Star sensor is a pose determination device based on stars, which has good measurement concealment, electromagnetic interference resistance, and extremely high pose determination accuracy. It is one of the indispensable attitude measurement devices for platform attitude and orbit control systems such as missiles, aircraft, airships, and ships. However, due to the influence of strong sky background radiation, it is difficult for star sensors to measure stars during daytime, which has always limited their promotion and application in the atmosphere. This is also the key technical bottleneck of current all-weather star sensors. Compared to the visible light band, the sky background radiation intensity in the near-infrared band is significantly reduced and there is a significant atmospheric window. Therefore, all day star sensors based on the near-infrared band have always been a research hotspot at home and abroad.
Astronomically, the near-infrared bands of the atmospheric window are divided into R, I, J, H, K and other bands, with distinct atmospheric absorption bands between different bands. Due to factors such as low quantum efficiency and high noise of early infrared sensors, as well as low transmittance of optical lenses in the infrared band, research on all day star sensors has mostly focused on high-altitude platforms with weaker atmospheric effects. Wide optical response bands are used to increase transmittance to shorten exposure time, thereby reducing the inherent noise of the sensor and improving star measurement results. In recent years, the hardware technology related to star sensors has developed rapidly, such as the 5 series produced by Sensors Unlimited in the United States μ A new InGaAs sensor with m pixels, dark current
At present, although there is a lot of research on all day star sensors, it mainly focuses on detector hardware design, star catalog establishment, star point extraction, and other aspects. There is less attention to the selection of sensitive bands for star sensors, and the actual bands used for star sensors are also different. In the BLAST high-altitude balloon experiment, Marie Rex et al. selected the 600-850nm band as the optical response band of the star sensor based on the brightness factor of the sky background radiation; Zhang Kaisheng et al. proposed using the 1100-1400nm band as the optical response band of the star sensor when designing the optical design of a large relative aperture all day star sensor, taking into account the signal-to-noise ratio of night time star measurements in various bands; Zhang Hailong et al. analyzed and believed that the H-band was the optimal detection performance of the star sensor based on the characteristics of stellar radiation flux during the overall design of the all-day star sensor; In the star observation experiment, Wang Wenjie believed that the performance of the 1300nm cutoff optical response band was better than that of the 1500nm cutoff band based on the signal-to-noise ratio of the same star. However, due to the non independent or linear correlation of factors such as atmospheric transmittance, sky background radiation, definition of 0-star radiation flux, and star distribution density, these single angle analysis methods are not rigorous and the conclusions obtained are not uniform.
(1) Magnitude reference
Astronomically, magnitudes are used to measure the brightness of stars, and different magnitudes correspond to different magnitudes of radiation flux. Usually, the Simpson formula is used to quantitatively calculate the difference in radiation flux between adjacent magnitudes. Magnitude is a quantity that represents the relative radiance of a star. Only by defining the 0 magnitude radiance of each band can the corresponding magnitude of each star in that band be calculated. Initial Astronomy
When Pogson established the Simpson formula, he proposed using Vega as the full band zero magnitude standard and referred to it as Vega magnitude. But with the improvement of astronomical observation technology, it has been discovered that Vega may be a Cepheid variable star and cannot meet the absolute calibration accuracy requirements of 0 magnitude. Therefore, different astronomical catalogs are usually based on factors such as project observation instruments, filter bandwidth, and observation positions to establish metering systems that are in line with project reality, such as Johson’s 11 color metering system, SDSS’s ugriz metering system, and 2MASS’s JHK metering system. But most of them are continuation of Vega magnitudes, and the 0 magnitude calibrated radiation flux is not significantly different from the corresponding band radiation flux of Vega. Table 1 shows the parameters of the USNO and 2MASS catalog photometry systems.
Tab. 1 Parameters of USNO and 2MASS photography systems
(2) Atmospheric influencing factors
The daytime sky background radiation and atmospheric transmittance are important factors that affect the star detection ability of star sensors throughout the day. The daytime sky background radiation in the shortwave band is mainly atmospheric background radiation generated by scattering sunlight from the atmosphere. Due to the high cost and low efficiency of measuring atmospheric background radiation and transmittance in all directions, modtran atmospheric calculation software developed by the US Air Force is widely used for simulation calculations both domestically and internationally, which has good accuracy and reliability. Therefore, this article uses modtran software to analyze the atmospheric background radiation and atmospheric transmittance. There are many factors that affect atmospheric background radiation and atmospheric transmittance, but the main factors are low aerosols, water vapor, and CO2 in the atmosphere. The first two are mainly concentrated in the lower atmosphere below 2 km, and CO2 hardly changes with altitude below 50 km. Therefore, for star sensors working in environments without clouds or rain, far from urban aerosols, the impact of meteorological factors such as aerosols, water vapor, and CO2 can be ignored. Using modtran software, the simulation parameters were set as no cloud or rain, with an altitude of 6 kilometers, and a 1976 standard atmospheric model in the United States. The zenith angle was observed at 45 °, and the observation direction was due south. The noon atmospheric background radiation brightness was simulated and used as a typical value for the whole day background brightness, as shown in Figure 1.
Fig. 1 Typical values of atmospheric background radiation
Under the same parameter conditions, the atmospheric transmittance of 30 °, 60 °, and 80 ° observation zenith angles was simulated using modtran as typical values of atmospheric transmittance in different sensitive directions of the star sensor, as shown in Figure 2.
Fig. 2 Atmospheric transitivity at different zenith angles
From Figures 1 and 2, it can be seen that the longer the wavelength, the smaller the atmospheric background radiation flux, and there are obvious atmospheric windows in different bands. But for all day star sensors, when the optical response wavelength exceeds 2um, the instrument’s own radiation will interfere with star detection. Therefore, considering the response characteristics of infrared sensors, atmospheric background radiation, atmospheric transmittance, and other factors, the optical response band of all day star sensors is generally selected between 600 and 1700 nm, namely the R, I, J, and H bands. To avoid the impact of different atmospheric window bandwidths in different bands, the system analyzes the relationship between the actual stellar detection power of the all-day star sensor and the response band. Within the R, I, J, and H band atmospheric window, the integrated mean of atmospheric background radiation intensity and atmospheric transmittance in the 100 nm bandwidth band are taken as shown in Table 2. The atmospheric transmittance is taken as the average of the three observed zenith angles as a typical value of the all-sky atmospheric transmittance.
It can be seen that the actual representation of constant stars corresponds to the magnitude of radiation flux in the corresponding band. For star sensors using different optical response bands, the higher the signal-to-noise ratio of the corresponding 0-class star, the higher the actual detection limit of the star sensor.
Tab.2 Mean values of atmospheric parameters in different bands
(1) Star sensor receives signal energy
For star sensors, stars can be regarded as point light sources, and the energy signals received by star sensors are related to the optical system’s aperture, sensor quantum efficiency, and exposure time. As a radiator, the energy received by a single pixel in the atmosphere is the product of the sphericity corresponding to the sensor pixel and the atmospheric radiation intensity,
(2) Satellite measurement signal-to-noise ratio model
Infrared sensor noise mainly includes dark current noise, readout noise, shot noise, quantization noise, etc. For the new generation of infrared sensors, dark current noise and quantization noise are small enough to be ignored, and readout noise can also be effectively suppressed through related technologies. However, shot noise is caused by the randomness of photon incidence and photoelectric conversion process, and there is no effective method to suppress it. Table 3 below shows the parameters of two typical infrared sensors.
Tab.3 Typical IR sensor parameters
From Table 3, it can be seen that for a strong daytime sky background, dark current noise, quantization noise, and readout noise can be ignored relative to photon shot noise. The photon shot noise caused by a strong sky background is the main noise source of the star sensor, which is affected by the photon number of the sensor signal and follows a Poisson distribution
According to formula derivation and analysis, the signal-to-noise ratio of the detector is not only related to the detection hardware itself, but also to the optical response band range, atmospheric transmittance, and the optical response band of stars
The radiation flux at the central wavelength is closely related to the atmospheric background radiation flux. Therefore, under certain hardware conditions of the detector, the selection of optical response bands will also significantly affect the signal-to-noise ratio of star sensor detection.
With the rapid development of infrared sensor technology, the daytime star observation effect of all day star sensors has significantly improved. The traditional wide optical response bandwidth design has caused a decrease in imaging color difference and signal-to-noise ratio, gradually becoming one of the bottlenecks limiting its further improvement in accuracy. Based on the atmospheric window characteristics of the near-infrared band, selecting an appropriate band as the optical response band of the all-day star sensor can effectively alleviate such problems. To this end, a star signal-to-noise ratio model with different optical response bands of 0 was established for the all-time star sensor. Based on this, the atmospheric parameters of the relevant bands were simulated and calculated using modtran software. Combined with the distribution density of stars in different bands, the number, distribution density, and attitude determination success rate of stars detected by the all-time star sensor in different optical response bands were analyzed, and validation experiments were conducted using a daytime star measurement platform. The results show that under the same hardware conditions, the H-band is the optimal optical response band for all day star sensors, and the overall star detection ability is about 17 times, 10 times, and 2 times that of the R, I, and J bands. The success rate of attitude determination also has a significant advantage.
In order to reduce the color difference of the all-day star sensor imaging and improve the daytime star measurement signal-to-noise ratio of the all-day star sensor, a comparative analysis was conducted on the star measurement capabilities of the all-day star sensor in various optical response bands. Through comparative analysis, it is found that the use of different optical response bands has a significant impact on the actual star measurement ability of all day star sensors; Under the same hardware conditions, the all-time star sensor has the most outstanding ability to detect stars in the H-band, with a detection number of stars about 17 times, 10 times, and 2 times that of the R, I, and J bands. The success rate of attitude determination also has a significant advantage. A daytime star observation platform was built, and the actual star measurement capabilities using different optical band platforms were experimentally calculated. The experimental results showed that the H-band all day star sensor has significantly stronger star measurement capabilities than other optical bands, making it the optimal optical response band for all day star sensors. Reasonably designing the optical response band of the star sensor throughout the day can effectively improve the daytime star measurement effect of the star sensor, and the exposure time and response bandwidth also affect the overall imaging quality of the red extraterrestrial image. Therefore, on the basis of solving the wavelength selection of all-day star sensors, comprehensive research is needed to optimize the optical response bandwidth and exposure time design of all-day star sensors.
Send us a message,we will answer your email shortly!
TY-Space Technology (Beijing) Ltd. is professional focusing on advanced attitude optical sensors, especially Star Trackers for space industry.
