Abstract:The autonomous path planning and landing of unmanned aerial vehicle in open scenes has been the focus of research in related fields. A method based on the point cloud data collected by the lidar to recommend the optimal landing address for the UAV is proposed. This method corrects and expands the original point cloud data by using the UAV pose information and maintaining the site selection window, then the RANSAC algorithm is improved to evaluate and select the generated candidate planes, and finally output the coordinate information of the optimal landing site. The experiments in the simulation environment show that the results of this method are stable and accurate, and the computational speed satisfies the requirements of the UAV to work in real time. At the same time, the space and time cost of this method is light, which meets the standard of the operation in real scenes.

Abstract:A fusion algorithm combining Mahony and extended Kalman filter (EKF) is proposed to solve the problem of low accuracy of micro inertial measurement unit and large error of traditional attitude calculation method. First, the initial attitude quaternion is obtained by fusing gyroscope, accelerometer and magnetometer data with Mahony filter. Then, the attitude quaternion of the Mahony filter is used as the measurement value of EKF. According to the size of the non-gravity acceleration, the measurement noise covariance matrix is automatically adjusted by the positive correlation. The EKF equation of state is established according to the angular velocity information measured by the gyroscope. Finally, the attitude estimation of UAV is obtained after EKF filtering. The simulation results show that the static attitude angle error is less than 0. 1° and the dynamic attitude angle error is less than 1°, both of which are better than the complementary filtering algorithm and the improved EKF algorithm. The fusion algorithm can effectively suppress the gyro drift error, filter out the high frequency noise mixed with the measured value of accelerometer and suppress the interference of non-gravity acceleration, and improve the attitude calculation accuracy.

Abstract:Aiming at the problems of insufficient data reliability and resource waste in the decision of redundant data of UAVs, a compensation algorithm for UAV IMU multi-sensor redundancy based on BP neural networks is proposed. The low-precision IMU sensor data is input to the BP neural network, and the non-linear fitting capability of the BP neural network is used to compensate for errors in low-precision IMU data, then use data arbitration algorithm based on confidence to arbitrate multiple higher-precision data and output the sensor data after data fusion. This process can also judge and locate sensor faults. The singularity problem can be solved by changing the installation method of similar sensors. The experimental results prove that after neural network error compensation, the error is reduced by 55. 2%. Furthermore, the error after neural network error competition is 53. 9% smaller than the error after using the kalman filter algorithm for error compensation. The algorithm takes full advantage of redundant sensor design, improves the reliability of the sensor system.

Abstract:Airborne antenna installed on the surface of UAV is easy to become a lightning attachment point. Lightning will pose a serious threat to the flight safety of the UAV. The purpose is to carry out the relevant design and verify the effectiveness of the design. A UHF airborne antenna is against to the research object. The segmented lighting diverter strips and RF port lightning suppressor were designed for protection. The initial leader attachment test, the induced transients in external mounted hardware test, the electrical performance test and the system evaluation calculation were successively carried out on the antenna to study the effect response of the antenna in different lightning environments. The test results show that damage is reduced. In addition, the influence of the installation of the segmented lighting diverter strips on the electrical performance of the antenna was 0. 3 dB at most, and the RF lightning suppressor will hardly affect the SWR and gain. Finally, the link margin meets the requirement of more than 3 dB. The effectiveness of the design was proved.

Abstract:Aiming at the discharge problem of transmission facilities in UAV patrol line, the ultraviolet discharge detection system is designed by adopting the ultraviolet discharge detection link model and using photomultiplier. Ultraviolet discharge detection system is used for discharge detection experiment. The experimental results are as follows, the detected signal of the system is highly coincident with the discharge signal, and the intensity of the detected signal of the system is linearly correlated with the intensity of the discharge signal. The relative error between the peak value of the detected signal and the average value of the detected signal is less than 3%, that is, the strength of the detected signal is related to the detected distance and has nothing to do with the detected position. The theoretical value calculated by the model under different detection distances is consistent with the actual detection value, and the maximum relative error is 15. 7%, which verifies the correctness of the model. The results of this paper have some guiding significance for UV discharge detection of UAV in line patrol. Keywords:UAV patrol; UV disch

Abstract:In recent years, the real-time kinematic (RTK) high-precision satellite positioning technology with carrier phase difference technology as the core has developed rapidly in the field of surveying and mapping. This article is based on the RTK positioning principle, aiming at light and miniaturization, high precision, stable and fast satellite positioning receivers, the focus is on multi-system and multi-band Global Navigation Satellite System (GNSS) signal processing technology, including the GNSS multi-frequency RF frontend processing and baseband signal processing key technologies of the UAV RTK receiver. Through professional simulation methods, it can be concluded that the designed RF front-end receiving sensitivity is higher than -130 dBm. The execution time for searching and capturing 6 758 sampling points is only 0. 68 s, and the capture frequency shift error is about 0. 932% of the Doppler frequency shift, and the frequency error after the GNSS signal carrier tracking stabilizes is basically concentrated below 0. 75 kHz. The simulation results show that the designed GNSS signal processing module meets the requirements of the actual multi-frequency RTK positioning receiver.

Abstract:Aiming at the problems of poor spatial recognition and excessive training sample requirements of traditional convolutional neural networks used in power line patrol foreign object detection, an improved capsule network model is proposed. The gray-scale data and three-dimensional block matching filter algorithm are used to preprocess the line survey data set. The adaptive contribution pooling is proposed to reduce the amount of data information loss. The foreign object data depth information extraction unit extracts the main features to filter out redundant information, reduce the number of data to improve model performance, improve the foreign object recognition of main capsule layer and dynamic routing structure to adapt to power line patrol for the second classification of line foreign object detection. For adaptive contribution pooling and maximum pooling, non-pooling, traditional structure capsule network and improved capsule network, improved capsule network and AlexNet, GoogLeNet were respectively compared with foreign object recognition experiment and improved capsule network spatial recognition performance. The experimental results show that under 3 700 small training samples, after 20 trainings, the average accuracy of the improved contribution network of adaptive contribution pooling is greater than the maximum pooling by 2. 7%, and the improved capsule network is better than the non-pooling, traditional structure capsules. The average accuracy of the network is increased by 3. 6%, and the improved accuracy of the improved capsule network is 21. 9% and 12. 6% higher than that of AlexNet and GoogLeNet, respectively, and the improved capsule network still has an average accuracy higher than 91% in test data of different sizes and angles. The improved capsule network has high foreign object recognition ability under the condition of complicated spatial recognition and few training samples, and realizes high-efficiency and high-accuracy automatic unmanned line inspection foreign object detection.

Abstract:Heart rate variability analysis can play an important role in emotion recognition. In order to establish an accurate model between ECG and emotion categories, a feature selection method based on maximum information coefficient (MIC) is proposed. In this paper, the Aubt database and the design of emotional induction experiments are used for research. First, 40 features based on heart rate variability in time domain, frequency domain, nonlinear and time-frequency domain were extracted, then emotion modelingwas conducted based on the MIC method combined with support vector machine, random forest and K nearest neighbor algorithm. The results show that based on the MIC feature selection method,the classification accuracy of the Aubt database for arousal, valence, and four emotions is 90%, 89%, and 84%, respectively. And further choose Pearson correlation coefficient, ANOVA feature selection method to compare with MIC. In the induced experimental data,the correct classification rate ofmultiple one-to-one emotion recognition is higher than 75%. It shows that the MIC feature selection method can significantly improve the classification accuracy, which is of great significance for emotion recognition based on ECG signals.

Abstract:To further study the reliability problems of capacitive RF-MEMS switches in practical applications, a multi-core relevance vector machine ( MKRVM) method based on differential evolution quantum particle swarm optimization (DE-QPSO) is proposed to predict the switch lifetime. First of all, bandwidth restricted empirical mode decomposition (BREMD) is used to denoise the life data obtained during the experiment to improve the data reliability; secondly, DE-QPSO is used to obtain the optimal sparse weight of MKRVM, and the MKRVM algorithm is used to predict the life of such switches; finally, the actual data obtained by experiment is used to test the accuracy of the methods. The experimental results show that MKRVM can obtain the prediction results within 0. 21 s. The root mean square of the data is 3. 104 3×10 6 s, which is the closest to the original data of 3. 065 7×10 6 s; DE-QPSO can be optimized within 0. 45 s, the variance is 7×10 -5 . At the same time, it is concluded that the switch life is the longest when the elastic coefficient is in the range of 4~ 16 N/ m.

Abstract:LCD electronic instrument has been widely used in industrial production because of its accurate indication and strong antiinterference ability. According to the display characteristics and display principle of LCD, proposes an adaptive color morphology preprocessing method based on information entropy, and enhances the LCD region segmentation based on HSV space and hue mapping V channel, and combines the region image enhancement algorithm of dual channel image weighted fusion to realize the segmentation of LCD position region. The experimental results show that, compared with the traditional segmentation scheme, the segmentation accuracy of the proposed method is improved by 8. 6% for monochromatic LCD and 12. 4% for color LCD.

Abstract:One of the paramount issues in the field of VLSI design is the rapid increase in power consumption. When the input signal is interfered and glitches occur, the power consumption of the double edge-triggered flip-flop (DETFF) will increase significantly. In order to effectively reduce the power consumption, this paper proposed an anti-interference low-power double edge-triggered flip-flop based on C-elements. The improved C-element is used in this DETFF. One side, it effectively blocks the glitches in the input signal, prevents redundant transitions inside the DETFF, and reduces the charge and discharge frequency of the transistor. The C-element also adds pullup and pull-down paths, reducing its latency. Compared with other existing DETFFS, the DETFF proposed in this paper only flips once on the clock edge, which effectively reduces power consumption. The HSPICE is used to simulate the proposed DETFF and the other 10 DETFFs, AILP-DET only increased the delay overhead by 7. 58%, the total power consumption is reduced by an average of 261. 28% without input glitches, and the average power consumption is reduced by 46. 97% with input glitches. Detailed voltage and temperature variations analysis indicate that the proposed DETFF features are less sensitive to voltage and temperature variations.

Abstract:To meet the needs of human-computer interaction, a wearable measurement system for angle between fingers is proposed in this paper. Firstly, based on spring sheet and strain gauge, a wearable sensor for the measurement of angle between fingers and a portable measurement circuit are designed. The measurement principle of angle between fingers and measurement circuit of sensor are analyzed in detail. Secondly, computer software is developed in LabVIEW. Data captured by the developed sensor is filtered and then converted to angle information by the computer software. And the angle between fingers is synchronous dynamic displayed by means of image segmentation and rotation. Finally, experiments are carried out to verify the effectiveness of the system. The experimental results show that the designed angle sensor has a nonlinear error of 2. 38%FS, a hysteresis error of 1. 94%FS, a repeatability error of 5. 29%FS and a total accuracy of 6. 11%FS within the measurement range of 0 to 30°.

Abstract:Wireless local area network (WLAN) based positioningis one of the main stream techniques for indoor positioning. In this paper, a WLAN based method is proposed based on a variational auto encoder ( VAE), which contains a positioning path and a reconstruction path. This structure has enabled the ability to position along with reconstruction of RSS radio map (RM). The proposed methodis validated by open-sourced dataset, which shows that the proposed method can increase the positioning accuracy by about 14% and 24% respectively compared to the compressive sensing (CS) based methodand the traditional kNN method in the positioning path. Moreover,from the reconstructed RM, the method can have accuracy enhancement of about 11% in positioning than that from the CS reconstructed RM. Keywords:variational

Abstract:Due to the important parameters of Loran-C signal amplitude, secondary phase delay and deformation are closely related to the ground electrical parameters and terrain fluctuations, the Loran-C signal simulator based on complex terrain is designed. The groundwave attenuation factor at different frequencies is calculated using the integral equation method, which multiplied by the standard signal in the frequency domain, and after IDFT, encoding and modulation, an actual signal containing ground electrical parameters, terrain features, and time information are generated. Taking Hexian County ( transmitting station) to Xi’ an ( observation point) as an example, the actual signal amplitude is reduced by 1. 47 dB, the secondary time delay is increased by 0. 407 μs, and the signal is widened by about 1. 5 μs relative to the uniform and smooth path. The experimental results show that the Loran-C signal simulator based on complex terrain can provide a navigation signal closer to the actual transmission path for the development, debugging, testing and maintenance of Loran-C receiving equipment.

Abstract:For traditional calibrating method of the antenna angular reference of landing guidance radar, a large amount of electromagnetic radiations is unavoidable, the capacity of resisting disturbance is lower and the work is inefficient. An improved method is proposed. The measured antenna received small signal instead of transmitting large signal during the testing procedure. Three axis programmable turntable and servo controller were adopted at far field testing to control the horn antenna move precisely, and working with the near-end operational software. The horn antenna can be remotely controlled between the distance of the 150 to 250 meters. Peak-seeking automatically and data logging and storage were realized as well. Compared with the traditional calibrating method, the electrical radiation reduced to 1 ∶ 850 000,the testing stuff reduced from 4 to 2 and the testing time reduced from 5 hours to 0. 5 hour. The improved calibrator which is proved by the experiments can replace the old one completely.

Abstract:Aiming at the problem of real-time condition monitoring of wind turbine, a real-time monitoring method based on incremental relative entropy was proposed. Firstly, based on the analysis of the characteristics of sliding window data, the formula of incremental relative entropy which is suitable for real-time calculation was derived. And the time complexity of incremental relative entropy is O(1), which is lower than O(n) of conventional calculation method. Next, a real-time wind turbine condition monitoring method based on datadriven and normal behavior modeling was proposed and the incremental relative entropy was used as the index of real-time residual analysis. The effectiveness of the proposed method was verified by the actual gearbox fault data of a 2 MW wind turbine. The results show that relative entropy residual analysis can realize fault warning at least 8~ 10 days in advance, which is better than the conventional statistics. The calculation time of incremental relative entropy was only 0. 4% ~ 1. 9% of the conventional calculation method, which has significant advantages in real-time performance.

Abstract:Aiming at the problem of cross interference in the analysis of bearing vibration signals by Wigner-Ville distribution (WVD) method, a time-frequency analysis method based on improved matching pursuit (MP) algorithm and WVD is proposed. Firstly, based on the analysis of bearing vibration signal, the basis function of dictionary atoms in the MP algorithm is determined, and atomic parameters are determined by correlation filtering method to complete the dictionary construction. Then, fast Fourier transform ( FFT) is used to calculate the cross-correlation spectrum between the signal and the atoms in the dictionary, and instead of the inner product operation in MP algorithm, the signal has the sparse representation, and the frequency parameters in the dictionary are updated according to the spectrum of the residual signal during the iteration process. Finally, the atoms in the sparse representation results are calculated by WVD, and the time-frequency representation of each atom is linear overlapping with the corresponding atomic coefficient as the weight to complete the time-frequency analysis of the signal. The experimental results show that it effectively improves the computing speed of MP algorithm, and compared with other three improved WVD algorithms, the time-frequency aggregation degree of the time-frequency analysis results of this paper method is higher, which can better overcome the cross interference term in the WVD method, and provide a new solution for the time-frequency analysis of the rolling bearing signal.

Abstract:Due to the limited battery capacity of smartphones, the power consumption of applications is one of its main consumptions, because applications may produce unexpected power consumption or energy bugs, most of which are design bugs. In order to understand the characteristics of application power consumption and traffic, a model-based method of power consumption and traffic analysis is proposed. Taking the hardware components of the mobile phone as the research object, the time automata model is extended, and the model checking tool UPPAAL is used to build the model. This method can be used in the early stages of development and provides a formal model of asynchronous power consumption for application design and analysis. Designs an example of WiFi power saving mode for research and analysis, verifying and analyzing the main attributes of the model, and then uses the mobile phone application QQ for experimental comparison. The analysis results show that the model conforms to the general model verification requirements. Compared with the measurement results of PowerTutor, its relative error is less than 7%, which can provide reference-based modeling and analysis idea for complex mobile phone systems.

Abstract:In order to locate the weak and small target accurately in the complex environment, according to the Gaussian shape characteristics of the target, the infrared dim small target detection algorithm based on intensity gradient mapping coupled with multi direction Median filter is designed in this paper. Firstly, according to the average intensity of infrared target in four different directions, the classical median filter was improved to effectively suppress the noise in complex background. Then, based on the central pixel of the small and weak target, the intensity information of the whole infrared image is obtained. The infrared image was divided into four sub blocks along the radius direction, and the polar coordinate system of each sub block was established to calculate its corresponding gradient value. According to the ratio of the maximum to the minimum gradient, the gradient information of the whole infrared image was obtained. Then, the intensity and gradient information were fused to get the background suppression image for enhancing the infrared dim target. Finally, the non-zero pixel mean value in intensity gradient mapping was used to calculate the threshold value for segmenting the background suppression image and locating the small and weak target accurately. The test data show that compared with the existing infrared dim small target detection technology, under the complex background interference, this algorithm has higher detection accuracy which can identify the target completely, and it presents a more ideal ROC curve

Abstract:In order to improve the accurate extraction rate of the traditional complete ensemble empirical mode decomposition with Adaptive Noise ( CEEMDAN) for motor bearing fault characteristic signals and reduce the distortion of the reconstructed signal, an improved CEEMDAN algorithm is proposed. The original signal is initially decomposed using traditional CEEMDAN to obtain several feature components (IMFs) and intrinsic modal components. Some IMFs are de-noised and extracted by entropy weight method. The filtered IMF components are secondary decomposed and secondary screened to obtain typical fault sensitive signals. Then the signal reconstruction is carried out by using SG ( Savitzky-Golay) smoothing filter and the motor bearing signal is de-noised. Finally, the performance of the improved algorithm is analyzed by using the data of Case Western Reserve University. The results show that the method can effectively reduce the signal noise of the motor bearing signal, and its SNR is improved by 2. 2 dB compared with the original signal.

Abstract:This study proposed a target azimuth estimation method of synthetic aperture radar ( SAR) images based on combination of sparse representation and collaborative representation. The sparse representation and collaborative representation reconstructed the test sample under the sparsity constraint and the minimum error constraint, respectively, which have good complementarity. The highly correlated training samples were selected by sparse representation and collaborative representation, respectively. And the two sets of training samples were intersected to find the most stable part. According to the true azimuths and coefficients of these samples, the target azimuth of the test sample can be estimated based on a proper weighting fusion algorithm. Experiments were investigated on SAR images of three targets from the MSTAR dataset and comparison was made with some present methods. The results showed that the estimation precision, stability and noise-robustness of the proposed method outperform some existing algorithms.

Abstract:In order to improve the accuracy and robustness of lane detection, as well as reduce the influence of illumination change and background interference, an improved Hough transform coupling density space clustering algorithm for lane line detection was proposed. Firstly, the lane line model is established, and the lane boundary is decomposed into a series of small line segments, which are represented by the least square method. Secondly, the improved Hough transform was used to detect the small line segments in the image. A noisy density based spatial clustering of applications with noise was introduced to cluster the extracted small segments, filter out the redundancy and noise in the image, and retain the key information of lane boundary. Then, the gradient direction of the edge pixels was used to define the direction of the small line segments, so that the small line segments on the same side of the boundary have the same direction, while the two small line segments on the opposite lane boundary have the opposite direction. Through the direction function of the small line segments, the candidate clusters of the lane segments were obtained. Finally, according to the candidate clusters, the vanishing point was used to fit the final lane line. It was tested in Caltech data set and the actual road, the data shows that compared with the current popular lane line detection algorithm, under the bad factors such as illumination change and background interference, this algorithm presents more ideal accuracy and robustness, which can accurately identify the normal lane line.

Abstract:Aiming at the problem that the complex and diverse background in insulator images makes it difficult to extract insulator regions in practice, proposes an improved PCM clustering algorithm to segment the insulator image. To improve the algorithm from two aspects,firstly, by defining the local correlation factor, introducing the spatial local information of the image enhances the anti-jamming ability and improves the segmentation accuracy, secondly, adding the repulsive term of class center to the loss function alleviates the central point overlapping problem of traditional PCM. In the experiment, using artificial data and images of insulator with complex background to compare the proposed algorithm with FCM, PCM, K-means, KFCM and IFCM clustering algorithm. The results show that, the improved PCM has stronger anti-interference ability and higher clustering accuracy, which has better segmentation performance for insulator images than other contrast methods, and the average segmentation error is 0. 153.

Abstract:If an elderly person lives alone and accidentally falls indoors or outdoors, she / he will not receive timely medical assistance and treatment, which will cause huge psychological and physical harm. In this paper, the attitude and heading reference system (AHRS) is used in fall detection, and the fall process is transformed into a change in attitude to judge the degree of fall. The AHRS attitude and heading system has been improved from the previous three experimental attitude angles, the pitch angle is increased by 2. 371%, the roll angle is increased by 9. 238%, the yaw angle is increased by 4. 682%, the maximum accelerometer accuracy is 0. 171g. The system fusion extended Kalman filter has improved the detection time calculation of the elderly by 1 second compared with the original experimental scheme in the past. In summary, the experiment proves that the AHRS attitude and attitude fusion can detect the fall of the elderly with more accurate data and less calculation time, which is a step closer to study the real fall posture of the human body.

Abstract:Estimation of lithium-ion battery remaining useful life (RUL) is the key to lithium-ion battery health. Achieving accurate and reliable remaining useful life prediction of lithium-ion batteries is very vital for the normal operation of the battery system. Proposes a lithium-ion battery RUL prediction method based on the combination of complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) and support vector machine-regression (SVR). First, a measurable health factor is extracted during the discharge process, and the correlation between health factor and capacity is analyzed by Pearson and Spearman methods. Then, the health factor is decomposed by CEEMDAN to obtain a series of the relatively stable components. Finally, the health factor decomposed by CEEMDAN is used as the input of SVR prediction model, and the capacity is used as the output, so as to realize lithium-ion RUL prediction. The lithium-ion battery data published by NASA PcoE is used to carry out simulation experiments, and compare it with the standard SVR model, the experimental results show that the proposed method can effectively verify the effectiveness of the proposed RUL prediction model, and the prediction error is controlled below 2%.