Search

The article presents a proposed customized genetic algorithm (CGA) to find the Pareto frontier for a bi-objective integer linear programming (ILP) model of routing in a dynamic network, where the number of nodes and edge weights vary over time. Utilizing a hybrid method, the CGA combines a genetic algorithm with dynamic programming (DP); it is a fast alternative to an ILP solver for finding efficient solutions, particularly for large dimensions. A non-dominated sorting genetic algorithm (NSGA-II) is used as a base multi-objective evolutionary algorithm. Real data are used for target trajectories, from a case study of application of a surveillance boat to measure greenhouse-gas emissions of ships on the Baltic sea. The CGA’s performance is evaluated in comparison to ILP solutions in terms of accuracy and computation efficiency. Results over multiple runs indicate convergence to the efficient frontier, with a considerable computation speed-up relative to the ILP solver. The study stays as a model for hybridizing evolutionary optimization and DP methods together in solving complex real-world problems.

Modern practical optimization problems are too often complex, nonlinear, large-dimensional, and sometimes dynamic making gradient-based and convex optimization methods too inefficient. Moreover, most such problems which must be solved for a reasonably approximate solution routinely in every few hours or every day must use a computationally fast algorithm. In this paper, we present a formulation of a dynamic vessel-to-vessel service ship scheduling problem. In a span of several hours, the service ship must visit as many moving vessels as possible and complete the trip in as small a travel time as possible. Thus, the problem is bi-objective in nature and involves a time-dependent traveling salesman problem. We develop a level-wise customized evolutionary algorithm to find multiple trade-off solutions in a generative manner. Compared to a mixed-integer programming (MIP) algorithm, we demonstrate that our customized evolutionary algorithm achieves similar quality schedules in a fraction of the time required by the MIP solver. We are currently developing an interactive decision support tool based on our proposed method for finding multiple trade-off schedules simultaneously and selecting a single preferred one.

Many algorithms have been proposed so far for pruning and sparse approximation of feedforward neural networks with random weights in order to obtain compact networks which are fast and robust on various datasets. One drawback of the randomization process is that the resulted weight vectors might be highly correlated. It has been shown that ensemble classifiers’ error depends on the amount of error correlation between them. Thus, decrease in correlation between output vectors must lead to generation of more efficient hidden nodes. In this research a new learning algorithm called New Sparse Learning Machine (NSLM) for single-hidden layer feedforward networks is proposed for regression and classification. In the first phase, the algorithm creates hidden layer with small correlation among nodes by orthogonalizing the columns of the output matrix. Then in the second phase, using $$L_1$$ -norm minimization problem, NSLM makes the components of the solution vector become zero as many as possible. The resulted network has higher degree of sparsity while the accuracy is maintained or improved. Therefore, the method leads to a new network with a better generalization performance. Numerical comparisons on several classification and regression datasets confirm the expected improvement in comparison to the basic network.

This study investigated the effects of ultrasound (US) and gas bubbling (GB) on permeate flux and membrane fouling. The ultrafiltration process of 1% skimmed milk was performed under various ultrasonic frequencies (37, 80 kHz, and tandem), irradiation modes (pulsed and sweeping), two-phase flow patterns (bubble and slug), and gas flow rates. The results showed that the US treatment improves the permeate flux up to 180%. The highest cleaning effect was obtained in the pulsed mode at 37 kHz. In the tandem frequency, the results showed that the fouling percentage significantly decreased with reduction in the frequency switching time. The GB injection treatment improved the mean of permeate flux up to 72% after 30 min. The slug pattern at medium gas flow rate was more effective than the other treatments. Best result was achieved in combination of the US (pulsed mode—37 kHz) and the GB (slug pattern—medium flow rate). In this condition, the permeate flux was significantly improved up to 384% in compari...

Fouling phenomenon is well known as a main obstacle in membrane separation technology. In this study, the effects of bubbling with various hydrodynamic factors and gas–liquid solubility were evaluated on fouling control and permeation flux in ultrafiltration of skimmed milk solution. Direct gas injection and carbonated feed, as a new bubbling method, were used for bubbling. In the direct gas injection technique, various two-phase flow patterns (slug and bubble), gas flow rates, and bubbling modes (continuous and intermittent) were investigated during a cross-flow ultrafiltration. The results showed that the both of gas bubbling methods improved the permeation flux during 30-min filtration. The permeation flux was enhanced up to 72 and 40% by direct injecting of N2 and CO2, respectively, while it was only enhanced up to 58% with carbonated feed. The evaluation of hydrodynamic resistance of membranes indicated that the gas bubbling by carbonated feed was more effective on the fouling resistance, whi...

Membrane processing has found many applications in various industries as an advanced method of separating materials in liquid or gas. The greatest obstacle in using membrane filtration technology is fouling, particularly in porous membranes. The fouling phenomenon is the result of organic and inorganic materials depositing on the surface and pores of membrane. In this study, the effect of ultrasound frequency (37, 80 kHz, and tandem) and amplitude of sonication (30, 60, and 90%) were evaluated on flux recovery during and after microfiltration. Results showed that the 37 kHz and tandem frequencies significantly improved the permeate flux at earlier minutes, particularly when tandem frequency was applied. In addition, the permeate flux was increased as the sonication power increased. However, the interaction effects between frequency and sonication power showed that the ultrasound frequencies were more effective than sonication power on flux recovery. Furthermore, the calculation of fouling percenta...

Several objective functions have been proposed in the literature to adjust the input parameters of a node in constructive networks. Furthermore, many researchers have focused on the universal approximation capability of the network based on the existing objective functions. In this brief, we use a correntropy measure based on the sigmoid kernel in the objective function to adjust the input parameters of a newly added node in a cascade network. The proposed network is shown to be capable of approximating any continuous nonlinear mapping with probability one in a compact input sample space. Thus, the convergence is guaranteed. The performance of our method was compared with that of eight different objective functions, as well as with an existing one hidden layer feedforward network on several real regression data sets with and without impulsive noise. The experimental results indicate the benefits of using a correntropy measure in reducing the root mean square error and increasing the robustness to noise.

Connected and automated vehicle applications are facilitated by Enhanced Digital Maps (EDMs) of the roadway environment. Due to the high numbers of roadway miles and signalized intersections, there is significant research interest in the automatic extraction of such maps from georectified LiDAR data. Most existing methods convert the LiDAR point cloud to a set of images for feature extraction and mapping. This rasterization step loses information that could be retained if new methods were developed that work directly on the LiDAR point could for feature extraction and mapping, without rasterization. This article presents one such approach that operates on a road surface point cloud, processing small patches at a time using a locally adaptive version of Otsu's method to discard low intensity reflections while retaining reflections from roadway markings. The main new aspect of the approach is a graph-based clustering algorithm implemented directly on the point cloud. A cluster growing method is used to group similar road markings into the same group to enable detection of the stop bars and lane edges. Finally, a SAE-J2735 map message is created from the extracted roadway features.

International audience; Exponential growth in demand for high bandwidth applications such as multimedia in cellular networks has triggered the need for development of new technologies capable of providing the required high bandwidth, reliable links in wireless environments. However, the available spectrum is a limited natural resource and the current technologies have been shown to have low spectrumefficiency. To improve the spectrum efficiency, heterogeneous networks are considered as a viable solution. They are composed of a mix of macro, pico, femto and relay base-stations that improve the spectrum efficiency and throughput performance of the system. The aim of this chapter is to provide an overview of these emerging technologies. This chapter describes the main characteristics and performance requirements that the next generation networks must fulfill. Particularly, the focus is on Long Term Evolution (LTE)/LTE-Advanced technologies where some possible improvements and challenges are explained. Subsequently, the analytical methods and simulations techniques to evaluate the performance of the next generation heterogeneous networks are discussed. Finally, the simulation results for some example scenarios are provided and discussed.

In this study, the effects of ultrasound (US) with carbonated feed were studied on ultrafiltration flux enhancement and membrane hydrodynamic resistances. A cross flow ultrafiltration of cherry concentrate was performed under US at 37 and 80 kHz with carbonated feed and noncarbonated feed, within the flat-sheet membrane module. Although, the US degassed the carbonated feed within the module, but no membrane cleaning effect was observed for the US with carbonated feed. Results showed that 37 kHz of US increases the permeation flux higher than 80 kHz, The lowest fouling resistance was obtained for gas bubbling by carbonated feed under 80 kHz, although there was no significant difference between 37 kHz and gas bubbling under 80 kHz. Practical Applications The most important limitation of ultrafiltration is flux decline during processing. Flux decline caused to higher operation cost and decrease the performance of separation. Using the ultrasound (US) treatment at low frequency (37 kHz) and carbonated feed with US at high frequency (80 kHz) can improve the flux permeation significantly. This technique can be applied to ultrafiltration to postpone the fouling and flux decline.

Fouling is one of the main obstacles to filtration process. In this study, the effects of gas bubbling by carbonated feed, gas bubbling by direct injection of CO2 (GB-DI-CO2) and N2 (GB-DI-N2) and ultrasound (US) were evaluated on flux enhancement during milk solution ultrafiltration and flux recovery during cleaning process of fouled hollow fibers membrane module. Results showed that the US and gas bubbling (GB) treatments significantly enhanced the performance of both ultrafiltration and cleaning processes, however, the effect of GB-DI-CO2 was not considerable. The best results were achieved when the GB-DI-N2 treatment was applied with US treatment in filtration and cleaning process, under this treatment, the permeation flux was remarkably enhanced up to 288.57% during ultrafiltration as well as the fouling percentage was reduced to 15.55% after cleaning the fouled membrane. Furthermore, the US was more effective than the GB during cleaning process. Practical Applications Filtration process is faced to flux decline as a result of fouling. The ultrasound at low frequency and liquid–gas two phase flow can affect the concentration polarization and cake layer significantly. The combination of these techniques can be used as an effective methods to handle the fouling and flux decline. Also artificial neural networks model is able to predict the flux permeation of hollow-fibers filtration under ultrasound and gas bubbling combination as a complex condition practically.