248 results about "Orbit determination" patented technology

The invention discloses a low-orbit target orbit determination method and device. The method comprises the following steps: acquiring observation data at different moments; determining a target position vector according to the observation data and a pre-established target position relation model; the target position vector comprises a first target position vector corresponding to an initial observation moment, a second target position vector corresponding to an intermediate observation moment and a third target position vector corresponding to a termination observation moment; the target position relation model is determined based on a second target position vector corresponding to the intermediate observation moment and a position vector of the observation platform relative to the earth center; and determining a predicted target orbit according to the initial observation moment, the first target position vector, the termination observation moment and the third target position vector. According to the method, the target orbit parameters can be determined only based on the observation data at different moments under the conditions that an initial value is not needed and perturbation errors exist, and the accuracy and stability of low-orbit target orbit determination are improved.

The invention discloses a space-based directional observation method and device based on orbit precision evaluation. The method comprises the following steps: acquiring fixed star angular distance measurement data information, a Ka-band inter-satellite link observation value and an L-band satellite-ground link observation value of a Beidou satellite; using the information to construct a multi-dimensional observation matrix and carrying out resolving processing to obtain satellite position parameters corresponding to the three space-based directional observation modes; and performing comparison processing on the satellite position parameters by taking an MGEX precise ephemeris as a comparison baseline to obtain an optimal observation mode. Therefore, three space-based directional observation modes are provided, the orbit precision is used as an evaluation criterion, an observation mode system with high adaptability and high practicability is formed, and the orbit determination precision of the Beidou satellite is improved.

The invention discloses an unknown target short-arc high-precision initial orbit determination method based on filtering enhancement, and the method comprises the steps: firstly, carrying out the adaption of an initial orbit determination method according to an observation data type, after obtaining a plurality of groups of initial orbit candidate solutions, determining an initial orbit optimal solution according to a measurement residual error minimum principle, and converting the initial orbit optimal solution into a position vector and a velocity vector; configuring a dynamical model and parameters by combining a target track type and physical characteristics; establishing a target state equation based on the model, and introducing a process noise compensation model error; and constructing an observation equation by combining observation platform position information. In order to guarantee effective starting of the filter, parameters such as the orbit state, the state error covariance, the process noise and the measurement noise need to be calculated and initialized in advance. And finally, in combination with the state equation and the observation equation, the target orbit state is predicted and updated through an EKF technology until all observation epochs are traversed, and high-precision target initial orbit information is obtained.

The invention provides a satellite Ka wave band inter-satellite link antenna phase center change calibration method and system, and the method comprises the steps: collecting L wave band observation data and Ka wave band inter-satellite link observation data of a ground observation station, carrying out the gross error elimination, cycle slip detection, epoch normalization and combination, and generating a clock error-free and geometry-free observation quantity; analyzing the variation range of a nadir angle and an azimuth angle of the observation data, constructing a piecewise linear original PCV function model taking the nadir angle and the azimuth angle as independent variables, and synchronously resolving original PCV parameters of a satellite L / Ka wave band in a joint orbit determination process; decomposing the original PCV into a smooth PCV and a corresponding Z-direction PCO by adopting a least square separation technology, performing interpolation on a PCV value based on a discrete grid, and establishing a final PCV grid correction model; joint orbit determination error correction is introduced into the PCV correction model, a new round of joint orbit determination is carried out, the satellite L / Ka waveband PCO correction amount is calculated, and high-precision in-orbit calibration of the satellite Ka waveband inter-satellite link antenna phase center change is achieved.

A method for calibrating a PCO and a hardware delay of a downlink navigation antenna of an LEO satellite Ls includes: calculating, by using a precise orbit determination and timing result and ground calibrations, a ground calibrated phase center orbit initial value of the downlink navigation antenna and a ground calibrated satellite clock bias initial value; calculating, based on the ground calibrated phase center orbit initial value and the ground calibrated satellite clock bias initial value, a correction amount of the downlink navigation antenna by separating or combining GNSS signals and LEO satellite downlink navigation signals; and correcting, based on the correction amount of the downlink navigation antenna, the PCO of the downlink navigation antenna and the hardware delay of the downlink navigation antenna, so as to realize in-orbit calibration of the PCO of the downlink navigation antenna and the hardware delay of the downlink navigation antenna.

The invention provides a low-orbit constellation autonomous orbit determination method and system based on Beidou B2b and space between satellites, and belongs to the technical field of low-orbit navigation constellation orbit determination. Comprising the following steps: acquiring Beidou B2b correction data of a low-orbit satellite and correcting the Beidou B2b correction data to obtain an observation data set; obtaining original inter-satellite distance measurement data between two low-orbit satellites by using a bidirectional observation mode, and obtaining an inter-satellite distance solution; a Kalman filtering model is constructed based on the observation data set and the inter-satellite distance solution, and a dynamic orbit is predicted and corrected; and converting the Kalman filtering model into a square root information filtering model, and outputting a high-precision orbit solution through real-time joint filtering. According to the invention, on the basis of reducing calculation delay and burden, completely autonomous navigation and high-precision orbit determination based on inter-satellite links can be realized.

The invention provides an earth rotation parameter determination method and system based on a GNSS data stream, and the method comprises the steps: achieving the purpose of increasing the estimation of earth rotation parameters in a process of achieving precise orbit determination through employing square root information filtering, regarding the acceleration values of polar shift and UT1 change as white noise in time updating, and forming a state equation; in the measurement updating process, a UT1 value in an ultra-fast ERP forecasting product provided by the IGS is added to serve as virtual observation to form a virtual observation equation, so that UT1 parameters and orbit parameters are decoupled and can be estimated, stable estimation of real-time earth rotation parameters is achieved, the precision is superior to that of earth rotation parameter forecasting products provided by the IGS and IERS, and the accuracy of the earth rotation parameter forecasting products provided by the IGS and the IERS is improved. And meanwhile, a precise track product with higher precision is obtained, and the real-time high-precision ERP application requirement can be better met.

The invention relates to the cross technical field of satellite navigation, satellite dynamics and the like, in particular to a dynamic orbit determination method based on integer ambiguity fixation, which can obviously improve orbit determination and extrapolation precision, and comprises the following steps: establishing a dynamic orbit determination model, establishing a GPS observation model, and then performing integer ambiguity fixation IAR, a single-difference IAR method adopting single receiver ambiguity correction specifically comprises the steps that firstly, GPS carrier phase observation is divided into continuous tracking segments for processing, for each tracking segment, an average Hatch-Melbourne-Wubbena (HMW) combination is constructed by means of double-frequency pseudo-range and carrier phase data, then cross-segment inconsistency is corrected, integers of wide lane ambiguity are achieved, and a single-receiver ambiguity correction result is obtained. And the corrected wide lane ambiguity is combined with an IF ambiguity floating point estimation value obtained by a least square method for use, and finally, the IF ambiguity difference is constrained as an integer value to realize the fixation of the single-difference ambiguity, and compared with the prior art, the orbit determination precision can be remarkably improved.

The invention discloses a method for carrying out correlation and orbit determination on a space target based on a double-star optical observation result, and the method comprises the steps: obtaining the observation vectors of two observation stars for a certain space target in the respective visual fields, calculating the shortest distance between the observation vectors, and if the distance is always smaller than a threshold value in a time sequence, determining that the space target does not exist. And if not, determining that the two observation stars observe the same space target, calculating a midpoint of two observation vector common vertical lines of the same space target by the two observation stars as an orbit determination point, and performing orbit determination on the space target based on an orbit determination point sequence. According to the method, a low-cost space-based optical orbit determination mode is adopted, space target association is completed through an algorithm with low calculation cost, target consistency is judged based on observation vector distance constraints, a complex model or a large amount of calculation power is not needed, efficient processing can still be performed when the number of targets is large and the data size is large, and the cost and efficiency advantages are remarkable.

The invention provides a space debris angle-measurement-only short-arc rapid initial orbit determination method and a space debris angle-measurement-only short-arc rapid initial orbit determination system. The method comprises the following steps: S1, acquiring space debris angle-only segmental arc observation data; and S2, performing polynomial fitting on the observation data. And S3, calculating the pointing vector of the space debris relative to the observation station in the J2000 geocentric celestial coordinate system at the midpoint moment of the arc section. And S4, calculating the change rate of the pointing vector at the midpoint moment of the arc section. And S5, determining the distance at the midpoint moment of the arc section and the initial value of the distance change rate. And S6, calculating an objective function corresponding to the initial value. Step S7, if the target function does not reach the convergence condition, calculating a target function gradient corresponding to the current distance and the distance variation rate, updating the current distance and the distance variation rate, and executing steps S6 to S7; and if the target function reaches the convergence condition, calculating the corresponding position and speed. The implementation process is simple and easy to implement, and the method can be effectively applied to the space debris angle-only short arc initial orbit.

The invention discloses a point cloud intersection-based space target optical segmental arc rapid association method. The method comprises the following steps: 1, acquiring optical segmental arc data and constructing a constraint admissible domain; 2, generating a track point cloud; 3, forecasting parallel orbit point cloud; 4, fast point cloud intersection is carried out; and 5, determining an initial orbit based on point cloud intersection. Through the above process, the space target rapid segmental arc correlation method based on point cloud intersection is provided, an admissible domain is constructed by two optical observation segmental arcs and orbit point clouds are generated, and then the point clouds of the two segmental arcs are forecasted to the same moment through a parallel orbit point cloud forecasting technology. Quickly locking an overlapping region of the two point clouds by adopting a minimum bounding box method, and judging the relevance between the point clouds through a mahalanobis distance; and finally, based on the Bayesian estimation principle, carrying out initial orbit determination on the successfully associated arc segments.

The invention discloses a space target multi-satellite collaborative observation task planning method based on a greedy algorithm, which belongs to the technical field of spaceflight, and comprises the following steps: firstly, carrying out visibility analysis on a space target, and comprehensively considering factors such as earth shadow shielding, illumination conditions, detection distance and target characteristics; designing a method for calculating the visibility of the optical satellite to the space target; secondly, designing a hybrid heuristic rule, and preprocessing the visible window set; finally, a task planning method based on conflict resolution and priority dynamic adjustment is designed, and efficient arrangement of observation tasks is guided through a preset hybrid heuristic rule. According to the method, observation requirements of space target orbit determination are strictly considered, a multi-satellite multi-target multi-frequency observation plan can be rapidly formulated, observation of important targets is preferentially guaranteed on the premise of limited resources, meanwhile, other targets are considered, the requirements of space situation awareness tasks are met, and the method has practical engineering application value.

The embodiment of the invention discloses a low-orbit navigation satellite clock correction calculation method and device, electronic equipment and a storage medium, and the method comprises the steps: carrying out the ionosphere-free combined precise orbit determination of observation data received by an LEO through the satellite orbit parameters and clock correction products of a medium-high-orbit GNSS satellite, and obtaining the satellite orbit parameters of the LEO and the clock correction of a signal receiving end; based on the double-frequency mixed observation data, jointly resolving the signal transmitting end clock error of the LEO in the ground visible segmental arc; mathematical models are established for the signal receiving end clock error of the LEO and the signal transmitting end clock error of the LEO in the ground visible arc section respectively, and the difference of the clock error models is extracted; and according to the difference of the clock difference model, converting the clock difference of the signal receiving end of the LEO into the clock difference of the full-arc signal transmitting end of the LEO.

The invention provides a space target cataloguing library maintenance method and device fusing post-event multi-source observation data, and belongs to the field of space situation awareness. The method comprises the following steps: acquiring post observation data of an observed space target and preprocessing the post observation data; calculating an observation arc section by using the preprocessed observation data, taking an obtained initial orbit determination result as an orbit feature of the observation arc section, further matching the observation arc section with orbit data of a target in a cataloguing library, and taking the successfully matched observation arc section as an associated arc section of the target; and performing precise orbit determination by fusing a plurality of associated arc segments of the target, and updating and maintaining the cataloguing library by using a precise orbit determination result. According to the method, data preprocessing can be rapidly carried out according to the input multi-source observation data, association matching of the data and the space target is completed, finally, precise orbit determination is carried out by fusing the multi-source observation data, the calculation efficiency and the process completeness are guaranteed, and maintenance of a space target cataloguing library is achieved.

The invention discloses a giant astronomical base partition centralized orbit determination and time synchronization method and system under the support of inter-satellite measurement, and relates to the field of giant astronomical bases, and the method comprises the steps: carrying out the space partitioning of all satellites in a giant astronomical base through employing a specific topological configuration, and obtaining a plurality of regional grids; node satellites in the current regional grid perform centralized Kalman filtering measurement updating by using all observation data to obtain a first updating result; sending the first updating result to each satellite in the current regional grid, and controlling each satellite to perform distributed Kalman filtering time updating to obtain a second updating result; the second updating result is sent to the node satellites in the current regional grid; and controlling the node satellites in the current regional grid to send all the received second update results to the node satellites in the adjacent regional grid, and repeating the steps to complete orbit determination and time synchronization of satellites of multiple batches. According to the invention, precise orbit determination and time synchronization of satellites in a giant satellite base are realized.

The invention discloses a wavelet noise reduction multi-path extraction method for improving a threshold function, which can be used for effectively extracting and weakening multi-path errors of a global navigation satellite system (GNSS), solves the problem that signals of wavelet coefficients are discontinuous or distorted at a preset threshold, improves the traditional wavelet noise reduction effect and multi-path error modeling precision, and improves the accuracy of multi-path error modeling. Therefore, the influence of multi-path errors on GNSS positioning and orbit determination precision is effectively weakened, a feasible solution is provided for realizing high-precision positioning and orbit determination in a complex environment, and a solid foundation is laid for establishing 1mm space-time reference in the future.

The invention discloses a GEO binocular orbit determination method and system based on inter-satellite communication, and belongs to the technical field of aerospace, and the method comprises the steps: predicting the priori estimation of the state variable of a binocular orbit determination system through unscented transformation according to an orbit dynamics model of the binocular orbit determination system; simulating and generating space-based observation data based on an observation model of a binocular orbit determination system; according to the principle of unscented Kalman filtering, the priori estimation of the state variable and the space-based observation data are fused, and the posteriori estimation of the state variable of the binocular orbit determination system is obtained; repeatedly iterating until the posteriori estimation orbit determination precision of the state variable of the binocular orbit determination system meets the task requirement; according to the method, target observation data is obtained through binocular observation, observation data between observation satellites is obtained by considering inter-satellite communication, prediction data and the observation data are fused through an unscented Kalman filtering algorithm, and a high-precision orbit determination method is provided for a GEO spacecraft.

The embodiment of the invention discloses a multi-platform common-view orbit determination method for an earth-moon space target, and relates to the technical field of space target orbit determination, and the method comprises the steps: enabling a plurality of observation platforms to observe the same earth-moon space target in the same observation time period, and obtaining a motion data sequence of the observation platforms in the observation time period, obtaining an observation data sequence obtained by observing the earth-moon space target; determining a relative position relationship between the observation platform and the earth-moon space target, constructing a first observation model for the earth-moon space target, and calculating the position and speed of the earth-moon space target at each moment in a corresponding observation time period; and constructing a second observation model, and calculating to obtain an orbit arc section of the earth-moon space target in the corresponding observation time period. The method is small in type limitation of earth-moon space targets, and can adapt to multiple types of orbits; after the initial orbit is successfully determined, orbit refinement can be carried out again by combining original observation data, so that the initial orbit precision is improved.

The invention discloses a simulation system and a simulation method for a deep space tracking measurement task, and relates to the technical field of deep space exploration tracking measurement. The system comprises a task simulation and visualization subsystem which is used for carrying out environment parameter configuration and task state control on a deep and far space simulation task and outputting operation information; the data set construction and verification subsystem is used for generating a navigation verification data set and performing functional performance simulation verification; and the algorithm performance evaluation subsystem is used for constructing a performance evaluation index system and evaluating the tracking measurement algorithm. According to the invention, integration of tracking measurement technologies such as space reference, one-way measurement, autonomous orbit determination, landing and patrol navigation and the like of deep and far space tasks and planet calendar, detection tasks and aircraft state simulation can be realized, test verification based on in-orbit actual measurement data is carried out, and a unified and standard evaluation index system is provided; the problem that an existing deep space task simulation system cannot provide accurate reference for a deep space ultra-large spatial-temporal scale simulation system is solved.

The invention belongs to the technical field of satellites. The invention provides a simulated in-orbit test method and system for low-orbit satellite time-frequency equipment. According to the embodiment of the invention, a reference atomic clock provides a reference basis for a GNSS navigation signal simulator, a time interval counter and a phase noise analyzer; the GNSS navigation signal simulator provides a Beidou / GNSS analog signal and an SSR (Simple Sequence Repeat) correction number for the satellite time frequency equipment; the satellite time frequency device completes real-time precise orbit determination and clock correction calculation of the low-orbit satellite by using the received Beidou / GNSS analog signal and the SSR correction number, and controls and regulates a built-in local atomic clock according to a calculated clock correction sequence; and respectively carrying out time precision and frequency stability test and analysis on the second 1PPS signal and the second 10MHz signal through a time interval counter and a phase noise analyzer. The low-orbit satellite in-orbit test can be simulated in the initial stage, and real-time calculation of in-orbit satellite test data can be realized.

At a satellite, in each of a plurality of time periods, a method of communicating with first and second ground stations, to exchange a first downlink message and a first uplink message with the first ground station and a second downlink message and a second uplink message with the second ground station, wherein the first and second downlink messages and the first and second uplink messages are time-stamped messages, and determining one or more orbit parameters indicative of an orbit of the satellite based on known positions of respective first and second ground stations relating to one or more of the plurality of time periods and based on first timing information relating to times of reception and transmission of respective first uplink and downlink messages and second timing information relating to times of reception and transmission of respective second uplink and downlink messages in the one or more time periods.

The application discloses a Beidou satellite maneuvering orbit recovery method, which comprises the following steps: S1, giving initial satellite state information; S2, detecting whether satellite orbit maneuvering occurs for each ephemeris based on current satellite state information; if yes, entering step S4; if no, entering step S3; S3, updating state parameter values of all satellites and returning to step S2; S4, processing the maneuvering orbit by adopting combined short-arc orbit determination and grouping filtering method, and realizing maneuvering orbit recovery. The combined short-arc orbit determination and grouping filtering method realizes quick recovery of the Beidou satellite maneuvering orbit, so as to guarantee the navigation, positioning and timing service performance of the Beidou system in the case that the Beidou satellite occurs orbit maneuvering. The method can not only effectively improve the orbit determination precision in the case that the satellite occurs orbit maneuvering, but also greatly improve the reliability of the global service capability of the Beidou No. 3 system.

The application discloses a target initial orbit determination method and device based on spatial filtering velocity measurement, which comprises the following steps: acquiring a star position image; determining an observation target according to the star position image, and establishing a mapping relationship between a measurement coordinate and an angle position of the observation target; modulating the star position image to obtain a distribution signal of target brightness of the observation target changing with time; acquiring a velocity of the observation target according to the distribution signal of the target brightness changing with time; obtaining an angular velocity and an acceleration of the observation target according to the velocity of the observation target through the mapping relationship between the measurement coordinate and the angle position of the observation target; and obtaining a position vector and an initial orbit velocity of the observation target relative to the earth center through an initial orbit determination algorithm. Compared with a traditional initial orbit determination method, the application directly measures the angular velocity of the target, reduces the coupling between the angle position error and the angular motion quantity, improves the initial orbit determination precision of the target under the observation error condition, and realizes the rapid orbit determination of the target.

An antenna and receiver design is described herein that allows a spacecraft to safely and accurately perform an autonomous flight safety system, an ascent, an entry, a re-entry, a landing, and / or an orbit determination. For example, a spacecraft can be equipped with 180-degree dual-opposed antennas and two receivers. A signal captured by a first antenna can be routed to a secondary radio frequency (RF) of the first receiver and to a combiner. A signal captured by a second antenna can be routed to a secondary RF of the second receiver and to the combiner. The mixer can perform the phase combination operation on the signals captured by the first receiver and the second receiver to produce a phase combined signal, and can route the phase combined signal to the primary RF of the first receiver and to the primary RF of the second receiver.

This invention provides a method for evaluating the collaborative tracking performance of high-low combination telescopes for GEO targets. The method includes: sampling data of basic indicators at the bottom level; for each basic indicator, first mapping its data to a set of evaluation levels, and then statistically obtaining a membership vector; aggregating from the bottom-level indicators until obtaining the membership vector of the top-level basic indicators; determining the score of the telescope combination scheme; and evaluating the scheme's performance based on the score. The basic indicators include full-load observation performance, observation performance, utilization rate, relay capability, orbit determination capability, guidance capability, relay success rate, orbit determination accuracy, orbit determination arc length, deviation, and response interval at different levels. This invention's method is based on D-S evidence theory for sampling, membership calculation, indicator aggregation, and evaluation, establishing a comprehensive performance evaluation index system. It can utilize performance weights to rank the merits of telescope combination schemes according to specific task requirements.

The application discloses a kind of only angle measurement lunar space Lagrange point orbit type intelligent identification method, belong to space technology field.The application establishes only angle measurement observation model, sets only angle measurement observation error covariance matrix;Establish the motion model of space Lagrange point orbit spacecraft in lunar space;According to the number of space Lagrange point orbit in lunar space, randomly generate space Lagrange point orbit in lunar space and recursion length, generate only angle measurement observation vector;The sample point required for training orbit type identification deep neural network is generated;Based on deep neural network, the fast and efficient identification of space Lagrange point orbit in lunar space is realized, and then the initial orbit determination and refined orbit estimation of space Lagrange point orbit in lunar space are realized.The application has the advantages of not depending on physical model, high identification accuracy, small amount of calculation, high efficiency and good robustness.The application is beneficial to improve the space situation awareness capability in lunar space, improve the determination accuracy and calculation speed of space Lagrange point orbit in lunar space.

This invention discloses a method for inverting pulse maneuvers in non-cooperative spacecraft based on minimizing orbit determination residuals. Addressing the difficulty in accurately inverting maneuver parameters for non-cooperative spacecraft under conditions of sparse observation data and unknown maneuver times, this invention constructs a sequence of historical orbits and the latest observed orbit states. Within a preset time window, candidate maneuver times are discretized. For each candidate time, orbit propagation is performed under the assumption of no maneuvering to obtain a predicted orbit. A residual vector is constructed based on the position difference between the observed and predicted orbits. A partial derivative matrix is ​​constructed through numerical differentiation. The least squares method is used to solve for the maneuver velocity increment. Candidate intervals and candidate times are selected based on the fitting error. Finally, the maneuver parameters are determined based on the criterion of optimal accuracy within the fit interval. This invention can achieve high-precision estimation of maneuver times and velocity increments under sparse observation conditions, reducing dependence on high-density observation data and prior maneuver information, and significantly improving the stability and reliability of maneuver inversion.

The present invention discloses a method, device, equipment and medium for precise orbit determination of GNSS satellites under the condition of regional stations. The method includes: obtaining GNSS observation data of regional ground stations and on-board GNSS observation data of low-earth orbit satellites, and performing combined orbit determination to obtain a preliminary orbit determination result of GNSS satellites in regional stations; solving to obtain spatial reference transformation parameters according to the preliminary orbit determination result of GNSS satellites in regional stations and the orbit determination result of GNSS satellites of IGS global stations; correcting the preliminary orbit determination result of GNSS satellites in regional stations on the current day based on historical spatial reference transformation parameters to obtain an orbit determination result of GNSS satellites in regional stations after compensating for spatial reference errors. The present invention uses combined orbit determination calculation to achieve centimeter-level high-precision orbit determination; for the first time, the idea of combining spatial reference transformation is proposed to analyze this effect, and the spatial reference errors caused by the overall rotation effect are corrected according to the spatial reference transformation parameters.

This invention discloses a method and system for monitoring and resetting the software heartbeat of a spaceborne receiver. The system includes a dual-channel GNSS positioning module, a dual-channel orbit determination module, and an antifuse FPGA chip. Both the GNSS positioning module and the orbit determination module have corresponding primary and backup FLASH memory. The antifuse FPGA acts as the heartbeat monitoring and control center, connecting to the GNSS positioning module and the orbit determination module via GPIO interfaces. It receives the software heartbeat signals from each module and sends reset commands to faulty modules via reset pins. This solution solves the problems of poor radiation resistance, low fault identification accuracy, and low recovery efficiency in existing spaceborne receiver software fault monitoring.

The application discloses a kind of space target ultra-short arc initial orbit intelligent determination method, belong to aerospace technology field.The application is based on artificial intelligence technology to realize space target ultra-short arc initial orbit determination, there is no strict requirement to complex nonlinear problem, and can develop more complex perturbation condition and initial orbit determination under the operation law of multiple body, coverage is wider;The application generates multiple groups of target orbits randomly and calculates the observation and state quantity of each visible arc segment, constructs the sample data set with ultra-large capacity, which is helpful for neural network to learn the nonlinear relationship between observation and state quantity using rich samples;The application builds and trains long short-term memory network model, determines the optimal network model weight parameter by designing and adjusting hyperparameter, can obtain the fast prediction solution of unknown space target ultra-short arc initial orbit according to observation;The application can meet more space target ultra-short arc initial orbit determination tasks by considering different inputs and outputs.