傳統非光學式氣體檢測器容易受到其他氣體干擾,限制了其在特定應用中的效能。光學式氣體檢測器因其高靈敏度、快速響應和專一性等優勢,逐漸成為市場關注的焦點。然而,光學氣體檢測器的高成本限制了其普及應用。為了解決這些問題,本研究設計並研發了一種基於非準直光源的濃度偵測系統。將非準直光路設計整合到光學濃度偵測系統的架構中,旨在降低成本的同時,保持其高靈敏度、快速響應和專一性等優點。我們計劃使用可見光系統,以墨水溶液作為可行性的研究樣本,評估這種新型設計的可行性和性能。 本研究使用ASAP光追跡軟體建構系統和光源的模型並模擬光路,透過實驗量測不同濃度樣本的光穿透率。透過模擬、數學擬合與斜率分析等方式對系統的部分參數進行簡化,找出光強度與濃度的關係式,並了解系統的可工作區和限制。實驗結果表明,在10-35ppm濃度範圍內,系統能穩定測量樣本濃度,且與實際濃度的殘差小於1ppm。通過這些結果,我們證實了非準直光源應用於濃度偵測的可行性,並提出了為了應用於實際環境氣體的檢測需要做的改變,以及提升系統性能的建議。;Traditional non-optical gas detectors are often limited by their susceptibility to interference from other gases, affecting their performance in specific applications. Optical gas detectors, due to their high sensitivity, rapid response, and specificity, have gained attention in the market. However, their high costs have restricted widespread use. To address these issues, this study designs and develops a concentration detection system based on a non-collimated light source. By integrating a non-collimated optical path design into the architecture of the optical concentration detection system, we aim to reduce costs while maintaining high sensitivity and fast response. We plan to use a visible light system and ink solution as a feasibility study sample to evaluate the feasibility and performance of this new design. We used ASAP ray-tracing software to model the system and the light source, simulating the optical path and measuring the light transmittance of samples with different concentrations. By simplifying system parameters through simulation, mathematical fitting, and slope analysis, we identified the relationship between light intensity and concentration, and explored the system′s working range and limitations. Experimental results show that the system can stably measure sample concentrations in the range of 10-35 ppm, with residuals less than 1 ppm from actual concentrations. These results confirm the feasibility of using non-collimated light sources for concentration detection, and we propose changes needed for real-world gas detection applications as well as suggestions for improving system performance.