SUBSTANTIATION OF THE NEURAL NETWORK STRUCTURE FOR DETERMINING AN ORBITAL OBJECT MOTION PARAMETERS BASED ON RESULTS OF ITS OBSERVATIONS FROM A SPACECRAFT
Annotation
The structure of a multilayer neural network of direct propagation, designed to determine the motion parameters of the mass center of a non-cooperating orbital object by results of measurements carried out by a spacecraft optical-electronic system, is substantiated. The orbital object zenith distances at characteristic points of the spacecraft's orbit and the time of the orbital object's passage between these points were chosen as the measured parameters. The shortest duration of training for a given accuracy of determining the of the orbital object's center of mass motion parameters is determined. The presented results can be used in justification and development of systems for autonomous determination of the parameters of motion of the orbital object center of mass based on a spacecraft onboard optical means.
Keywords
Постоянный URL
Articles in current issue
- COMPUTING PROCESS VERIFICATION IN INFORMATION SYSTEM
- ACCOUNTING FOR NET DELAY TIME IN DIRECT CIRCUIT WHEN CALCULATING DISCRETE CORRECTION OF DIGITAL TRACKING SYSTEM WITH MINIMUM TIME OF FREE PROCESS TOTAL ATTENUATION
- DEVELOPMENT OF HIGH-APERTURE MICROSCOPE OBJECTIVES WITH AN ENLARGED FIELD OF VIEW AND IMAGE RECORDING ON CCD DETECTORS
- AUTOMATIC DETECTION OF FILAMENT BURR IN THREE-DIMENSIONAL PRINTING DEVICES
- DEVELOPMENT OF A SYSTEM OF MAGNETIC INDUCTION SENSORS FOR DIAGNOSING WEAR OF THE ROLLING STOCK WHEELSET CREST
- METHODOLOGY FOR ESTIMATING THE DURATION OF THE TECHNOLOGICAL RUN OF TELECOMMUNICATION EQUIPMENT
- STUDY OF THE LATEST DENTAL MATERIALS BY STRETCHING TO OBTAIN THE TENSILE STRENGTH PARAMETER