| 引用本文: | 范俞超,孙青林,董方酉,陈增强.基于粒子群算法–反向传播神经网络自适应的氧调器控制系统[J].控制理论与应用,2020,37(3):687~695.[点击复制] |
| FAN Yu-chao,SUN Qing-lin,DONG Fang-you,CHEN Zeng-qiang.Control system of oxygen regulator based on particle swarm optimization-back propagation neural network adaptive control algorithm[J].Control Theory and Technology,2020,37(3):687~695.[点击复制] |
|
| 基于粒子群算法–反向传播神经网络自适应的氧调器控制系统 |
| Control system of oxygen regulator based on particle swarm optimization-back propagation neural network adaptive control algorithm |
| 摘要点击 2513 全文点击 911 投稿时间：2019-01-03 修订日期：2019-06-04 |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| DOI编号 10.7641/CTA.2019.90008 |
| 2020,37(3):687-695 |
| 中文关键词 氧气面罩 氧气调节器 氧气控制 粒子群算法 反向传播神经网络 PSO-BP神经网络自适应算法 参数自适应 |
| 英文关键词 oxygen mask oxygen regulator oxygen control particle swarm optimization back propagation neural network PSO-BP neural network adaptive algorithm parameter adaptive |
| 基金项目 国家自然科学基金,其它 |
|
| 中文摘要 |
| 氧气面罩中的核心部件是氧气调节器(氧调器).针对当前氧气调节器的大流量、低吸气阻力、快速响应的性能需求, 本文在分析了电子氧气调节器工作原理的基础上, 介绍了氧气调节器的数学模型, 采用了反向传播(BP)神经网络自适应控制算法, 并使用粒子群算法(PSO)对BP神经网络自适应控制算法的初值进行筛选. 最后, 对算法的性能进行了仿真. 仿真结果表明, 系统具有鲁棒性, 且与传统的比例-积分-微分(PID)控制方法和自抗扰控制(ADRC)方法相比, PSO-BP 神经网络自适应控制方法实现了更精确的吸气阻力调节, 更快的响应速度. 此外, 当呼吸频率变化或者外界干扰变化时, 相比于常规PID算法和ADRC算法则需要人工调整控制参数, PSO-BP神经网络自适应算法则可以自动在线学习训练并调整控制参数, 应用前景广阔. |
| 英文摘要 |
| The core component of the oxygen mask is the oxygen regulator (OR). Aiming at performance requirements of high flow rate, low suction resistance, and fast response for current oxygen regulators, this paper introduces the mathematical model of the oxygen regulator based on the analysis of the working principle of the electronic oxygen regulator. Then, the back propagation (BP) adaptive neural network control algorithm is adopted. Third, the particle swarm optimization (PSO) algorithm is used to screen the initial values of BP adaptive neural network control algorithm. At last, this paper simulates the performance of the algorithm. The simulation results show that the system is robust and compared with the traditional proportion integral derivative (PID) control method and active disturbance rejection control (ADRC) method, PSO-BP neural network adaptive control method achieves more accurate suction resistance adjustment and faster response speed. In addition, when the respiratory frequency changes or the external disturbance changes, the control parameters need to be manually adjusted compared with the conventional PID algorithm and the ADRC algorithm. The PSO-BP neural network adaptive algorithm can automatically learn and adjust the control parameters online, and the application prospect is broad. |
|
|
|
|
|