Overview of Applications and Developments in the Harmony Search Algorithm - Music-Inspired Harmony Search Algorithm: Theory and Applications - page 23

that it experiences. Several soil layers may be present with different properties. The
location of slip surface is found by minimising the factor of safety.
Two-dimensional slip surfaces of arbitrary shape were analysed using HS in [62],
modified HS methods in [63, 64], a hybrid GA-HS algorithm in [11, 65, 66], three
new Chaos HS algorithms in [67] and hybrid PSO-HS in [28]. A hybrid PSO-HS
method was also used for three-dimensional slope stability analysis in [68]. Cheng et
al [19] presented a comprehensive comparison of six meta-heuristic optimisation
methods for slope stability analysis, including two HS-based methods. These modi-
fied HS methods are discussed in more detail in [69, 70].
3.5 Information Technology Applications
Several IT applications tackle data clustering. The aim of clustering is to divide a data
set into groups such that there is a high level of similarity for members within a group,
but a low level of similarity between different groups. Clustering of web documents was
studied in [71, 15] using three novel hybrid HS - K -means clustering algorithms. Fuzzy
classification of the Fisher Iris data set was investigated in [72]. Initial classification was
performed using the Fuzzy C -Means (FCM) method then optimisation of the fuzzy
membership functions was accomplished by a new hybrid HS - Clonal Selection Algo-
rithm (CSA) method. Malaki and co-workers [73] developed two hybrid IHS-FCM
clustering algorithms, which were tested on a ~58,000 element NASA radiator data set.
Forsati et al [16, 74] studied multi-cast routing, which refers to the transmission of
the data from a sender through a network to multiple recipients. Two modified HS al-
gorithms were developed to solve the least cost multi-cast routing problem subject to
maximum bandwidth and delay time constraints.
Cruz et al [75] used a new hybrid optimisation algorithm inspired by GA, Simu-
lated Annealing (SA) and HS for a parameter estimation problem in the development
of virtual urban environments.
3.6 Transport-Related Problems
A 20-city TSP was solved in [1, 7] and a 51-city TSP was also presented in [1], both
using a modified HS algorithm. A school bus routing problem was investigated in [76,
77] and briefly in [32]. The problem was to find the required number of school buses
and the best route for each bus subject to constraints on the bus seating capacity and
maximum allowable journey time. A generalised orienteering problem, for the best
touring in China, was solved using a modified HS method in [78]. The aim was to
find the tour route that maximised the collective tourism opportunities offered by each
city, subject to a constraint on the maximum tour length.
A parameter estimation problem for the annual energy demand of the Turkish
transportation sector was reported in [79].
3.7 Thermal and Energy Applications
Geem and Hwangbo [12] performed multi-objective optimisation for the design of a sat-
ellite heat pipe. The problem involved finding the heat pipe dimensions and operating
temperature to minimise the heat pipe's mass while maximising its thermal conductance.