Genetic Algorithm of Energy Consumption Optimization for Reorientation of the Spacecraft Orbital Plane

Abstract

The paper is dedicated to the problem of finding optimal spacecraft trajectories. The equations of spacecraft motion are written in quaternion form. The spacecraft moves on its orbit under acceleration from the limited in magnitude jet thrust. It is necessary to minimize the energy costs for the process of reorientation of the spacecraft orbital plane. The equations of spacecraft motion are written in orbital coordinate system. It is assumed that spacecraft orbit is circular and control has constant value on each part of active spacecraft motion. In this case the lengths of the sections of the spacecraft motion are unknown. We need to find the length of each section, their quantity and value of control on each section. The equations of the problem were written in dimensionless form. It simplifies the numerical investigation of the obtained problem. There is a characteristic dimensionless parameter in the phase equations of the problem. This parameter is a combination of dimension variables describing the spacecraft and its orbit. Usually the problems of spaceflight mechanic are solved with the maximum principle. And we have to solve boundary value problems with some kind of shooting method (Newton’s method, gradient descent method etc.) Each shooting method requires initial values of conjugate variables, but we have no analytical formulas to find them. In this paper spacecraft flight trajectories were found with new genetic algorithm. Each gene contains additional parameter which equals to " True" , if the gene forms the control and equals to "False" otherwise. It helps us determine the quantity of spacecraft active motion parts. The input of proposed algorithm does not contain information about conjugate variables. It is well-known that the differential equations of the problem have a partial solution when the spacecraft orbit is circular and control is constant. The genetic algorithm involves this partial solution and its speed is increased. Numerical examples were constructed for two cases: when the difference between angular variables for start and final orientations of the spacecraft orbital plane equals to a few (or tens of) degrees. Final orientation of the spacecraft plane of orbit coincides with GLONASS orbital plane. The graphs of components of the quaternion of orientation of the orbital coordinate system, the longitude of the ascending node, the orbit inclination and optimal control are drawn. Tables were constructed showing the dependence of the value of the quality functional and the time spent on the reorientation of the orbital plane on the maximum length of the active section of motion.