Low-thrust spacecraft trajectory optimization via a DNN-based method

Yin, Shanshan; Li, Jian; Cheng, Lin

doi:10.1016/j.asr.2020.05.046

Cited by 26 publications

(11 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…where λ f ¼ ½ν 1 , ν 2 , ν 3 , 0; 0; 0 T . Then stationarity conditions for saddle-point controls α * i and β * i are in equations ( 25)- (28).…”

Section: Free-time Eopegmentioning

confidence: 99%

“…When getting the root of equation ( 31), the saddle-point controls of two players will be given by equations ( 25)- (28).…”

Section: Free-time Eopegmentioning

confidence: 99%

“…Equation ( 45) indicates that the Euclidean norm of λ does not change the optimal controls, which can also be derived from the equations ( 25)- (28). As a conclusion of the analysis, the terminal position costates are restricted to an unit-sphere.…”

Section: Normalization Of Costatesmentioning

confidence: 99%

“…Since that the DNNs could be trained off-line and the multi-scale cooperation strategy was introduced, the DNNs-based controller had remarkable performance in efficiency and robustness. Yin et al 28 studied on the low-thrust orbit transfers via the DNNs. The homotopy technique was further applied to the fuel-optimal problem.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Numerical solution for elliptical orbit pursuit-evasion game via deep neural networks and pseudospectral method

Zhang

Zhu

Lee

et al. 2022

Proceedings of the Institution of Mechanical Engineers, Part G:

View full text Add to dashboard Cite

This paper presents an efficient and stable DNNs-based Radau pseudospectral method for the free-time elliptical orbit pursuit-evasion game based on the equivalent reconstruction of the game model. Firstly, the relative dynamics equations are established by adding the nonlinear terms caused by the eccentricity to the Hill–Clohessy–Wilshire equations. Then the original pursuit-evasion problem is deduced to a 4-dimensional one-sided optimal control problem (OCP) based on the equivalent reconstruction. Secondly, in order to apply the deep neural networks (DNNs) to map the relationship between the OCP and the solution, the normalization of costates is introduced to eliminate the non-uniqueness of solutions when generating samples for training DNNs. Thirdly, the DNNs-based Radau pseudospectral method is proposed where the DNNs output the guesses of solutions to the derived OCP and the Radau pseudospectral method optimizes the histories of controls obtained by the guesses to the convergence. The simulation results demonstrate that the proposed method converges more stably and decreases the calculation time greatly compared with the traditional indirect method.

show abstract

“…where λ f ¼ ½ν 1 , ν 2 , ν 3 , 0; 0; 0 T . Then stationarity conditions for saddle-point controls α * i and β * i are in equations ( 25)- (28).…”

Section: Free-time Eopegmentioning

confidence: 99%

“…When getting the root of equation ( 31), the saddle-point controls of two players will be given by equations ( 25)- (28).…”

Section: Free-time Eopegmentioning

confidence: 99%

Section: Normalization Of Costatesmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Numerical solution for elliptical orbit pursuit-evasion game via deep neural networks and pseudospectral method

Zhang

Zhu

Lee

et al. 2022

Proceedings of the Institution of Mechanical Engineers, Part G:

View full text Add to dashboard Cite

show abstract

“…This study proposes to acquire a fuel-optimum path of LSL with variable propulsion. This trajectory optimization (TO) issue is designed into constrained optimum control complexity linked to key features of the problem [ 8 – 12 ]. The Japanese lunar exploration plan is a robot and human travel program for the moon; its principal purpose is to clarify the origin and the moon's evolution and its future use for the moon [ 13 – 16 ].…”

Section: Introductionmentioning

confidence: 99%

Hybrid Approach Named HUAPO Technique to Guide the Lander Based on the Landing Trajectory Generation for Unmanned Lunar Mission

Latif

Mehedi

Iskanderani

et al. 2022

Computational Intelligence and Neuroscience

View full text Add to dashboard Cite

This manuscript proposes a hybrid method for landing trajectory generation of unmanned lunar mission. The proposed hybrid control scheme is the joint execution of the human urbanization algorithm (HUA) and political optimizer (PO) with radial basis functional neural network (RBFNN); hence it is named as HUA-PORFNN method. The HUA is a metaheuristic method, and it is used to solve several optimization issues and several nature-inspired methods to enhance the convergence speed with quality. On the other hand, multiple-phased political processes inspire the PO. The work aims to guide the lander with minimal fuel consumption from the initial to the final stage, thus minimizing the lunar soft landing issues based on the given cost of operation. Here, the HUAPO method is implemented to overcome thrust discontinuities, checkpoint constraints are suggested for connecting multi-landing phases, angular attitude rate is modeled to obtain radical change rid, and safeguards are enforced to deflect collision along with obstacles. Moreover, first, the issues have been resolved according to the proposed HUAPO method. Here, energy trajectories with 3 terminal processes are deemed. Additionally, the proposed HUAPO method is executed on MATLAB/Simulink site, and the performance of the proposed method is compared with other methods.

show abstract