with the characteristics of jump glide.The work in this section lay the foundation for the follow-up trajectory tracking method research.2)In order to improve the trajectory tracking accuracy of the boost-glide vehicle in the boost phase,the working principle of the boost phase tracking sensor-space-based infrared early warning satellite was analyzed,the measurement noise of which was proposed as quantized noise.A quantized noise model was established and analyzed.The quantization noise characteristics are analyzed.Traditional filtering methods are based on Gaussian noise assumptions and face the problem of poor tracking accuracy under quantization noise conditions.Based on the Kalman filter framework,this paper first derives the state expectation and covariance calculation method under quantized measurement,and proposes the quantized extended Kalman filter method;then based on the Bayesian estimation framework and particle filter idea,combined with the quantized measurement probability Density function,a nearly optimal quantized genetic resampling particle filter method is proposed,which introduces the idea of natural genetic evolution for resampling to avoid particle degradation and ensure particle diversity.Aiming at the difficulty of multi-dimensional Gaussian integration caused by quantization noise,a multi-dimensional Gaussian numerical integration method is proposed by combining Genz transformation and pseudo Monte Carlo integration method.Finally,mathematical simulations verify the superiority of the proposed algorithm in terms of tracking accuracy.3)In the scene of space-based infrared early warning satellite network tracking boost-glide aircraft,distributed filtering methods based on hybrid consistency are proposed for angle measurement and pixel coordinate measurement,respectively.Among them,a distributed filtering framework is designed based on the hybrid consensus method,which collects measurement information and estimates the target state in a decentralized manner,and achieves high system accuracy because of the hybrid consensus method that combines measurement consensus and information consensus methods.In order to solve the strong nonlinear problem of the angle measurement equation,from the perspective of measurement equation conversion,the nonlinear measurement equation is converted into a pseudo measurement equation,which is linearly related to the aircraft state,and the conversion error is compensated.Combining with the framework of hybrid consensus distributed information filtering,a filtering method of angle measurement hybrid consensus distributed pseudo-measurement information filtering is proposed.Under the pixel coordinate measurement,the quantization noise is processed based on the Sheppard correction method,the system nonlinearity is processed by the fifth-order spherical-radial volume rule,and the hybrid consensus distributed information filtering framework is integrated,therefore,proposes the pixel coordinate measurement hybrid consensus distributed information filtering.A trajectory tracking simulation of the boost phase in the sensor network is designed to verify the consensus speed and accuracy of the algorithm.4)The boost-glide vehicle has strong maneuverability and various maneuvering forms in glide phase,and the active radars are vulnerable to false data injection attacks,resulting in poor accuracy of traditional trajectory tracking algorithms.To this end,a variable structure multi-model filtering method considering false data injection attacks is proposed.Firstly,a false data injection attack model and an active radar measurement model under attack conditions are established.A pseudo-measurement equation for attack parameters is designed to make the attack parameters estimable.The processing idea of“attack detection-attack parameter estimation-attack parameter compensation” is proposed to deal with scenarios where false data injection attacks exist for a long time.Then,based on the idea of merging the model set activation and model set termination operations,the model set adaptive strategy was designed according to the edge model probability,and the attack parameter compensation was completed by modifying the innovation information.A new variable structure multi-model method is proposed to reduce the model competition in the traditional variable structure multi-model method and compensate the impact of attack parameters.Finally,the superiority of the proposed algorithm in trajectory tracking accuracy is analyzed,and the influence of attack intensity on the proposed algorithm is analyzed.