Spatial Network Generation Tool for Performance Analysis of Water Distribution Networks - Pattern Recognition for Reliability Assessment of Water Distribution Networks - page 80

Step 7 : The IDs for each edge in the generated sets are detected representing two connected
nodes, a start node ('NodeFrom') and an end node ('NodeTo').
Step 8 : The screening of sub-graphs takes place in this step; the layouts of duplicated pipes,
isolated nodes or islands of pipes will be eliminated.
Step 9 : The attributes of finally selected sub-graphs are saved in a text file, after assigning
the values for length, diameter and roughness from selected range of values (or pipe
coordinates for the lengths).
Steps 5 to 9 are repeated until the process of generation has been completed. In non-random
generation, the process ends when all elements in the rows and columns have been visited, or
the (lesser) specified number of networks to be generated has been reaches. In random
generation, that is the moment when the selected number of networks has been generated.
Step 10 : The *.inp files are created from the generated text file for each network according to
the required EPANET format.
4.5.1 Screening of Sub-Graphs
The need to eliminate the generated sub-graphs exists in the following three cases:
1. If the number of edges is less than n -1, some of the n vertices will not be connected, e.g.
in case of combination {1,2} on Figure 4.15, left.
2. The number of edges can be equal or greater than n -1, and still some of the n vertices will
not be connected, e.g. in case of combination {1,3,5} on Figure 4.15, middle.
3. The sub-graph may contain the number of isolated edges creating islands, e.g. in case of
combination {2,5,7,8} on Figure 4.14, right.
Figure 4.14 Faulty sub-graphs
The flow chart for the screening of sub-graphs mentioned above as Step 8 of the NGT
algorithm described above is given in Figure 4.15. The screening process is similar to the
Breadth First Search algorithm (BFS) discussed by Sempewo et al. (2008). The algorithm
applied in NGT can be described in six steps:
Step 1 : After a (non)-random generation process has been completed, a set of links that need
to be screened on the above features will be stored in an array, which is a new input along
with the array of nodes.