Swarm Intelligence Algorithms And Their Applications

Swarm intelligence algorithm is an algorithmic approach, inspired by the foraging behavior of the real animals, which had been applied to many problems. The dissertation focuses on the principles, theory, and applications of Ant Colony Optimization algorithm (ACO) and Particle Swarm Optimization (PSO), especially, an in-deep and systemic study on how to improve the basic ACO and PSO algorithm, parallel implementation of ACO and PSO, solving the problems such as large scale TSP and continuous space optimization, constrained optimization, wireless sensor network routing protocol. The main works can be summarized as follows:1. To tackle the large scale TSP, An Ant Colony Algorithm in pheromone Diffusion model based on Modularity Clustering is proposed. Firstly, the large scale TSP is divided into several clustering based on modularity measure. Secondly, all the clustering will be solved in parallelization by an improved ant colony algorithm based on the pheromone diffusion model, this algorithm formulate the new rule of pheromone diffusion, and a disturbance mechanism is added to ant colony optimization for improving the new algorithms'performance. The simulation of TSP problem shows that this algorithm has a satisfied global convergence. Lastly, all the solutions of each clustering will be merged into the solution of the large scale TSP by some rules. Simulation results show that the convergence of the proposed algorithm has been improved.2. A hybrid optimization algorithm, in which Alopex algorithm is embedded into the improved ant colony optimization algorithm, is proposed for searching continuous space optimization. In the algorithm, the new pheromone updating rule and the moving strategy of ants are defined. The algorithm is with the rapid search capability of the improved Alopex algorithm and the good search characteristics of the improved ant colony optimization algorithm. Simulation results show that the algorithm is effective.3. A hybrid algorithm is proposed by combining particle swarm optimization(PSO)with Alopex algorithm that is a stochastic optimization method, for solving constrained optimization problems. In the algorithm, inertia weight is zero, and the position of the particle whose evolution has been stopped is produced by Alopex algorithm to improve the global search ability. Simulation results show that the algorithm is effective. 4. A co-evolutionary particle swarm optimization algorithm to solve constrained optimization problems is proposed. Firstly, A new co-evolutionary PSO (CPSO) is constructed. In the algorithm, a deterministic selection strategy is proposed to ensure the diversity of population. Meanwhile, based on the theory of extrapolation, the induction of evolving direction is enhanced by adding a co-evolutionary strategy, in which the particles make full use of the information each other by using gene-adjusting and adaptive focus-varied tuning operator. Secondly, infeasible degree selection mechanism is used to handle the constraints. A new selection criterion is adopted as tournament rules to select individuals. Also, the infeasible solution is properly accepted as the feasible solution based on a defined threshold of the infeasible degree. This diversity mechanism is helpful to guide the search direction towards the feasible region. Our approach was tested on six problems commonly used in the literature. The results obtained are repeatedly closer to the true optimum solution than the other techniques.5. To make efficiently use of the limited energy resource of the wireless sensor network (WSN) and prolong the lifetime is an important problem in the study of the WSN. An energy efficient routing algorithm based on Ant Colony Optimization for wireless sensor network is presented. The algorithm forms a clustering structure based on real network structure of wireless sensor network first and we use a new parameter-Modularity Measure to evaluate whether the clustering fits for the real network structure. Based on the above steady cluster structure, when we select the cluster head in each cluster, the residual energy of the nodes and the energy distributing in the cluster are both considered. The algorithm constructs a novel probabilistic model; the model considers both the overhead on the route and the residual energy of the node. Simulation results show that compared with other algorithms like LEACH, our approach is able to obtain a more reasonable and steady distribution of clustering, and can effectively prolong the sensor network lifetime.