초록 ▼

A method, Kinematic Algorithm for Rocket Motor Apperception (KARMA), for processing radar returns for identifying the type of a missile target includes generating tracks representing the missile, and applying the tracks to a set of plural template-based filters, each representing one missile hypothe

A method, Kinematic Algorithm for Rocket Motor Apperception (KARMA), for processing radar returns for identifying the type of a missile target includes generating tracks representing the missile, and applying the tracks to a set of plural template-based filters, each representing one missile hypothesis, to generate plural sets of missile states, one set for each hypothesis. The missile states are processed to generate kinematic parameter likelihood values (LLHs). The LLH values for each filter hypothesis are normalized and weighted. A weighted maximum likelihood value (WMLH) is calculated for each hypothesis. The correct hypothesis is deemed to be the one having the maximum WMLH, thus identifying the missile type.

대표청구항 ▼

1. A method for identifying the type of a boosting missile, said method comprising: receiving radar returns from a boosting missile to generate a plurality of position measurements constituting a target missile track;providing a plurality of nominal kinematic missile rocket parameters including at l

1. A method for identifying the type of a boosting missile, said method comprising: receiving radar returns from a boosting missile to generate a plurality of position measurements constituting a target missile track;providing a plurality of nominal kinematic missile rocket parameters including at least nominal missile speed, altitude, and flight path angle, said nominal kinematic missile parameters defining templates for each of a plurality of missile types;applying each of said nominal kinematic missile rocket parameters for each missile of said plurality of missile types to a separate boost phase filter of a set of plural boost phase filters;applying said target missile track to each boost phase filter of said set of plural boost phase filters, to generate a plurality of state estimates, one from each filter; andproviding the output of each of said boost phase filters to at least one computer processor, said at least one computer processor performing the steps of:processing an output of each of said boost phase filters with the nominal kinematic missile rocket parameters applied to the boost phase filter, to generate a maximum likelihood value for each filter parameter and filter output; andselecting as a correct missile type that missile type associated with the filter having the maximum likelihood value. 2. A method according to claim 1, wherein said step of processing the output of each of said boost phase filters with the nominal kinematic missile rocket parameters applied to the boost phase filter comprises the steps of: calculating a speed likelihood parameter LLH for each filter hypothesis j of a plurality of filter hypotheses by MLH⁢Speed(j)=12⁢π⁡(PSpeed+PSTemplate)⁢ⅇ(-12⁢delSpeed2(PSpeed+PSTemplate)) where: PSpeed=Z.′_⁢Pz.⁢z.⁢Z_.Z._′⁢Z_. is a covariance of a magnitude of Ż; PSTemplate=(S⁢Template⁡(TALest-ⅆt)-S⁢Template(TALest)ⅆt)2⁢PTALest is a covariance of a speed from a nominal template lookup; delSpeed=Speed(j)−STemplate(TALest) is a speed difference between the filter value and the template value; anddt=0.1 is a small time increment. 3. A method according to claim 1, wherein said step of processing the output of each of said boost phase filters with the nominal kinematic missile rocket parameters applied to the boost phase filter comprises the steps of: calculating an altitude likelihood parameter MLHAlt for each filter hypothesis j of a plurality of filter hypotheses by MLHAlt⁡(j)=12⁢π⁡(PAlt+PAltTemplate)⁢ⅇ(-12⁢delAlt2(PAlt+PAltTemplate)) where: PAlt=Z′_⁢Pzz⁢Z_Z′⁢Z is a covariance of a target altitude; PAltTemplate=(Alt⁢Template⁡(TALest-ⅆt)-Alt⁢Template⁡(TALest)ⅆt)2⁢PTALest is a covariance of a speed from a nominal template lookup; PAltTemplate=(Alt⁢Template⁡(TALest-ⅆt)-Alt⁢Template⁡(TALest)ⅆt)2⁢PTALest is a covariance of a speed from a nominal template lookup; and dt=0.1 is a small time increment. 4. A method according to claim 1, wherein said step of processing the output of each of said boost phase filters with the nominal kinematic missile rocket parameters applied to the boost phase filter comprises the steps of: calculating a flight path angle likelihood parameter MLHFpa for each filter hypothesis j of a plurality of filter hypotheses by MLHFpa⁡(j)=12⁢π⁡(PFpa+PFpaTemplate)⁢ⅇ(-12⁢delFpa2(PFpa+PFpaTemplate)) where: PFpa is a covariance of a magnitude of a flight path angle calculated using PFpaTT*(T′pPZZTp+Tv′PŻŻTv+T′pPZŻTv+Tv′PŻZTp) where: TT=1(1-cos2⁢θ)*⁢Z_*⁢Z._;cos⁢⁢θ=Z_′⁢Z_.Z_⁢Z._TP=(Z_·Z._)⁢Z_Z_·Z_-Z_.;andTV=(Z_·Z._)⁢Z._Z_.·Z._-Z_ PFpaTemplate=(Fpa⁢⁢Template⁡(TALest-ⅆt)-Fpa⁢⁢Template⁡(TALest)ⅆt)2⁢PTALest is a covariance of a nominal flight path angle template lookup; delFpa=Fpa(j)−FpaTemplate(TALest) is a flight path angle difference between the filter value and the template value; anddt=0.1 is a small time increment. 5. A method according to claim 1, wherein said step of selecting as the correct missile type that missile type associated with the filter having the maximum likelihood value comprises the steps of: normalizing the likelihood parameters across the entire filter set to produce normalized kinematic likelihood parameters;weighting said normalized kinematic likelihood parameters to produce weighted normalized kinematic likelihood parameters; andcombining said weighted normalized kinematic likelihood parameters so that each filter hypothesis of the plurality of filter hypotheses has a normalized total maximum likelihood value. 6. A method according to claim 5, wherein said step of normalizing the kinematic likelihood parameters across the entire filter set comprises the steps of: computing a sum σk of the LLH values for each parameter across all filter hypothesis σk=∑j=1N⁢⁢LLHj⁡(k)where:k=LLH parameter, (speed, altitude, fpa, etc.);j=Filter hypothesis; andσk=sum of LLH values for parameter k for N filter outputs; andnormalizing the LLH values for each filter hypothesis j by LLHj⁡(k)⟸LLHj⁡(k)σk. 7. A method according to claim 5, wherein said step of weighting said normalized kinematic likelihood parameters comprises the steps of: selecting weighting parameters for each LLH parameter such that a sum of the weights equals unity or 1 by ∑i=1k⁢⁢Wi=1computing a weighted maximum likelihood value WMLH for each filter hypothesis WMLH⁡(j)=∑i=1k⁢⁢MLHⅈ⁡(i)⁢Wi where: j corresponds to filter hypothesis j; andi corresponds to LLH parameter i for filter hypothesis j. 8. A method according to claim 7, further comprising the step of identifying that filter hypothesis associated with the maximum value of weighted maximum likelihood to be a correct hypothesis. 9. A method according to claim 8, further comprising the step of identifying a missile type to be the missile type associated with said correct filter hypothesis. 10. A method for identifying a missile, said method comprising the steps of: receiving radar returns from a missile to generate a plurality of position measurements constituting a target missile track;providing a plurality of nominal kinematic missile parameters including at least nominal missile speed, altitude, and flight path angle, said nominal kinematic missile parameters defining templates for each missile type of a plurality of missile types;applying each of said nominal kinematic missile parameters for each missile type to a separate boost phase filter of a set of plural boost phase filters;applying said target missile track to each boost phase filter of said set of plural boost phase filters, to generate a plurality of state estimates, one for each missile type; andproviding the output of each of said boost phase filters to at least one computer processor, said at least one computer processor performing the steps of:processing an output of each of said boost phase filters with the nominal kinematic missile parameters applied to the boost phase filter, to generate at least one maximum likelihood value for each filter parameter and missile type;for each of said maximum likelihood values, computing a sum across all said missile types;normalizing the likelihood values for all missile types to generate normalized likelihood values;selecting weighting parameters for each likelihood value, such that the sum of the weights for each likelihood value equals unity;determining a weighted maximum likelihood value for each missile type; andselecting as a correct missile type that missile type associated with the filter having the maximum likelihood value. 11. A system for identifying a missile, said system comprising: a non-transitory machine-readable medium upon which is stored instructions anda computer processor executing said instructions for performing the following steps:receiving radar returns from a boosting missile to generate a plurality of position measurements constituting a target missile track;providing a plurality of nominal kinematic missile rocket parameters including at least nominal missile speed, altitude, and flight path angle, said nominal kinematic missile parameters defining templates for each of a plurality of missile types;applying each of said nominal kinematic missile rocket parameters for each missile of said plurality of missile types to a separate boost phase filter of a set of plural boost phase filters;applying said target missile track to each boost phase filter of said set of plural boost phase filters, to generate a plurality of state estimates, one from each filter;processing an output of each of said boost phase filters with the nominal kinematic missile rocket parameters applied to the boost phase filter, to generate a maximum likelihood value for each filter parameter and filter output; andselecting as a correct missile type that missile type associated with the filter having the maximum likelihood value. 12. A system according to claim 11, wherein said step of processing the output of each of said boost phase filters with the nominal kinematic missile rocket parameters applied to the boost phase filter comprises the steps of: calculating a speed likelihood parameter for each filter hypothesis of a plurality of filter hypotheses. 13. A system according to claim 11, wherein said step of processing the output of each of said boost phase filters with the nominal kinematic missile rocket parameters applied to the boost phase filter comprises the steps of: calculating an altitude likelihood parameter for each filter hypothesis of a plurality of filter hypotheses. 14. A system according to claim 11, wherein said step of processing the output of each of said boost phase filters with the nominal kinematic missile rocket parameters applied to the boost phase filter comprises the steps of: calculating a flight path angle likelihood parameter for each filter hypothesis of a plurality of filter hypotheses. 15. A system according to claim 11, wherein said step of selecting as the correct missile type that missile type associated with the filter having the maximum likelihood value comprises the steps of: normalizing the likelihood parameters across the entire filter set to produce normalized kinematic likelihood parameters;weighting said normalized kinematic likelihood parameters to produce weighted normalized kinematic likelihood parameters; andcombining said weighted normalized kinematic likelihood parameters so that each filter hypothesis of the plurality of filter hypotheses has a normalized total maximum likelihood value. 16. A system according to claim 15, wherein said step of normalizing the kinematic likelihood parameters across the entire filter set comprises the steps of: computing a sum of the speed likelihood values for each parameter across all filter hypotheses. 17. A system according to claim 15, wherein said step of weighting said normalized kinematic likelihood parameters comprises the steps of: selecting weighting parameters for each likelihood value such that the sum of the weights equals unity or 1. 18. A system according to claim 17, wherein the processor executes instructions for performing the further step of identifying that filter hypothesis associated with the maximum value of weighted maximum likelihood to be a correct hypothesis. 19. A system according to claim 18, wherein the processor executes instructions for performing the further step of identifying a missile type to be the missile type associated with said correct filter hypothesis. 20. A method according to claim 1, wherein said plurality of kinematic nominal missile rocket parameters includes acceleration. 21. A method according to claim 1, wherein said step of processing the output of each of said boost phase filters with the nominal missile rocket parameters applied to the boost phase filter comprises the steps of: calculating a speed likelihood parameter for each speed template of a plurality of speed templates,calculating an altitude likelihood parameter for each altitude template of a plurality of altitude templates, andcalculating a flight path angle likelihood parameter for each flight path angle template of a plurality of flight path angle templates. 22. A method for identifying the type of a boosting missile, said method comprising: receiving radar returns from a boosting missile to generate a plurality of position measurements constituting a target missile track;providing a plurality of nominal missile rocket parameters for each of a plurality of missile types;applying each of said nominal missile rocket parameters for each missile of said plurality of missile types to a separate boost phase filter of a set of plural boost phase filters;applying said target missile track to each boost phase filter of said set of plural boost phase filters, to generate a plurality of state estimates, one from each filter; andproviding the output of each of said boost phase filters to at least one computer processor, said at least one computer processor performing the steps of:processing an output of each of said boost phase filters with the nominal missile rocket parameters applied to the boost phase filter, to generate a maximum likelihood value for each filter parameter and filter output; andselecting as a correct missile type that missile type associated with the filter having the maximum likelihood value,wherein said step of processing the output of each of said boost phase filters with the nominal missile rocket parameters applied to the boost phase filter comprises the step of:calculating an altitude likelihood parameter for each filter hypothesis of a plurality of filter hypotheses. 23. A system for identifying a missile, said system comprising: a non-transitory machine-readable medium upon which is stored instructions, anda computer processor executing said instructions for performing the following steps:receiving radar returns from a boosting missile to generate a plurality of position measurements constituting a target missile track;providing a plurality of nominal missile rocket parameters for each of a plurality of missile type;applying each of said nominal missile rocket parameters for each missile of said plurality of missile types to a separate boost phase filter of a set of plural boost phase filters;applying said target missile track to each boost phase filter of said set of plural boost phase filters, to generate a plurality of state estimates, one from each filter;processing an output of each of said boost phase filters with the nominal missile rocket parameters applied to the boost phase filter, to generate a maximum likelihood value for each filter parameter and filter output; andselecting as a correct missile type that missile type associated with the filter having the maximum likelihood value,wherein said step of processing the output of each of said boost phase filters with the nominal missile rocket parameters applied to the boost phase filter comprises the step of:calculating an altitude likelihood parameter for each filter hypothesis of a plurality of filter hypotheses.