This paper introduces a human-computer interaction methodology for a marine propeller design optimization problem. While the outcome of most optimization algorithms is a frontier with the best solutions, in blade desi...
详细信息
ISBN:
(数字)9781624105982
ISBN:
(纸本)9781624105982
This paper introduces a human-computer interaction methodology for a marine propeller design optimization problem. While the outcome of most optimization algorithms is a frontier with the best solutions, in blade design the engineers by taking into consideration all objectives and constraints want to focus on a part of this frontier and try to improve it and enhance it, in order to have more design alternatives. Thus there is a need for guiding the optimization to a specific direction, enabling the involvement of the designer in the design process. A proposal to achieving this goal is the use of interactive evolutionary computation, which is an optimization methodology based on genetic algorithms whereby the blade designer is called during the intermediate steps to visualize and assess specific areas of interest of the Pareto plot. The results have shown that the blade designer can steer the optimization to a specific direction and a more refined Pareto frontier is obtained. At the current stage there is a modest improvement of the design fitness, but a clear reduction in cost.
An algorithm to calculate optimal trajectories of e-buses to fulfil on-demand requests for transportation in a smart city environment is presented. Instead of solving the transportation problem once when all transport...
详细信息
An algorithm to calculate optimal trajectories of e-buses to fulfil on-demand requests for transportation in a smart city environment is presented. Instead of solving the transportation problem once when all transportation requests are registered, we solve the transportation problem at any virtual or public station during the taxi. This approach enables gathering transportation requests dynamically and also consider changes in the environment (e.g. weather conditions, roadblocks, traffic jams, etc) during the transportation. The algorithm is tested with real data. Some results of the simulation are presented.
In recent years, China's logistics industry has developed rapidly, and the logistics network has become increasingly large. How to optimize the vehicle route is one of the key issues to reduce the cost of logistic...
详细信息
In recent years, China's logistics industry has developed rapidly, and the logistics network has become increasingly large. How to optimize the vehicle route is one of the key issues to reduce the cost of logistics and distribution. With the continuous progress of new energy technology, scholars at home and abroad have given a lot of attention to the EVRP problem model and its solution method. With continuous in-depth research, many branch problems with more constraints have been developed, and their solution methods have become more and more diverse. In order to further sort out the research status of EVRP problem at home and abroad, the basic model, variant problem and its solving algorithm type of EVRP problem are firstly introduced.
The vibration method is usually used for field measurement of cable tension of cable system bridges. The cable tension evaluation method is mostly based on the simple taut string theory. However, the simple theory may...
详细信息
ISBN:
(纸本)9783037857755
The vibration method is usually used for field measurement of cable tension of cable system bridges. The cable tension evaluation method is mostly based on the simple taut string theory. However, the simple theory may cause unacceptable errors in many applications especially for the cables with big bending stiffness and two ends fixed boundary conditions. In this paper a cable tension estimation method based on iterative algorithm and optimization algorithm is presented and implemented using finite element method and ANSYS soft ware. Compared with the analytical method and empirical formulas the method presented in this paper is more convenient and the application range is more extensive. The accuracy of the method has been verified by a set of test. In the end, the method is used to estimate the cable tension of a tied-arch bridge's suspenders.
Structure of convolutional neural network (CNN) applied for image recognition requires large numbers of tuning for designated datasets in practice. It is a time-consuming process to finally come up with a feasible str...
详细信息
Structure of convolutional neural network (CNN) applied for image recognition requires large numbers of tuning for designated datasets in practice. It is a time-consuming process to finally come up with a feasible structure for specific requirement. This paper proposes a method based on Taguchi method which can efficiently determine the optimal structure of hyperparameters combination. Five hyperparameters with four levels are defined as control factors and two indicators are chosen to measure the performance of CNN structure. L16(45) orthogonal array is used to arrange the experiment. S/N ratio and main effect plot are used to identify the optimal structure (hyperparameter combination) of CNN. The classic case of MNIST is employed to verify the practicability of the proposed method. Results show that the proposed method can identify the optimal CNN structure efficiently and also rank the significance priority of hyperparameters.
In the present work, a duct geometry is optimized to reduce the flow distortion using adjoint method. The S-duct geometry is from AIAA Propulsion and Aerodynamics Workshop (PAW [1]). ANSYS Fluent Adjoint solver is use...
详细信息
ISBN:
(数字)9781624105982
ISBN:
(纸本)9781624105982
In the present work, a duct geometry is optimized to reduce the flow distortion using adjoint method. The S-duct geometry is from AIAA Propulsion and Aerodynamics Workshop (PAW [1]). ANSYS Fluent Adjoint solver is used to run the design optimization study. The workflow involves a 4-step process: CFD run, adjoint calculation, design calculation based on the sensitivity data and environment constraints, and mesh update. This process is repeated until the optimum geometry is obtained. The variance of total pressure at the Aerodynamic Interface Plane (AIP) is chosen as the quantity of interest. It correlates well with the circumferential and radial distortion index (CDI & RDI) as defined in [2], which indicates flow distortion at the inlet/exit plane. After 3 design iterations, the variance of total pressure and the average CDI at AIP was reduced by 20.91% and 8.36% respectively, while the maximum CDI and RDI showed a drop of 2.15% and 12.37% respectively.
This paper presents a novel robust fault reconfiguration scheme based on optimized Takagi-Sugeno (T-S) fuzzy model and nonsingular terminal sliding mode observer (NTSMO) with adaptive law for micro-gas turbine (MGS). ...
详细信息
This paper presents a novel robust fault reconfiguration scheme based on optimized Takagi-Sugeno (T-S) fuzzy model and nonsingular terminal sliding mode observer (NTSMO) with adaptive law for micro-gas turbine (MGS). An optimized T-S fuzzy model is introduced because it can approximate any nonlinear model with arbitrary precision, and an improved imperial competition algorithm (ICA) using adaptive reform probability is proposed to improve the accuracy of the model. A linear transformation method is introduced to decouple the fault and disturbance of the system. The nonsingular terminal sliding mode observer is designed to reconstruct actuator fault and disturbance with unknown upper bound of a change rate, in which an adaptive law is introduced to update the sliding mode gain in real-time to eliminate the influence of fault, disturbance and modeling uncertainty. Simulations in Matlab/Simulink show high reconfiguration accuracy and high-speed of the proposed method despite of the presence of fault, disturbance and modeling uncertainty.
In the hot rolling process, the shifting strategy of downstream work rolls has been explored for a long time, as much as possible to guarantee uniformity of roll wear along the width of the strip, thereby improving th...
详细信息
In the hot rolling process, the shifting strategy of downstream work rolls has been explored for a long time, as much as possible to guarantee uniformity of roll wear along the width of the strip, thereby improving the steel strip profile. So a smart-shifting strategy based on differentiated step size in the cycle with the sine function is proposed to make the roll wear distribute more evenly, in which a new hybrid optimization algorithm is adopted to find the optimal control parameters. A large number of industrial experiments for the production of non-oriented silicon steel demonstrated that the new strategy was able to improve the wear contour of work roll, and increased the uniformity of the thickness distribution of the strip along the width direction. This has important implications for the improvement of the dimensional accuracy of the strip.
optimization methods are of great importance for the efficient training of neural networks. There are many articles in the literature that propose particular variants of existing optimizers. In our article, we propose...
详细信息
optimization methods are of great importance for the efficient training of neural networks. There are many articles in the literature that propose particular variants of existing optimizers. In our article, we propose the use of the combination of two very different optimizers that, when used simultaneously, can exceed the performance of the single optimizers in very different problems. We propose a new optimizer called ATMO (AdapTive Meta Optimizers), which integrates two different optimizers simultaneously weighing the contributions of both. Rather than trying to improve each single one, we leverage both at the same time, as a meta-optimizer, by taking the best of both. We have conducted several experiments on the classification of images and text documents, using various types of deep neural models, and we have demonstrated through experiments that the proposed ATMO produces better performance than the single optimizers.
暂无评论