Global Journal of Computer Science and Technology, D: Neural & Artificial Intelligence, Volume 22 Issue 1

Figure 2: Process Flow of the Proposed Solution [Reference number] As shown in the above figure 2, Ant Colony Optimization (ACO) algorithm is a probabilistic technique based on the above phenomenon to find the optimal path. With the inclusion of K-Means Clustering, this modified approach has solved the constraints of the MPMDVRPTWHF, which has resulted in shorter time consumption, delivery within the time window and lower transportation costs along with the inclusion of multiple pickup and delivery nodes wherein a pickup point might or might not have multiple delivery locations. a) Parameter initialization Looking at the research done in [28], the following parameters are set as follows: Number of ants = 22 , = 2, = 5, = 0.80, = 80, = 3. In the graph = ( , ) , each arc ( , ) has been assigned a variable called pheromone trail . The probability of better solution is directly proportional to the pheromone intensity. This means that when an ant wants to go to another node from its current node, it will choose the one with the maximum pheromone intensity. to make this work, a fixed quantity of pheromone is allocated to every arc. To decide which node to proceed to (node ), the ℎ ant will use the pheromone trail which is showcased below in equation 14: Initially, all probabilities are set to 1. = 1 is a heuristic value, pheromone concentration on the edge when the ant travels from node to nodev is denoted by and relative influence of the pheromone concentration and the heuristic value is shown by and . If we go into the specifics, then denotes how much favourable is the next node while implies how much better is the next node relatively. b) Solution Construction In this scenario, the solution is generated when an artificial ant takes vehicles from the vehicle set and constructs a path, starting from the warehouse or depot, by choosing those nodes that satisfy the set of constraints. The ant continues to build the route until the limit of route length has been reached or when the time window constraint has been disobeyed. so, in forming the route, the ant will check each node whether it fulfils all the constraints and if it finds such a node, it will Global Journal of Computer Science and Technology Volume XXII Issue I Version I 53 ( )D Year 2022 © 2022 Global Journals Vehicle Routing Problem with Time Window Constrain using KMeans Clustering to Obtain the Closest Customer