A multi-objective service composition method based on cost-benefit optimization

Abstract

The invention discloses a multi-objective service combination method based on cost benefit optimization. The method comprises two parts of a multi-objective service combination model and an elite guidance based multi-objective artificial bee colony EMOABC algorithm. According to the multi-objective service combination model, a service combination problem is modeled into the multi-objective service combination model with two optimization objectives of maximizing quality of service and minimizing cost. According to the EMOABC algorithm, a search for a service combination scheme is simulated by employing foraging of a bee. According to the multi-objective service combination model based on cost benefit optimization provided by the method, the complicated demand of a user can be satisfied, and compared with other algorithms, the EMOABC algorithm has clear advantages of operation efficiency and solution quality.

Description

Technical field

[0001] The invention relates to a multi-objective service combination method based on cost-benefit optimization, and belongs to the technical field of service-oriented computing. Background technique

[0002] In the service-oriented computing model, service, as a resource that can be accessed at any time, greatly facilitates the use of users. The loose coupling and high reusability of services make it possible to combine them to provide complex functions when a single service cannot meet user needs. Service composition technology can realize resource sharing, so it has become a research hotspot in recent years. However, the promotion of cloud computing has led to a sharp increase in the number of services in the network, which requires service composition methods to be more effective.

[0003] Service composition is the process of selecting appropriate service components in each service group for binding, and then combining the service components into a new servic...

Examples

Embodiment

[0165] Randomly generate 500,000 simulated service data, and the evaluation value of each service for each quality attribute is evenly distributed in the range (0,1). The experimental environment is: Intel Core i3-2370M (2.4GHz), 6.0GB RAM, Windows 7 (64bit), MATLAB R2010b. Compare the EMOABC algorithm with similar algorithms. In the experiment, each algorithm uses the same control parameters, the population number is 50, and all the experimental results are the average of 30 experiments. The parameters of the comparison algorithm are set as follows:

[0166] 1) NSGA-II: The crossover probability is set to 0.9, the mutation probability is 0.1, and the simulation binary crossover and polynomial mutation strategy is adopted. The distribution indexes of the crossover and mutation operators are both 20;

[0167] 2) MOPSO: The size of the repository is the number of populations, the inertia weight w is 0.4, and the individual learning coefficient c 1 And the global learning coefficien...

Technical field

[0001] The invention relates to a multi-objective service combination method based on cost-benefit optimization, and belongs to the technical field of service-oriented computing. Background technique

[0002] In the service-oriented computing model, service, as a resource that can be accessed at any time, greatly facilitates the use of users. The loose coupling and high reusability of services make it possible to combine them to provide complex functions when a single service cannot meet user needs. Service composition technology can realize resource sharing, so it has become a research hotspot in recent years. However, the promotion of cloud computing has led to a sharp increase in the number of services in the network, which requires service composition methods to be more effective.

[0003] Service composition is the process of selecting appropriate service components in each service group for binding, and then combining the service components into a new servic...

Examples

Embodiment

[0165] Randomly generate 500,000 simulated service data, and the evaluation value of each service for each quality attribute is evenly distributed in the range (0,1). The experimental environment is: Intel Core i3-2370M (2.4GHz), 6.0GB RAM, Windows 7 (64bit), MATLAB R2010b. Compare the EMOABC algorithm with similar algorithms. In the experiment, each algorithm uses the same control parameters, the population number is 50, and all the experimental results are the average of 30 experiments. The parameters of the comparison algorithm are set as follows:

[0166] 1) NSGA-II: The crossover probability is set to 0.9, the mutation probability is 0.1, and the simulation binary crossover and polynomial mutation strategy is adopted. The distribution indexes of the crossover and mutation operators are both 20;

[0167] 2) MOPSO: The size of the repository is the number of populations, the inertia weight w is 0.4, and the individual learning coefficient c 1 And the global learning coefficien...

Images