Research On Basic Theories And Key Technologies Of System Reconfiguration For Distributed Opportunistic Array Radar

Distributed opportunistic array radar(DOAR)is a new system radar.In DOAR,the platform stealth is taken as the core and the digital array is regarded as the base.The array elements are placed opportunistically at the electromagnetic open 3-D area over the entire platform.DOAR has flexible work modes and multiple tactical functions,and takes an opportunistic work mode by battlefield situation awareness.The high degree of digitization and modularization and the opportunistic array arrangement enable DOAR to flexibly conduct the system reconfiguration and resource management.Nevertheless,the complexity and changefulness of environment,the randomness and fuzziness of target information,the opportunity of arrangement of array elements,etc.will continuously bring the uncertainty to DOAR,and dynamically affect the demand that the tasks to be executed have to the radar resource.Under uncertain conditions,how to reasonably reconfigure and manage the hardware and software resources with the opportunistic work mode is the key to improve the detection and tracking performance of ODAR.Under this background,combined with the uncertainty theory,this thesis mainly studies the problems of antenna array aperture resource management,time resource management,power resource management and joint time and power resource management.The detailed work and contributions are summarized as follows:1.The source of various uncertain factors in ODAR is studied and analyzed.The uncertainty theory is introduced into ODAR,and the uncertain variables to characterize various uncertain factors and the uncertain programming to allocate system resource are elaborated.The general forms of radar system resource management model based on chance-constrained programming(CCP)conditioned on uncertain conditions are summarized.2.The antenna array aperture resource management algorithms of DOAR are studied.Firstly,due to the uncertainty of areas and numbers of working elements resulting from the opportunistic distribution of array elements,and the opportunity of choosing work modes and tactical functions,the fuzzy random variable is used to characterize the uncertainty of one-dimensional antenna array in pattern synthesis,and the CCP,which aims to minimize the mainlobe width error and peak sidelobe level under the condition that the number of working elements satisfies the chance constraint,is built.To solve the multi-objective model,a fuzzy random simulation is embedded into a fast and elitist nondominated sorting genetic algorithm(NSGA_II)to produce a hybrid intelligent algorithm,and then the Pareto optimal solution set is obtained.Compared with the one-dimensional antenna array,the two-dimensional antenna array aperture resource management algorithm builds the CCP model based on two antenna arrays,in order to minimize the total number of working elements conditioned on satisfying the constraints of beam parameters.Further on,the beams produced by pattern synthesis are applied to the target tracking process.This algorithm derives the Bayesian Cramér-Rao lower bound(BCRLB)of multiple target tracking,and builds the CCP model of resource management to minimize the maximal tracking BCRLB of all the targets.This algorithm can optimally allocate the antenna aperture,and accomplish the target tracking with as few array elements as possible.3.A time resource management algorithm based on the adaptive fuzzy logic priority is proposed for multiple target tracking.This algorithm adopts the fuzzy logic inference system to intelligently imitate the human decision-making process to calculate the priority of targets.Due to the time-varying environment and uncertain target information,the target radar cross section(RCS)is viewed as a random variable.Combined with the priority of targets,the time resource management model based on the CCP is built.A random simulation is embedded into genetic algorithm to produce a hybrid intelligent algorithm to predict the optimal dwell time allocation of each target at next sampling moment,and the predicted values are applied to unscented Kalman filter(UKF)to estimate the target states.With the decrease of the confidence level of the CCP,the time resource saving rate increases.After considering the priority of targets,the dwell time can be allocated to the targets more reasonably,and the total tracking time is brought down further.4.The power resource management algorithms of DOAR are studied for diffierent target tracking cases.Firstly,the target RCS is considered as a fuzzy variable,and the fuzzy CCP model of power resource management is built.This algorithm determines the range of the fuzzy variable by historical data and relavant experience,in order to overcome the disadvantage that the random variable results in the deviation due to the insufficient sample data.Secondly,compared with the aforementioned resource allocation algorithm in ideal conditions,a joint beams and power allocation scheme for multiple target tracking in cluttered environment is proposed.The radar system selects the appropriate tracked targets in terms of prior CRLB,and the measurement origin uncertainty(MOU)is measured by introducing information reduction factor(IRF).Considering the randomness of target RCS,the joint beams and power resource management model based on random CCP is finally built.This algorithm can accomplish the adaptive scheduling of beams,and optimally allocate the power over all the beams.Finally,a joint sampling interval and power allocation algorithm for maneuvering target tracking in multiple DOAR system is developed.This algorithm replaces the multimodal Markovian switching dynamics with best-fitting Gaussian(BFG)approximation,and derives the BCRLB-like of target tracking error.The radar system determines the adaptive sampling interval according to prior CRLB-like,and accomplishes the reasonable allocation of power among the DOARs in line with BCRLB-like.This algorithm can adaptively adjust the sampling interval of maneuvering target,and reasonably allocate the transmitting power.5.A joint time and power management algorithm of DOAR in the search and tracking is studied.Initially,the relevant parameters which can be optimized are analyzed and studied for the search task,and then a concrete example is cited for support.Based on this,a joint beam dwell time and transmitting power resource management algorithm in the search and tracking is proposed.In the search process,this algorithm uses Riemannian manifolds to represent the beam dwell time of each direction,and it can compensate for the gain loss caused by the increase of the scanning angle.Minimizing the transmitting power of the search beam conditioned on satisfying the detection performance and search data rate can spare more power for the tracking process to minimize the tracking BCRLB.This algorithm can reasonably allocate the time and power resource between the search and tracking,and guarantee the equal power detection of each direction in the search process,and reasonably allocate the power over the targets in the tracking process.