本論文提出一個新的軟硬體架構從系統分析，模擬，設計電路，軟體控制可以獨立達成，因此成本相較其他作者更為低廉。本論文要實現的旋轉式倒單擺，是不假其他的助力，自行購置材料，製作電路，利用比例積分(PD)控制法則編寫軟體，嘗試實現角度的平衡。但在開始測試階段因為結構的關係，遇到多次失敗，後續分析而分析原因，主要是各個零件的連結是有縫隙所造成，使得讀取的數據有誤差，經過一段時間累積後，就造成系統崩潰。另外考量成本，所有的連接配件，並沒有使用金屬製作，而是先將配件使用3D建模繪製完成，使用3D建模軟體可以節省很多時間製作配件，主要是繪製立柱、連桿都可以配合較容易買的到的螺絲螺帽，也因為如此連接處會更加牢靠，本系統是完全採用市面上可以購得的零件，也因為3D件模軟體製作的配件，剛好可以將所有零件連接的恰到好處。另外實作過程中，也可以記錄即時的資料，採用了MATLAB的GUI介面去編寫一個串口的介面，此目的也可以將馬達定位控制及倒單擺的平衡控制，完成相關資料的紀錄。如此也可以比較模擬的結果與實作的誤差，為往後相關研究打好基礎。

This paper proposes a novel hardware-software architecture that allows for independent achievement of system analysis, simulation, circuit design, and software control, thereby resulting in lower costs compared to other authors. The rotating inverted pendulum aimed to be realized in this paper is intended to be constructed without additional assistance, with materials procured independently, circuitry designed, and software control implemented using a proportional-derivative (PD) control algorithm to achieve balance. However, during the initial testing phase, multiple failures were encountered due to structural issues. Subsequent analysis revealed that the discrepancies arose from gaps in the connections between various components, resulting in erroneous data readings. Over time, these discrepancies accumulated, leading to system collapse. Furthermore, considering cost constraints, metal components were not utilized for the connecting accessories. Instead, all accessories were first modeled using 3D software. This approach saved considerable time in accessory production, with pillars and connecting rods designed to accommodate readily available screws and nuts. Moreover, because the connections were more robust, solely market-available components were utilized. The accessories produced using 3D modeling software seamlessly integrated and connected all components. In addition, real-time data recording was facilitated during the implementation process. We employed MATLAB’s GUI interface to develop a serial interface, enabling the recording of motor positioning control and inverted pendulum balance control. This allowed for the comparison of simulated results with implementation errors.