| 研究課題/領域番号 |
18K04320
|
|---|
| 研究機関 | 横浜国立大学 |
|---|
研究代表者 |
シリンゴリンゴ ディオン 横浜国立大学, 先端科学高等研究院, 特任教員(准教授) (60649507)
|
|---|
| 研究期間 (年度) |
2018-04-01 – 2021-03-31
|
| キーワード | system identification / sparse representation / seismic responses / long-span bridge / structural assessment |
|---|
| 研究実績の概要 |
Objectives of the research is to develop structural system identification using Sparse Representation model with the application for structural assessment of long-span bridge. The model consists of several sub-models: Compressed Sensing (CS), Data Fusion and Cleansing (DFC), Features Extraction and Unsupervised Learning Novelty Detection (ULND). The sub-models are data-driven and can be implemented for structural assessment purpose. Based on these sub-models, pattern recognition and diagnosis on structural condition will be performed. The developed mathematical models will be trained, verified and validated using numerical model of suspension bridge, in this case Hakucho-Bridge. For this purpose, a finite-element model of the bridge will be developed, and seismic responses of the bridge will be generated using various scenarios of earthquakes. The simulated seismic responses will be used as input to train the model. Several realistic seismic-induced damage scenarios that commonly observed on the suspension bridge will be utilized to generate structure responses and validate model. Information of structural responses will be generated using several sensor arrangements including the current sensor arrangement. Based on these scenarios the efficacy of model in assessing structural condition will be evaluated.
|
| 現在までの達成度 (区分) |
現在までの達成度 (区分)
2: おおむね順調に進展している
理由
In the first year the research focuses on: (1) Development of structural system identification using sparse representation technique. The model consists of several sub-models: Compressed Sensing (CS), Data Fusion and Cleansing (DFC). (2) Comparative study and validation of the sparse-based structural system identification using numerical model and (3) Preparation of real data from long-term bridge monitoring to fit within the purpose of sparse-based structural system identification.
|
| 今後の研究の推進方策 |
In the second year, the research will focus on: (1) Verification of sparse-based structural identification method using FEM-generated data from a benchmark long-span bridge structure under various scenarios related to damage, (2) Developing Features Extraction and Unsupervised Learning Novelty Detection (ULND). In these tasks, we utilize the previously-defined models in simulations using FEM analysis of benchmark bridge to evaluate the performance of the models and the parameters used in the models. In this simulation, we use conditions of seismic response established in previous task and features associated with bridge condition. The parameters used in the Unsupervised Learning Method need training to obtain optimum performance.
|
| 次年度使用額が生じた理由 |
1. Software related expenses for conducting numerical simulations using extensive computer model of long-span bridge.
2. Attending conferences and workshop for dissemination of research
3. Site visit and gathering of data related to research.
|
