In order to meet the requirements of the miniaturization of the new large-format high-resolution uncooled infrared detector（1024×768 pixels）and the infrared thermal imaging system, the continuous zoom structure with mechanical positive group compensation is adopted. An uncooled infrared continuous zoom imaging optical system with focal length of 25--75 mm is designed through the reasonable distribution of optical focal degree of each lens. The operating band is 8--12 m, and the F-number is 1.2.The system consists of five lenses with a total mass of only 255 g and optical length of 125 mm. The design result shows that the system is small,simple and compact, and good in imaging. The modulation transfer function value is greater than 0.3 at the corresponding characteristic frequency（36 lp/mm）of the detector, which can meet the application requirements.

In order to improve the reliability of the product, an optical passive non-thermal infrared lens working in the 8--14 m waveband is designed, which can image clearly in a wide temperature range without focusing. The system selects a new type of chalcogenide glass that can be processed by pressing and is suitable for mass production. The optical lens adopts three-piece structure. It is a passive non-thermal lens which is based on the principle of refraction/diffraction.The focal length of the system is 42 mm and the F-number is 0.9. The lens matches an uncooled long-wave infrared detector with pixel size of 14 m and pixel number of 640×512. The lens can clearly image in the temperature range of -40 ℃--50 ℃ and in the full field-of-view. The modulation transfer function(MTF)is greater than 0.38@36 lp/mm at the operating temperature in the full field-of-view, which is close to the diffraction limit.

Aiming at the problems in the active terahertz (THz) imaging such as the poor image quality,the variety of hidden objects and the scarcity and imbalance of training samples, the objects segmentation networks based on the conditional generative adversarial networks′model Mask-CGANs and the objects detection and recognition networks based on the RetinaNet are built, which realizes the multi-object segmentation, detection and recognition of hidden objects in the THz imaging. The constraint loss functions and the networks structures proposed for the segmentation task make the model keep the balance between the recall rate and the false alarm rate, and the requirement of training sample size is reduced. The loss functions used for the detection task improve the detection accuracy under the condition of unbalanced training samples.

To denoise the optical time domain reflectometer (OTDR) data, a denoising algorithm based on variational mode decomposition (VMD) and singular spectrum analysis (SSA) is proposed. First, perform VMD decomposition on the OTDR data; then, determine the effective component according to the correlation coefficient criterion, perform singular spectrum analysis on the component with the smaller correlation coefficient to the original signal, and then sum all the components to obtain the reconstructed signal. So signal-to-noise ratio of the data is improved. It is verified by experiments that the algorithm has good denoising effect, simple implementation and great application value.

The stochastic resonance phenomenon for a fractional-order oscillator with multiplicative signal and random characteristic frequencies is studied. Under the condition of low driving frequency, the long-term analytical expression for the system output signal of the oscillator is derived by using the characteristics of the fractional-order operator. The results show that the resonance phenomenon of system output signal is observed, due to the interaction of multiplicative signal, noise and fractional-order oscillator. This phenomenon appears on the relationship curves of the output signal versus the noise intensity, versus the noise correlation rate, versus the fractional exponent, versus the viscous damping coefficient, as well as versus the system characteristic-frequency. In addition, the non-monotonous dependence of the output signal on the friction coefficient is also discussed.

There is a large temperature difference in the tow cross-section during the graphitization of carbon fibers via laser, which leads to a serious fuzz phenomenon of carbon fiber tows. When the laser exceeds a certain power, the phenomenon of tow breaking may even occur, seriously affecting the continuous graphitization of the carbon fiber via laser. The use of microwaves and lasers to synergistically graphitize carbon fibers can reduce the temperature difference and thermal stress of the tow during graphitization. In view of the particularity of microwave heating, a heating cavity that can achieve continuous and efficient temperature rise is designed; then, a multiphysics model of electromagnetic heating--laser heating--thermal radiation--flow heat transfer--solid mechanics is established for simulation. The simulation results show that the introduction of microwaves can improve the distribution of the temperature field of the tow during laser graphitization and effectively reduce the thermal stress during tow heating.

Whether Infrared Thermography (IRT) can be used as a method to evaluate the therapeutic effect of rheumatoid arthritis (RA) is discussed. A self-control study is conducted on 30 wind-cold-dampness type RA patients according to the traditional Chinese medicine syndrome differentiation. Before and after the treatment,HIR--2000A infrared thermal imager is used to scan the palms of the patients, and the measured body surface temperature is recorded. The differences of body surface temperature of the patients′palms before and after the treatment are statistically significant (P < 0.01).The early application of IRT technology in the evaluation of the therapeutic effect of RA has important theoretical and practical significance for the traditional Chinese medicine treatment of RA.