Abstract:The recent system construction, data acquisition and application, system status monitoring and fault prediction, verification and validation of the PHM technology are summarized. Then, according to the current engineering application of equipment PHM technology, the development and application bottlenecks of military aircraft PHM technology are analyzed from the aspects of equipment test design and verification, multisource fault information acquisition, interaction and management, remaining service life prediction and maintenance decision of aviation parts, verification and validation and standard development. And it is pointed out that the research on the technologies of testability demonstration, intelligent information processing and fusion, standardized management of aviation equipment and information, data mining and platform construction and so on should be carried out. Finally, some research ideas and development trends of domestic PHM are given.

Abstract:A realtime online estimation algorithm on the adhesion performance parameters of the rail surface is presented for the problems such as low adhesion utilization, easy idling and easy slipping of the locomotive in the operation of heavy duty locomotive. Firstly, based on the analysis of locomotive adhesion behavior, Kiencke adhesioncreep model is selected as the identification model. Then, the algorithm uses model parameter identification framework under the significance of the maximum likelihood to transform the parameter estimation into solving quadratic programming problem, and the iterative algorithm for identification is constructed. At the same time, considering that the rail environment mutation cannot be measured, the timevarying forgetting factor is introduced into the identification algorithm to adapt to the switching of rail surface environment. The simulation results show that the algorithm is able to track the change of wheel rail environment timely, and identify the parameters of adhesion performance effectively.

Abstract:To reduce the time complexity of the traditional SIFT algorithm related to the feature points detection and matching, an optimized image registration algorithm is proposed. Namely the feature points are extracted by Trajkovic algorithm, and the SIFT algorithm distribution descriptor method is adopted to allocate feature points descriptor parameters. Then sparse dimension reduction principle is used to reduce the feature points descriptor dimensions. Finally, the similarity measure algorithm based on bidirectional matching is used to match them. Six evaluation criteria including the feature points detection of image, matching pair, correct matching pair, matching accuracy, registration time and registration time drop rate are adopted in the simulation experiments. The simulation results show that the optimization algorithm in comparison with traditional SIFT algorithm is equivalent in the accuracy of feature point matching. However, it has obvious improvement in terms of feature points matching speed.

Abstract:Compared GJB151B RE102 test method lower than 30 MHz frequency band with that of GJB151A, the rod antenna grounding network and the height are changed greatly, which directly influences the accuracy of measurement results. To research the influence of the changed test method on measurement results, simulation and experimental methods are utilized, the different measurement results caused by different grounding network and the height of rod antenna are analyzed based on the circuit theory and electromagnetic theory. Finally, the paper concluded that the radiated emission results by GJB151A standard are 5~10 dB higher than that of GJB151B in 2~30 MHz frequency band. These measurement results also show that the GJB151A test setup will lead to resonance phenomenon. Furthermore, the power cord is the main reason that RE102 measurement results exceed the limit line at low frequency. The research will promote the implementation of GJB151B, help EMC engineers to understand the impact of grounding for EMI test and analyze RE102 measurement results of electronic equipment. It also helps them to handle results exceeded limit line and abnormal results, and to avoid giving misjudgment about radiated emission of electronic equipment.

Abstract:The requirement for the reliability of the chip in the area of biochemical analysis is very strict.In order to ensure the precision of complex biochemical experimental analysis results, the contamination should be cleaned.In this paper,the online contamination removal based on Maxmin ant colony algorithm has been proposed to realize reducing the time of washing contamination.Establishing MTSP model for digital microfluidic biochips contamination,establishing tabu judgement strategy based on fluid and time constraints and the selection strategy of the ant colony algorithm to achieve the purpose of optimizing the path planning of wash droplet and cleaning up contamination.The experiment results show that the scheme can reduce the time of clearning contamination,comparing with MCC algorithm it can reduce the used cells effectively.

Abstract:Aiming at the problem that the accuracy of localization algorithm based on RSSI ranging of indoor positioning is low due to the obstructions indoor, this paper proposes new localization algorithm based on distance ranging and nonranging. Based on the results using the original positioning algorithm based on RSSI ranging, it corrected the distance vector from signal reader to signal source, which was calculated by original location algorithm, and the characteristic information on the interference of the room internal barrier was imported to the signal propagation and took it as reference. Considering the relationship between the obstacle and the position of the label, it also corrected the positioning result by improving weighted centroid location algorithm based on potential obstacles iterations. The positioning accuracy of the method is increased by 50% compared to the method of scheduled, in which the influences of the information of obstacles is not consider.

Abstract:With the wide application of nonlinear loads, the disadvantages of the method of decentralized harmonic suppression at the terminal are gradually revealed. To achieve the configuration of active power filter in the middleandsmall sized grid, the isolated niche particle swarm optimization is adopted in this paper. Firstly, the configuration model of APF is built, the objective of which is to minimize the rms of node THD. Besides, the constraint conditions are executed by the penalty function. The modified algorithm has a better performance for global and local optimization and avoids the problem falling into local optima. The steps of allocation are elaborated, and the effectiveness of the proposed method is proved by the examples, the method can optimize the location and capacity of APF synchronously and achieve the power quality improvement for single or multi harmonics sources system. As a result, the THD of node voltage is less than 4% and the harmonic additional losses reduced by more than 90% after the improvement.

Abstract:A novel parallel Laguerre blind equalization algorithm for common data link (CDL) is proposed to overcome the problems of traditional algorithms such as long tap coefficients and the high bit error rate (BER) for deep fading channels. By analyzing CDL signals, the new algorithm replaces FIR filter with Laguerre filter so as to approach longpulse response sequence with shorter tap coefficients. Meanwhile, the convergence is accelerated by using the decisiondirected algorithm and fractionally spaced structure. The selfadaptive step module directly adopts the expected value from the square of equalized output as a modified variable to further promote its equalization effect. The experimental results show that the proposed method can accelerate convergence and reduce BER in short tap coefficients.

Abstract:Islanding detection is an important guarantee for the safe operation of photovoltaic grid connected system. At present, there are a lot of improvement researches on slidemode frequency(SMS) and OFP/UFP islanding detection method（IDM）, but the disadvantage of large nondetection zone (NDZ) of OFP/UFP is not changed when the matching degree of load consumption power and inverter output power is high, the disadvantage of large influence to power quality of SMS is not changed. A mixed islanding detection method based on SMS and OFP/UFP is proposed. The NDZ of OFP/UFP and the influence to power quality of SMS is decreased. The malfunction of the system which is caused by the abrupt change of the system was avoided. It can switch different IDM algorithms when the load is changed. Islanding detection experiment is carried out in MATLAB＼Simulink and photovoltaic microgrid experiment system. The experimental result verifies the correctness of theoretical analysis.

Abstract:To realize the automatic adjustment controlling the height of the roller in the cutting process of the shearer, a dynamic identification method of coalrock interface based on acoustic emission signal is proposed to test and extract the acoustic emission characteristic signals. Using timedomain analysis method gets the time domain indexes under the different conditions, using wavelet analysis method extracts the coalrock cutting acoustic emission signal characteristic value. The different coalrock cut ratio under the condition of acoustic emission signal optimal membership function is established to dynamic monitoring coal and rock ratio, then the purpose of coalrock interface identification is achieved. The results of laboratory tests show that there is little difference between the trajectory of the roller and the coalrock, and the experimental results verify the accuracy and reliability of the recognition system.

Abstract:Aiming at the problem of opencircuit fault arising in diodeclamped threelevel inverter, a new fault diagnosis method based on decision tree support vector machine (DTSVM) is proposed. Taking the inverter state as an example, firstly, the operation conditions of main circuit in inverter are analyzed to classify faults. Then, in terms of the multiscale decomposition of wavelet analysis, the middle, upper and down bridge voltages are selected to extract the fault features, respectively. Moreover, particle swarm clustering algorithm is built to construct the DTSVM classify model, and the multimodel fault diagnosis of power component in threelevel inverter is finally accomplished. The simulation results show that this method in case of less classification model to complete fault diagnosis, comparing to other methods such as back propagation neural network, oneversusone support vector machine and extreme learning machine, the diagnostic accuracy up to 98.46% for multimode fault diagnosis of threelevel inverter in 10% white noise, which indicate that the algorithm has better accuracy and robustness.

Abstract:Aiming at the problem that the equipment of wireless sensor network ranging distance is short and the positioning accuracy is not high, MDSMAP positioning algorithm based on pseudo random code correlation is proposed. Firstly, the ultrasonic signal emitted from the node is coded by using the pseudo random code correlation detection technology, the node distance and the accuracy are effectively increased. For a node failed to measure the distance, which is processed by using the Euclidean and the shortest path fusion algorithm, and then relative coordinate of node is generated by using the relative coordinates of MDSMAP algorithm. Finally, the final coordinate of node is obtained by using the plane transform model. The experiment results of simulation show that the improved algorithm can achieve a higher positioning accuracy and smaller positioning error under different network size and distance error conditions.

Abstract:To achieve high precision of linear frequency modulation continuous wave (LFMCW) radar ranging, a kind of LFMCW radar ranging method is presented based on the ImprovedRife (IRife) algorithm. At first, it introduces the principles and implementation steps of this method. And then, it implements this algorithm on FPGA of EP4CE55EF23A7N model. Finally, the algorithm embedded into the LFMCW radar ranging system is performed in field test. The simulation and experiment results verifying the effectiveness of this method show that IRife algorithm reveals its small rootmeansquare error (RMSE) and high ranging precision and stability compared with the Rife algorithm under the condition of the same signaltonoise ratio (SNR).

Abstract:For paralleled active power filter (APF) into the power grid, the control strategy of double closed loop soft start is proposed to solve the problem that the capacitor voltage of DC side is started and the current impact of compensation end is presented. The new PI fuzzyPI control method adopting a DC output voltage deviation square values ring in the outer voltage is used for more accurate and stable buffer starting the capacitor voltage. The proportional resonant (PR) control method is used in the current loop, and an improved harmonic detection method is also used. Finally, a simulation experiment is carried out in the MATLAB software, and the method is used to test the feasibility of the 60 kVA prototype. Comparing with the conventional control algorithm, the method obtaines better control result on the inhibition of impulse voltage and impulse current end in the APF grid, it has good prospects for engineering applications.

Abstract:The problem of multiparameter, nonlinearity and complex mechanism in process industry limits the performance of shallow algorithm, and therefore the deep learning is introduced into the processing industry predictive modeling. However, feature mining is insufficient for diversity data by single deep network, so an improved stacked autoencoder is proposed. Firstly, attribute of input is divided into several classes by clustering algorithm, and then the parallel sparse autoencoder is entered to detect feature locally. The parallel output is integrated into the following deep networks to extract feature layer by layer and get the fitting results. To overcome predictive error due to the overfitting, the “dropout” technique is introduced. In the prediction modeling of hydrocracking, the presented algorithm has better prediction level and generalization ability.

Abstract:Existing fitting algorithms for photovoltaic (PV) module correction IV curve under standard condition suffer from several problems, such as overcomplicated calculation and lack of accuracy. To improve fitting accuracy under condition deviated from STC, a fitting algorithm based on solar cell fourparameter model was proposed. The module’s IV mathematical model for curves fitting is calculated by the algorithm through selecting six points as fitting points near the maximum power point. A selfadaption fitting point selection method also was presented for accuracy improvement because different PV modules have different optimal interval among fitting points. Moreover, a type of IV tracer employing capacitors array as load to increase number of data points near the MPP was developed to improve fitting point’s accuracy. Several MATLAB simulations and actual measurement were carried out, which show that the proposed algorithm fitting results were superior to that of polynomial fitting algorithm. And the designed IV tracer not only has excellent performance but is suitable for engineering surveying.

Abstract:In order to monitor the freshness of banana in the process of storage and transportation, an electronic tag type banana releasing gas sensor which includes tag antenna and gas sensitive film is designed. At resonance frequency of 915 MHz, a bowtie tag antenna is designed with electromagnetic simulation software. The multiwalled carbon nanotube is selected as the gas sensitive material by comparing the complex impedance of five kinds films. With the incorporation of a small amount of conductive carbon paste, the impedance regulation of the gassensitive film is successfully resolved, and the impedance matching between the antenna and the gas sensor is achieved. The gas sensitive testing platform is established to test the response characteristic of the sensor for banana releasing gas at different maturity. The experimental results indicate that the tag gas sensor has a good response and recovery characteristic for banana releasing gas. The resonant frequency change of sensor is 9, 36, 18 MHz when banana is in fresh, ripe, and rotten period respectively. Therefore, the freshness of banana can be qualitatively identified.

Abstract:There are few existing studies which are directly suitable for the mine locomotive wireless network communication model. According to the characteristics of mine locomotive constantly moving, the mine locomotive dynamic power scheduling method based on the interference management is proposed. First, according to the traditional mine locomotive communication process, a dynamic power scheduling model based on SIC is established. Tomake the model can solve in polynomial time, SIC set and maximum SIC set concept is proposed, and an algorithm for solving all the biggest SIC set is designed to effectively reduce the complexity of the solving problem. The simulations show that the improvement rate of the optimization effect of the methodthan ordinary SIC method isof between 20%~35%, andthe improvement rate isof between 40%~50% compared to traditional methods of no SIC.

Abstract:To improve the ability of robot path planning in complex environment, a hybrid path planning method based on improved QPSO and Morphin algorithm is proposed. Firstly, using grid map establishes the environmental model and determines the starting point and the target point. Through the introduction of adaptive search strategy and crossover operation, the QPSO algorithm is improved to plan an optimal global path, the robot is walking according to the global path. When the unknown or dynamic obstacles are found, the Morphin algorithm is used to carry out the local path planning to avoid the obstacles and return to the original global path to continue walking for the target point. The simulation of Matlab and practical application verify the effectiveness and feasibility of the hybrid path planning method.

Abstract:Aiming at the imperfections of traditional artificial potential field (APF) method such as vibration, dead zone, local minimum, improved artificial potential field (IAPF) method with a combination of robot position, speed, accelerated speed, and obstacles’ position is presented. Taking advantage of the Morphin algorithm, which is high efficiency and easy to combine the global path planning algorithm, multilayer Morphin search tree algorithm is presented. First, IAPF method is used to complete the path of global planning. On this basis, the multilayer Morphin searching tree algorithm is used for path planning nearby obstacle. MATLAB simulation tests show that the mixed algorithm with IAPF and multilayer Morphin search tree algorithm can effectively shorten the path length and improve the efficiency to reach the target in the autonomous mobile robot path planning, efficiently complete path planning.

Abstract:To solve the location problem of the timefrequency sound signal of equipment fault under the high noise environment when using algorithm of GCC, the timefrequency subspace delay (TFSD) estimation algorithm which combine shorttime Fourier transform(STFT) domain and image analysis is proposed. The characteristics of noise and signal in STFT domain is used, meanwhile the correlation information of different microphone is considered, the principle signal subspace is formed using the method of filter or spectral subtraction to separate area of signal and noise subspace. Based on signal subspace, the signal in time domain is inversed, then the endpoint detection is undertaken, the delay is estimated and the sound source is located. The simulation results demonstrate that the time delay can be estimated accurately even the amplitude of additive noise is more than two times of signal, the estimation error of delay and location in area of 5 m×5 m, this results are proved by measurement.

Abstract:An online adaptive balancing system is described in this paper. This system is mainly used for measuring and correcting imbalances of load adaptively in high throughput screening platform. Firstly, an online adaptive efficient balancing scheme for this system is proposed according to the characteristics of load of high throughput screening platform dynamic changes. Then, considering the system composition and structure, the method of the improved narrow band tracking filter and dynamic balance correction is proposed, meanwhile the adaptive control strategy is investigated. Finally, the experimental device is developed and the system is debugged to illustrate the effectiveness of the proposed method. The testing results show that the system of experimental device is feasible and valid.

Abstract:The modulation on stator current signal of induction motor under the situation of load torque vibration and gap eccentricity is analyzed respectively. In view of that the modulation results in these two cases are difficult to be distinguished in frequency domain while the vibration frequency and rotor frequency are consistent, the Wigner timefrequency domain analysis method is used to calculate the phase difference of sidefrequency current. The work in this paper points out the difference of modulation effects between load torque vibration and gap eccentricity in timefrequency domain, which provides theoretical basis for motor airgap eccentric fault diagnosis.