Cross-Reference to Related Applications

[0001] This application claims benefit of and priority to Indian Patent Application No. 202211008374 filed on 17 February 2022.

Field of the Invention

[0002] The field of the invention relates to a spacecraft with a reaction system to limit the impact of environmental conditions and a method for controlling the spacecraft.

Background of the invention

[0003] New spacecraft operational domains are becoming more prevalent and relevant. These include the Cis-Lunar domain, near-Sun domain, and interplanetary space domain. The requirements for spacecraft in these domains vary significantly due to environmental differences (e.g., radiation characteristics) and propulsion needs, for both reaching these domains, and maintaining position.

[0004] In order for the next generation of spacecraft to be able to fulfil these requirements economically and efficiently, a standardized solution is required which enables the spacecraft to adapt and respond autonomously to different environmental conditions. In doing so, the spacecraft will be able to protect components, such as, but not limited to critical life-limited components resulting in an extension of the spacecraft lifetime and a more economic mission execution. Lifetime extensions are also possible as the spacecraft can be capable of responding to hazards it encounters.

Summary of the invention

[0005] A method for controlling a spacecraft is described. The method comprises the step of monitoring environmental conditions of the spacecraft using at least one first measurement device. The method further comprises the steps determining a health condition of at least one component of the spacecraft and of controlling at least one reaction-system depending on the environmental conditions and the health condition. [0006] The step of monitoring the environmental conditions of the spacecraft can comprise measuring at least one first parameter characterizing the environmental conditions of the spacecraft using a first measurement device.

[0007] The first parameter can be related to cosmic radiation.

[0008] The step of determining the health condition of the at least one component can comprise predicting material degradation using a prediction model and the first parameter as an input to the prediction model.

[0009] The prediction model can be generated using supervised learning algorithms such as support-vector machine, logistic regression, or convolution neural network.

[0010] The step of determining the health condition of the at least one component can further comprise measuring at least one second parameter influencing the health condition of the at least one component. Measuring the second parameter can be performed using at least one second measurement device. The second parameter can be a further input to the prediction model for predicting material degradation of the at least one component.

[0011] The second parameter can be related to one of temperature and radiation level.

[0012] The reaction system of the spacecraft can be controlled to limit the impact of the environmental conditions on the health condition of the at least one component.

[0013] The step of controlling the reaction system of the spacecraft can comprise controlling a propulsion unit to alter movement of the spacecraft.

[0014] The step of controlling the reaction system of the spacecraft can comprise controlling a radiation shielding unit to provide a magnetic field to shield the at least one component.

[0015] The step of controlling the radiation shielding unit can comprise adapting the shape and strength of a magnetic field by altering the current supplied to at least a first high- temperature superconducting coil and a second high-temperature superconducting coil of the radiation shielding unit.

[0016] The step of controlling the reaction-system of the spacecraft can be performed autonomously without control from outside of the spacecraft.

[0017] A spacecraft is further described in this document. The spacecraft comprises a first measurement device for monitoring environmental conditions of the spacecraft and at least one component. The spacecraft further comprises a reaction system for limiting the impact of the environmental conditions on a health condition of the at least one component. The spacecraft further comprises a data processing unit for determining the health condition of the at least one component and autonomously controlling the one reaction system depending on the environmental conditions and the health condition.

[0018] The reaction system can be a propulsion unit or a radiation shielding unit comprising of at least two high-temperature superconducting coils.

Description of the figures

Fig- 1 illustrates an overview and relevance of platforms for a smart spacecraft.

Fig- 2 illustrates a structure of the smart spacecraft.

Fig- 3 illustrates an example machine learning training and prediction methodology schematic for monitoring health of components of the smart spacecraft.

Fig. 4A illustrates a double helix superconducting coil.

Fig. 4B illustrates direction of currents in the superconducting coils and resultant magnetic field.

Fig. 5 illustrates the operational domain of different propulsion systems.

Detailed description of the invention

[0019] Fig. 1 illustrates three broad categories of platforms for a smart spacecraft 10. A nano/cube/small spacecraft 110 with mass less than 400 kg can be used for near-Earth missions. A medium-sized spacecraft 120 with mass less than 3000 kg can be used for Geosynchronous Orbit satellites, space weather observing payloads or on-orbit servicing, assembly, and manufacturing (OSAM) applications. A large spacecraft 130 with up to and above 100000 kg mass can be used for the purpose of Lunar, interplanetary, asteroidal transfers of crew and cargo. The smart spacecraft 10 can be operated in a Cis-Lunar domain (the region between Earth and Moon, where the annual radiation dose of about 3 to a few hundreds of sieverts is received by the smart spacecraft 10), a near-Sun domain (less than about 0.5 ALT from sun, where the solar energetic particles (SEP), solar flares and coronal mass ejections can interfere with the electronics onboard the smart spacecraft 10), and an interplanetary space domain (orbital transfers between planets, e.g. Earth to Mars, where single event effects can be caused by galactic cosmic radiation (GCR) dosage). The smart spacecraft 10 needs to provide a number of capabilities that need to be implemented on board the smart spacecraft 10 to fulfil the requirements for the smart spacecraft 10 to be operated in these domains.

[0020] Fig. 2 illustrates the structure of the smart spacecraft 10 to provide the following capabilities:

[0021] Spacecraft Sensing: the smart spacecraft 10 comprises a first measurement device 30 and a second measurement device 35 (also called embedded sensors 30 and 35) throughout the smart spacecraft 10 to monitor the surrounding (harmful) environmental conditions 20 (e.g. radiation environment, micrometeoroids, space debris, solar storms) and health conditions (e.g. thermal stress, sputtering (a form of space weathering that changes the physical and chemical properties of the surface through ejection of microscopic surface particles due to electric potential differences between the surface of the smart spacecraft 10 and its surroundings), total radiation dose) of key sensitive components / subsystems (e.g. of a power processing unit (PPU) 90, optical instruments, a power source 700), respectively.

[0022] Sensor Network Frame: An integrated system network which supports on-board data generation based on the embedded sensors 30, 35 with ground-based data (e.g., CDM, solar storm information), hence providing inputs required for autonomous responses to different environmental conditions. The network frame is an integrated network of 3D printed and embedded commercial off-the-shelf sensors, connected through 3D printed wiring. In Fig. 2, this network frame operates in the following manner: the embedded sensors 30 and 35 sense environmental conditions 20 and a first parameter 22 and a second parameter 42 indicating the health condition (as described above) of a component 15. The first parameter 22 and the second parameter 42 (for example pressure and temperature values) are transferred to the on-board high-performance computer 400 and memory storage device 600, for continuous processing through use of algorithms and storage of the data, respectively. The details of the sensors, data processing and all other components and operations of this network frame are described in detail throughout this description section. [0023] Autonomy: Artificial Intelligence algorithms that perform predictive analytics of the smart spacecraft 10 operations based on sensor measurements.

[0024] High Performance Computer (HPC) 400: Significant computational power, for example using technologies such as central processing units (CPUs), graphics processing units (GPUs), field programmable gate arrays (FPGAs) and edge computing, in order to manage, process, and interpret this wealth of data, and enable the smart spacecraft 10 to respond autonomously to the first parameter 22 and the second parameter 42. The first parameter 22 and the second parameter can be data such as surface temperature, pressure and radiation dose generated through the embedded sensors 30 and 35 described later in the document. The first parameter 22 and the second parameter 42 need to be processed for machine learning and optimization algorithms 320 (both described later in detail), which shall be performed through HPC 400 to have computation time in the order of seconds.

[0025] Cosmic Radiation Protection: through shielding of key sensitive components from harmful environmental conditions 20 using a radiation shielding unit 80.

[0026] In order to deploy these capabilities, certain supporting technologies/features are required. The High-Temperature Superconductor (HTS)-based radiation shielding unit 80 is used to protect key components 15 against the harmful environmental conditions 20. The radiation shielding system 80 utilizes a first high-temperature superconducting coil 82 and a second high-temperature superconducting coil 84 to vary the electromagnetic shield for protection of the components 15 against radiation. One example of such a radiation shielding unit 80 is the European Union’s SR2S project (Evaluation of superconducting magnet shield configurations for long duration manned space missions / Ambroglini, F.; Battiston, R.; Burger, W. J. - In: FRONTIERS IN ONCOLOGY. - ISSN 2234-943X. - 6(2016), pp. 97.1-97.21. [10.3389/fonc.2016.00097]), which studied the use of HTS-based radiation shielding systems to protect from radiation, SEP and GCR effects, which as mentioned above are relevant for the multiple domains of space. The first high-temperature superconducting coil 82 and the second high-temperature superconducting coil 84can be in, but not limited to, toroidal, helical, or quadruple configurations to allow for multiple magnetic field topologies. One such example configuration, namely helical configuration, is discussed in detail towards the end of this detailed description section, along with description of the first high-temperature superconducting coil 82 and the second high- temperature superconducting coil 84 and a diagram shown in Fig. 4. Note that the radiation shielding for the components 15 is autonomously controlled based on the detected health condition of the component and subsystem(s) and the detected harmful environmental conditions 20. For example, large amounts of radiation detected will lead to an increase of the electromagnetic field generated by the radiation shielding unit 80. [0027] HTS-based Power Management and Distribution system 900 (PMAD), to efficiently manage and distribute the large amounts of power needed for the above capabilities.

[0028] Nanotechnology-based graphene networks, 3D printed into the systems of the smart spacecraft 10, to connect the embedded sensors 30, 35 together. Alternatively, Molybdenum-based networks could be employed depending on the trade-off between electrical conductivity vs mechanical strength.

[0029] The design and operations of the above-mentioned capabilities of the smart spacecraft 10 platform are described in the document hereafter.

[0030] The embedded sensors 30, 35, their working, sensor network frame and integrated circuit manufacturing are described in detail in the following paragraphs. Some of the embedded sensors 30 and 35 on board the smart spacecraft 10, can be sensors such as, but not limited to, Faraday Probes, Hall Probes, temperature sensors, radiation sensors, ExB probes, magnetometers, thermometers, spectrometers, ultrasound sensors, mass spectrometers, will measure parameters including, but not limited to, temperature, pressure, decomposition gases, frequencies, and radiation. These embedded sensors 30, 35 will perform sensing at operating temperatures of the components 15 for which the embedded sensors 30, 35 are used.

[0031] The embedded sensors 30, 35 will be located outside on all six sides of the smart spacecraft 10 to generate accurate data of the environmental conditions 20 on all surfaces of the smart spacecraft 10. The embedded sensors 30, 35 can for example be located in a thermal management system (TMS) of the smart spacecraft 10. For a TMS on the smart spacecraft 10, is the TMS being composed of layers of materials, more than two embedded sensors 30, 35 per layer can be used to accurately measure temperature at different locations inside the thermal management system.

[0032] The embedded sensors 30 and 35 are connected to the data storage 600 and the HPC 400, through a printed network of embedded circuits on the body of the smart spacecraft 10. The sensor network is based on microcircuits built with traditional deposition techniques including, but not limited to, Pulsed Laser Deposition (PLD), Electron Beam Physical Vapour Deposition (EB-PVD), and chemical vapour deposition (CVD). These circuits shall be embedded on each layer of TMS (in the example described in the previous paragraph). [0033] The graphene/carbon nanotubes circuit architecture will be built based on standard 3D printing including, but not limited to, atmospheric plasma spraying, selective laser melting, and selective laser sintering. Graphene materials show an excellent thermal and electrical conductivity among other properties, with these properties in mind the graphene conductor network will allow interconnection of the different sensors within the structure of the corresponding critical component for which the health conditions must be monitored. The manufacturing and fabrication process of such a network involves use of multi-material 3D printing systems, allowing the direct integration of 3D printed dielectric structures with electronics components fabricated together using a single non-assembly build sequence (C. Shemelya et al., “Multi-functional 3D printed and embedded sensors for satellite qualification structures,” 2015 IEEE SENSORS, 2015, pp. 1-4, doi: 10.1109/ICSENS.2015.7370541.).

[0034] The autonomous health prediction capability for the health conditions and prediction methodologies are presented in the following sections. The following paragraphs respectively describe inputs, outputs, predictor training methodology, machine learning algorithms, details and applications of prediction as illustrated in Fig. 3.

[0035] The integrated system network operates by determining S120 a health condition 40 of the component 15 by generating analytic predictions of material degradation due to radiation, sputtering and thermal degradation. In Fig. 3 embedded sensors 30, 35 include, but are not limited to, convolution-neural -network (CNN)-enabled Timepix detectors 301 (M. Ruffenach, S. Bourdarie, B. Bergmann, S. Gohl, J. Mekki and J. Vaille, “A New Technique Based on Convolutional Neural Networks to Measure the Energy of Protons and Electrons With a Single Timepix Detector,” in IEEE Transactions on Nuclear Science, vol. 68, no. 8, pp. 1746-1753, Aug. 2021, doi: 10.1109/TNS.2021.3071583), piezoelectric sensors 302 (example: a flexible piezopolymeric transducer based on lead-free piezoelectric zinc oxide thin films (Laurenti, Marco, Denis Perrone, Alessio Verna, Candido F. Pirri, and Alessandro Chiolerio. 2015. “Development of a Flexible Lead-Free Piezoelectric Transducer for Health Monitoring in the Space Environment” Micromachines 6, no. 11 : 1729-1744. https://doi.org/10.3390/mi6111453)), and off-the-shelf high temperature pyrometric sensors 303 which can measure over 2200 degree Celsius. The data from the embedded sensors 30, 35 indicates radiation levels 311, sputtering 312, thermal stress 313 onboard the smart spacecraft 10, respectively. The data 311, 312 and

313 are the inputs for the prediction model 320.

[0036] The data 311, 312 and 313 is further taken in and processed in real-time using realtime HPC 400 (https://en.wikipedia.org/wiki/Real-time_computing).) and is fed into the pre-trained Artificial Intelligence (Al) prediction model 320 on-board the smart spacecraft 10, to generate relevant predictions of material degradation 330 using real-time computing paradigm. Description of this Al training, algorithm of example method and use of the predictions are described in the following.

[0037] The training of the prediction model 320 would have been carried out using ground and space experiments data to relate the environmental conditions 20 and their effects on the components 15. These trainings employ supervised machine learning (ML) algorithms 260, including, but not limited to, support-vector machine, logistic regression, and CNN. The prediction models 320 will be trained using material degradation databases 240, generated through non-destructive testing techniques including, but not limited to, infrared thermography 231, ultrasonic testing 232, and liquid penetrant testing 233. For example, material degradation data will be collected from literature 201 (example sources: Herr, J.L. and McCollum, M.B., 1994. Spacecraft environments interactions: Protecting against the effects of spacecraft charging. NASA STI/Recon Technical Report N, 95, p.19780.; Koons, H. & Mazur, J. & Selesnick, R. & Blake, J. & Fennell, J. & Roeder, Jim & Anderson, P. (1998). The Impact of the Space Environment on Space Systems. Paper presented at 6 ^th Spacecraft Charging Technology. -1. 7-11.; Minow, J.I., Parker, L.N. and Jacobs, E.S.S.S.A., Spacecraft Charging: Anomaly and Failure Mechanisms.), experimental measurements 202, and in-flight measurements related to space environment’s effects on previously flown spacecraft (presented by dotted line connecting predictions and training zones in Fig. 3. Ground experiments 202 on space components in vacuum chambers simulating radiation, electric arcs and temperature conditions will be conducted to simulate space environment. These components will then be tested, using the tests mentioned above, to generate data of the damage caused (that is, material degradation data). The degradation will be assessed by , inter alia, tabulating number of cracks, depth of cracks (e.g. using potential probe method (H. Cost, V. Deutsch, P. Ettel, M. Platte - Wuppertal, 1996, Crack Depth Measurement - Modern Measuring Technique for a well- known method, NDTnet Vol. 1 No.06)), surface smoothness changes (e.g. employing laser based roughness meters (Vladimir Ostafiev, Sergey Sakhno, Grigoriy Tymchik, Sergey Ostafiev, Laser diffraction method of surface roughness measurement, Journal of Materials Processing Technology, Volume 63, Issues 1-3, 1997)) and sputtering (e.g. using quartz crystal microbalance (G. Hayderer, M. Schmid, P. Varga, HP. Winter, and F. Aumayr , “A highly sensitive quartz-crystal microbalance for sputtering investigations in slow ionsurface collisions”, Review of Scientific Instruments 70, 3696-3700 (1999) https://doi.org/10.1063/L 1149979)) caused by the vacuum chamber experiments. These are only examples of material degradation data, and more types of data can be collected. The vacuum chamber conditions (example the radiation levels 211 and the temperature data 213) would act as features 220 inputs for the ML algorithms 260 and the material degradation data 240 (example cracks, internal crevices, smoothness changes) would serve as the training labels 250. The literature 201 (example sources mentioned above) data will be used in an identical manner to feed to the ML algorithms 260. Thus, through this training, the ML algorithms 260 would have learned the relationship between the environmental conditions 20 and its damaging effects on the components 15 of the smart spacecraft 10. The prediction model / trained classifier 320 will be loaded on the smart spacecraft’s 10 on-board data storage 600 (see Fig. 2) (the data storage 600 is a radiation hard solid-state recorder with 100s of Gigabits memory capacity), which will be able to make accurate predictions of the health condition 40 of the components 15 of the smart spacecraft 10 using the data from the embedded sensors 30, 35described above. The working of the ML algorithm 260 and more details on prediction are described in the following three paragraphs.

[0038] The ML algorithms 260 to be used are chosen depending on the model selection algorithm described in Gianoglio, Christian, Edoardo Ragusa, Paolo Gastaldo, Federico Gallesi, and Francesco Guastavino. 2021. “Online Predictive Maintenance Monitoring Adopting Convolutional Neural Networks” Energies 14, no. 15: 4711. https://doi.org/10.3390/enl4154711. As an example, multiple linear regression is a straightforward and effective algorithm that can be used. Internal computations have shown this method to be accurate for more than 95% simulations, with low computing power costs. Linear regression finds a linear relationship between the inputs (x) and outputs (y). It aims to minimize a cost function, which is given by (1/n) X (predicted value - true value) over all the data points. This method can be used for accurate predictions. Other ML algorithms 260, which work on this similar principle, include but are not limited, to support-vector machine, logistic regression and convolution neural network can be used for making predictions, using inputs and training methodology described in previous paragraphs.

[0039] Once accurate prediction of the health condition 40 is made, the smart spacecraft 10 can protect itself in the changing environmental conditions 20. For example, the smart spacecraft 10, through the procedure described above, predicts the relation between the total radiation dose received and the health condition 40 of the components 15 in real-time. This information is then used to control S140 at least one reaction system 50 to autonomously modify mission characteristics (e.g., employing an on-board multi-mode propulsion unit 70 to modify one orbital parameter, such as orbital altitude and inclination, depending on the mission constraints defined by the payload operator) or control another reaction-system S142, such as increasing shielding for particular ones of the components using the HTS coils 82 and 84 in the radiation-shielding unit 80. This decision comes from life-time predictor present in the prediction model 320. The life-time prediction involves the knowledge of component radiation rating, erosion rating (this information comes from the ratings provided by the component manufacturer based on the intrinsic mechanical properties of materials including, but not limited to, impact strength, elasticity, plasticity, cohesion) and polynomial extrapolation of received total dose during the mission to compare the expected and the nominal lifetimes. Thus, the shielding is increased accordingly to sustain the component 15 for the mission duration provided by the payload 18 provider. For example, after predicting the degradation of the power processing system 900, the smart spacecraft 10 will prioritise protection of this critical system. The method employed for this protection include increasing radiation shielding (using the radiation shielding unit 80) at or near the components 15 and reducing power consumption to extend lifetime by sending only critical data to Earth, as reduced power of a RF Deployable Antenna 500 would mean more power available for critical operations, such as operating HTS coils 82 and 84.

[0040] Thus, the material degradation predictions 330 are aided by the high-performance computing 400 that can run numerical codes incorporating Artificial Intelligence algorithms. The radiation dose, temperature, pressure, and other sputtering related data (from the embedded sensors 30, 35) will act as inputs for the high-performance computing 400. After determining S120 the health condition 40, the final output would be control signals for controlling S140 the at least one reaction system and controlling S142 another reaction system, for example the magnetic field value required to sustain the component 15 for its desired lifetime. Thus, the high-performance computing 400 will send a signal to the HTS coils 82 and 84 to change the magnetic field configuration accordingly. Thus, the HTS coils 82 and 84 generate the required electromagnetic field topologies to shield a volume of space around the smart spacecraft 10 against the environmental conditions 20. Multiple configurations of the HTS coils 82, 84 are possible (Md. Abdullah Al Zaman, H.M.A.R. Maruf, M.R. Islam, Neelufar Panna, Study on superconducting magnetic shield for the manned long termed space voyages, The Egyptian Journal of Remote Sensing and Space Science, Volume 24, Issue 2, 2021, Pages 203-210, ISSN 1110-9823, https://doi.Org/10.1016/j.ejrs.2021.01.001.), including but not limited to, toroidal, helical, and quadruple. Fig. 4 shows the example of helical configuration of coils, based on SR2S project referred above. The HTS coils 82, 84 and how they generate different magnetic field topologies is also explained by Fig. 4 and also later in text. Collision avoidance capability of the platform are now described in detail.

[0041] The Al-enabled smart spacecraft 10 will be able to autonomously perform collision avoidance manoeuvres when transiting areas of the space with dense asteroid or space debris traffic. Repositioning collision avoidance manoeuvres shall be conducted to alter the trajectory of the spacecraft sufficiently to reduce the chance of a collision below some given collision probability value, as indicated by a payload provider. A first order simplistic idea is to use constant tangential (low) thrust, as soon as a conjunction data message (CDM) is received, and collision risk is deemed high. These risks are readily determined by information in the CDMs through use of collision probability and miss distance information given in CDMs. The risk can simply be the collision probability parameter in the received message. (CDMs (ISO 19389:2014) are data in standard message format for use in exchanging spacecraft conjunction information between originators of Conjunction Assessments (CAs) and satellite owner/operators and other authorized parties.). The CDMs are generated by object tracking databases, as also described in the next few paragraphs.

[0042] In literature, this approach for electric propulsion systems has shown to be efficient to alter the orbit sufficiently to avoid collision (Juan Luis Gonzalo, Camilla Colombo, Pierluigi Di Lizia, Introducing MISS, a new tool for collision avoidance analysis and design, Journal of Space Safety Engineering, Volume 7, Issue 3, 2020, Pages 282-289). The relatively high thrust from the propulsion unit such as a hybrid propulsion system (a first high-temperature superconducting coil 72 and a second high-temperature superconducting coil 74 based electric and chemical propulsion in a single system, using the same infrastructure and propellants) compared to conventional electric propulsion systems, ensures more effective avoidance, and also permits performing efficient manoeuvres with shorter warning times. Such as thruster is described in Betancourt, M. La Rosa, et al. "Applied-Field Magnetoplasmadynamic Thrusters (SUPREME™) as an Enabling Technology for Next-generation of Space Missions." Journal of the British Interplanetary Society 72 (2019): 401-409.).

[0043] It is sufficient to use the relatively low thrust of the electric propulsion of the hybrid propulsion system for some of the collision avoidance manoeuvres. The electric propulsion of the hybrid propulsion system can deliver thrust for longer durations. There are, however, other types of the collision avoidance manoeuvre which require a rapid repositioning of the smart spacecraft 10. The relatively high thrust of the chemical propulsion of the hybrid propulsion system can enable the smart spacecraft 10 to perform such rapid repositioning collision avoidance manoeuvres.

[0044] Such a hybrid propulsion system can operate on shared feeding systems, tanks, and propellants (e.g., hydrazine or ammonia) for the chemical system and electric systems and reach specific impulses of over 4000 seconds. The hybrid propulsion system can, however, also use different propellants for the chemical systems and the electric systems, as especially the electric systems are flexible with regards to the used propellant.

[0045] Fig. 5 shows the operational domains of different propulsion systems. The hybrid propulsion system has, for example, a chemical-electromagnetic architecture enabled using Applied Field Magnetoplasmadynamic (AF-MPD) electric propulsion. AF-MPD thrusters overcome many of the drawbacks of the traditional hybrid or multi-mode systems and can operate using, operating on a shared propellant (e.g., the aforementioned hydrazine or ammonia) with the chemical system. Hence, the hybrid propulsion system can offer a simple, powerful, and flexible propulsion system which will simultaneously reduce complexity by using the shared propellant tanks and feeding systems. The hybrid propulsion system is a light, simple system with a wide range of thrust levels from a few mN/kW to more than 600 mN/kW as shown in Fig. 5 and specific thrust from a few seconds to more than 9000s as also shown in Fig 6. The hybrid propulsion system can offer both high thrust of up to more than 600 mN/kW as shown in Fig. 5 for rapid manoeuvres, as well as efficient operation for routine uses such as for small adjustments in the orbit of the satellites and maintenance of the orbit. This technology finds its use in defence and non-defence domains such as: small satellite constellations, LEO, MEO, GEO communications satellites, space situational awareness, cis-lunar cargo transfers, solar weather monitoring satellites and majorly in on-orbit servicing and manufacturing (OSAM) missions.

[0046] The operational domain of the hybrid propulsion systems such as the hybrid chemical system and electrical systems is much larger compared to chemical only propulsion systems. The enhanced capabilities of real-time processing of CDM and optimization (described later) along with the use of the multi-mode thruster 70 on the smart spacecraft 10 can enable greater responsiveness to such collisions. These enhanced capabilities are due to increased speed of decision making (because HPC 400 is used for data processing, optimization, and command generation) and high-thrust chemical propulsion capability, thus enabling both prompt computation and effective-autonomous execution of required manoeuvres to avoid collisions. The manoeuvre computation and details on the optimization algorithms are described in the following paragraphs in detail.

[0047] The manoeuvre computations are made, by a piece of embedded software, if the risk of collision is greater than 1 in 10000. The manoeuvres are calculated using off-the- shelf semi -analytical methods (Javier Hernando-Ayuso, Claudio Bombardelli, Low-Thrust Collision Avoidance in Circular Orbits, 2021, Journal of Guidance, Control, and Dynamics, 983-995). Further, weighted optimization algorithms (such as genetic algorithms, particle swarm optimization, evolutionary algorithms) are employed to minimize propellant consumption, time required for the manoeuvre and aberrations to the nominal mission. The mission aberrations could be, for example that the smart spacecraft 10 is not able to make all the planned observations, due to change of trajectory forced by the collision avoidance manoeuvre. Thus, one of the variables to be optimized by the algorithm would be to maximize the possibility of performing all planned observations. That is, along with reducing the fuel and propellant mass required for the manoeuvre, the algorithm would consider as well, maximizing the mission yield. Input data used for such optimizations involve orbital characteristics (position and velocity of the smart spacecraft 10) 16, CDMs, power available, collision risk and propellant mass available. The optimization algorithms are freely available as open-source python libraries (e.g., pymoo or pygmo libraries). These codes would be installed on board the smart spacecraft 10 to perform the optimizations. The exact choice of algorithm to be used is dependent on the smart spacecraft’s 10 operation domain (some algorithms are described later). A predetermined set of instructions will be saved on the on-board data storage 600 to determine which algorithm to use and when. For example, if the compute power and time available are low, the set of instructions will allow the computer to choose the least computationally costly algorithm, with loss of accuracy. A trade-off table of algorithms will be prepared through extensively research. This will involve numerical simulations of orbits ranging from low-earth orbit to inter-planetary, randomized debris simulation, and the outputs would be computation time required to reach convergence and percentage reduction in collision risk for these diverse conditions. Such analysis will allow relating the electric power available (PA) for the onboard computer to the optimum optimization algorithm. The trade-off table will have data for all the optimization algorithms mentioned previously. For each of the algorithm, computation time (tcomp) and processing power required (P _re q) will be stored in table, on data storage 600, for varying number of data points, domain of space and accuracy achieved by the algorithm. A simple example of set of instructions involve: a) checking all the algorithms for which PA > Preq; b) Among the algorithms chosen in a), identifying those which can be employed on less number of data points than those available for the particular scenario; c) Among the algorithms shortlisted in b), choosing the ones that are relevant for the space domain the satellite is currently in, using the telemetry data 16 to identify satellite domain; d) Among the algorithms chosen in c), identifying those whose computation time (tcomp) is at least one order of magnitude less than the potential collision time; e) Among the algorithms chosen in d), the optimum algorithm can then be selected by identifying the one having maximum accuracy.

[0048] Input parameters would include the information from CDMs, including, but not limited to, relative position, relative velocity, miss distance, and collision probability. These CDMs would come from object surveillance databases, provided by external bodies such as the US Space Surveillance Network or Combined Space Operations Center (CSpOC) or companies like LeoLabs, OKAPEOrbits. Other input parameters would come from the telemetry data 16 of the spacecraft, to accurately predict its orbital characteristics. [0049] Optimization algorithms such as non-dominated sorting genetic algorithm II (NSGA-II) (Deb, K., Pratap, A., Agarwal, S. and Meyarivan, T.A.M.T., 2002. A fast and elitist multi -objective genetic algorithm: NSGA-II. IEEE transactions on evolutionary computation, 6(2), pp.182-197.) and R-NSGA-II (Kalyanmoy Deb and J. Sundar. 2006. Reference point based multi -objective optimization using evolutionary algorithms. In Proceedings of the 8th annual conference on Genetic and evolutionary computation (GECCO '06). Association for Computing Machinery, New York, NY, USA, 635-642. DOI: https://doi.org/10.1145/1143997.1144112) can be used from pymoo for the multiobjective problem described above. These are well known and fast sorting evolutionary algorithms. These can provide convergence of solutions while deal with non-linearities and complex interactions. The basic steps in the implementation of NSGA-II are: population initialization, non-dominated sorting, crowding distance assignment, selection of individuals using binary tournament selection with crowded-comparison and genetic operators, recombination of generations and selection.

[0050] The radiation shielding unit 80 and HTS coils 82, 84 are described in detail in the following paragraphs. The cosmic radiation shielding unit 80 will be designed in a way, based on the SR2S project referred previously, that magnetic fields generated by the HTS coils 82 and 84 can be adjusted and modified in response to the environmental conditions 20. For example: the smart spacecraft 10 will operate in higher-thrust mode (using propulsion unit’s 70 chemical mode) when passing through the Van Allen belt to reduce the time spent in the harmful environmental conditions 20. Simultaneously, the magnetic field of the radiation shielding unit 80 may be increased aiming to protect against the higher radiation dosages in that area. Due to the lightweight and compactness of the HTS coils 72, 74, 82 and 84, the electromagnets can be designed in a way that they can be adapted to the spacecraft architecture with minimal mass penalties and design constraints. This is also shown in Fig. 1, where the mass of the medium- 120 and large-sized 130 spacecraft platform are identical, while the payload carrying capacities of the two are 3000 and 100000 kg, respectively.

[0051] During operation, the embedded sensors 35 and 30 measure in-situ operating conditions of the components 15 and the environmental conditions 20. The high- performance computing 400 executes numerical codes and software that perform determining SI 20 the health condition 40 of the components 15/ smart spacecraft 10 performance and communicate with the cosmic radiation shielding unit 80 to adapt magnetic fields, aiming to increase lifetime of the components 15 and the smart spacecraft 1010.

[0052] The HTS coils 82 and 84 are designed in configurations of coils with different geometries and orientations. In this way, different variations of current can be supplied in order to alter and customize the shape of the resulting magnetic field for the specific conditions. The embedded sensors 30, 35 will monitor the environment conditions 20 and use relevant stimuli, such as flux of particle radiation, to sense regions and periods of increased environmental hazards. The embedded sensors 32, 35 are also be used in combination with the trained classifier 320 to predict the predicted material degradation 330. The current supply to the HTS coils 82, 84 can then be adapted in order to produce a magnetic field shape which provides increased shielding in the region of the harshest environmental conditions 20.

[0053] The HTS coils 72, 74, 82 and 84 are produced of a rectangular cross section with a superconducting layer being formed of any type of superconductor. Examples of the superconductor include, but are not limited to, type 2G high-temperature superconductors such as Yttrium Barium Copper Oxide, Lanthanum Barium Copper Oxide and other Rare- Earth Barium Copper Oxides, Magnesium Diboride, Bismuth Strontium Calcium Copper Oxide (Bi2223 or Bi2212). The use of very high-temperature superconductors, including those which require higher pressures for operation, and those which could be operated at room temperature, are also considered as potential materials.

[0054] The HTS coils 72, 74, 82, 84 can be kept cool by a cryogenic system. Such a system uses cooling technologies such as, but not restricted to, Pulse Tube Tactical Cooling; Pulse Tube Miniature Tactical Cooling; Joule-Thompson Coolers; Reverse Turbo-Brayton Coolers; Stirling Cryocoolers; The coolers are integrated and connected with the coils within a cryostat which maintains the operational temperature for the coil operation. In an alternative aspect, a radiatively cooled superconductors is envisaged as a possibility which do not require a cryogenic system.

[0055] The HTS coils 72, 74, 82, 84 in this aspect are loaded with electrical current either through a physical coil loading connection, such as, but not limited to ohmic current leads, joints, or connectors, or through a non-physical coil loading connection, such as, but not limited to, inductive loading using a device such as a flux pump. (Flux pump, an alternative to current leads, is a device used for imparting current to superconducting coils through generating changing magnetic field. These include but are not limited to travelling wave flux pumps (for example using rotating permanent magnets or linear electromagnets) and transformer-rectifier flux pumps (for example switched or variable resistance pumps). Detailed information about these flux pumps can be found in literature such as T.A. Coombs et al (2016) ‘An Overview of Flux Pumps for HTS Coils’ and L. Fu, K. Matsuda, and T.A. Coombs (2016) ‘Linear Flux Pump Device Applied to HTS Magnets: Further Characteristics on Wave profile, Number of Poles, and Control of Saturated Current’).

[0056] The use of a single HTS coil 72, 74, 82, 84 does only allow for the control of the magnitude of the B-field but not its topology. The magnetic field topology of a ring shape coil decreases fast as we move away from the coil centre. Changing the coil configuration, areas of large B -field can be extended.

[0057] The use of HTS based double and triple helix coils enable the control of the B-field topology as shown in Fig. 4A shows a superconductor magnet with a double helix winding of two HTS coils 82 and 84 with a common axis. Two different types of windings of the HTS coils 82 and 84 are shown in Fig. 4B and lead to different topologies of the magnetic field depending on the direction of the electric current flowing in the HTS coils 82 and 84. The smaller arrows on the HTS coils 82 and 84 show the direction of the magnetic field and the larger arrow on the right-hand side of the figure shows the resultant magnetic field. The use of a solenoid instead of a ring will increase the length of the region with high magnetic fluxes.

[0058] Two other relevant components, namely the RF Deployable Antenna 50 and the power source 700of the smart spacecraft 10 are described in the following paragraphs. The receiving and sending of data, including, but not limited to, telemetry, CDM, commands data, requires a robust and reliable communication with the ground stations. For this purpose, the smart spacecraft 10 is equipped with multiple communication devices, including, but not limited to, the RF deployable antenna 500 and a laser interlink 550. The RF deployable antenna 500 can transmit radio frequencies ranging from L-band to Ka- band (i.e., 1 to 40 GHz). The laser interlink 550 provides laser interlinking as another potential mechanism of communication, with data rates as high as up to and above 10 Gigabits per second and data transmission ranges being in the order of thousands to millions of kilometres. For example, entities building such laser terminals include, but are not limited to, Aquarian Space, Archangel Lightworks and Mynaric. This laser interlinking communication with other satellites is important for situations where a direct link to Earth may not be practical, for example, certain eclipse periods, or long-distance interplanetary missions. In this case, the communications infrastructure can be supported through laser interlink 550, in addition to the conventional RF deployable antenna 500 data transfers.

[0059] The power required for operating the propulsion unit 70, HTS coils 72, 74, 82 and 84, their subsystems, HPC 400, a payload 18 and the components 15, RF deployable antenna 500 and laser interlink 550, and all the other components on board the smart spacecraft 10, comes from the power source 700. The power source 700 could be formed by solar arrays, nuclear reactors, or a combination of both, depending on the mission requirements. For example, a near-Earth mission could use solar arrays, whereas a mission to Pluto would require a nuclear power source onboard the smart spacecraft 10. The solar arrays which can be used include, but are not limited to, silicon cells covered in thin glass arrays, multi -junction cells (e.g., made of gallium arsenide), flexible solar arrays such as Roll Out Solar Arrays used on International Space Station. The nuclear power sources include, but are not limited to, small fission reactors such as TOPAZ nuclear reactor, radioisotope thermoelectric generators and fusion reactors.

Reference Numbers

10 Spacecraft

15 Component

16 Telemetry

18 Payload

20 Environmental conditions

22 First parameter

30 First measurement device

35 Second measurement device

40 Health condition

42 Second parameter

50 Reach on- system

70 Propulsion unit

72 First high-temperature superconducting coil in propulsion unit

74 Second high-temperature superconducting coil in propulsion unit

80 Radiation shielding unit

82 First high-temperature superconducting coil

84 Second high-temperature superconducting coil

90 Power Processing Unit (PPU)

110 Nano/Cube/Small spacecraft

120 Medium-sized spacecraft

130 Large spacecraft

200 Training

201 Literature

202 Vacuum Chamber Experiments

211 Radiation Level

212 Sputtering data

213 Temperature

220 Features

231 Infrared Thermography

232 Ultrasonic Testing

233 Liquid Penetration Testing

240 Material Degradation 50 Labels 60 Machine Learning Algorithm 00 Predictions 01 Timepix Detector 02 Piezoelectric Sensor 03 High Temperature Pyrometric Sensor

311 Sensed Radiation Level

312 Sensed Sputtering data

313 Sensed Temperature

320 Trained Classifier

330 Predicted Material Degradation

400 High performance computer

500 RF Deployable Antenna

550 Laser interlink

600 Data Storage

700 Power Source

900 Power management and distribution system

S100 Monitoring environmental conditions

S120 Determining a health condition

S140 Controlling at least one reaction-system

S142 Controlling another reaction-system