For hard optimization problems, it is difficult to design heuristic algorithms which exhibit uniformly superior performance for all problem instances. As a result it becomes necessary to tailor the algorithms based on the problem instance. In this paper, we introduce the use of a cooperative problem solving team of heuristics that evolves algorithms for a given problem instance. The efficacy of this method is examined by solving six difficult instances of a bicriteria sparse multiple knapsack problem. Results indicate that such tailored algorithms uniformly improve solutions as compared to using predesigned heuristic algorithms. 相似文献
Very low pressure photolysis (VLPØ) of chlorine nitrate was performed in a quartz Knudsen cell. The light source was a 2500 W high-pressure xenon lamp, and a modulated molecular-beam mass spectrometer was used to monitor the concentration of ClONO2 and photolysis products. Because of the low pressures used (? 10?3 torr) and the short residence time in the cell (≈1 s), secondary reactions were unimportant and the primary products could be directly identified. The primary photolysis products (λ ≈ 2700 Å) are atomic chlorine and NO3 free radical. Chlorine atoms were identified both by the appearance of Cl2 (wall recombination product; the walls were not poisoned) and by HCl produced when C2H6 was added to the cell. Nitrate free radical was directly identified as a mass peak at m/e = 62, as well as by chemical titration with nitric oxide: NO3 + NO → 2NO2. It was verified by direct tests that the peak at m/e = 62 did not arise from possible HNO3 contamination or from N2O5, a possible secondary product. This titration reaction was used to measure quantitatively a lower limit to the primary quantum yield, φ ? 0.5 ± 0.3. This represents a lower limit because of the unknown extent of the secondary photolysis of NO3 under our conditions. We believe this to be the first observation using mass spectrometry of the NO3 free radical. The quantum yield for atomic chlorine is φ = 1.0 ± 0.2. N2O was used to test for O(1D) according to the reaction, O(1D) + N2O → products; none was observed. Triplet oxygen, O(3P) was observed to the extent of ≈ 10% by the reaction O(3P) + NO2 → NO + O2, but this yield can also be due to the photolysis of NO3 free radical produced in the primary step. We conclude that the predominant reaction pathway is . 相似文献
The angular dependence of photoemission from oxygen chemisorbed on the (100) face of aluminum is calculated using a molecular cluster model. The cluster contains five aluminum atoms with one oxygen atom located in the four-fold site; two A1O distances are considered. The calculations employed the Xα scattered wave formalism and are the first results to be obtained for a chemisorption problem in which both initial and final states are based on a cluster model. 相似文献