3D Piping Route Design Including Branch and Elbow Using Improvements for Dijkstra's Algorithm

Abstract

Abstract-In ship design problem, piping route design is the most complex stage because the internal spaces inside the ship are limited and the ship, in general, has many systems and subsystems that consist a large number of pipes. In order to optimize the piping cost as well as the construction time, so many methods have been considered and applied. But, most of methods just consider to problem of shortest path while the piping cost and the construction time also depend on the branch and number of bend & elbow. Dijkstra's algorithm is one of the most famous method of deterministic approach methods because it is always for the best solution. But it takes a long calculation time, occupies much memory and just consider to problem of shortest path. To fix that disadvantage, in this paper, the Dijkstra's algorithm with some improvements is used to reduce the calculation time, occupied memory and the shortest paths which are exported by this method will have a minimum number of bend & elbow. An example in a 3D simple engine room was used to show the results what are exported by using our method.

A Distance-Field-Based Pipe-Routing Method

“…Piping algorithms can be categorized into two major groups, the automatic methods and the semi-automatic ones [ 1 ]. The former group includes deterministic and heuristic methods [ 3 , 4 , 5 , 6 , 7 , 8 , 9 , 10 , 11 , 19 ], while the latter is mainly composed of expert systems and immersive virtual reality procedures [ 2 , 18 ]. Deterministic piping procedures always produce the optimum solution when routing a single pipe.…”

“…In Lee’s work, the problem domain is divided into cells, and the pipe path is routed upon the cells. His domain-decomposition method has had profound influences on later researches, including [ 2 , 3 , 4 , 5 , 7 , 8 , 12 , 13 , 14 , 15 ] and this work. In [ 12 ], Kim et al propose to split the domain into regions first.…”

“…However, the safety regulations and spatial constraints which should be met by a pipe have been well studied and published in the literature. They can be briefly summarized as follows [ 1 , 7 , 8 ]: The pipe path must be collision-free and connect the sources and destinations. The pipe length and number of elbows should be minimized.…”

“…Conditions 1 and 2 are set for producing a functionable and low-cost pipe. These two conditions can be met if Dijkstra’s shortest path method or the A* algorithm is used to lay out the pipe [ 1 , 7 , 8 , 9 ]. Optimizing the total cost of a group of pipes is NP-hard and can be approximated by heuristic methods, including the particle swarm intelligence, ant colony, and genetic algorithms [ 3 , 4 , 10 , 11 , 12 , 13 ].…”

Railway route optimisation from Romania to Poland based on an analysis of China's 'One Belt and One Road' initiative

“…Chen (2013) established a dynamic network model based on the Dijkstra algorithm to find optimal paths in emergency logistics networks. Nguyen et al (2016) defined the route scheduling problem as the most difficult due to the complexity of the factors. To discover the optimal, lowest cost route, he altered the Dijkstra algorithm to examine a simple engine room.…”