Search

Your search keyword '"Ma, Yunyi"' showing total 8 results
Start Over Author "Ma, Yunyi"
8 results on '"Ma, Yunyi"'

1. TraffNet: Learning Causality of Traffic Generation for What-if Prediction

Author

Xu, Ming, Ai, Qiang, Li, Ruimin, Ma, Yunyi, Qi, Geqi, Meng, Xiangfu, and Jin, Haibo

Subjects
Computer Science - Machine Learning, Computer Science - Artificial Intelligence
Abstract

Real-time what-if traffic prediction is crucial for decision making in intelligent traffic management and control. Although current deep learning methods demonstrate significant advantages in traffic prediction, they are powerless in what-if traffic prediction due to their nature of correla-tion-based. Here, we present a simple deep learning framework called TraffNet that learns the mechanisms of traffic generation for what-if pre-diction from vehicle trajectory data. First, we use a heterogeneous graph to represent the road network, allowing the model to incorporate causal features of traffic flows, such as Origin-Destination (OD) demands and routes. Next, we propose a method for learning segment representations, which models the process of assigning OD demands onto the road network. The learned segment represen-tations effectively encapsulate the intricate causes of traffic generation, facilitating downstream what-if traffic prediction. Finally, we conduct experiments on synthetic datasets to evaluate the effectiveness of TraffNet. The code and datasets of TraffNet is available at https://github.com/iCityLab/TraffNet.

Published
2023

2. A study of lexical variation, comprehension and language attitudes in deaf users of Chinese Sign Language (CSL) from Beijing and Shanghai

Author

Ma, Yunyi

Abstract

Regional variation between the Beijing and Shanghai varieties, particularly at the lexical level, has been observed by sign language researchers in China (Fischer & Gong, 2010; Shen, 2008; Yau, 1977). However, few investigations into the variation in Chinese Sign Language (CSL) from a sociolinguistic perspective have previously been undertaken. The current study is the first to systematically study sociolinguistic variation in CSL signers’ production and comprehension of lexical signs as well as their language attitudes. This thesis consists of three studies. The first study investigates the lexical variation between Beijing and Shanghai varieties. Results of analyses show that age, region and semantic category are the factors influencing lexical variation in Beijing and Shanghai signs. To further explore the findings of lexical variation, a lexical recognition task was undertaken with Beijing and Shanghai signers in a second study looking at mutual comprehension of lexical signs used in Beijing and Shanghai varieties. The results demonstrate that Beijing participants were able to understand more Shanghai signs than Shanghai participants could understand Beijing signs. Historical contact is proposed in the study as a possible major cause for the asymmetrical intelligibility between the two varieties. The third study investigated signers’ attitudes towards regional varieties of CSL and Signed Chinese via a questionnaire. The findings demonstrate that older signers tended to have a conservative attitude towards their comprehension of regional signs of CSL, and that participants of both regions tended to ascribe high solidarity to their own varieties and high social status to Signed Chinese. This study has expanded our knowledge of sociolinguistic variation in Beijing and Shanghai signing varieties, and lays the groundwork for a future comprehensive study of the regional varieties in CSL. This study may also serve as a useful reference for official sign language planning in China including such issues as promoting a standardised lexicon across China and offering qualifications for CSL learners and interpreters.

Published
2020

3. TraffNet: Learning Causality of Traffic Generation for Road Network Digital Twins

Author

Xu, Ming, Ma, Yunyi, Li, Ruimin, Qi, Geqi, Meng, Xiangfu, and Jin, Haibo

Subjects
FOS: Computer and information sciences, Computer Science - Machine Learning, Artificial Intelligence (cs.AI), Computer Science - Artificial Intelligence, Machine Learning (cs.LG)
Abstract

Road network digital twins (RNDTs) play a critical role in the development of next-generation intelligent transportation systems, enabling more precise traffic planning and control. To support just-in-time (JIT) decision making, RNDTs require a model that dynamically learns the traffic patterns from online sensor data and generates high-fidelity simulation results. Although current traffic prediction techniques based on graph neural networks have achieved state-of-the-art performance, these techniques only predict future traffic by mining correlations in historical traffic data, disregarding the causes of traffic generation, such as Origin-Destination (OD) demands and route selection. Therefore, their performance is unreliable for JIT decision making. To fill this gap, we introduce a novel deep learning framework called TraffNet that learns the causality of traffic volumes from vehicle trajectory data. First, we use a heterogeneous graph to represent the road network, allowing the model to incorporate causal features of traffic volumes. Next, inspired by the traffic domain knowledge, we propose a traffic causality learning method to learn an embedding vector that encodes OD demands and path-level dependencies for each road segment. Then, we model temporal dependencies to match the underlying process of traffic generation. Finally, the experiments verify the utility of TraffNet. The code of TraffNet is available at https://github.com/mayunyi-1999/TraffNet_code.git.

Published
2023
Full Text
View/download PDF

4. Optimal allocation of distributed generation in distributed network with genetic algorithm

Author

Crebbin, Gregory, Ma, Yunyi, Crebbin, Gregory, and Ma, Yunyi

Abstract

With the higher and higher demand by people for the quality and reliability of the power supply, the centralized power grid cannot meet the requirements. Distributed generation access to the distributed network is an inevitable trend of the development of power industry. It is also one of the important aspects of the development of a smart grid. Distributed Generation is a small scale, low investment and clean power resource. Installing distributed generation in the network can also have positive effects of the networks, such as lower losses, stable voltage and so on; these effects are related to the location and sizing of distributed generation. So this thesis is about optimal location and sizing of distributed generation in distributed network. At first, I introduce the definition of distributed generation and some common types of DG, such as wind power and PV power. Using Power factory simulate IEEE14 system was simulated to find the how the DG affects the power losses and voltage quality of the network. According to the analysis above, a model was built with the lowest investment and power losses and the highest voltage quality based on IEEE 14 buses system. The model was simulated in Power Factory; calculated the power flow using Power factory and used adaptive genetic algorithm to solve the optimal problem. The results of the case study will prove that installing distributed generation can decrease the power losses and improve the voltage quality. Finally, from the practical applications aspect, some directions for future research in this problem have been indicated.

Published
2015

5. A Simplified Lumped Parameter Model for U-Tube Steam Generator

Author

Ma Yunyi and Zhang Yongsheng

Subjects
Engineering, business.industry, Boiler (power generation), System dynamics, Energy conservation, Dynamic simulation, Control theory, Real-time simulation, Thermal, Feedforward neural network, business, MATLAB, computer, computer.programming_language
Abstract

A simplified lumped parameter model for U-tube steam generator (UTSG) is presented, according to its working principle and the mass and energy conservation theory as well as the principle of thermal system dynamics. A three-element controller of water level is designed to verify the simplified model. Matlab/Simulink module is used for the dynamic simulation of UTSG, result shows that the presented model can well reflect the dynamic trends of main parameters of UTSG and constitute the real time simulation system of nuclear powered secondary circuit with other models.

Published
2010

Catalog

Books, media, physical & digital resources
See catalog results