Designing the trajectory of microsatellite swarms from the macro-micro perspective

As an rising multi-satellite cooperative flight mode, the microsatellite swarm has turn out to be an essential future analysis concern for distributed space methods. It provides low value, fast response, and collaborative decision-making. To handle the coordination of swarms for autonomous brokers, a probabilistic steering method has been investigated, which contained sub-swarms with totally different mission targets.

Probabilistic swarm steering permits autonomous microsatellites to generate their particular person trajectories independently in order that the whole swarm converges to the specified distribution form. Nonetheless, it’s important to keep away from crowding for lowering the opportunity of collisions between microsatellites, which provides challenges to the design of the collision avoidance algorithm.

In a analysis paper not too long ago revealed in House: Science & Know-how, Bing Xiao, from Faculty of Automation, Northwestern Polytechnical College, proposed a Centroidal Voronoi tessellation (CVT) and Mannequin Predictive Management (MPC) primarily based synthesis technique, aiming to realize macro-micro trajectory optimization of a microsatellite swarm.

The creator formulated the switch mannequin of swarm microsatellites in 3D space and launched the probabilistic swarm steering regulation. Afterwards, because it was important to keep away from crowding for lowering the opportunity of collisions between microsatellites, the security evaluation of collision avoidance wad performed primarily based on discovering the decrease certain of the minimal distance between all microsatellites at any time.

To find out the collision-free steering trajectory of every microsatellite from the current position to the goal space, a collision avoidance algorithm was mandatory. Nonetheless, with high-level coordination that used the macroscopic fashions, collision-free trajectories had been very arduous to generate. Therefore, the creator introduced a synthesis technique, the place the trajectory planning was divided into macro-planning and micro-planning.

Then, the creator introduced the small print of macro-planning and micro-planning of the microsatellite swarm, respectively. Within the Macro-planning of microsatellite swarm, the goal place of every microsatellite was decided by the centroid generated by the CVT algorithm, and all microsatellites moved to the corresponding centroid till the algorithm converges.

The ultimate distribution of the microsatellite swarm within the space was obtained in line with the placement of the centroid. Within the Micro-planning of the microsatellite swarm, MPC was adopted to generate the optimum trajectories for every step and at last reached the required place within the goal dice.

Particularly, the creator established the orbital dynamics mannequin contemplating J₂ perturbation and applied the convexification of collision avoidance constraints within the means of swarm reconfiguration. To realize the real-time trajectory planning, model predictive control was launched, which used a receding horizon to replace the optimum trajectories primarily based on the present state info. Considerably, the proposed technique can’t solely notice collision avoidance of microsatellite swarm maneuvering on the macrolevel, but additionally supplied optimum trajectories for every microsatellite of swarm people on the micro-level.

Lastly, the numerical simulation was carried out to confirm the proposed macro-micro trajectory planning technique of microsatellite swarm. The creator gave a digital central microsatellite and designed a large-scale (300) microsatellite swarm with an omnidirectional flight configuration. The CVT algorithm was used to divide areas, and in order to find out the place of the microsatellites to be transferred on the subsequent second.

Then, one of many cubes was chosen within the switch course of and carried out CVT on it to find out the switch place of the microsatellite. After 50 iterations, a steady configuration was obtained, and the place the place the microsatellite moved on the subsequent second was decided. Because of the massive scale of the microsatellite swarm, the method of reaching the ultimate configuration required many transitions.

To confirm the proposed trajectory optimization primarily based on mannequin predictive management, one of many microsatellites was chosen from the preliminary level to the following desired goal level at a sure second. The person microsatellites can attain the specified level properly. After the specified level was reached, the following iteration can be carried out, and because of the affect of orbital dynamics, the microsatellite could not stay the goal level with out management constraints.

To make the mission of microsatellite swarm extra sensible, MPC was utilized in micro-planning to enhance the efficiency of microsatellite swarm by way of gasoline consumption and useful resource utilization. Thus, simulation outcomes in regards to the collision-free steering trajectory of microsatellites verified the advantages of the planning scheme, which accorded properly with engineering apply.

Offered by
Beijing Institute of Know-how Press Co., Ltd