Pneumatic artificial muscles (PAMs), also known as McKibben artificial muscles, are used in a variety of applications owing to its many positive features, such as simple and lightweight structure and high force to weight ratio. Sensing is essential to precisely control PAMs. Previous studies have proposed various embeddings of PAMs with sensors. However, these often require sophisticated fabrication processes and dedicated setups, which can cause production to be laborious and costly. This letter presents a method to create sensorized PAMs using a dielectric elastomer sensor (DES). In these self-sensing actuators, which we call PAM-DES, the sensor is consisting of a thin, highly compliant dielectric and conductive membranes, which allow the length of the actuator to be detected by reading the capacitance. The sensor directly wraps around the inner tube of the actuator. Therefore, the entire structure remains simple. This feature allows the fabrication of actuators to be relatively effortless and inexpensive. As a proof of concept, a PAM-DES with a length of 100 mm and diameter of 9 mm is fabricated. An analytical model is constructed to design and predict the sensor response. Characterization results show that the sensor properly detects actuator deformation, and low hysteresis and high robustness are observed. The prediction model shows good agreement with the experimental data. Furthermore, the presence of the sensor has almost no influence on the actuator performance. These results validate the concept and suggest that PAM-DES is a highly effective strategy to sensorize actuators of the same type.

中文翻译：

---

嵌入介电弹性体传感器的自感应 McKibben 人造肌肉


气动人工肌肉（PAM），也称为 McKibben 人工肌肉，由于其结构简单、重量轻、力重比高等优点而被广泛应用。传感对于精确控制 PAM 至关重要。先前的研究提出了 PAM 与传感器的各种嵌入方式。然而，这些通常需要复杂的制造工艺和专用设置，这可能导致生产既费力又昂贵。这封信介绍了一种使用介电弹性体传感器 (DES) 创建传感 PAM 的方法。在这些我们称为 PAM-DES 的自感应执行器中，传感器由薄且高度顺应的介电和导电膜组成，可以通过读取电容来检测执行器的长度。传感器直接缠绕在执行器的内管上。因此，整个结构保持简单。这一特征使得致动器的制造相对轻松且便宜。作为概念验证，制造了长度为 100 毫米、直径为 9 毫米的 PAM-DES。构建分析模型来设计和预测传感器响应。表征结果表明，传感器可以正确检测执行器变形，并且观察到低滞后和高鲁棒性。预测模型与实验数据吻合良好。此外，传感器的存在对执行器性能几乎没有影响。这些结果验证了这一概念，并表明 PAM-DES 是一种非常有效的策略来感知相同类型的执行器。