A vehicle routing problem involves finding a set of optimal route for a fleet of capacitated vehicles which are available at a location to service the demands of a set of customers. In its simplest form, a customer is required to be visited once and the capacity of a vehicle must not be exceeded. In this paper, a review of four common and applicable variants of the vehicle routing problem, namely, capacitated vehicle routing problem, vehicle routing problem with time windows, periodic vehicle routing problem and the dynamic vehicle routing problem, are considered based on formulation techniques, methods of solution and areas of application. A summary table is presented for each variant to emphasis some key features that represent direction of current research.

A survey of recent advances in vehicle routing problems

In the past one decade there has been significant increase in the growth of digital data. Therefore, good data mining techniques are important for the better decision making. Clustering is one of the key element in the field of data mining. K-means is a very popular algorithm present in the literature which is widely used for the clustering purpose. However k-means algorithm suffers from the problem of stucking into local optimum solution because of it’s dependency on the random initialization of initial cluster center. In this paper a novel variant of Bat algorithm based on dynamic frequency is introduced. Further the proposed variant is hybridized with K-means to present a new approach for clustering in distributed environment. Since evolutionary computation is very computation intensive, traditional sequential algorithms are not able to provide satisfactory results within the reasonable amount of time for the large scale data problems. To mitigate this problem the proposed variant is parallelized using the MapReduce model in the Hadoop framework. The experimental results show that the proposed algorithm has outperformed K-means, PSO and Bat algorithm on eighty percent of the benchmark datasets in terms of intra-cluster distance. Further DBPKBA has also achieved significant speedup for dealing with massive datasets with increase in the number of nodes.

Dynamic frequency based parallel k-bat algorithm for massive data clustering (DFBPKBA)

Unsupervised data classification (data clustering) is one of the mostly used data analysis methods which groups the unlabeled data into identical clusters (groups). Classical clustering methods do not perform effectively while clustering high dimensional datasets viz micro array datasets. Therefore, a novel clustering method based on Biogeography based optimization is proposed to extend the capabilities of traditional clustering methods. Performance of proposed method has been tested on the four micro-array datasets. Experimental results validate the effectiveness of proposed method.

Unsupervised data classification using improved biogeography based optimization

Direct torque control of dual star induction motor using a fuzzy-PSO hybrid approach

Efficient workflow scheduling in modern cloud environment involves optimization of various conflictive objectives like execution performance (time), reliability, energy consumption etc. Despite this trend, numerous heuristics have been devoted to workflow scheduling mainly focused on the optimization of makespan (execution time) only without giving much attention on other important objectives. Reducing energy consumption is the major concern as it brings several important benefits like reduction in the operating costs, increase in the system reliability and environmental protection. Moreover, the compute processors in cloud are not failure free. Any kind of failure can be critical for an application. Hence in this paper, we proposed the multi-objective NSGA-II based scheduling algorithm for workflow applications with the aim to optimize three conflicting criterion simultaneously: makespan\execution time, reliability and energy consumption for executing the workflow application in cloud environment. In order to reduce the computation complexity of the algorithm, we used the efficient non-domination level update mechanism rather than applying the non-domination sorting from the scratch each time. The simulation analysis of the proposed algorithm on CloudSim toolkit shows that the Pareto optimal solutions obtained have good convergence, uniform diversity and computational efficiency.

Meta-heuristic based reliable and green workflow scheduling in cloud computing

Particle swarm optimization (PSO) is a popular population based approach used to solve nonlinear and complex optimization problems. It is simple to implement and swarm based probabilistic algorithm but, it also has drawbacks like it easily falls into local optima and suffers from slow convergence in the later stages. In order to reduce the chance of stagnation, while improving the convergence speed, a new position updating phase is incorporated with PSO, namely fitness based position updating in PSO. The proposed phase is inspired from the onlooker bee phase of Artificial Bee Colony (ABC) algorithm. In the proposed position updating phase, solutions update their positions based on probability which is a function of fitness. This strategy provides more position updating chances to the better solutions in the solution search process. The proposed algorithm is named as fitness based particle swarm optimization (FPSO). To show the efficiency of FPSO, it is compared with standard PSO 2011 and ABC algorithm over 15 well known benchmark problems and three real world engineering optimization problems.

Fitness based particle swarm optimization

Particle Swarm Optimization (PSO) is a stochastic strategy used to solve complicated nonlinear optimization problems. But, like other swarm intelligence based algorithms, it also suffers from slow convergence and stagnation problems. These problems cause disproportion in the exploration and exploitation capabilities of the algorithms. Thus to make equilibrium between diversity and convergence, a strategy named L;evy Flight PSO (LFPSO) is developed. In the proposed strategy, a l'evy flight random walk inspired local search approach is hybridized with PSO. Further, it is shown that the proposed strategy provides competitive solutions as compared to standard PSO 2011 algorithm when tested on 10 well known benchmark problems.

Varsha Chhamunya

Papers

Fitness based particle swarm optimization

L'evy flight particle swarm optimization (LFPSO)