During hypersonic wind tunnel tests， the pose information of a model supports the reliability of the experimental data， and the measurement accuracy of these data has a significant impact on the test results. Binocular vision measurement systems can measure the attitude parameters of free flight test models in a hypersonic wind tunnel. The system is usually placed outside of the wind tunnel test section and observes the model through a glass window situated on a wall of the test section. However， the glass window distorts the image and reduces the measurement accuracy. This paper proposes a measurement method to correct the glass window distortion. By modeling the image surface distortion caused by refraction through the glass window， a correction method is proposed based on the linear fitting of marker points to eliminate the image distortion and improve the measurement accuracy. Moreover， a binocular vision measurement system is developed at the scene of the Φ 1 m hypersonic wind tunnel test， and a six degrees of freedom attitude of a free flight model is successfully measured. The results show that the measuring accuracy was greater than 0.5 mm when the measuring range was 1 m × 1 m × 1 m. Therefore， the system satisfies the requirements of subsequent pneumatic data analysis.

To solve the unnatural restoration of the sky area and imprecise estimation of haze density， a dehazing algorithm for sky segmentation and haze density estimation is proposed. First， to improve the precision of transmission estimation and the quality of image dehazing， the thresholds of gradient and brightness are used to segment the sky region. Next， an adaptive dark channel prior and quadratic tree subdivision method are utilized to estimate the atmospheric light. Finally， different transmission estimation methods are used for the sky and non-sky regions； a bright channel prior is used in the sky region， and a linear haze density estimation model is proposed in the non-sky region. The final transmission is obtained by combining the probability distribution of the pixel and edge refinement using guided filter， and the recovered image is attained using the atmospheric scattering model. Experimental results show that the dehazed images perform well in terms of subjective and objective quality evaluation. The proposed dehazing algorithm can restore a more natural sky and dehaze more thoroughly to improve the clarity of image details. The operating speed of the proposed algorithm is similar to that of the current algorithms. Furthermore， the proposed algorithm is more stable compared to traditional algorithms for different hazy scenes.

To address the problems of low calibration accuracy， difficulties in large target fabrication， and complicated operation in engineering fields， a two-step camera calibration method for vision measurement system with a large field of view is herein proposed and implemented based on industrial close range photogrammetry. The mathematical model of the perspective model of camera imaging is investigated. First， in the close range， the calibration of the internal parameters at the front section is realized using a small-scale cross target and the pyramid method. Second， in the far range， several coding mark points are arranged in the measured space， and the external parameters are calculated based on the principle of single image intersection. Finally， all the internal and external parameters are optimized using bundle adjustment. To verify the feasibility and accuracy of the proposed method， a vision experiment with a large field of view is carried out. Experimental results show that the re-projection error is less than 0.08 pixels， the maximum absolute error of the three-dimensional measurement is 0.43 mm， and the pitch angle error of the rotor with a diameter of 10 m is less than 0.1°. It is therefore proved that the method can be used to achieve internal as well as external parameter calibration in the external field separately.

This paper proposes a 3D vehicle detection algorithm for unmanned driving systems to solve the problem of low accuracy in environmental perception based on lidar. First， according to statistical filtering and a random sampling consensus algorithm （RANSAC）， the ground point cloud segmentation was analyzed in order to eliminate the redundant points and outliers of the lidar data. Second， we improved the 3DSSD deep neural network to extract vehicle semantic and distance information from the point cloud through fusion sampling. According to the feature information， the candidate point position was adjusted twice to generate a center point. The 3D center-ness assignment strategy was adopted to create a 3D vehicle detection box. Finally， we divided the KITTI dataset into different scenes， to be used as experimental data， by comparing various current 3D vehicle detection algorithms. The experimental results showed that the proposed method could detect vehicles quickly and accurately. The average detection time was 0.12 s， and the highest detection accuracy was 89.72%.

To solve the problem of interfacial bonding strength between diamond-like carbon （DLC） films and metal substrates， this study uses direct current plasma enhanced chemical vapor deposition （DC-PECVD） technology to deposit composite DLC films under the same duration but with different negative bias voltage conditions on 45 steel substrates. A scanning electron microscope （SEM） and an atomic force microscope （AFM） were used to observe the morphology of the films， a Raman spectrometer was used to analyze their compositions， and a coating adhesion automatic scratch tester was used to determine the bonding strength between the films and substrates. The results show that when the bias voltage changes from -600 V to -1 200 V， the surface roughness of the films increases， the total film thickness increases （up to 16.3 μm）， the hydrogen content decreases， the graphite relative content increases， and the structural composition difference between the transition layer DLC and the top DLC film decreases. Meanwhile， the residual stress increases， the binding force decreases， and the maximum value is approximately 54.5 N. When the bias voltage of the top DLC layer is between -600 V and -800 V， the comprehensive performance improves.

This paper presents a multidimensional disturbance measurement platform based on a sensor array distribution to measure the multidimensional disturbance forces of heavy equipment. Based on a piezoelectric sensor， the platform adopts the vibration measuring strategy of a redundant array， which meets the measurement requirements of large loads with a high stiffness， and avoids the losses of measurement accuracy associated with structural coupling. In order to overcome the redundant measurement errors caused by array measurements， the measurement accuracy is optimized based on the generalized inverse method presented in this paper. Sensors in various positions are used as measuring units for the different vibration sources. A linear decoupling algorithm， using a full regression model， is used to obtain a more accurate expression of the three-dimensional force. This method avoids the systematic error introduced by redundant measurements and reduces the influence on the measurement results of the platform from vibration sources with different mechanical characteristics. Finally， the prototype system of the arrayed multidimensional disturbance force measuring platform was built， and the feasibility of the platform was verified by experiments. Experimental results showed that the system can guarantee a high load capacity and stiffness （the fundamental frequency of the prototype system is 1 174 Hz， with a load capacity of 416 kN）. The dynamic relative error of the three-dimensional generalized forces， within a frequency range of 8-800 Hz， is less than 5%. The device meets the increased precision measurement requirements for large loads with a high stiffness.

Current deviation decoupling control based on sliding mode active disturbance rejection （SADRC-CDDC） was designed for permanent magnet linear synchronous motors （PMLSMs）， which were affected by current decoupling and parameter variation. The d-q axis cross-coupled branch was introduced at the difference point between the reference and actual currents. The current deviation decoupling controller （CDDC） was then designed by establishing the current control dynamics with the coupling term and calculating the coupling term for system compensation， which weakened the influence of the d-q axis current coupling. However， the decoupling value cannot be eliminated when the value of inductance varies. Therefore， the sliding mode active disturbance rejection controller （SADRC） was used to solve the disturbance caused by the parameter variation and to compensate for the system； thus， the system can realize approximately complete decoupling. From the theoretical analysis， the controller satisfied asymptotic stability， and the robustness of the system was verified. The effectiveness of the designed SADRC-CDDC scheme was proved by the system experiment result. Compared with CDDC， SADRC-CDDC can achieve a more robust performance as the maximum oscillation amplitude of the d-axis current is reduced by 34.88%-54.76%， and the maximum oscillation amplitude of the q-axis current is reduced by 47.83%-71.43% when the system is affected by current coupling and parameter variation.

As the aperture of large ground-based optical telescopes increases， wind disturbance has become one of the most significant dynamic factors degrading the performance of telescopes. To investigate the influence and interaction principles of wind disturbance on the performance of telescopes， a time-history simulation of the wind disturbance response and performance prediction for telescopes were performed in the time domain. First， the structural parts of the telescope were simply introduced， and a dynamic model of the telescope was established using the finite element method. Using modal transformation， the dynamic model， expressed in nodal coordinates， was transformed into modal coordinates， thus greatly decreasing the model dimension and greatly improving the computational efficiency. Second， a time-history simulation method for wind speed was presented based on two-dimensional stochastic fields. The wind speed field in the doom was expressed as a two-dimensional stochastic field that varied according to the temporal and spatial frequencies. The numerical instability occurring in the Cholesky decomposition of the cross-power spectrum matrix in the spectral method was avoided by introducing the wave-number spectrum. The simulation efficiency was further improved by applying the fast Fourier transform （FFT） technique. Finally， using a ground-based telescope with a 2 m aperture as a case study， a time-history simulation of the wind speed， wind disturbance response， and system performance prediction for the telescope was performed. The simulation results showed that wind disturbances with a mean speed of 10 and 15 m/s would result in a maximum surface root mean square error of the primary mirror of approximately 45 and 70 nm， respectively. In addition， the wind disturbance acting on the secondary mirror and truss would mainly cause optical axis angle and position errors.

Optical cavity ring-down spectroscopy with a Hz-level response rate is introduced to detect trace gases in the atmosphere. By loading a 100 MHz modulated sine wave signal on the electro-optic phase modulator that generates sidebands and using the mixer to extract the first harmonic generated as the error signal by the beat frequency of the carrier and sidebands after passing through the 3 m gas absorption cell， the frequency locking of a 1 572 nm distributed feedback laser on the 6 361.25 cm-1 hyperfine transition line of carbon dioxide gas molecule was achieved. The ring-down time of the cavity without gas and the ring-down time with gas absorption were simultaneously measured using the wavelength-division multiplexing method， and the detection limit of the system obtained was 4.82×10-10 cm-1 on a 330 mm optical resonator. The system has a good linear response in a large carbon dioxide concentration range， and the linear correlation coefficient is greater than 0.999. The long-term observation results of the system are highly consistent with the data from Picarro commercial instruments， and the deviation between the two is less than 1.0%. The system proved the feasibility of locking the laser frequency to the molecular hyperfine transition line using the first harmonic and applying it to the optical cavity ring-down spectroscopy system to achieve a fast trace gas detection.

A dual energy imaging inspection method is proposed， which uses dual layer flat panel detector （FPD） to acquire X-ray dual energy images under single exposure， and dual energy subtraction is performed. Single layer， dual layer FPD structures are introduced， and the principle of dual layer FPD imaging is detailed as well. Different and same tube voltage with different filters induced X-ray spectra are simulated， and dual energy image characteristics of chest phantom are analyzed by performing kVp switching， dual layer FPD imaging tests. The results show that dual energy images captured by kVp switching， dual layer FPD can represent attenuation difference of soft tissue， bone with respect to low， hight energy X-rays. Dual layer FPD method owns advantages of lower exposure dose and no motion artifacts compared to kVp switching. Finally， dual energy image registration method for dual energy FPD is proposed， and bone enhanced and bone suppressed images are acquired by performing dual energy subtraction after registration. Experiment results show dual energy subtraction images using dual layer flat panel detector can realize better visualization for doctors’ diagnosis， and the subtracted images present better contrast than kVp switching method does.

In order to solve the problem that the schlieren method cannot accurately measure the weak signal region of the sidelobe beam in large laser devices of the national large scientific facility， a new schlieren method based on the diffraction inversion of the sidelobe beam is proposed to measure the far-field focal spot for high-power laser. The key point of this method is that an indirect measurement approach is used based on reverse deduction， while deducting along the reverse direction of the optical path propagation. The diffraction intensity image and phase image of the sidelobe beam are the inputs to calculate the far-field focal spot distribution of the front sidelobe beam， which is not shielded. Compared with the traditional far-field focal spot measurement based on the schlieren method， the improvements and optimizations proposed in this paper are as follows. First， the mathematical model of far-field focal spot measurement using the schlieren method is improved to reveal the rationality of the model， theoretically based on the principle of diffraction inversion of the sidelobe beam and the indirect measurement approach. Then， the feasibility of this method is verified by simulating the whole experimental process of high power laser far-field focal spot measurement， which consists of sidelobe beam diffraction， denoising， sidelobe beam diffraction inversion， and focal spot reconstruction. Finally， the improved DnCNN algorithm is used to remove the noise of different levels （0-75 dB） of 12 bit scientific CCD images in the mainlobe and sidelobe beams， and the reconstruction accuracy of far-field focal spot is improved. The experimental results show that this method not only eliminates the influence of the schlieren sphere on the diffraction of the sidelobe beam but also obtains the real intensity distribution of the sidelobe beam in the weak signal region， including the important parameters of far-field focal spot measurement， such as the amplitude and position of each peak of the side lobe beam， the dynamic range ratio， the amplitude and position of each peak of the sidelobe beam， and the ratio of dynamic range. The error between the reconstructed and theoretical focal spots of the dynamic range ratio is 3.20%. It is significant to improve the reliability and experimental accuracy of the far-field focal spot of high-power laser measurement using the schlieren method.

The controlled dispersion of a magnesium fluoride crystal resonator was studied to optimize the performance of the whispering gallery mode （WGM） resonator and the quality of generated optical frequency combs （OFCs） based on the WGM optical resonator. Firstly， the effect of the edge shape of a MgF2 crystal resonator on the cavity mode field and total dispersion was investigated by theoretical simulation. Thereafter， two kinds of MgF2 crystal resonators were fabricated according to the simulation results， namely plane edge and unilateral wedge types， respectively. The resonator performance detection and crystal resonator optical frequency comb generation systems were established， and the quality factor （Q factor） of the processed MgF2 crystal resonator samples was recorded up to 1.1×108. Finally， the Kerr optical frequency comb， with a wide spectral range of more than 200 nm， was excited in both the planar crystal resonators. The experimental results verified that the edge-wedged micro-resonator structure could effectively compress the mode field and regulate the total dispersion. In addition， a wider spectral range of the OFC could be produced in comparison with the edge-flat micro-resonator.