Water network design (Program screen shot ~18KB)

This is a more complicated example which uses optiGA to solve the Optimal Water Distribution Network Design problem.
The sample uses a method, partially described by Dr. Dragan Savic and Dr. Godfrey Walters, and uses the hydraulic solver "EPANET Toolkit" developed by Mr. Lewis Rossman from the U.S. Environmental Protection Agency.

In this example you will find the implementation of many of optiGA's features. See the source code!!!

This example is taken from the article "Design of optimal water distribution systems" by E. Alperovits and U. Shamir, 1977.

Consider the network shown below, which has eight pipes of 1000 meters length each, arranged in two loops and is fed by gravity from a constant head reservoir.

The demands are given in the data table below and the head at each node is to be at least 30 meters above the ground elevation of the node.

The pipes costs, per unit length, are given in the table below.

Chw is 130 for all pipes.

The demo program can solve this problem with discrete or continues diameters. You also have control on the number of generation the genetic algorithm will run, the population size and the mutation probability.
Try changing these parameters and see how they effect the network cost. For good results use 0.03 as mutation probability for discrete diameters and about 0.15 for continues diameters. For continues diameters, you will get good result if you let the genetic algorithm run more generation (about 500).

NODE	DEMAND	ELEVATION
1	-1120	210
2	100	150
3	100	160
4	120	155
5	270	150
6	330	165
7	200	160

Pipes costs:

Diameter (")	Unit Cost
1	2.0
2	5.0
3	8.0
4	11.0
6	16.0
8	23.0
10	32.0
12	50.0
14	60.0
16	90.0
18	130.0
20	170.0
22	300.0
24	550.0