This paper presents a Hybrid system for numerical global optimization problems based on the geneticalgorithms (GAs) and modified GRID-point search. Experimental results indicate that the Hybrid system outperforms the...
详细信息
This paper presents a Hybrid system for numerical global optimization problems based on the geneticalgorithms (GAs) and modified GRID-point search. Experimental results indicate that the Hybrid system outperforms the classical GAs as the modified GRID can (i) speed up the search, (ii) further improve the fine tuning capabilities of GAs, and (iii) overcome the premature termination. The Hybrid system not only improve the searching capabilities of classical GAs but it also preserves the randomization of the searching space. In addition, the effectiveness of the genetic operators is addressed in this paper.
The chosen shape representation defines the subset of the shape space that the genetic Algorithm can search. If the fitness function can not correctly evaluate all parts of the shape space then we must ensure that the...
详细信息
The chosen shape representation defines the subset of the shape space that the genetic Algorithm can search. If the fitness function can not correctly evaluate all parts of the shape space then we must ensure that the shape representation only defines that subset which can be correctly evaluated. This paper details a parametric description which was used after earlier less restrictive shape representations produced totally impractical designs.
Test data were automatically generated for different procedures. The tests were derived from the program's structure and the aim was to execute every branch in the software code (structural testing). genetic Algor...
详细信息
Test data were automatically generated for different procedures. The tests were derived from the program's structure and the aim was to execute every branch in the software code (structural testing). geneticalgorithms (GAs) are used to generate these test data. The emphasis of this paper is to investigate the usage of different fitness functions and representations for different types of input variables. GAs were chosen because they are able to handle input data which may be of complex data structure and to execute branches whose predicates may be a complicated and unknown function of the input data.
This paper describes an investigation into using a genetic algorithm to evolve the optimum set of inputs for a neural network. The network is to be used in a novel way for the prediction of nuclear reactor parameters ...
详细信息
This paper describes an investigation into using a genetic algorithm to evolve the optimum set of inputs for a neural network. The network is to be used in a novel way for the prediction of nuclear reactor parameters under fault conditions. The development of transients is calculated in a recursive manner. The previous work and the next stage of research are described. The procedure and genetic algorithm options, including fitness, are discussed along with explanations. Finally an outline of the remaining work is introduced.
A structural genetic algorithm is proposed to optimize the neural network topology and connection weightings. This approach is to partition the genes of chromosome into control genes and connection genes in a hierarch...
详细信息
A structural genetic algorithm is proposed to optimize the neural network topology and connection weightings. This approach is to partition the genes of chromosome into control genes and connection genes in a hierarchical fashion. The control genes represented in bits are used to govern the layers and neurons activation and considered to be the higher level genes. Whereas the connection genes in the form of real values are the weightings and bias representations, regarded as the lower level genes. This inherent genetic variations enable multiple changes in lower level genes by a single change at the higher level genes. Such formulation of chromosome is found to be a phenomenal improvement over the traditional GA approach that without genes classification. As a result, the learning technique of the neural network is greatly improved. Simulation results have indicated that the proposed learning scheme requires the least iteration steps to reach a optimum network as compared to the uses of backpropagation and traditional non-structural geneticalgorithms.
This paper first discusses problems existing in neural network design using mathematically guided training methods. It then presents a genetic algorithm based design technique to train the network, which overcomes all...
详细信息
This paper first discusses problems existing in neural network design using mathematically guided training methods. It then presents a genetic algorithm based design technique to train the network, which overcomes all these problems. The paper also presents suitability conditions for using the genetic algorithm based design methods and develops, under these conditions, direct neurocontrollers with a novel structure inspired by proportional plus derivative control. Techniques are also developed to select the architectures in the same process of parameter training. The proposed methods are validated by several examples, including one with plant transport-delay.
The relative positions of the antennas of a direction finder are critical regarding the efficiency and the robustness of the indications of the apparatus. The state of the art in direction finder optimization is limit...
详细信息
The relative positions of the antennas of a direction finder are critical regarding the efficiency and the robustness of the indications of the apparatus. The state of the art in direction finder optimization is limited with classical methods to the case of direction finders with 3 antennas, with naive trial-and-error methods for more complex cases. The use of geneticalgorithms brings a tremendous increase to that domain, allowing the optimization of direction finders with up to 10 antennas in a reasonable computing time. Moreover, as GAs provide multiple optimal solutions on multi-modal problems, a thorough numerical investigation of all possible optima provides a new insight in the underlying optimization problem, finally leading to the derivation of a formula for the optimal direction finders.
A genetic algorithm has been compared with a semi-automatic experimental design approach in sequential elimination of levels (SEL). While the GA is a very general method, SEL needs some human input before it can be se...
详细信息
A genetic algorithm has been compared with a semi-automatic experimental design approach in sequential elimination of levels (SEL). While the GA is a very general method, SEL needs some human input before it can be set up for a particular problem. Otherwise, it proceeds in a more automatic manner than traditional experimental design methods. Several areas have been identified which require further study including the nature of fitness landscape in a particular problem which need some more exploration in order to determine just how challenging a problem it is.
We present a genetic algorithm (GA) whose population possesses a spatial structure. The GA is formulated as a probabilistic cellular automaton: The individuals are distributed over a connected graph and the genetic op...
详细信息
We present a genetic algorithm (GA) whose population possesses a spatial structure. The GA is formulated as a probabilistic cellular automaton: The individuals are distributed over a connected graph and the genetic operators are applied locally in some neighborhood of each individual. By adding a self-organizing acceptance threshold schedule to the proportionate reproduction operator we can prove that the algorithm converges to the global optimum. First results for a multiple knapsack problem indicate a significant improvement in convergence behavior. The algorithm can be mapped easily onto parallel computers.
This paper presents a genetic-based approach to mobile robot motion planning with a distance-safety criterion. A wave front method is used to build the numerical potential fields for both the goal points and the obsta...
详细信息
This paper presents a genetic-based approach to mobile robot motion planning with a distance-safety criterion. A wave front method is used to build the numerical potential fields for both the goal points and the obstacles by representing the workspace as a grid. A computationally efficient genetic algorithm is proposed to search for near optimal paths, where a combined global and local optimization approach is employed to speed up the search process while considering the imposed requirements. Various simulation results show the effectiveness of the presented algorithm, including a comparison with the A* method.
暂无评论