近年來，人工智慧的議題逐漸受到重視，越來越多人以此求解最佳化的問題。而基因演算法為人工智慧的分支之一，它是以達爾文進化論為理念發展出來的：仿照天擇，利用族群的整體資訊，將適應力高的基因遺傳給下一代，並經由突變避免遺漏掉更好的基因。換成最佳化的問題來討論，遺傳優良基因可視為快速收斂至最佳解，而突變則是避免掉入局部最佳解。也因此，基因演算法對於找出全域最佳解的效率相當高，值得探討研究。

本篇論文將介紹基因演算法的緣由、步驟程序及應用。其中，我們需要搭配打靶法、最小平方法來解決最佳化問題。特別是軌道最佳化，我們可以了解到如何藉此找出適當的初始值，並與兩篇論文的結果做比較：一篇是於西元2017年由王璿豪先生等人共同著作的期刊論文``A full-space quasi Lagrange-Newton-Krylov algorithm for trajectory optimization problems" (簡稱為WLHH) ，另外一篇論文於西元2018年由連政杰先生所著的``A parallel full-space Lagrange-Newton method for
low-thrust orbit transfer trajectory optimization problems" (簡稱為CCL) 。而後我們欲證明初始值在可行解區域上或非可行解區域上的效益是否有差別，相關數據結果將會於最後呈現。

More recently, people put more and more emphasis on the artificial intelligence (AI) issue and solve optimization problems by it. However, genetic algorithms (GAs) is one branch of AI. It is based on the theory of evolution from Darwin: imitating the natural selection and using the total information of groups, the chromosomes with high adaptability are inherited to the new generation. In addition to this, the chromosomes may be mutated due to avoid missing the greater genes. In other words, inheritance is considered as converging to the optimal solution rapidly and mutation is preventing from falling into the local extrema. Consequently, the efficiency of finding the global extrema is excellent for GAs. It is worth to explore and research.

We will introduce the summary and the applications of GAs in this paper. Besides, we need to solve the optimization problem by shooting method and the least square method. Especially trajectory optimization, we can understand how to find the suitable initial guess, and compare the result with two papers: one is a journal paper ``A full-space quasi Lagrange-Newton-Krylov algorithm for trajectory optimization problems" written by Hsuan-Hao Wang et al. in 2017 (called WLHH), another is ``A parallel full-space Lagrange-Newton method for low-thrust orbit transfer trajectory optimization problems" written by Cheng-Chieh Lien in 2018 (called CCL) . If the initial guess is one of the feasible solutions, is the effect whether better or not? The numerical results will be presented at the end.

[1] D.R. Jones. Direct global optimization algorithm direct global optimization algorithm. In Encyclopedia of optimization, pages 431–440. Springer, 2001.

[2] M.S. Arumugam, M.V.C. Rao, and A. Chandramohan. A new and improved version
of particle swarm optimization algorithm with global-local best parameters. Knowl
Inform Syst, 16:331–357, 2008.

[3] K.C. Goh, M.G. Safonov, and G.P. Papavassilopoulos. A global optimization approach for the BMI problem. In Decis Contr, volume 3, pages 2009–2014, December
1994.

[4] Y.T. Kao and E. Zahara. A hybrid genetic algorithm and particle swarm optimization
for multimodal functions. Appl. Soft Comput., 8:849–857, 2008.

[5] R. Chelouah and P. Siarry. Genetic and Nelder–Mead algorithms hybridized for a
more accurate global optimization of continuous multiminima functions. Eur. J.
Oper. Res., 148:335–348, 2003.

[6] M. Kelly. An introduction to trajectory optimization: How to do your own direct
collocation. SIAM Rev., 59:849–904, 2017.

[7] Y.-C. Ma. Energy aware flow scheduling for data center network using genetic algorithms. Master’s thesis, National Chiao Tung University, 2012.

[8] F.-T. Hwang. Aeroacoustic antenna modeling for the near field. Master’s thesis,
University of Minnesota, 1996.

[9] F.W. Glover and G.A. Kochenberger. Handbook of Metaheuristics. Springer Science
& Business Media, 2006.

[10] A. Homaifar, C.-X. Qi, and S.-H. Lai. Constrained optimization via genetic algorithms. Simulation, 62:242–253, 1994.

[11] Ö. Yeniay. Penalty function methods for constrained optimization with genetic algorithms. Math. Comput. Appl., 10:45–56, 2005.
[12] F.-T. Lin. Evolutionary computation Part I: Three theoretic models of evolutionary
algorithms. Journal of Taiwan Intelligent Technologies and Applied Statistics, 3:1–28,
2005.

[13] F.-T. Lin. Evolutionary computation Part II: Genetic algorithms and their three
applications. Journal of Taiwan Intelligent Technologies and Applied Statistics, 3:29–
56, 2005.

[14] L.-C. Chang. Optimal planning of integrated groundwater management. Master’s
thesis, National Chiao Tung University, 2004.

[15] A. Chipperfield, P. Fleming, H. Pohlheim, and C. Fonseca. Genetic algorithm toolbox
for use with MATLAB. Technical report, 1994.

[16] T. Strutz. Data Fitting and Uncertainty: A Practical Introduction to Weighted Least
Squares and Beyond. Vieweg and Teubner,Wiesbaden, 2010.

[17] C.T. Kelley. Iterative Methods for Optimization. SIAM,Philadelphia, 1999.

[18] P.J.G. Teunissen. Nonlinear least squares. Manuscripta geodaetica, 15:137–150, 1990.

[19] B.A. Boukamp. A nonlinear least squares fit procedure for analysis of immittance
data of electrochemical systems. Solid State Ionics, 20:31–44, 1986.

[20] J.E. Dennis, D.M. Gay Jr, and R.E. Walsh. An adaptive nonlinear least-squares
algorithm. ACM Trans. Math. Softw., 7:348–368, 1981.

[21] S. Timothy. Numerical Analysis, 2nd Edition. Pearson, Boston, 2011.

[22] S.N. Ha. A nonlinear shooting method for two-point boundary value problems. Com-
put. Math. Appl., 42:1411–1420, 2001.

[23] H.-H. Wang, Y.-S. Lo, F.-T. Hwang, and F.-N. Hwang. A full-space quasi Lagrange-
Newton-Krylov algorithm for trajectory optimization problems. Electron. Trans.
Numer. Anal., 2017.

[24] C.-C. Lien. A parallel full-space Lagrange-Newton method for low-thrust orbit transfer trajectory optimization problems. Master’s thesis, National Central University,
2018.