Higher-Order Nonlinear Discrete Approximate Iteration on the Continuing Dynamic Programming

Abstract:

In order to solve the “dimension curse” and higher-order nonlinear, the paper proposes the discrete approximate iteration, and uses it to solve the continuum dynamic programming. Firstly, according to the network method, discretizes the state variables and transforms the model into multiperiod weighted digraph. Secondly, uses the max-plus algebra to solve the minimal path that is the admissible solution. At last, based on the admissible solution，we continues iterating until the two admissible solution is near. The paper also proves the convergence of the method.

• Periodical: Advanced Materials Research (Volume 459)
• Pages: 571-574
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.459.571
• Online since: January 2012
• Authors: Peng Zhang, Jing Yi Zhou
• Keywords: Discrete Approximate Iteration, Dynamic Programming, Max-Plus Algebra, Pivoting Algorithm

References

[1] Bellman R. . Dynamic programming [M]. Princeton University Press, Princeton, (1957).

[2] Qin Yuyuan. Optimum Path Problems in Networks [M]. Hubei Education Press, 1992. (in Chinese).

[3] S. Senthil Kumar, V. Palanisamy. A dynamic programming based fast computation Hopfield neural network for unit commitment and economic dispatch [J]. Electric Power Systems Research 2007 (77) 917–925.

DOI: 10.1016/j.epsr.2006.08.005

[4] Sitarz S., Hybrid methods in multi-criteria dynamic programming[J], Applied Mathematics and Computation 2006 (180): 38–45.

DOI: 10.1016/j.amc.2005.11.164

[5] Trzaskalik T., Sitarz S.,. Discrete dynamic programming with outcomes in random variable structures, European Journal of Operational Research, 2007(177/3): 1535-1548.

DOI: 10.1016/j.ejor.2005.10.019

[6] Sitarz S. Ant algorithms and simulated annealing for multicriteria dynamic programming [J]. Computers & Operations Research, 2007: 1-14.

DOI: 10.1016/j.cor.2007.09.011

[7] D. P. De Farias, B. Van Roy. The linear programming approach to approximate dynamic programming[J]. Operations Research, Vol. 51, No. 6, November–December 2003, p.850–865.

DOI: 10.1287/opre.51.6.850.24925

[8] Zhang Peng, The research On the Multiperiod M-SV Portfolio Selection Optimization by Discrete approximate iteration [J]. Economic Mathematics, 2008, 3: 124-129. (in Chinese).

[9] Zhang Peng, Zhang Zhong-zhen, Yue Chao-yuan. Optimization of the Mean Semi - absolute Deviation Portfol io Selection Model with the Restricted Short Sell ing Based on the Pivoting Algorithm [J]. Chinese Journal of Management Science，2006, 1: 7—11. (in Chinese).

[10] Zhang Peng, Zhang Zhong-zhen, Zeng Yong-quan. The Optimization of The Portfolio Selection with The Restricted Short Sales [J]. Application of Statistics and Management, 2008, 1: 124—129. (in Chinese).

[11] Zhang Peng. The Comparison Between Mean-variance and Mean-VaR Portfolio Models Without Short Sales [J]. Chinese Journal of Management Science，2008, 4: 30—35. (in Chinese).

[12] Heidergott B, Olsder G. J. der Woude J. V. Max Plus at Work—Modeling and Analysis of Synchronized Systems: A Course on Max-Plus Algebra and its Applications [M]. Princeton University Press, (2006).

DOI: 10.1515/9781400865239