We present Vertiport-informed Surrogate-Based Optimization with Machine Learning Surrogates (VinS), a novel framework for solving the vertiport location problem for urban air mobility operations. The primary focus of this work is on the optimization of vertiport locations to facilitate efficient urban air transportation. Our framework helps choose not only the optimal vertiport locations but also the optimal number of vertiports. We develop a simulation model to analyze the impacts of various vertiport location configurations on the efficiency of the transportation network. To expedite the simulation process, a surrogate model is trained using machine learning algorithms, effectively reducing the computational time for evaluating a given vertiport location configuration. With the machine learning surrogate models, we apply a genetic algorithm to find the optimal set of vertiport locations. An empirical study was performed in the San Francisco Bay Area, from which we found that given the optimal set of vertiport locations, we further reduced the total travel time in the entire transportation system in the Bay Area compared with the sampled sets by 0.05% (400 h) on average. We release our framework at: https://github.com/Xuan-1998/LPSim .

Traffic simulation, a tool for recreating real-life traffic scenarios, acts as an important platform in transportation research. Considering the growing complexity of urban mobility, various large-scale regional simulators are designed and used for research and applications. Calibration is a key issue in the traffic simulation: it finds the optimal system pattern to decrease the gap between the simulator output and the real data, making the system much more reliable. This paper proposes DRBO, a calibration framework for large-scale traffic simulators. This framework combines the travel behavior adjustment with black box optimization, better exploring the structure of the regional scale mobility. The motivation of the framework is based on the decomposition of the regional scale mobility dynamic. We decompose the mobility dynamic into the car-following dynamic and the routing dynamic. The prior dynamic imitates how vehicles propagate as time flows while the latter one reveals how vehicles choose their route according to their own information. Based on the decomposition, the DRBO framework uses iterative algorithms to find the best dynamic combinations. It utilizes the Bayesian optimization and day-to-day routing update to separately calibrate the dynamic, then combine them sequentially in an iterative way. Compared to the prior arts, the DRBO framework is efficient for capturing multiple perspectives of traffic conditions. We further tested our simulator on SFCTA demand to further validate the speed distribution from our simulation and observed data.

Studies have been using microscopic traffic simulation to evaluate the efficiency of bicycle infrastructure in multi-modal urban road networks. However, compared to car and pedestrian flow, the models used for the simulation of bicycle movement in state-of-the-art traffic simulation tools still fall short of capturing realistic cycling behavior and require substantial improvement. This work implements a non-lane-based bicycle simulation model in SUMO. Lateral movements, including overtaking and aligning, are determined by a decision-making mechanism proposed based on insights from naturalistic observations. Cyclists’ lateral movement distance and longitudinal acceleration are computed by the model and integrated into the sublane environment. The model is able to (i) achieve staggered queue formation through a modified gap function based on a hexagonal representation of bicycles, (ii) introduce an efficient queueing behavior to ensure the great lateral space utilization of bicycle flow, and (iii) avoid collisions by predicting the future positions of the ego-cyclist and its surrounding cyclists. In addition, the simulation becomes more realistic by considering cyclists’ decision-making frequency and reaction time lag. We verify the model by analyzing queue formations and intersection capacities in test scenarios of signalized bike lanes with different widths. Future work will calibrate the model parameters using bicycle trajectory data collected from real-world road environment. It is expected that bicycle traffic flow dynamics in conditions with mass cycling demand can be better reproduced using the proposed model.

As an important anatomical structure, the mediastinum can develop a variety of lesions. Endoscopic ultrasonography-guided fine needle aspiration is a novel auxiliary diagnostic method for these diseases at present. Since the images presented by endoscopic ultrasonography are not direct views, even experienced doctors may have certain difficulties in identifying the exact location of lesions in clinical diagnosis. Therefore, deep learning can be used to classify endoscopic images to assist doctors in the positioning of endoscopic ultrasound (EUS). In this paper, a hybrid model combining the pre-trained deep learning network-ResNet and the metaheuristic algorithm called Slime Mould Algorithm (SMA) is proposed to identify the structure of the aortic arch by distinguishing the aortic arch from non-aortic arch to assist doctors in positioning rapidly. The hybrid model utilizes ResNet50 for a raw feature extraction, adopts SMA for a further feature selection based on the extracted features, and finally uses the Support Vector Machine classifier to classify the filtered features. The hybrid model takes advantage of the excellent feature extraction ability of the deep network and the good search characteristics in the solution space of the SMA. This model achieves an average accuracy of 92.21% for classification on practical data of 14 patients’ EUS images. Experimental results show that the proposed model outperforms other deep networks and metaheuristic algorithms. It shows that this model has achieved good results in the aortic arch classification problem.

Genome-wide prediction methods establish a genetic statistical model using molecular marker genotypes distributed throughout the whole genome and biological trait data to predict phenotype values, in order to providing a guide for individual selection breeding. Most of the existing genome-wide selection methods focus on the prediction of a single trait, which can only examine each trait one by one when predicting multiple traits that would translate into neglect of the correlation among traits. In this paper, we proposed a joint prediction method for multiple biological traits based on feedback stacking deep networks, which stack orderly the output of the model to the non-shared layer so as to introduce the correlation among traits into the learning network. The method selected the single nucleotide polymorphism (SNP) loci obtained from whole genome sequencing as molecular markers, extracted the genotype of each SNP allele to constitute sample features, and adopted a two-stage feature selection method to pick up the SNP loci highly correlated with target, and then constructed a deep network model with feedback stacking for joint prediction of multiple traits. The proposed approach was compared with other models by using oyster and corn datasets, and the results showed that the method on this work could achieve better prediction effect than the others.