Abstract:Based on the characteristics of non - contact high measurement accuracy and ease of use, laser triangulation distance measurement has been widely used in scientific research and industrial production practice. Displacement measurement is determined from the shift of laser spot on an imaging system′s photo sensitive surface . Different application normally requires selection of different polynomial fitting coefficient terms between the two. Proper selection of polynomial terms is very important for realizing high speed and high accuracy in measurement. In this paper, regarding the non-linear relationship between the measured surface displacement y of laser triangulation displacement measurement and the displacement x of the imaging spot on the photosensitive surface, systematically analyzed the error sources due to polynomial fitting of y as a function of x. A miniature laser triangulation displacement sensor were built, Based on the experimental measurement data, it is verified that the laser triangle displacement measurement device satisfies the Scheimflug condition. According to the experimental measurement values and the displacement values calculated by polynomial fitting, a quality factor Q is defined to judge the quality of the fitting, and the optimal number of polynomial terms is determined by calculating the Q factor.

Abstract:Proposes an approach to measure the numerical aperture (NA) of the infinite conjugate objective. It utilizes the total internal reflection (TIR) phenomenon occurring on the air-coverslip interface to generate a reference on the back focal plane (BFP), and the NA is measured based on Abbe sine condition. According to the principle, a home-designed measurement system is developed, containing the optical system, the control system and the whole image-processing software to extract the NA value from experimentally acquired BFP images. The algorithm of the image-processing software utilizes the two-dimensional Fourier correlation analysis and grayscale statistics. The NA values of several objective lenses are measured experimentally. The results indicate that the relative error of the proposed method is no more than 1% , much lower than that of the other method compared in this work. The proposed method features on the simple measurement configuration and fast and automatic measurement of the NA of practical oil-immersion objective lenses. The results show high precision and good repeatability. The approach works for the NA measurement of the infinite conjugate oilimmersion objective lenses.

Abstract:Aiming at the problem of traceability and calibration of the laser inertial confinement fusion capsule profile high-precision measurement system, proposes a calibration and traceability method based on the principle of laser differential confocal measurement. Based on the property that the zero-crossing point of the axial response curve of the laser differential confocal capsule measurement system accurately corresponds to the focus point of the measurement system, the method firstly uses the laser differential confocal capsule measurement system to measure the roundness of the standard elliptical blocks calibrated by National Institute of Metrology, China. Secondly, through comparing the roundness measurement value with the roundness metrological value of the standard elliptical blocks, the measurement transfer coefficient of the system is obtained, which is 1. 03. Finally, through multiple verification measurement methods, the high-precision calibration of the system is completed. The experiment results show that using the calibrated system to conduct the comparison measurement of the laser fusion target metal capsule, the standard deviation is 37 nm; the measurement repeatability of this calibration method is 17 nm, which lays a solid foundation for the high-precision measurement of the surface profile of the capsule.

Abstract:The conventional NIR spectrometers based on fixed grating and array detectors have disadvantages of high cost and narrow spectral range. To address these issues, the dual single detectors NIR micro spectrometers are proposed, which are based on integrated scanning grating micromirror. This system employs the integrated scanning grating micromirror as the core component, which can realize the scanning and diffraction simultaneously. The spatial layout of two focusing mirrors and dual InGaAs single detector are utilized to avoid the mutual interference between different optical paths in the optical system design. It can realize the simultaneous and independent operation of two optical paths. In addition, a band-pass filter with different cutoff wavelengths is established in front of the dual-path detectors to eliminate spectral overlap. A theoretical model is formulated by the ray tracing method to calculate the structural parameters of the optical system. The optimized design of the optical system is achieved by using ZEMAX and the optimal structural parameters are given. Test results show that the operating wavelength range is 800~ 2 532 nm. Resolution is smaller than 12 nm (800~ 1 600 nm) and 17 nm (1 600~ 2 532 nm), and the overall size is 145 mm×135 mm×75 mm. This instrument has the advantages of wide spectrum, small volume, low cost and other advantages.

Abstract:Building prediction model to predict and compensate thermal error is a common method to solve the problem of thermal error of machine tools. In this method, the prediction accuracy and robustness of the model are easily affected by the environmental temperature, so a robust thermal error modeling algorithm based on partial least square method is proposed. Firstly, the correlation coefficient method is used to screen the temperature sensitive points, and the partial least squares regression prediction model of thermal error is established. Then, based on the multi batch thermal error experimental data under the annual ambient temperature, the optimal number of temperature sensitive points is analyzed. Finally, the partial least squares regression model of thermal error is established and compared with the ordinary multiple linear regression model. The results show that the average prediction accuracy of the proposed algorithm is 5. 7 μm, and the robustness of the model is 0. 56 μm. Compared with the ordinary multiple linear regression algorithm, the prediction accuracy and robustness are improved by 13. 8% and 49. 5% respectively. It shows that the thermal error robust modeling algorithm proposed in this paper can maintain high prediction accuracy and high robustness when the ambient temperature changes greatly.

Abstract:A large amount of heat is generated during the working process of high-speed motorized spindle, which leads to thermal deformation of the front end of the spindle. The machining accuracy of the spindle is affected seriously. In this study, the thermal error prediction model of the motorized spindle with variable load preload is proposed. The heat transfer theory calculation and the bond diagram model are combined. In this way, the thermal elongation of the motorized spindle with similar structure can be predicted accurately in real time. The thermal-solid coupling model of the spindle is formulated, and the distribution of temperature field of the spindle under coupling action is achieved by the finite element analysis. According to the distribution law of temperature field and the flow direction of heat energy, the integral model of the spindle is divided into nodes by the thermodynamic theory. The thermal parameters of the nodes are calculated. The 20-sim platform is used to establish the thermal bond diagram model of the spindle, and the real-time temperature monitoring module of the key nodes is associated with the thermal elongation module. According to the thermal elongation of the nose end, the thermal deformation of each key point is calculated and the thermal error modeling is completed. Experimental results show that the prediction model error of the formulated bond graph is within 0. 5 μm. Research results can provide a theoretical basis for real-time thermal error compensation of intelligent motorized spindle under different working conditions.

Abstract:When the measuring distance is selected, it is difficult to balance the accuracy and efficiency. To address this issue, an optimization method of measuring point based on error source for hole shaft straightness detection is proposed. The error source of real part is analyzed. Then, the dimensional error of surface shape of part is obtained. According to the engineering semantics of the error source and straightness tolerance, the simulation function of the surface shape of the part is established. Based on the principle of error theory, the relationship between the measured distance and the error evaluation value is analyzed, which is based on the simulation function. And the point set accuracy function is formulated. For the given tolerance and machining method, a series of simulated surfaces are randomly generated based on the Monte Carlo method simulation principle. By using the precision function of the set of measured points, the best measuring distance of each simulated surface is analyzed and counted. In further, the best measuring distance of real parts is analyzed. Finally, two engineering examples are implemented to evaluate the accuracy and effectiveness of the proposed method. Compared with the theoretical method, results show the number of axial and radial measuring points in the first example is reduced by 1 994 (226) and 42 (396), respectively. The proposed method improves the measuring efficiency under the condition of satisfying the accuracy.

Abstract:Firstly the indication error main source of the angle measurement system of circular grating ( eccentricity and tilt of grating disk) is studied theoretically, and the coaxial installation method of grating disk and rotating shaft is proposed. The continuous full circle angle standard device in NIM ( the measurement uncertainty is 0. 05″) is used to directly perform traceability test of the measured circular grating angle measuring system, which avoids the instrument errors introduced in the installation process of the circular grating and the indirect traceability test process ( the calibration using combined polyhedral prism and photoelectric autocollimator) . Secondly, the distribution of eccentricity, tilt and other order errors in the spectrum is analyzed using harmonic theory. Finally, the harmonic compensation is carried out for the errors caused by installation eccentricity and tilt. The experiment results show that the angle indication error of the angle measurement system is reduced from 100“ before compensation to 10” after compensation, which effectively eliminates the stable and repeatable error harmonic components caused by the installation eccentricity and tilt of grating disk.

Abstract:Geometric error and thermal error of the five-axis CNC machine tool are two important factors that affect the machining accuracy of workpiece. The analysis of these error factors can effectively improve the machining accuracy of thin-walled workpiece. Firstly, based on the homogeneous coordinate transformation method, the geometric error model of rotating axis of five-axis CNC machine tool with double turntable is formulated. Then, on the basis of standard ball, the machine contact measurement is carried out. Twelve geometric errors of the two rotating axles are identified. These errors consider the mutual influence between the two rotating axles and the influence of thermal error. Finally, the geometric error field in the machining space of five-axis CNC machine tool is analyzed. In this machining space, the geometric error gradually increases from the center to the outside. When the rotation angle of axle increases, the maximum error also increases. Compared with other position error identification methods, the measurement accuracy of the proposed method meets the processing requirements. The measurement time only needs 30 min.

Abstract:Hysteresis is one of the main sources of measurement errors for magnetostrictive displacement sensors ( MDS ). The measurement accuracy of the sensor can be significantly improved by reducing hysteresis. According to the hysteresis characteristics of ferromagnetic materials, the formation process of the displacement hysteresis is analyzed. To address large hysteresis in the short stroke of MDS, the Preisach model is used to describe the input-output relationship of hysteresis system which represents MDS. Then, a software compensation method for hysteresis on this basis is proposed, which linearizes the hysteresis curve piecewise to calculate initial iteration value. The search range of convergence value is determined according to the steady-state error. To solve the problem of the large deviation between the theoretical model and reality within a short distance after the magnetic ring turns, the convergence condition is adjusted near the turning point to improve the accuracy of the compensation method in this area. Experimental results show that the method requires only a few iterations. After compensation, the hysteresis of MDS is reduced to about 1 / 3 of the original value. The nonlinearity is also slightly improved. Without changing the existing sensor structure, this method can quickly and effectively reduce the displacement hysteresis of the sensor. It provides a new solution for the hysteresis compensation of MDS.

Abstract:Presents a high precision multi-turn absolute time-grating angular displacement sensor, which is based on the re-modulation scheme. The sensor consists of a single-turn absolute sensor and a multi-turn counting module. Firstly, a two-stage single row timegrating angular displacement sensor is used to form a single-turn absolute sensor. Four orthogonal traveling wave signals are constructed through the first stage sensor. And two differential traveling wave signals are directly reflected as precise measurement signals. Meanwhile, the first four orthogonal traveling wave signal is utilized as the excitation signal of the second sensor for re-modulation to achieve the whole cycle single cycle signal, which is used as the rough measurement signal. The precise measurement signal is combined with the rough measurement signal to realize the high-precision single-turn absolute angular displacement measurement. On this basis, combined with Wiegand multi-turn counting module to record the turn number information, high-precision multi-turn absolute angular displacement measurement is realized. The principle prototype is made by PCB technology. The experimental platform is established, and the experiment is carried out. Experimental results show that the original accuracy of the sensor in the whole cycle range can reach ±8″. And the accurate memory of the number of turns is realized in the case of power on and power off.

Abstract:In order to avoid the effects of speed and current sensor faults on the speed loop, current loop and suspension control in the vector control system of bearingless asynchronous motor (BAM), a fault-tolerant control strategy of the BAM sensor faults is proposed. Firstly, the extended observer is designed to realize the observation of the speed and error of the BAM, and the data fusion strategy is used to achieve smooth switching of the speed when the fault occurs. Then a reasonable threshold is designed to achieve fault diagnosis and fault tolerance of the speed sensor according to the speed observation error. Secondly, the current observation is realized according to the state equation, and the expansion observer is used to compensate the observed disturbance of the state equation. Then a reasonable threshold is designed to realize the fault diagnosis and fault tolerance of the current sensor according to the current observation error. Finally, the simulation and experiment results show that the proposed fault-tolerant control system for the BAM sensors can accurately achieve fault diagnosis and fault-tolerant control of speed and current within 0. 025 s, and the errors are within ±63 r/ min and 58 r/ min, respectively. In addition, the rotor displacement fluctuation amplitudes are within ±38 μm throughout the process, which indicates that the rotor can achieve good suspension.

Abstract:In view of the application requirement of surface nondestructive testing of ferromagnetic workpiece such as hot rolled steel pipe and oil well pipe in service, a magnetic bridge displacement sensing method and a sensor array are proposed to realize magnetic imaging detection of surface morphology defects with certain spatial resolution. In this method, the sensor and the workpiece form the magnetic circuit together. The sensor uses a special compact array structure of magnetization, which could obtain better performance due to varied specifications. Firstly, with a time-sharing method, it reduces the interference between adjacent measuring points, and pulse magnetizes each single displacement measuring magnetic circuit in the array. Secondly, the magnetic field in each measuring magnetic circuit is sampled at intervals. The displacement value of corresponding positions is calculated. Through theoretical calculation and simulation, this study analyzes the resolving power under different basic lift-off, in the array sensor testing morphology distortions like grooves. The spatial resolution of imaging the measured surface is analyzed under the influence of the system′s scanning and sampling parameters. By experiment evaluation, the measuring range of this sensor array is 0. 2 ~ 7. 0 mm, and the B-y sensing characteristics fitting curve′s linearity is about 1% . In comparison, the pre-existing H-shaped magnetic bridge displacement sensor is in the range of 1 ~ 5 mm, and the linearity is about 2% .

Abstract:As a core component of in-vitro coagulation sensor structure, the sensor failure is easily caused by the deformation of its elastic support. To address this issue, this paper proposes a reliability life assessment method, which is based on the fatigue life and breakage mode of the sensor. Failure analysis is performed on the sensor reliability according to the operating principle and structural features of the in-vitro coagulation sensor. By formulating the mechanical model for theoretical reliability analysis, the expression of determination criterion for the life limit of coagulation sensor is derived, which is attributable to the strain transfer coefficient of elastic support. The harmonic response of the elastic support is analyzed by the finite element analysis. An elastic support-based method for assessing the ultimate sensing life is proposed by taking the 15th level of sensor fatigue life as the failure threshold. By using this method, 108 million stress fatigue tests are implemented on the dual-channel in-vitro coagulation sensor. The full-scale relative effective errors of spiral clips for the elastic support are 3. 89% , 3. 79% and 4. 75% , respectively. Under the 4. 5% boundary of the full-scale relative effective error, the results indicate that a clip has reached its stress life limit, while other two clips are yet to reach the safety threshold level. Thus, the fatigue life reliability of the sensor is approximately 108 million times, which basically agrees with the simulation data. These results verify that the proposed life limit assessment method for the in-vitro coagulation sensor conforms to the requirements of sensor failure analysis, which also offers a core technique for improving the development procedure of such sensors.

Abstract:Proposes an on-orbit non-uniformity correction method for complex scenes, in order to achieve high-frequency and highprecision relative radiation calibration for mechanically interleaved stitching time delay integration charge coupled device ( TDICCD) cameras on orbit. This method performs orthogonal secondary imaging before and after the camera rotates around the yaw axis, and uses the side slither imaging data and conventional imaging data to respectively calibrate the probe elements within and between arrays. Combining the characteristics of the wide radiation dynamic range of complex scenes, the method of multi-point curve fitting is used to calculate relative calibration parameters and finally achieve the purpose of non-uniformity correction. By using simulation imaging data to perform relative radiometric calibration experiments on uniform scenes and complex scenes, these results show that the image non-uniformity NU coefficient which is processed by the orthogonal secondary imaging calibration of the complex scene(OSICCS) in the non-uniform scene is 0. 499 1% lower than that after the common calibration method, meanwhile, the noticeable striping artifact and residual noise are removed effectively.

Abstract:To solve the problem of low efficiency and error prone of manual feeding of the crankshaft bearing caps (CBCs), the visual location and detection method of CBC feeding robot based on attention mechanism is studied to realize automatic feeding. Aiming at the unapparent image features, the attention mechanism is introduced into the feature extraction network of Faster R-CNN to map the weights of the CBC image at different positions to the feature channel, so that the deep learning model can pay more attention to the edge and center semantic information of the CBC. To further improve the location accuracy, this paper also improves the candidate box generation method and loss function. Experiment results show that compared with those of traditional machine learning methods and classic deep learning target detection models, the detection speed of this method reaches 0. 419s, the location accuracies are the best (IOU and GIOU are 0. 941 3 and 0. 940 9, respectively). In addition, the proposed method possesses good robustness. On site test shows that the success rate for the method guide the feeding robot to grasp and place the CBCs reaches 95. 14% , which improves the efficiency of the engine assembly line.

Abstract:The hand-crafted features in the traditional anomaly detection algorithms can not represent the appearance and motion patterns in a unified way for different scenes. In this paper, we propose a novel anomaly detection algorithm based on the convolutional variational auto-encoder (ConVAE). Firstly, a ConVAE which takes the raw frames series as input is constructed to extract the deep features of the scene. These deep features can represent the appearance and motion patterns more specifically. And then multiple Gaussian models are employed to fit the deep feature vectors of the corresponding receptive fields. The fitted Gaussian models which correspond to the receptive fields are used to decide the deep feature of the corresponding receptive fields from the test sample is anomalous or not. The proposed anomaly detection algorithm is evaluated the UCSD anomaly detection datasets. Experimental results show that the area under curve (AUC) of the proposed method are 95. 7% and 69. 9% in frame-level and pixel-level, respectively.

Abstract:To solve the problem of the deviation between the maximum position of the multi-scale correlation response map and the true position of the target, a polygon centroid based position correction method for correlation filter tracking is proposed. First, we proposed a correlation response map evaluation index to evaluate the quality of the response map, and make a credible judgment of the tracking result of the current frame. Then, we use the centroid of the polygon vertex to correct the tracking result that is judged to be unreliable to reduce the tracking result and target deviation from the correct position. Finally, the performance evaluation experiment of the algorithm in this paper is carried out on the three benchmark video sets of OTB50, OTB- 2015 and UAV20L. The Area Under Curve of this algorithm in OTB50, OTB- 2015 and UAV20L reached 0. 625, 0. 668 and 0. 429 respectively, and the tracking accuracy reached 0. 844, 0. 885 and 0. 578 respectively. The results show that compared with the mainstream tracking algorithms in recent years, our algorithm has achieved better success rate and tracking precision in a variety of complex scenarios.

Abstract:In order to realize accurate simulation and real-time adjustment of the road surface adhesion coefficient (RSAC) during vehicle braking simulation test, an automobile braking simulation test-bench (ABSTB) was designed which is capable of simulating different RSACs in real time by controlling the excitation current of an magnetic powder clutch. Then, a simulation model of a single-wheel automobile braking system based on road surface recognition (RSR) was built, and braking simulations were performed under the single road surface and the jumping road surface, respectively. An experimental system for the automobile braking simulation was developed to carry out the automobile braking simulation experiment on a single road surface and the tracking control experiment of the RSAC under the jumping road surface, respectively. Research results show that the braking distance reduces by 3. 1% under the RSR-based optimal slip ratio condition compared with that under the fixed target slip ratio condition for braking on the wet asphalt road surface at an initial speed of 120 km/ h. The phenomenon becomes more obvious under the road surface with a low RSAC. Moreover, experimental values of vehicle speed and wheel speed basically consist with simulation values under the single road surface. The maximum tracking error of the RSAC is only 6. 2% , which proves that the tracking control effect is satisfactory under the jumping road surface condition.

Abstract:The performance-based logistics is a new support strategy, which has been widely applied in aviation, defense, energy, etc. Based on the functional check model for the single component system with two-stage failure process, this paper establishes the revenue rate and profit rate function through associating the supplier′s revenue and availability. The maintenance decision model is formulated based on the performance-based logistics. Then, the effectiveness of the model is evaluated through a numerical example of steam turbine blades. Compared with the cost minimization, results show that the profit and availability of the proposed policy are increased by 2. 19% and 0. 079% , respectively. Compared with the availability maximization, the profit is increased by 26. 61% and the cost is reduced by 13. 35% . These results prove that the proposed policy can achieve multi-objective requirements at the same time. The profit could be maximized and the system performance could be improved at a lower cost. Finally, the sensitivity analysis of maintenance cost and downtime is carried out, which is helpful to determine the optimal inspection interval and profit rate.

Abstract:The contact force fluctuation of the pantograph-catenary system greatly affects the current quality when the train is running at high speed. To address this issue, an active control method of pantograph is proposed, which is based on contact force prediction of long short-term memory network. The advantage of long short-term memory network for time series prediction is fully utilized. Firstly, the pantograph model with two lumped masses is taken as the research object. Its dynamic equation is established, and the contact force fluctuation data is obtained by simulation of the model. Then, the contact force data obtained from simulation are utilized to train the long short-term memory network. In this way, a prediction model is formulated to predict the contact force at the next moment. Finally, the difference between the predicted value and the expected value of the contact force is taken as the target control force output to the magnetorheological damper. The magnetorheological damper provides the control force to act on the pantograph. The dynamic fluctuation of the contact force is suppressed and the current quality of the train is improved. Experimental results show that the proposed method is more accurate in the control of the contact force. The standard deviation of the fluctuation of the contact force is significantly reduced by more than 70. 13% . The offline situation of the pantograph-catenary system is avoided, which verifies the stability and superiority of the proposed method for improving the current quality of the pantograph.

Abstract:Electric vehicles have advantages of simple operation and sustainable development, which have been widely used. Obstacle avoidance is a very important part of the autonomous driving of electric vehicles. It is of great significance to the development of autonomous vehicles. This paper proposes an autonomous obstacle avoidance control algorithm based on the genetic algorithm (GA), which solves the problem that traditional algorithms can only avoid obstacles horizontally. Based on the genetic algorithm and cluster behavior rules, three corresponding velocity terms are designed by kinematics and dynamics. The obstacle avoidance ability of the proposed algorithm is evaluated by the fitness function designed in this paper. Finally, simulation results of Ubuntu show that the final fitness is 0. 92, 0. 87 and 0. 8 when the corresponding self-starting speed is 4, 6 and 8 m/ s, respectively. Four of the six fitness functions are exactly 1 (completely collision free). It is concluded that the proposed autonomous obstacle avoidance control algorithm has advantages of high stability, simple operation and strong obstacle avoidance ability.

Abstract:Aiming at the characteristics of high reliability, long service life, complex working environment, and single test stress factor of airborne fuel pump, a fuel pump degradation test platform under complex stress conditions was designed and built. And the influence of stress factors was analyzed. Firstly, through the analysis of the failure mechanism of the fuel pump, the electrical stress and mechanical vibration which affect bearing wear were selected as the main stress to study its performance degradation. Secondly, the degradation test platform and vibration test device of fuel pump were built based on the selected stress. And the pressure sensor and flow sensor were selected. The signal acquisition and control system was introduced, the fuel pump clamping device was designed and its dynamic performance was analyzed. Finally, the experimental scheme was designed based on the orthogonal experimental idea, and the test results were analyzed through range analysis and variance analysis. The number of tests was reduced by 2 / 3. It was concluded that voltage has more significant influence on the reliability of fuel pump and the confidence level was up to 99% .

Abstract:Considering the complex nonlinear characteristics of chemical process faults and the underlying structural characteristics of samples, a fault identification method based on manifold regularized stochastic configuration network is proposed. Based on classical stochastic configuration network, this method randomly selects hidden parameters under the supervision mechanism of embedded manifold constraints to add hidden nodes one by one. Then, the output of hidden layer weights is calculated by manifold regularized least square method. It keeps the important geometric characteristics of data. The information redundancy is avoided and the relevant characteristics of different from different categories could be identified. Experimental results on test set show that the identification accuracy values of TE fault and semiconductor fault are 87. 72% and 84. 27% , respectively, which are higher than those of random vector function connection network and stochastic configuration network. In addition, for most fault types, the precision and recall rates of the proposed method are high. Results prove that the proposed method can effectively identify faults. The generalization ability of fault identification model is improved.

Abstract:The high correlation between single fault and composite fault samples, resulting in misclassification. Moreover, rotating machinery often works at different speeds, which further increases the difficulty of composite fault diagnosis of rotating machinery. Aiming at the above problems, a composite fault diagnosis method of rotating machinery at different speeds with one-dimensional depth subdomain adaptation was proposed. Firstly, frequency domain signals of composite faults of rotating machinery are used as the input of the network to get rid of the dependence on signal processing and professional knowledge; Secondly, a domain shared onedimensional convolutional neural network was built to learn the frequency domain signal characteristics of composite faults of rotating machinery at different speeds; Then, the local maximum mean difference is added to form the sub-domain adaptation layer, which aligns each pair of sub-domain distribution to avoid the feature mixing of single fault and compound fault, and reduces the feature distribution difference of the two subdomains by minimizing the local maximum mean difference to reduce the interference caused by different speeds. Finally, softmax classification layer is added after the sub-domain adaptation layer to realize fault state identification of the target data. The effectiveness of the proposed method is proved by the composite fault diagnosis experiments of rotating machinery at different speeds.

Abstract:The building structure and household behavior of end-users are different. The heating datasets of end-users have features of large amount, strong nonlinearity, long response time, etc. In the original data space, it is hard to implement anomaly detection by the clustering analysis. The problem is the serious data crossing that greatly reduces the accuracy. In this paper, the high dimensional Gaussian mixture clustering (HGMM) is proposed to map datasets in original space to high-dimensional space for clustering. Kernel function mapping, inner product, decomposition of high-dimensional feature space are used to improve clustering accuracy and avoid dimensional disaster. Industrial big data ingestion and analysis platform ( IBDP) is established. The clustering and anomaly detection accuracy of K-Means, Gaussian mixture model (GMM), constant false alarm rate, and HGMM are compared. The proposed method could improve the clustering accuracy to 90. 72% and reduce the detection error rate to 5. 92% . Four types of abnormal heating patterns are identified and analyzed. The proposed HGMM could be used to effectively analyze the residential heating characteristics, detect the abnormal datasets, help reduce the heating energy consumption, and realize the building energy saving.

Abstract:The identification of adhesion state of pipeline with anti-corrosion coating is a hot research direction of pipeline state pre diagnosis. The curved pipe is divided into several closely connected micro body element plate structures, the propagation model of nonlinear ultrasonic guided wave in micro body element structure is established, the dispersion characteristics and energy transfer characteristics of guided wave in micro body element and between adjacent micro body elements are analyzed, and the echo signal is analyzed by spwvd time-frequency analysis and wavelet packet decomposition algorithm, In order to extract the characteristic quantity which can represent the different bonding state of pipeline. In this paper, the PMMA aluminum double-layer bonding structure with similar performance to the pipeline with anti-corrosion coating is taken as the experimental object. The ultrasonic echo signals of the intact bonding state, the weak bonding state based on the density change and the partial debonding state based on the thickness change are collected respectively, and the corresponding relationship between the material parameters and the bonding state is analyzed, The recognition rate is 92. 31% .

Abstract:Aiming at the problem that the ultrasonic Lamb wave excited by the electromagnetic ultrasonic transducer ( EMAT) has multiple modes, a method is proposed to excite a single mode of the Lamb wave by optimizing the structure of the transducer. First, according to Biot Savart‘s law and Maxwell′s equations, the relationship between the shape of the permanent magnet and the magnetic field distribution is studied, and a new type of magnetizable “arch bridge” permanent magnet structure is designed. Then, the simulation experiment and the actual measurement experiment were performed on the EMAT before and after the optimization. The experimental results show that the S0 modal amplitudes of the Lamb wave excited by the EMAT after optimization are 1. 05 and 1. 1 before the optimization, and the A0 modal amplitudes are respectively It is 0. 22 and 0. 12 before optimization; the ratio of S0 mode to A0 mode is also increased from 2. 60 and 1. 82 before optimization to 12. 48 and 16. 17 respectively. It can be seen that, compared with the ordinary EMAT before optimization, the optimized EMAT can excite the S0 mode and suppress the A0 mode well, thus verifying that the new EMAT can effectively excite the single S0 mode of the Lamb wave.

Abstract:For mobile robot using ultrasonic sensors detect interference existing in the environmental outline and the data uncertainty problems, based on the analysis of working principle of the ultrasonic sensors and the adjacent position after the correlation characteristics of detecting data, three position is proposed based on ultrasonic environmental detection method. First by using ultrasonic sensors to build the interior environment figure; Then the improved strong tracking UKF-SLAM method filters the ultrasonic measurement data and the driving model of mobile robot and gets more accurate pose information and map features after fusion optimization. In this study, setting up a simulation environment and assembling an Omni-directional mobile robot equipped with ultrasonic sensor in an indoor experimental to verify the feasibility and accuracy of algorithm. The simulation results show that the error in simultaneous localization and mapping was reduced by 58. 058% that compared with the other algorithm. Furthermore, the average error of the robot′s acquisition of environmental features is reduced by 50. 2863% . The feasibility and effectiveness of the improved algorithm proposed in this paper are further verified, and the method has certain reference value for Simultaneous Localization and Mapping.

Abstract:In the production of polyurethane sandwich insulation board, void defects are easy to form in the polyurethane foam layer due to uneven foaming, and it directly affects the heat conservation effect. For this reason, a rapid detection method implementing electromagnetic ultrasonic SH guided wave inspection from the outside of the metal board is proposed. Through theoretical analysis, analytical derivation, simulation calculation and experiment verification, the relationship between the transmission reflection coefficient ratio of the SH guided wave in polyurethane sandwich insulation board and the excitation parameters is established. It is found that when the magnet spacing is 4. 5 mm and the incident angle of the SH0 guided wave is about 75°, the best detection effect can be achieved. Furtherly, it is verified that the amplitude of the received signal of the SH0 guided wave has linear relationship with both the defect area and defect depth basically. On this basis, a defect equivalent size evaluation method that could be applied in actual inspection is proposed. With the evaluation formula, the problem of defect signal evaluation fluctuation caused by the change of the distance between the receiving and transmitting probes can be eliminated, which has a good instruction for the later realization of defect location.