Research Of Three-channel Fiber-optic Displacement Sensor

In recent years, the advantages of optical fiber sensor have small volume, light weight, high sensitivity, strong resistance to electromagnetic interference and so on. It is widely used in the sensing field. The displacement measurement is one of the important applications of the optical fiber sensor. The conventional non-contact displacement sensor is limited by the reflection area, when it need to analyze displacement and vibration conditions of the small size objects, the resulting measurement is not very satisfactory. But fiber-optic displacement sensor can be made very small probe (minimum0.2mm diameter). It also can be made of a straight line in the form of emission and reception, which can calculate the displacement values by measuring the degree of occlusion of the fiber objects in the process of displacement. In many cases, it is able to complete the task that the traditional displacement sensor is difficult or impossible to complete. Reflective fiber-optic displacement sensor which is constituted by two of the multi-mode fiber is a typical structure, but the structure exists light peak phenomenon. The use of two-channel structure can solve this problem, and system output is the ratio of two receiving fiber. This can eliminate the impact of the equipment or the environment, which improve the measurement accuracy. However, the two-channel structure is affected by the reception range of the fiber-optic probe, so the measuring range of the sensor is also limited.In order to avoid these shortcomings, this paper proposes a three-channel fiber-optic displacement sensor. The system consists of three optical paths as the perception of the light path. The system of measurement results are normalized, which can solve the interference of the external environment and improve the precision of the system.Hardware circuit portion of the system are consist of a light modulation circuit, optoelectronic detector circuit, filter circuit and truth value converter circuit. The system selects the LD as the light source of system, the use of the optoelectronic detector circuit can convert an optical signal into an electrical signal so that the output signal can be sampled. Regression analysis and radial basis function neural network (RBF) in the software section are used for data processing. At the same time, there also discuss the advantages and disadvantages of the two algorithms. The algorithm has been run into the sampler. The results show that the three-channel fiber-optic displacement sensor combined with the appropriate data fusion method can reduce the instability of the system and improve measurement accuracy and dynamic range of the sensor.