Search

Your search keyword '"Wang, Zhuo"' showing total 11,100 results
Start Over Author "Wang, Zhuo"
11,100 results on '"Wang, Zhuo"'

228. Data-driven state estimation of large-scale battery systems

Author

Wang, Zhuo, Gladwin, Dan, and Foster, Martin

Abstract

The system states of a grid-connected battery energy storage system (BESS), state of charge (SOC) and state of health (SOH), are essential for its control to trade energy and provide services such as frequency response. There is significant work in estimating these states at cell-level, however, for a large-scale BESS these methods have not been examined before to see whether they can scale. There are a number of challenges with large BESS that need to be considered. The first is that these often contain 10,000-100,000s of cells where interconnections make the system more complex. The second is that estimation methods rely on accurate and reliable measurements of voltage and current, for large BESS where the range of the sensors is larger the errors will be higher. This thesis also considers the real-world scenario where data granularity, accuracy and quality is variable. In this work it is shown how cell-level state estimation techniques can be utilised on large-scale BESSs using experimental data from a 2MW, 1MWh BESS. The results show how a Dual Sigma Point Kalman Filter (DSPKF) SOC estimation can provide improved accuracy over the integrated commercial battery management system SOC estimation. It is then demonstrated how the DSPKF parameters can be tuned by a genetic algorithm to simplify selection to generalise the application of the method for different BESSs. Using system round-trip efficiency (RTE) measurements, validation on the accuracy of the methodology is provided. This thesis also proposes how the improved SOC estimation can be combined with a total least-squares (TLS) method for capacity estimation to less than 1% error. To achieve this an approach is presented for data selection that is required to minimize the error. Finally, parameters of the equivalent circuit model (ECM) of BESSs are estimated in the weight filter of the DSPKF and the results are validated by a voltage simulation process. Throughout the thesis online system state estimation is demonstrated using both designed test and real-world operational data where the BESS has provided contracted frequency response services to the GB National Grid.

Published
2022

230. Scalable Rule-Based Representation Learning for Interpretable Classification

Author

Wang, Zhuo, Zhang, Wei, Liu, Ning, and Wang, Jianyong

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

Rule-based models, e.g., decision trees, are widely used in scenarios demanding high model interpretability for their transparent inner structures and good model expressivity. However, rule-based models are hard to optimize, especially on large data sets, due to their discrete parameters and structures. Ensemble methods and fuzzy/soft rules are commonly used to improve performance, but they sacrifice the model interpretability. To obtain both good scalability and interpretability, we propose a new classifier, named Rule-based Representation Learner (RRL), that automatically learns interpretable non-fuzzy rules for data representation and classification. To train the non-differentiable RRL effectively, we project it to a continuous space and propose a novel training method, called Gradient Grafting, that can directly optimize the discrete model using gradient descent. An improved design of logical activation functions is also devised to increase the scalability of RRL and enable it to discretize the continuous features end-to-end. Exhaustive experiments on nine small and four large data sets show that RRL outperforms the competitive interpretable approaches and can be easily adjusted to obtain a trade-off between classification accuracy and model complexity for different scenarios. Our code is available at: https://github.com/12wang3/rrl., Comment: Accepted by NeurIPS 2021; Interpretable ML; Neuro-Symbolic AI

Published
2021

231. Depth Quality-Inspired Feature Manipulation for Efficient RGB-D Salient Object Detection

Author

Zhang, Wenbo, Ji, Ge-Peng, Wang, Zhuo, Fu, Keren, and Zhao, Qijun

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

RGB-D salient object detection (SOD) recently has attracted increasing research interest by benefiting conventional RGB SOD with extra depth information. However, existing RGB-D SOD models often fail to perform well in terms of both efficiency and accuracy, which hinders their potential applications on mobile devices and real-world problems. An underlying challenge is that the model accuracy usually degrades when the model is simplified to have few parameters. To tackle this dilemma and also inspired by the fact that depth quality is a key factor influencing the accuracy, we propose a novel depth quality-inspired feature manipulation (DQFM) process, which is efficient itself and can serve as a gating mechanism for filtering depth features to greatly boost the accuracy. DQFM resorts to the alignment of low-level RGB and depth features, as well as holistic attention of the depth stream to explicitly control and enhance cross-modal fusion. We embed DQFM to obtain an efficient light-weight model called DFM-Net, where we also design a tailored depth backbone and a two-stage decoder for further efficiency consideration. Extensive experimental results demonstrate that our DFM-Net achieves state-of-the-art accuracy when comparing to existing non-efficient models, and meanwhile runs at 140ms on CPU (2.2$\times$ faster than the prior fastest efficient model) with only $\sim$8.5Mb model size (14.9% of the prior lightest). Our code will be available at https://github.com/zwbx/DFM-Net., Comment: accepted in ACM MM 2021

Published
2021

Catalog

Books, media, physical & digital resources
See catalog results