Harmony Search Algorithms for Water and Environmental Systems - Music-Inspired Harmony Search Algorithm: Theory and Applications - page 122

selection, a reserve that represents the only occurrence of a species, is selected uncon-
ditionally. Also, because each decision variable has binary value, there is no pitch ad-
justment.
Table 6 shows the computational results of SA [49], HS [46], and the exact method
(Opt) [50]. As seen in the table, the modified HS reached global optimum in 15 cases
out of 24 cases. When compared with SA, HS reached better solutions in 14 cases
while it reached worse solution only once.
Although HS did not reach the global optimum (all 426 species covered) with P =
23, it was able to find many different near-optima (426) with P = 24. As a practical
matter, the identification of alternate optima is valuable in planning and decision mak-
ing; if a particular reserve became unavailable for some reason, another solution that
utilized different parcels but had the same payoff could (potentially) be implemented.
6 Summary and Conclusions
This chapter has reviewed various HS applications in the areas of water resources and
environmental planning and management. Specific topics include design of water dis-
tribution networks, scheduling of multiple dam system, parameter estimation of flood
routing model, and determination of ecological reserve locations.
The computational results showed that HS has comparative advantages in clean
water distribution, irrigation, hydroelectric power generation, flood prediction, and
nature conservation. For clean water distribution, HS found the least-cost design with
the least number of function evaluations when compared with other methods. In
terms of maximal benefit from irrigation and power generation, HS found multiple
global optima while others only reached near-optima. To better predict flood amount
using hydrologic model, HS calibrated three hydrologic parameters and minimized
the difference between observed and routed data, and outperformed other methods.
Finally, for nature conservation HS was modified by adopting several problem-
specific operators and in many cases obtained more effective conservation plans than
another existing heuristic method (SA) in literature.
With the success of these applications, our future research direction focuses on fur-
ther promoting the HS algorithms to more research areas. Particularly, we will look
into how to integrate HS with various practical technologies such as Web, GUI, DB,
GIS, and API.
References
1. BBC News, Sanitation 'best medical advance' (2007) (accessed October 20, 2008),
http://news.bbc.co.uk/2/hi/health/6275001.stm
2. Rosing, K.E., ReVelle, C.S., Williams, J.C.: Maximizing species representation under lim-
ited resources: a new and efficient heuristic. Environmental Modeling and Assessment 7,
91-98 (2002)
3. Cesario, L.: Modeling, analysis, & design of water distribution systems. American Water
Works Association, Denver (1995)
4. Geem, Z.W., Tseng, C.L., Kim, J., et al.: Trenchless water pipe condition assessment using
artificial neural network. In: Najafi, M., Osborn, L. (eds.) Pipelines 2007: advances and
experiences with trenchless pipeline projects. ASCE, Reston (2007)