Showing papers on "Multi-swarm optimization published in 1970"

Cognitive and social information based PSO

Abstract: In 1995 swarm intelligence based PSO (Particle Swarm Optimization) has been designed and implemented for solving optimization problems. Since then many researchers have developed many versions, based upon its theoretical concept, technical aspects and parameters involve in the algorithm. In broad sense, every swarm updates its position based upon the knowledge of its initial velocity and accelerating components such as cognitive and social information. In this paper we have presented a reformed and modified concept of PSO with the thought that every swarm updates its position based upon cognitive and social environment knowledge only and the key aspect used here is that these parameters are no longer assumed to be accelerating components rather position components. This algorithm is termed by us as Cognitive and Social Information based PSO (CSIPSO). The performance of CSI-PSO is validated by 23 benchmark functions and the empirical results clearly support the effectiveness of our concept. Keywords : Particle Swarm, PSO, Swarm Theory, Benchmark functions

Hybrid computer optimization of systems with random parameters

Abstract: A hybrid computer Monte Carlo technique for the simulation and optimization of systems with random parameters is presented The method is applied to the simultaneous optimization of the means and variances of two parameters in the radar-homing missile problem treated by McGhee and Levine

Multiple-objective, Shape Optimal Design ViaGenetic Optimization

Abstract: A multiple objective, goal programming formulation is coupled to a hybrid genetic optimization nonlinear programming algorithm and applied to the design of structures. The combination provides a unique design environment which can generate solutions not attainable by a traditional approach to structural optimization. Crossectional, geometric and topological change are considered in the formulation. Design goals include weight, cost and robust character. Both hard and soft constraints are included. The genetic optimizer controls the topology of the structure, while a gradient based, nonlinear programming method refines the local geometry and crossectional properties. A specific example involving a ten bar truss is presented which highlights the benefit of the approach over previously reported results.

Nonlinear Multi-objective Optimization OfMachining Processes Parameters

Abstract: Multi-objective optimization applications in manufacturing of structural components have been mostly in the area of scheduling and process planning. Until now most well developed multi-objective optimization codes are linear. Due to the high level of nonlinearity, few applications have been reported in the area of multi-objective optimization of manufacturing processes parameters. This paper presents a general nonlinear multi-objective optimization model for machining processes parameters. The developed nonlinear multi-objective optimization model allows different priority level for each manufacturing objective. The resulting nonlinear optimization problem is solved using unconstrained optimization techniques. This reduces the effort required to model the manufacturing process problem since linearization is not required. The optimum solution can also be achieved from any starting point, since no feasibility conditions are required. The application and efficiency of the developed model is studied using a machining process test case with different manufacturing priority.

Enhanced Performance of Particle Swarm Optimization with Generalized Generation Gap Model with Parent-Centric Recombination Operator

Abstract: Particle Swarm Optimization (PSO) algorithm has recently gained more attention in the global optimization research due to its simplicity and global search ability. This paper proposes a hybrid of PSO and Generalized Generation Gap model with Parent- Centric Recombination operator (G3PCX) [25], a well-known real-coded genetic algorithm. The proposed hybrid algorithm, namely PSPG, combines fast convergence and rotational invariance of G3PCX as well as global search ability of PSO. The performance of PSPG algorithm is evaluated using 8 widely-used nonlinear benchmark functions of 30 and 200 decision variables having different properties. The experiments study the effects of its new probability parameter Px and swarm size for optimizing those functions. The results are analyzed and compared with those from the Standard PSO [14] and G3PCX algorithms. The proposed PSPG with Px = 0.10 and 0.15 can outperform both algorithms with a statistical significance for most functions. In addition, the PSPG is not much sensitive to its swarm size as most PSO algorithms are. The best swarm sizes are 40 and 50 for unimodal and multimodal functions, respectively, of 30 decision variables.

Optimization Algorithm Of Parameters In The TankModel

Abstract: This paper proposes the optimization technique to identify the parameters in Sugawara's Tank model that has been widely employed for long-term runoff analysis. Researchers have developed some mathematical optimization methods to identify unknown parameters, notably, the Newton method, the Powell method and the Davidon-Fletcher-Powell method. Computational burden is too stringent in application of the Powell method to parameter optimization in the Tank model, because a great number of runoff computations are required. In this paper, the unknown parameters are identified essentially by the Newton method. The efficiency of optimization performance resulting from use of the Newton method clearly depends on how effectively the sensitivity coefficients can be derived. An important feature of the proposed approach is the theoretical derivation of sensitivity coefficients which can directly be used in the optimization scheme of the Newton method. A vector differential equation in terms of storages in the tanks is numerically integrated using the transition matrix which is computed by expanding the matrix exponential. The elements of this transition matrix play a significant role in eliminating additional computations involved in the solution of sensitivity coefficients. On the basis of simulation results, the Newton method coupled with sensitivity coefficients appears to have potential performance advantages for optimizing the parameters in the Tank model.