Figure 3 presents the general structure of the proposed capacity-restricted KMC algorithm (CKMC).
(i) In the Constructive Phase, the CKMC is executed only once for two random K values which will be used by the [mu]GA to determine the lower and upper limits ([K.sub.min], [K.sub.max]) for K.
(ii) In the Local Search Phase, the CKMC is iterated P times, and at each iteration, different values of K (within [K.sub.min] and [K.sub.max]) and the ratio of acceptance V (which has an upper limit given by X) are considered.
In this case, the individuals of the population of the [mu]GA consist only of pairs of values ([K.sub.min], [K.sub.max]) that can define the lower and upper limits of a range that may contain the K value that can lead to a total minimum distance on a single execution of the CKMC algorithm.
As presented in Figures 1 and 3, the best solution found by the CKMC in the Local Search Phase is improved by a decision algorithm which performs insertion, deletion, and exchange of points between clusters.
Thus, small values were considered for X (the upper limit for the number of nearest points to each cluster), the executions of the CKMC in the Local Search Phase (P) and the number of pairs of points to be considered for exchange, deletion, or insertion (Y).
(iv) Integrating the use of Artificial Neural Networks (ANNs) to dynamically determine the number of clusters to improve speed and convergence of the CKMC algorithm.