榮譽榜

本研究探討營建業如何依據全球永續性報告協會（GRI）準則編寫永續報告書，並利用機器學習分析這些準則與企業財務績效之間的關聯。以中鼎工程公司為案例，研究其永續報告書的撰寫方法及內容，並整理 52 家台灣上市營建企業依循 GRI 準則的情況，分析 GRI 準則與財務績效（EPS、ROE、ROA 及 Income）之間的關聯。研究方法包含兩部分：首先，針對中鼎工程永續報告書進行深入解析，整理其 GRI 準則索引及報告內容與 GRI 準則的對照；其次，搜集 52 家營建企業的 GRI 準則依循數據及財務績效數據，利用線性回歸、K-近鄰演算法及隨機森林等機器學習模型進行分析。
研究結果顯示，隨機森林模型在預測財務績效方面表現最佳，並鑑別出對財務
績效有顯著影響的十大 GRI 準則特徵。這些特徵包括：確定給付制義務與其他退休計畫、氣候變遷風險與機會、組織內部的能源消耗量等。最後，本研究建議企業在撰寫永續報告書時，應重點關注上述特徵，以提高報告書的品質並促進財務績效提升。總結來看，依循 GRI 準則編寫的永續報告書能夠對營建業公司的財務表現產生正向影響，並有助於企業達成永續發展的目標。

關鍵字：營建業、永續報告書、GRI 準則、財務績效、機器學習分析

With the development in technology, the point cloud data has high potential usage in the construction field. A prerequisite of most point cloud applications is first to register the point cloud segments in the same global coordinate system. However, manually registering point clouds is time-consuming and labor-intensive work. This research addresses the point cloud registration problem in the built environment by using line geometry, which is considered a salient feature in the building environment. A framework, including a line extraction module, a line feature matching module, an outlier pruning module, and a 3-linebased registration module is proposed in this research to perform point cloud registration automatically. The experiment showed that the registration recall could be up to 50% in the WHU-TLS dataset even in a highly contaminated search space, which contained only 10% inlier correspondences. The experiment results proved the feasibility of the line-based registration framework and identified several performance indicators to perform a successful point cloud registration.

Keywords: Point Cloud Registration; Deep Learning; Building Area; Point Cloud Application

Predicting building thermal loads is critical for optimizing building energy management systems and enhancing energy efficiency. The high-dimensional characteristics of HVAC loads necessitate the selection of an optimal number of features to improve model accuracy. In this research, a novel approach is introduced that utilizes weighted feature selection methods. Four different techniques, including Pearson, Spearman, and Kendall correlations, as well as a dimensionality reduction method, Principal Component Analysis, are used to determine four distinct sets of features based on the dimensionality of feature numbers. After that, the combination of Neural Networks (NN) and Bidirectional Gated Recurrent Units (BiGRU) is used with the Optical Microscope Algorithm (OMA), an optimization algorithm inspired by microscopic mechanisms, to fine-tune the NN-BiGRU framework. This Opticalinspired Bidirectional Machine Learning model (OMA-NN-BiGRU) predicts the outcomes from the four data methods. Additionally, OMA is applied to optimize the weight combinations of the predictions from the individual feature selection methods, based on their contributions to the final prediction. The results demonstrate that the proposed model achieves significantly higher accuracy than the single methods. The weighted feature selection model demonstrates superior performance, with a Root Mean Squared Error
(RMSE) of 0.046, a Mean Absolute Error (MAE) of 0.032, a Mean Absolute Percentage Error (MAPE) of 12.6%, and R² values of 0.95 and 0.90, respectively. The OMA-NNBiGRU model outperforms other compared models in building thermal load prediction. This study offers fresh insights into creating an integrated approach based on different input
variable dimensions.

Keywords: Thermal Load Prediction; Weighted Feature Selection Approach; NeuralNetwork; Bidirectional Gated Recurrent Unit; Optical Microscope Algorithm

臺灣地區河川及溪流的沖蝕現象嚴重，經常受到颱風、地震等天然災害的影響，再加上山坡地的過度開發，導致水土流失的問題更加嚴重。為了減少災害發生，興建大量水土保持設施；然而，許多構造物已年久失修，急需進行檢查和整修。為了改進巡查效率、提高決策客觀性，以及優化相關單位維護資源的配置。本研究致力於建立一套水土保持構造物延壽評估系統，這套系統以互動網頁之方式整合了無人機(UAV)、人工智慧(AI)和資料庫等技術，實現水土保持設施的巡檢、劣化樣態辨識、耐久性評估和維護方式建議的多功能合一，以供水土保持設施維護人員參考。研究結果顯示，所訓練出的模型在辨識水土保持構造物劣化樣態方面具有優異的成效。耐久度評估模式能夠精確判斷構造物的狀況，並提供相應的延壽修繕建議，整個評估系統的建置不僅能夠有效地協助維護人員，同時確保水土保持構造物能夠更長久地發揮其機能。

關鍵字：無人機、水土保持構造物、遷移式學習

在建築物和橋梁的生命週期中，營運階段佔用了絕大部分的時間和資源，尤其是後續的維護和定期檢修。因此，如何更有效地管理和監測它們成為一個重要議題。傳統的檢測方法依賴目視檢測，需要大量時間和人力成本。
隨著 AI 技術和深度學習算法的發展，無人機 (UAV) 和影像辨識的結合提供了
解決方案。無人機具高機動性和大面積偵查能力，能有效解決位置不便和工作效率問題。然而，傳統的無人機應用僅限於拍攝影像，後續的影像處理和辨識需額外人力和時間成本，且缺乏即時性。
本研究利用 ROS (Robot Operating System) ，結合無人機和影像辨識模型，建立即時的裂縫辨識系統。此系統將無人機捕獲的影像匯入開發板上的 ROS 系統，經過一系列影像處理步驟後，輸入已訓練好的影像辨識模型進行裂縫辨識。辨識結果即時儲存並向使用者發出警示通知，實現裂縫檢測和通報功能的即時化和高效化。
本研究通過先進技術手段，達成即時裂縫影像辨識系統，提高建築物和橋梁監
測作業的效率和安全性，滿足社會需求。

關鍵字：ROS (Robot Operating System)、深度學習、影像辨識

In energy research, metaheuristic optimization is crucial for refining the
hyperparameters of artificial intelligence (AI) models. As data science evolves, the complexity of datasets, input variables, and patterns has increased, posing significant challenges in developing efficient AI models. This research introduces the Age of Exploration-Inspired Optimizer (AEIO), a novel metaheuristic technique designed to enhance performance by fine-tuning hyperparameters in AI models for energy applications.The AEIO algorithm is rigorously evaluated using benchmark functions across various scales and dimensions. Its effectiveness is further demonstrated through three case studies: green energy production, residential energy consumption, and regional energy demand forecasting. In each scenario, the AEIO algorithm consistently outperforms existing methods when combined with machine and deep learning models to predict discrete and time-series data in the energy sector. This study also pioneers integrating density-based spatial clustering of applications with noise (DBSCAN) into a metaheuristic optimization algorithm. It comprehensively assesses this novel approach to optimizing AI models within the energy domain, underscoring its practicality and significance.

Keywords: Energy Generation; Energy Consumption; Metaheuristic Algorithm; DBSCAN; AI Models