University of Twente Student Theses

Login

Mattress Loading plan generator

Korbee, W. (2017) Mattress Loading plan generator.

[img] PDF
4MB
Abstract:The Calamity Jane is a support vessel from the Allseas fleet, which can install concrete structures, also called mattresses on the seabed to support and/or protect pipelines. To get the desired height, a stack often exists from multiple standardized mattresses. The Calamity Jane has two cranes which can lift multiple mattresses at the same time to complete the task faster. Most of the mattresses are however not in the reach of the cranes. A skidding system transports the mattresses horizontally, to get them under the crane. An accurate loading plan with a predefined unloading sequence is necessary to lift them as efficient as possible. It is almost not possible to rearrange the cargo on the deck. To increase the efficiency for a full and efficient loading for the vessel, an application is created which can automatically calculate loading plans for mattresses while minimizing the amount of lifts required for unloading for both cranes. An unloading sequence is made which also take care that skidding load, deck load and lifting loads are not too high. Also it deals with neat ordered stacks, maximum stacking height and that mattresses are always fully supported. Possible lifting configurations are combined with the coordinates from a discretized grid. This grid represents the small skidplates (3x3m) on which the mattresses are stored. Together they form the decision variables from a binary vector. Relations between the variables are made with (in)equality constraints in matrix vector form. The used solvers are mixed integer linear solvers from Matlab and Cplex, and a linear binary solver from Cplex. The solution is always binary, because a spot is fully occupied or not. To reduce the computation time several methods are tried which includes usage of a local grid (selection of decision variables), ordering the constraints from difficult to easy, fewer smarter constraints, hybrid constraints, try several objective functions and neglecting the smaller lifting configurations. When the system of equations is made it can be solved with two strategies: An iterative method which first solves the lifting configurations and then create the loading plan. This method can use a reduced system of equations but may be repeated several times. The full solver solves the loading plan and the lifting plan at the same time, but it has a large system of equations. It is highly dependent on the scenario which method is faster. The user of the program can insert the desired stacks on the seabed in an user friendly GUI which makes the constraint, selects the best option and get the result is a few seconds. The solution is given to the user in a convenient way with the orders of the mattresses, the location of the mattresses, mattress type and lift numbers. Also a flat printout on excel can be generated. The order of magnitude to obtain a feasible solution is for a large problem about half a minute.
Item Type:Internship Report (Master)
Faculty:ET: Engineering Technology
Subject:52 mechanical engineering
Programme:Mechanical Engineering MSc (60439)
Keywords:Optimization, Mattresses, Loading plan, Mixed integer solver, Binary solver, Linear, Runtime
Link to this item:http://purl.utwente.nl/essays/84017
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page