The strain of the ring gear can reflect the dynamic characteristics of planetary gearboxes directly, which makes it an ideal signal to monitor the health condition of the gearbox. To overcome the disadvantages of traditional methods, a new approach for the dynamic measurement of ring gear strains using fiber Bragg gratings (FBGs) is proposed in this paper. Firstly, the installation of FBGs is determined according to the analysis for the strain distribution of the ring gear. Secondly, the parameters of the FBG are determined in consideration of the accuracy and sensitivity of the measurement as well as the size of the ring gear. The strain measured by the FBG is then simulated under non-uniform strain field conditions. Thirdly, a dynamic measurement system is built and tested. Finally, the strains of the ring gear are measured in a planetary gearbox under normal and faulty conditions. The experimental results showed good agreement with the theoretical results in values, trends, and the fault features can be seen from the time domain of the measured strain signal, which proves that the proposed method is feasible for the measurement of the ring gear strains of planetary gearboxes.

Diagnosing the fault of planetary gearboxes primarily depends on the acceleration signal obtained from the gearbox casing. While the casing acceleration signal is easy to acquire, it is susceptible to interference from vibrations of other components. As a result, the strain signal, which offers a high signal-to-noise ratio, has gained prominence in recent years. However, there is limited physical knowledge of the strain signal for creating distinct fault diagnosis techniques. This paper aims to explore the characteristics of the strain signal and develop a corresponding strain signal processing approach to identify multiple faulty teeth in the sun gear. The circumferential strain of the ring gear is collected by the fiber Bragg grating mounted at the end face of the ring tooth root. The frequency distribution of the measured strain signal is revealed by a phenomenological strain model, which is established by summarizing the experimental phenomena. Based on the model, a specific strain preprocessing method, which comprises comb filtering and frequency band selection, is proposed to enhance the fault features in the strain signal. Then, a strain separation strategy is presented to obtain the strain segments linked to each tooth of the sun gear for multi-fault identification. Both simulation and experimental studies are conducted. The frequency distribution of the simulated strain signal closely matches that of the measured strain signal, validating the credibility of the established strain model. Furthermore, the necessity of the steps in the proposed strain processing approach is verified, and a comparison with other relevant methods demonstrates its superior performance.

Fault diagnosis of planetary gearboxes using acceleration signals is facing difficulties at present. In order to solve the problem, the strain signals of ring gear are adopted in this paper. First, the features of two typical faults are revealed in the outer surface strain and tooth root strain of ring gear respectively. Then, a fiber Bragg grating wavelength demodulation system with a high resolution is established to measure the strain of the ring gear. Finally, the fault diagnosis experiments are carried out. After signal processing, the feature frequency of the run-out error appears in the frequency spectrum of the surface strain signal, and the second derivative of the tooth root strain signal near the crack tooth gets larger amplitude, which shows the effectiveness of the fault diagnosis method based on strain signals.

Hyper-redundant manipulators have a wide range of applications in narrow environments. However, due to the difficulty of obtaining the joint state with multiple degrees of freedom (DoFs), the motion of the manipulator is generally controlled by the open-loop strategy, which is susceptible to interference. To provide precise joint-level closed-loop control, this work presents an optical fiber sensor for monitoring the joint bending state. The framework of the sensor is a beam with a circular cross section, and the sensing unit comprises three identical fiber Bragg gratings (FBGs). In addition, a multifidelity bending state measurement model is established. The model takes a physics-informed neural network as the basic structure. During the modeling process, the calibration data are constrained by the sensing mechanism using the transfer learning technique. The experimental results demonstrate average measurement errors of 1.22% for the joint bending angle and 1.09% for the bending direction, confirming the effectiveness of the proposed FBG sensor and the multifidelity measurement model. However, the training results with different sample sizes demonstrate that the proposed measurement model consistently achieves the highest measurement accuracy across all sample sizes, with a smaller variation in error compared to other models. This indicates that the proposed multifidelity model exhibits superior robustness.