<?xml version="1.0" encoding="UTF-8"?>
<article xmlns:xlink="http://www.w3.org/1999/xlink" article-type="research-article" dtd-version="1.3" xml:lang="zh-CN">
  <front>
    <journal-meta>
      <journal-title-group><journal-title>南太湖新工科产业学报</journal-title></journal-title-group>
      <issn>3136-2249</issn>
      <publisher><publisher-name>求真学术出版</publisher-name></publisher>
    </journal-meta>
    <article-meta>
      <article-id pub-id-type="publisher-id">1b1a0d49-6111-4104-95c6-f60f4abb2c14</article-id>
      <article-id pub-id-type="doi">10.68249/jstnei.2026.0015</article-id>
      <elocation-id>0015</elocation-id>
      <title-group><article-title>欠驱动自行车机器人视觉误差级联平滑与安全约束循迹控制</article-title></title-group>
      <contrib-group>
        <contrib contrib-type="author" corresp="yes"><name><surname>关皓轩</surname></name><aff>北华航天工业学院电子与控制工程学院</aff></contrib>
        <contrib contrib-type="author"><name><surname>段雨薇</surname></name><aff>北华航天工业学院电子与控制工程学院</aff></contrib>
        <contrib contrib-type="author"><name><surname>李思奇</surname></name><aff>北华航天工业学院电子与控制工程学院</aff></contrib>
      </contrib-group>
      <pub-date publication-format="electronic"><year>2026</year><month>8</month><day>29</day></pub-date>
      <volume>1</volume><issue>2</issue>
      <fpage>13</fpage><lpage>17</lpage>
      <abstract><p>针对欠驱动自行车机器人单目摄像头视觉车道线识别循迹过程中横向偏差信号易受阴影变化、车道线短时缺失和检测噪声影响，同时可能通过转向执行引发姿态振荡，导致机器人车体失衡，本文提出一种视觉误差级联平滑与安全约束循迹控制方法。该方法不改变既有车道识别与底层平衡控制结构，而在二者之间构建视觉识别与执行接口层：首先，利用短时滑动平均抑制高频随机噪声，随后通过相邻差值约束削弱大幅突变，并结合输出饱和限制转向指令幅值；同时，采用高速UART实现平滑指令的实时下发。实车在室外400 m跑道开展直道及左右弯道闭环测试。实车实验结果表明，该方法可在STM32平台上实现实时处理，在直道、左转和右转场景下分别将横向RMSE降低40.2%、39.85%和32.35%。</p></abstract>
      <kwd-group><kwd>自行车机器人</kwd><kwd>视觉识别</kwd><kwd>级联滤波</kwd><kwd>平衡控制</kwd><kwd>抗扰动</kwd></kwd-group>
      <permissions><license><license-p>CC BY-NC-ND 4.0</license-p></license></permissions>
      <self-uri xlink:href="https://www.qiuzhenpress.com/articles/jstnei-2026-0015" />
    </article-meta>
  </front>
  <back>
    <ref-list>
      <ref id="R1"><mixed-citation>[1]Li S, Guan H, Ge J, et al. Mathematical Modeling and Vision-Guided Triple-Loop Control of an Underactuated Bicycle Robot[J]. Mathematics, 2026, 14(12): 2160.</mixed-citation></ref>
      <ref id="R2"><mixed-citation>[2]Li J. PID Line-Tracking Vehicle Based on OpenCV with Object Recognition and Response Function[C]. AIP Conference Proceedings. 2024, 3144:050009.</mixed-citation></ref>
      <ref id="R3"><mixed-citation>[3]Zhai S, et al. An algorithm for lane detection based on RIME optimization and optimal threshold[J]. Scientific Reports, 2024, 14(1):27244.</mixed-citation></ref>
      <ref id="R4"><mixed-citation>[4]Meijaard J P, Papadopoulos J M, Ruina A, et al. Linearized dynamics equations for the balance and steer of a bicycle: a benchmark and review[J]. Proceedings of the Royal Society A, 2007, 463(2084): 1955-1982.</mixed-citation></ref>
      <ref id="R5"><mixed-citation>[5]Aström K J, Klein R E, Lennartsson A. Bicycle dynamics and control: adapted bicycles for education and research[J]. IEEE Control Systems Magazine, 2005, 25(4): 26-47.</mixed-citation></ref>
      <ref id="R6"><mixed-citation>[6]KARAPET B, SAVITSKYI R, VAKALIUK T. Method of comparing and transforming images obtained using UAV[J], Radioelectronic and Computer Systems, 2024(1):99.</mixed-citation></ref>
      <ref id="R7"><mixed-citation>[7]Canny J. A computational approach to edge detection[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1986, 8(6): 679-698.</mixed-citation></ref>
      <ref id="R8"><mixed-citation>[8]Duda R O, Hart P E. Use of the Hough transformation to detect lines and curves in pictures[J]. Communications of the ACM, 1972, 15(1): 11-15.</mixed-citation></ref>
      <ref id="R9"><mixed-citation>[9]Bradski G. The OpenCV Library[J]. Dr. Dobb&apos;s Journal of Software Tools, 2000, 25(11): 120-126.</mixed-citation></ref>
    </ref-list>
  </back>
</article>
