Multi-Variate Optimization of Polymer Electrolyte Membrane Fuel Cells in Consideration of Effects of GDL Compression and Intrusion

This study applies a comprehensive surrogate-based optimization techniques to optimize the performance of polymer electrolyte membrane fuel cells (PEMFCs). Parametric cases considering four variables are defined using latin hypercube sampling. Training and test data are generated using a multidimensional, two-phase PEMFC simulation model. Response surface approximation, radial basis neural network, and kriging surrogates are employed to construct objective functions for the PEMFC performance. There accuracies are tested and compared using root mean square error and adjusted R-square. Surrogates linked with optimization algorithms, i.e., genetic algorithm and particle swarm optimization are used to determine the optimal design points. Comparative study of these surrogates reveals that the kriging model outperforms the other models in terms of prediction capability. Furthermore, the PEMFC model simulations at the optimal design points demonstrate that performance improvements of around 56–69 mV at 2.0 A/cm2 are achieved with the optimal design compared to typical PEMFC design conditions.

Kriging Model for Reliability Analysis of the Offshore Steel Trestle Subjected to Wave and Current Loads

Offshore steel trestles (OSTs) are exposed to severe marine environments with stochastic wave and current loads, making structural safety assessment challenging and difficult. Reliability analysis is a suitable way to consider both wave and current loading intensity uncertainties, but the implicit and complex limit state functions of the reliability analysis usually imply huge computational costs. This paper proposes an efficient reliability analysis framework for OST using the kriging model of optimal linear unbiased estimation. The surrogate model is built with stochastic waves, current parameters, and the corresponding load factors. The framework is then used to evaluate the reliability of an example OST subjected to wave and current loads at three limit states of OST, including first yield (FY), full plastic (FP), and collapse initiation (CI). Three different distributions are used for comparison of the results of failure probability and reliability index. The results and the computational cost by the proposed framework are compared with that from the Monte Carlo sampling (MCS) and Latin hypercube sampling (LHS) method. The influences of sample number on the prediction accuracy and reliability index are investigated. The influence of marine growth on the reliability analysis of the OST is discussed using MCS and the kriging model. The results show that the reliability analysis based on the kriging model can obtain the reliability index for the OST efficiently with less calculation time but similar results compared with MCS and LHS. With the increase of the number of samples, the prediction accuracy of the kriging model increases, and the corresponding failure probability fluctuates greatly at first and then tends to be stable. The reliability of the example OST is reduced with the increase of marine growth, regardless of the limit state.

Stochastic Bayesian optimization for predicting borderline knock

Engine knock is an undesirable combustion that could damage the engine mechanically. On the other hand, it is often desired to operate the engine close to its borderline knock limit to optimize combustion efficiency. Traditionally, borderline knock limit is detected by sweeping tests of related control parameters for the worst knock, which is expensive and time consuming, and also, the detected borderline knock limit is often used as a feedforward control without considering its stochastic characteristics without compensating current engine operational condition and type of fuel used. In this paper, stochastic Bayesian optimization method is used to obtain a tradeoff between stochastic knock intensity and fuel economy. The log-nominal distribution of knock intensity signal is converted to Gaussian one using a proposed map to satisfy the assumption for Kriging model development. Both deterministic and stochastic Kriging surrogate models are developed based on test data using the Bayesian iterative optimization process. This study focuses on optimizing two competing objectives, knock intensity and indicated specific fuel consumption using two control parameters: spark and intake valve timings. Test results at two different operation conditions show that the proposed learning algorithm not only reduces required time and cost for predicting knock borderline but also provides control parameters, based on trained surrogate models and the corresponding Pareto front, with the best fuel economy possible.

Multi-objective Optimisation of Torsion Beam Structure of Driverless Vehicle Chassis for Improved Fatigue Resistance

The fatigue resistance of the torsion beam is the keyway to prolong the service life of the chassis of the driverless vehicle. The rigid-flexible coupling finite element model of the chassis is constructed using anti-fatigue algorithm. In this model, the stress time history of the torsion beam is obtained by modal stress recovery. The nominal stress method is used to analyze the fatigue life of the structure. It is known that the structure weight affects the fatigue life, so the algorithm aims at lightening the structure to realize the improved fatigue resistance of the torsional beam structure. The parametric model of torsion beam is constructed with mass and fatigue life as optimization objectives, first-order torsion mode frequency and torsion stiffness as constraints. Multi-objective particle swarm optimization (MPSO) based on the Kriging model is used to achieve improved fatigue life of the torsion beam. After optimization, the structural weight of the torsion beam is reduced by 19.20%, and the light-weight and anti-fatigue effect are better than the baseline design.

Aerodynamic Shape Optimization Method of Non-Smooth Surfaces for Aerodynamic Drag Reduction on a Minivan

To reduce aerodynamic drag of a minivan, non-smooth surfaces are applied to the minivan’s roof panel design. A steady computational fluid dynamics (CFD) method is used to investigate the aerodynamic drag characteristics. The accuracy of the numerical method is validated by wind tunnel test. The drag reduction effects of rectangle, rhombus and arithmetic progression arrangement for circular concaves are investigated numerically, and then the aerodynamic drag coefficient of the rectangle arrangement with a better drag reduction effect is chosen as the optimization objective. Three parameters, that is, the diameter D of the circular concave, the width W and the longitudinal distance L among the circular concaves, are selected as design variables. A 20-level design of an experimental study using a Latin Hypercube scheme is conducted. The responses of 20 groups of sample points are obtained by CFD simulation, based on which a Kriging model is chosen to create the surrogate-model. The multi-island genetic algorithm is employed to find the optimum solution. The result shows that maximum drag reduction effects up to 7.71% can be achieved with a rectangle circular concaves arrangement. The reduction mechanism of the roof with the circular concaves was discussed. The circular concaves decrease friction resistance of the roof and change the flow characteristics of the recirculation area in the wake of the minivan. The roof with the circular concaves reduces the differential pressure drag of the front and rear of the minivan.

Advances in Regression Kriging-Based Methods for Estimating Statewide Winter Weather Collisions: An Empirical Investigation

Winter conditions create hazardous roads that municipalities work hard to maintain to ensure the safety of the travelling public. Targeting their efforts with effective network screening will help transportation managers address these problems. In our recent efforts, regression kriging was found to be a viable and effective network screening methodology. However, the study was constrained by its limited spatial extent making the reported results less conclusive and transferrable. In addition, our previous work implemented what has long been adopted in most of conventional studies—the Euclidean distance; however, use of the road network distance would, intuitively, result in further improving kriging estimates, especially when dealing with transportation problems. Therefore, this study improves upon our previous efforts by developing a more advanced kriging model; namely, network regression kriging using the entire state of Iowa with the significantly expanded road network. The transferability of the developed models is also explored to investigate its generalization potential. The findings based on various statistical measures suggest that the enhanced kriging model vastly improved the estimation performance at the cost of greater computational complexity and run times. The study also suggests that regional semivariograms better represent the true nature of the local variances, though an overall model may still function adequately if higher fidelity is not required.