Wisanlaya Pornprakun.. Optimal mathematical modelling of sugarcane harvesting in Thailand. Doctoral Degree(Applied Mathematics). King Mongkut's University of Technology North Bangkok. Central Library. : King Mongkut's University of Technology North Bangkok, 2020.
Optimal mathematical modelling of sugarcane harvesting in Thailand
Abstract:
The sugar industry is of great importance to the Thai economy. In general,
the government sets sugarcane prices at the beginning of each sugarcane crop
year based on type (fresh or fired), sweetness (sugar content) and gross weight.
The main aims of this thesis are to use three optimization methods and two
optimal control methods to find ''optimal cutting patterns' for harvesting fresh
and fired sugarcane for the four sugarcane producing regions of Thailand, namely
North, Central, East and North-east, for the crop years 2012/13 to 2018/19. The
three optimization methods used are the bi-objective e-constraints method, a
constrained quasi-Newton method and linear programming, and the two optimal
control methods used are discrete and continuous optimal control. For the bi-
objective and constrained quasi-Newton methods and linear programming, a crop
year was divided into 15-day periods and an optimal cutting pattern was denied as
the amount of each type of sugarcane harvested and delivered to the sugar factories
in each period that maximized the profit to the farmers subject to constraints on
the maximum amount that could be cut in each period. For linear programming
and discrete and continuous optimal control, optimal cutting patterns were found
for 1-day periods. Optimal cutting patterns were computed by all methods for
a range of values of cutting constraints and costs and compared with the actual
profits and actual cutting patterns. The results showed that all methods gave
similar optimal cutting patterns and profits. However, there was a big difference in
computation times with the linear programming and optimal control methods being
orders of magnitude faster than the bi-objective and quasi-Newton methods, even
though the linear programming and optimal control methods used approximately
180 1-day cutting periods, whereas the bi-objective and quasi-Newton methods
used only 12 15-day cutting periods. Further, the optimal control methods have
the advantage over linear programming that they can be used for both linear
and nonlinear problems. Finally, we found that the programming for the discrete
optimal control was much simpler than for the continuous optimal control and
that the computation times were also shorter. The data on prices, costs and actual
cutting patterns used in this thesis were obtained from the Ministry of Industry
and the Ministry of Agriculture and Co-operatives of the Royal Thai government.