Near-optimum Parameters Research Articles

Optimization and determination of the parameters for a PID based ventilation system for smoke control in tunnel fires: Comparative study between a genetic algorithm and an analytical trial-and-error method

A PID-based longitudinal ventilation system for smoke control in tunnel fires is numerically illustrated to yield excellent performance. However, the three parameters in the PID algorithm significantly affect the performance. Two methods are proposed to optimize the parameters: (1) an analytical trial-and-error method (ATEM), based on an order-of-magnitude analysis; (2) a heuristic method: system identification and genetic algorithm (SI-GA). An SI-GA without a differential controller was also used for the comparative study, labeled as SI-GA*. The results show that all methods can obtain near-optimum parameters under their optimization goals. The ATEM is more time consuming, but it is more straightforward to manipulate the performance metrics directly than with the SI-GA method. The optimized parameters are tested with different fire sizes and different tunnel dimensions. The results show that the system with the optimized parameters can control the smoke well for large enough fire. In general, the system performs better for the stabilization time of the smoke front when using the parameters obtained from ATEM, but it achieves shorter maximum back-layering distances with the parameters obtained from SI-GA and SI-GA*. The latter two also give stronger fluctuations, but lower average values, of the ventilation velocity. Nevertheless, SI-GA* has more application potential due to the absence of a differential controller.

A new genetically optimized tensor product functional link neural network: an application to the daily exchange rate forecasting

The training speed for multilayer neural networks is slow due to the multilayering. Therefore, removing the hidden layers, provided that the input layer is endowed with additional higher order units is suggested to avoid such problem. Tensor product functional link neural network (TPFLNN) is a single layer with higher order terms that extend the network’s structure by introducing supplementary inputs to the network (i.e., joint activations). Although the structure of the TPFLNN is simple, it suffers from weight combinatorial explosion problem when its order becomes excessively high. Furthermore, similarly to many neural network methods, selection of proper weights is one of the most challenging issues in the TPFLNN. Finding suitable weights could help to reduce the number of needed weights. Therefore, in this study, the genetic algorithm (GA) was used to find near-optimum weights for the TPFLNN. The proposed method is abbreviated as GA–TPFLNN. The GA–TPFLNN was used to forecast the daily exchange rate for the Euro/US Dollar, and Japanese Yen/US Dollar. Simulation results showed that the GA–TPFLNN produced more accurate forecasts as compared to the standard TPFLNN, GA, GA–TPFLNN with backpropagation, GA-functional expansion FLNN, multilayer perceptron, support vector regression, random forests for regression, and naive methods. The GA helps the TPFLNN to find low complexity network structure and/or near-optimum parameters which leads to this better result.

A methodology correlating code optimizations with data memory accesses, execution time and energy consumption

The advent of data proliferation and electronic devices gets low execution time and energy consumption software in the spotlight. The key to optimizing software is the correct choice, order as well as parameters of optimization transformations that has remained an open problem in compilation research for decades for various reasons. First, most of the transformations are interdependent and thus addressing them separately is not effective. Second, it is very hard to couple the transformation parameters to the processor architecture (e.g., cache size) and algorithm characteristics (e.g., data reuse); therefore, compiler designers and researchers either do not take them into account at all or do it partly. Third, the exploration space, i.e., the set of all optimization configurations that have to be explored, is huge and thus searching is impractical. In this paper, the above problems are addressed for data-dominant affine loop kernels, delivering significant contributions. A novel methodology is presented reducing the exploration space of six code optimizations by many orders of magnitude. The objective can be execution time (ET), energy consumption (E) or the number of L1, L2 and main memory accesses. The exploration space is reduced in two phases: firstly, by applying a novel register blocking algorithm and a novel loop tiling algorithm and secondly, by computing the maximum and minimum ET/E values for each optimization set. The proposed methodology has been evaluated for both embedded and general-purpose CPUs and for seven well-known algorithms, achieving high memory access, speedup and energy consumption gain values (from 1.17 up to 40) over gcc compiler, hand-written optimized code and Polly. The exploration space from which the near-optimum parameters are selected is reduced from 17 up to 30 orders of magnitude.

Signal design for displaced left-turn intersection using Monte Carlo method

Unconventional intersection designs have been receiving increased consideration at intersections that serve high traffic demands. Although operational performance of unconventional intersections has been evaluated in a number of studies, little research has been conducted regarding methodologies to obtain optimal signal timings for them. Several studies have reported that commercial signal timing packages have not been able to be utilized to develop optimal signal timings due to the unique design features of the unconventional intersections. In this paper a Monte Carlo simulation method to minimize intersection delay is described as well as a bandwidth maximization method for a Displaced Left-Turn (DLT) intersection, one of the most popular unconventional intersections. The proposed method is applied to a representative intersection design for two different traffic demand scenarios. It is seen that the two optimization approaches may result in different optimal solutions and the Monte Carlo simulation is able to provide near-optimum parameter selection ranges for the given traffic demands. Also, it is found that the Monte Carlo method is sufficiently flexible to provide intersection-tailored solutions with intersection specific design parameters.

Robust multi-objective calibration strategies – possibilities for improving flood forecasting

Abstract. Process-oriented rainfall-runoff models are designed to approximate the complex hydrologic processes within a specific catchment and in particular to simulate the discharge at the catchment outlet. Most of these models exhibit a high degree of complexity and require the determination of various parameters by calibration. Recently, automatic calibration methods became popular in order to identify parameter vectors with high corresponding model performance. The model performance is often assessed by a purpose-oriented objective function. Practical experience suggests that in many situations one single objective function cannot adequately describe the model's ability to represent any aspect of the catchment's behaviour. This is regardless of whether the objective is aggregated of several criteria that measure different (possibly opposite) aspects of the system behaviour. One strategy to circumvent this problem is to define multiple objective functions and to apply a multi-objective optimisation algorithm to identify the set of Pareto optimal or non-dominated solutions. Nonetheless, there is a major disadvantage of automatic calibration procedures that understand the problem of model calibration just as the solution of an optimisation problem: due to the complex-shaped response surface, the estimated solution of the optimisation problem can result in different near-optimum parameter vectors that can lead to a very different performance on the validation data. Bárdossy and Singh (2008) studied this problem for single-objective calibration problems using the example of hydrological models and proposed a geometrical sampling approach called Robust Parameter Estimation (ROPE). This approach applies the concept of data depth in order to overcome the shortcomings of automatic calibration procedures and find a set of robust parameter vectors. Recent studies confirmed the effectivity of this method. However, all ROPE approaches published so far just identify robust model parameter vectors with respect to one single objective. The consideration of multiple objectives is just possible by aggregation. In this paper, we present an approach that combines the principles of multi-objective optimisation and depth-based sampling, entitled Multi-Objective Robust Parameter Estimation (MOROPE). It applies a multi-objective optimisation algorithm in order to identify non-dominated robust model parameter vectors. Subsequently, it samples parameter vectors with high data depth using a further developed sampling algorithm presented in Krauße and Cullmann (2012a). We study the effectivity of the proposed method using synthetical test functions and for the calibration of a distributed hydrologic model with focus on flood events in a small, pre-alpine, and fast responding catchment in Switzerland.

Selection of Embedding Dimension and Delay Time in Phase Space Reconstruction

A new algorithm is proposed for computing the embedding dimension and delay time in phase space reconstruction. It makes use of the zero of the nonbias multiple autocorrelation function of the chaotic time series to determine the time delay, which efficiently depresses the computing error caused by tracing arbitrarily the slop variation of average displacement (AD) in AD algorithm. Thereafter, by means of the iterative algorithm of multiple autocorrelation and Γ test, the near-optimum parameters of embedding dimension and delay time are estimated. This algorithm is provided with a sound theoretic basis, and its computing complexity is relatively lower and not strongly dependent on the data length. The simulated experimental results indicate that the relative error of the correlation dimension of standard chaotic time series is decreased from 4.4% when using conventional algorithm to 1.06% when using this algorithm. The accuracy of invariants in phase space reconstruction is greatly improved.

Selection of genetic algorithm operators for river water quality model calibration

Model calibration is the most important step in the overall model development. Genetic algorithm (GA) was used in this study to optimise model parameters of river water quality models. In general, the efficiency of using GA depends on the proper selection of GA operators, which are the components that make up the overall GA process. A comprehensive investigation on the importance of GA operators on model parameter optimisation was conducted in this study (and presented in this paper) based on the hypothesis that the selection of GA operators depends on the sensitivity of model parameters to model output. A numerical experiment on GA operators was conducted first considering a hypothetical river network water quality model with both insensitive and sensitive model parameters and were later validated using Yarra River Water Quality Model (YRWQM). It was found that a robust GA operator set obtained from the literature was capable of achieving a near-optimum model parameter set for the sensitive model but not for the insensitive model. However, due to insensitivity of water quality model parameters on model output, the ‘not so’ near-optimum parameter set did not contribute a great difference to the overall water quality predictions in the insensitive model. Therefore, based on these numerical experiments, it was concluded that a GA operator set obtained from the literature is adequate for calibration of model parameters of river water quality model using GA.

Low noise DC SQUIDs fabricated in Nb-Al/sub 2/O/sub 3/-Nb trilayer technology

The authors have designed, fabricated and tested all-refractory DC SQUIDs in Nb-Al/sub 2/O/sub 3/-Nb trilayer technology that have noise performance comparable to the best previously reported for any technology. A variety of SQUID designs were incorporated as part of a trilayer process development test vehicle. SQUID inductance, junction area, and resistive shunt geometry were varied in matrix fashion to give SQUIDs with near-optimum parameter values for a factor of five range in Josephson current density and shunt sheet resistance. The devices were fabricated using a selective niobium anodization with a minimum feature size of 2 mu m. The base electrode and Nb wiring were patterned with dry etching, and the junction areas were defined by anodization: the Ti resistors were patterned with a lift-off process. Current density on different wafers was varied from 400 to 1000 A/cm/sup 2/ with typical junction V/sub m/'s of 60 mV. The shunt sheet resistance was varied in the 1-5- Omega / Square Operator range. The noise was measured with an RF SQUID direct small-signal readout scheme. A 50-pH SQUID with 3- mu m/sup 2/, 16- mu A junctions and 14- Omega shunt resistors was shown to have an ideally low white noise of 1*10/sup -13/ Phi /sub 0//sup 2//Hz, a white to 1/f crossover frequency at 7 Hz, and a noise level less than 6*10/sup -12/ Phi /sub 0//sup 2//Hz at 0.1 Hz.

A large unit cell semiempirical molecular orbital approach to the properties of solids. II. Covalent materials: diamond and silicon

For pt.I see ibid., vol.12, p.2487 (1979). Some properties of diamond and silicon crystals have been calculated using a large unit cell-complete neglect of differential overlap (LUC-CNDO) crystal orbital approach. The dependence of the predictions on the parameter set chosen, on the size and shape of the large unit cell and on the cut-off distance for atomic interactions has been investigated. The results obtained for the perfect solid with a near-optimum parameter set are, in general, in good agreement with experimental values.

Rhenium and rhenium-tungsten deposition by thermochemical reduction of the hexafluorides—A preliminary study

Near-optimum parameters were determined by the Bureau of Mines for vapor deposition of rhenium on heated copper substrates by hydrogen reduction of rhenium hexafluoride (ReF 6). High-purity deposits of nearly 90% theoretical density were obtained with the reaction chamber at 250 °C, with a H 2 to ReF 6 volume ratio of 25 to 1, and with a ReF 6 flow rate of 5 grams per hour. Some rhenium deposited as a loose, black powder. When the established near-optimum parameters were carefully followed, this could be minimized but never completely avoided. Varying the reaction chamber pressure, both above and below atmospheric, yielded excessive quantities of rhenium powder. Deposition efficiencies, based on only the coherent metal, ranged between 70 and 75%. Massive deposits contained only spectrographic traces of impurities. The average microhardness (Knoop, 100-gram load) of vapor-deposited rhenium was 907. Limited research on the deposition of rhenoium-tungsten alloys by reduction of the mixed hexafluorides produced deposits in which linear composition gradients were pronounced; this nonhomogeneity was attributed to the wide variation in optimum hydrogen-reduction temperatures of the two hexafluorides (250 °C for ReF 6 and 550 °C for WF 6). Using deposition temperatures between 250 ° and 550 °C did not eliminate the composition variations.

Engineering approach to the design of tapered dielectric-rod and horn antennas

The taper profile of optimized dielectric-rod and horn antennas is synthesized as a series of non-interacting planar radiating apertures. The method is semi-empirical, straightforward to apply, enables the dielectric-rod antenna to be satisfactorily optimized and provides a means of evaluating and optimizing a dielectric-horn antenna with variable wall thickness. The optimum profiles are taken as those which smoothly transform the surface-wave power from the launcher to the radiating aperture. The optimization of the dielectric-rod antenna considerably improves the radiation pattern while computations supported by measurements confirm earlier reports that a dielectric-horn can have a higher gain than a metal horn of similar dimension but side-lobe level is seen to be an important issue. Wide flare-angle horns give ideal E-plane patterns at the expense of a high side-lobe level in the H-plane; for small flare angles the dielectric horn gives similar patterns to the tapered rod antenna and thus preserves rotational symmetry. Calculations throughout are restricted to cylindrical geometry but other geometries and variations on the dielectric-horn principles are described. Useful engineering design data have been compiled for both the dielectric-horn and rod antennas and curves are given which determine near-optimum parameters for gains up to about 20 dB which is seen to be a practical operating limit for these surface wave devices. A unified impression of dielectric antennas emerges with the important conclusion that, when optimized, dielectric-rod and horn antennas are in fact competitive with small metal horns for some applications; furthermore the dielectric-horn antenna, used singly or in arrays, is an ideal device for producing a low side-lobe level in the E-plane.

A preliminary study of vapor deposition of rhenium and rhenium-tungsten

Coherent, high-purity rhenium deposits were obtained by the hydrogen reduction of ReF 6 from the vapor phase, although some rhenium powder was always produced. The reaction temperature and the ratio of H 2 to ReF 6 significantly influenced the amount of coherent rhenium deposited. The near-optimum parameters are temperature, 250 °C; H 2 to ReF 6 gas volume ratio, 25. The deposition efficiency for coherent rhenium deposits was 70–75%. The average Knoop microhardness of vapor-deposited rhenium was 907, measured under a 100 g load. Intermittent deposition is undesirable for forming coherent deposits because it produces discontinuities in the metal. Codeposits of rhenium and tungsten were obtained, but they were of non-uniform composition, owing to the large difference in optimum reduction temperatures of their hexafluorides.