Lasers are widely used in traditional interferometry due to their high coherence. However, good coherence leads to multi-surface crosstalk, and optical components such as parallel plates, prisms and thin films cannot be directly measured. Using short-coherence lasers as light sources can isolate non-measurement surfaces and realize direct measurement of such special optical components. Compared with the wavelength stability of 0.0001 nm of He-Ne lasers, the stability of short-coherence semiconductor laser is still not good enough, limiting its application in high-precision interferometry scenarios. The temperature of short coherence semiconductor laser directly affects its wavelength, spectral width, and power, influencing the accuracy of the measurement system. At the same time, the slow response speed and poor resistance to environmental disturbances also make the use of short coherence semiconductor lasers poor. The performance can be improved by proportional-integral-derivative (PID) algorithm, but the parameter adjustment process takes a lot of time, and the temperature control accuracy is difficult to reach below 0.01 ℃. The temperature-current dual closed-loop control can achieve a temperature control accuracy of 0.002 ℃, but the structure is complex and difficult to implement. Here we use fuzzy sliding mode control to improve the temperature control performance and enhance the stability and response speed of short coherence semiconductor lasers.

Aiming at the stability of temperature control, the principles of semiconductor coolers and thermistors are studied, and the physical model of the temperature control system is established. Sliding mode control is used to improve the control effect and robustness of the temperature control system, and fuzzy control is combined to change the approach rate to speed up the response speed. Two semiconductor coolers are used to independently control the upper and lower surface temperatures of the semiconductor laser diode to further speed up the response speed and enhance the robustness. The simulation results show that compared with sliding mode control and PID control, fuzzy sliding mode control has fast response speed, small overshoot, and high stability. A verification experiment is built to collect temperature information through a negative temperature coefficient thermistor, and the control voltage is outputted after running the fuzzy sliding mode control algorithm on the microcontroller control unit. After the control voltage passes through the 20 bit digital-to-analog converter, the analog signal is transmitted to the semiconductor cooler driver, so that it drives the semiconductor cooler to control the laser temperature. There is a monitoring thermistor in the laser, and the temperature information is collected and recorded as the temperature measurement result through a digital multimeter and a computer. The laser is connected to the power meter and spectrometer to measure its power, central wavelength, and spectral width stabilities.

fuzzy sliding mode control, sliding mode control, and PID control. The temperature measurement results are shown in Fig. 9. The experiment shows that the rise time of fuzzy sliding mode control is 6.1 s, the steady-state time is 3.3 s, and the overshoot is 0.04%. Fuzzy sliding mode control algorithm has the characteristics of fast response speed and small overshoot. In contrast, the steady-state time and rise time of sliding mode control are longer, and the response speed is slow. PID control has more serious oscillations at the beginning, and its overshoot is larger. The comprehensive performance of fuzzy sliding mode control is the best. After the laser using fuzzy sliding mode control enters the steady-state temperature range, the temperature is continuously monitored for 6 h. The temperature control accuracy is measured to be ±0.003 ℃, and the temperature fluctuation range is 0.004 ℃, indicating the fuzzy sliding mode control algorithm has good temperature stability. Spectrum and power measurements are performed on the short coherence semiconductor laser using fuzzy sliding mode control. The central wavelength stability of the short coherent semiconductor laser is better than 0.00032%, the spectral width stability is better than 0.31%, and the power stability is 0.18%, indicating it has good performance stability.

Based on the principles of semiconductor refrigerators and thermistors, this study establishes a physical model of the temperature control system. Fuzzy sliding mode control is used to improve the stability and response speed of the temperature control system, and an experiment is built based on the single-chip microcomputer control unit. The experiment verifies that fuzzy sliding mode control has good temperature control response speed and stability, and the short coherent semiconductor laser using fuzzy sliding mode control has good spectral and power stabilities. The fuzzy sliding mode control algorithm proposed in this study provides a reference for the design and optimization of semiconductor laser temperature control algorithms.