Abstract:In view of the shortcomings of the single variable prediction method of photovoltaic (PV) power, a new multivariable phase space reconstruction prediction method of PV power is designed. Firstly, based on the correlation analysis, the historical PV power and meteorological factors time series of the actual PV power plant are selected to form multivariate time series. Then, the multivariable phase space of PV power prediction is reconstructed by C-C method and false nearest neighbors ( FNN) method, and its chaotic characteristics are identified by small data method. Finally, combined with the powerful nonlinear fitting ability of radial basis function (RBF) neural network, a PV power prediction model based on multivariate phase space reconstruction and RBF neural network is established. The example analysis shows that the proposed multivariate phase space reconstruction prediction method has better performance than the single variable prediction method.

Abstract:Aiming at the problem of improper ash blowing operation on the heating surface of coal-fired boilers, this paper proposes a soot blowing optimization method based on Gamma acceleration soot deposition model with the aim of minimizing soot blowing cost. In this paper, firstly, in the case that the ash accumulation state on the heating surface of the boiler is continuously monotonically increasing, a continuous non-negative random variable obeying the exponential distribution is used to express the ash velocity of the boiler heating area, and an accelerated ash accumulation model is established; Secondly, under the circumstance the ash process is regarded as a similar process to the Gamma model. An optimized model for ash blowing on the heating surface of the boiler is established to minimize the average cost. Taking the ash state of the heating area of a 300 MW coal-fired economizer as an example, the PSO algorithm is used to optimize the average cost of ash blowing, obtaining the best soot blowing times and soot blowing thresholds on the heating surface of the boiler, so that making the average cost of the soot blowing on the heating surface of the boiler smallest. It also illustrates the feasibility of the minimum cost rate optimization model proposed in this paper.

Abstract:In order to solve the problem that mobile robots fall into local convergence and cannot achieve the optimal path in the path planning of ant colony algorithm, this paper proposes an improved variable-step ant colony algorithm to enable it to achieve the path with fewer convergence iterations optimal. According to the relevant characteristics of ant colony algorithm that applied in path planning, it optimizes the allocation of pheromone, reduces the impact of local pheromone content on the algorithm, avoids the ant colony from falling into the local optimum when searching the path, adds the weighting factor in the transition probability formula and increases the probability of the mobile robot moving in the direction of the end point, it effectively reduces the number of ant colony convergence iterations, changes the mobile robot’s moving step length, enables it to move freely and without collision within 360 °, and effectively shortens the path length. The simulation results show that: in the simple environment, the convergent iteration times and the optimal path length of the improved variable step ant colony algorithm are 2 times and 28. 042 m respectively, while the convergent iteration times and the optimal path length of the traditional ant colony algorithm are 25 times and 29. 213 m respectively. In the complex environment, the convergence iteration times and the optimal path length of the improved variable step ant colony algorithm are 2 times and 43. 960 2 m respectively, and the convergence iteration times and the optimal path length of the improved potential field ant colony algorithm are 16 times and 45. 112 7 m respectively. The simulation results demonstrate the effectiveness and superiority of the improved variable step ant colony algorithm.

Abstract:In the traditional sliding mode observer based sensorless field oriented control scheme for permanent magnet synchronous motor (PMSM), the fundamental wave of the observed back electromotive force ( back-EMF) has high frequency harmonics and chattering problem, and the observed rotor position is not accurate enough. Focused on these problems, a PLL adaptive sliding mode observer based sensorless control method for PMSM is proposed. Firstly, under the premise of Lyapunov stability, adaptive rule of back EMF is established and deduced. It makes the error of the observed back-EMF decline rapidly by constructing the adaptive sliding mode observer. Meanwhile, the rotor position is obtained by introducing the rotated influence immune phase-locked loop, which eliminates the influence of rotation speed change and further improves the observation accuracy. Finally, the experimental results based on a 2. 9 kW PMSM show that the proposed method can effectively suppress the chattering of the sliding mode, reduce the high frequency harmonics of the back-EMF, and improve the observation accuracy of the rotor position.

Abstract:The initial resource allocation optimization problem of automatic manufacturing system is to complete the preset production task under the premise of the minimum use of initial resource. To solve this problem, a method based on labeled Petri nets and integer linear programming is proposed. First, according to the structural characteristics of labeled Petri nets, the constraint relationship between the proposed task and the initial resource is given, and the initial resource allocation optimization problem is abstracted into an integer linear programming problem. Secondly, the software Lingo is used to solve the integer linear programming problem. Finally, an example is given to verify the proposed method. The experimental results show that the proposed method is simple and efficient, which can be used for reference in the initial resource allocation of the actual automatic manufacturing system.

Abstract:The method of multiple inverters paralleled is presented to meet the need of large current and fast control for the plasma vertical displacement power supply in the experimental advanced superconducting Tokamak. The current sharing control strategy based on gray prediction model in voltage mode is proposed to overcome the inconsistent output current of each branch, which caused by the different parameters of each branch. Current rising rate of each branch of next time can be predicted by this model and adjust PWM duty cycle dynamically in real time, thus current sharing control in voltage mode is achieved. Simulation and experimental results show that the current sharing control strategy based on gray prediction model can get relatively consistent output current of each branch, and the feasibility and the effectiveness of this strategy in EAST fast control power supply are verified.

Abstract:In order to solve the problem of global planning of multi-unmanned vehicle coverage paths in irregular areas, a regional coverage method based on stepwise genetic algorithm is proposed. First, the target area is rasterized according to the size of the cleaning vehicle, and the multi-vehicle area coverage path planning problem is transformed into a multi-travel agent (MTSP) problem. Then, the multi-traveler problem is solved by using the stepwise genetic algorithm. The first step is to transform the multi-traveler problem into the multi-traveler (TSP) problem by using the fuzzy C-means clustering method. In the second step, a stepwise genetic algorithm is used to solve each single traveling salesman problem, and the selection mechanism of the genetic algorithm is improved by using the idea of neutron parent coexistence of weed invasion algorithm. Finally, simulation experiments are carried out in the simulated campus scene and community scene respectively. The experimental results show that the proposed method in the two scenarios can achieve multi-unmanned cleaning vehicles to complete the regional path coverage, and the proposed step-genetic algorithm has a faster convergence rate than the grouping genetic algorithm. In campus scenarios, the proposed stepwise genetic algorithm is 54% less time-consuming and 38% less optimal solution path length than the grouped genetic algorithm. In the cell scenario, the proposed stepwise genetic algorithm reduces the time consumption by 55% and the optimal solution path length by 44% compared with the grouped genetic algorithm.

Abstract:The parameter uncertainty of permanent magnet synchronous motor (PMSM) drive system and the nonlinearity of inverter will reduce the robustness of motor speed and lead to the drastic fluctuation of electric vehicle speed. In order to improve the response performance and robustness of PMSM drive system for electric vehicles, a model-free adaptive backstepping control ( MF-ABC) is proposed. First, based on the algebraic estimation method of model-free control, an ultra-local model disturbance estimator is established. Second, based on the adaptive backstepping control framework, MF-ABC control suitable for high reliability operation of PMSM drive system is constructed. The results show that the proposed controller can observe, estimate and eliminate all kinds of uncertainties, including unmodeled dynamics and external disturbances. The disturbance value estimated by MF-ABC can be used to describe uncertainty and analysis error. The proposed scheme improves the calculation efficiency in real time and achieves the research goal of high performance and robust operation of the vector controlled PMSM drive system MTPA mode. Through dSPACE bench test, compared with traditional backstepping control, the proposed MF-ABC achieves 22% improvement in the torque disturbance suppression capability of PMSM and 0. 02 s increase in the step response speed. The experimental results show that MF-ABC has strong disturbance suppression capability and fast response performance.

Abstract:Aiming at the problem of correlation among multiple degradation variables in the reliability evaluation of fiber optic gyro (FOG), a storage reliability evaluation method of fog based on Copula function is proposed. Firstly, the performance degradation data of fog is obtained by accelerated stress degradation test, and the Wiener process is used to model the performance degradation data, and their degradation model and life distribution are obtained. Then, the copula function is used to fuse the life data of each performance parameter to get the reliability index. Finally, the method is verified by combining the performance degradation data of fog. In ambient environment of 55 ℃ , the storage life of fog is 38 795 h when considering the correlation between parameters and 17 485 h and assuming the performance parameters are independent. The results show that the reliability index obtained by considering the correlation of performance parameters is more reasonable. At the same time, the method effectively solves the problem of multi parameter reliability evaluation of products, and has better practicability and popularization.

Abstract:To investigate the problem of small flow measurement, a micro Coriolis mass flowmeter is introduced which is designed with diversion pipe crossed. Liquid in two different tubes flows in the opposite direction. The measurement value of Coriolis mass flowmeter is the difference between the two tubes’ flow. A phase-frequency matching method for frequency estimation, phase difference estimation is discussed in detail. Smoothing method for the estimation results is also put forward. Experimental research is carried out under three different basic flows. On this basis, the zero stability, linearity and measurement error of the flowmeter are analyzed. The experimental results show that the zero point of the differential flowmeter has a great relationship with the basic flow. The instability of the zero point is estimated to be 0. 539 g / min, and the accuracy of the full range is within 0. 5%. The mass flow has a remarkable linearity with the time difference.

Abstract:Aiming at the problem that the geometric features used for 3D point cloud registration, a combination of DAISY descriptor and LBP descriptor is used to extract the image features of the overlapping area of adjacent measurement stations, and the positions between the coordinate systems of adjacent measurement stations are solved. The transformation matrix is used to transform the 3D point cloud data of multiple measurement stations into the same coordinate system. First, the edge detection and the construction of DAISY descriptors are introduced. Then, the feature points of adjacent pictures are matched by Euclidean distance, and the coordinate transformation relationship at different stations is calculated based on the relationship between the matching points. The experimental results show that the method can well realize the coarse stitching of 3D point cloud data without using other auxiliary tools, and provide a theoretical basis for the application of point cloud registration technology in the fields of 3D reconstruction and reverse engineering.

Abstract:Broadband network measurement is an important means to obtain the electrical characteristics of microwave and millimeter wave devices, and the TRM calibration method is widely used in the calibration of broadband network measurement systems. In order to effectively evaluate the accuracy of the measurement data, it is necessary to analyze the uncertainty of the TRM calibration and measurement. The principle of small-signal calibration and error correction based on 8-term on wafer measurement is explained. Based on this, the uncertainty propagation formula of the TRM calibration method is pushed, which can effectively evaluate the non-ideality of each calibration kits. Effect of calibration results. The verification experiment was performed by using a transmission line with known parameters as a non-ideal straight-through calibration kit. The experimental results show that the uncertainty propagation formula can accurately evaluate the error of the measurement data, and the maximum difference does not exceed 0. 02 dB.

Abstract:In order to timely and accurately estimate the number of occupants in a room, and then to understand the occupancy distribution for various areas in a building, an addressable active infrared intrusion detector using an IP address as its identification was designed in this paper. Differently from the traditional infrared detectors that were always attached to the host, the detector was achieved to communicate with the host computer through TCP sockets, rendering one host could receive messages from multiple detectors through network configurations. Our designed addressable active infrared intrusion detector composed of active infrared intrusion detector, microprocessor and wireless communication module. The active infrared intrusion detector collected the data during the occupants’ entering and exiting process; the microprocessor processed the perceived signals; the data was transmitted to the host computer by the wireless communication module. Furthermore, based on the detector, an occupant counting program accounting for the entering and exiting was developed. Experiments results shown that the addressable active infrared intrusion detector could accurately and timely detect changes in occupants and transmitted messages reliably to the host computer, and the occupant counting program could accurately estimate the number of occupants in a given area.

Abstract:Aiming at the problem of high eddy current loss of in-wheel permanent magnet synchronous motor with high power density,a kind of surface mounted permanent magnet synchronous motor with novel magnetic pole structure is designed. The outer arcs of magnetic pole is eccentric and the inner and outer arcs have different arcs. The influence of the new eccentric magnetic pole structure on air gap flux density and eddy current loss is analyzed by analytical method. The 10-pole / 12-slot three-dimensional motor model is built in the finite element software to simulate the electromagnetic and steady-state temperature fields. The results show that the eddy current loss of permanent magnet per cubic centimeter is reduced by 17. 20% and the steady-state temperature of permanent magnet is reduced by 3 ℃ , therefore, the structureimproves the running performance of in-wheel permanent magnet synchronous motor.

Abstract:In order to increase the amplitude of the ultrasonic guided wave (UGW) signal at the receiving end of the rail and improve the identification of the guided waves, the coded excitation technique is applied to UGW broken rail detection system in this paper. At the transmitters, the system uses a long duration 13-bit Barker-coded to generate UGW excitation signal. At the receivers, the matched filtering method is used to make pulsed compression of the echo signal. In this paper, the simulation analysis and experimental verification on 1 meter 60 type rail are carried out. The results show that the experimental results are in good agreement with the simulation results. The signal amplitude of Barker code excitation after pulsed compression by matched filtering is much higher than the received signal of single pulse excitation, compared with the received signal of the single pulse excitation, the amplitude of the Barker code excitation received signal increases to more than three times of that. The normalized waveforms of the two signals are almost completely coincident in the time domain, the similarity between the two signals in the simulation and experiment results is reached 99. 39% and 99. 29% respectively. The conclusion can be drawn that using Barker code as excitation to obtain UGW is feasible and the Barker code excited UGW can be applied in broken rail detection.

Abstract:In order to solve the problem that quantifying the area of tank defects detected by ultrasonic wave, an improved quantitative model of tank corrosion defect area is proposed. This model uses the feature importance of XGBoost to initialize the parameters of artificial neural network (ANN) a priori to improve the ANN model. The model can converge faster and improve the accuracy. Design an experimental platform according to national standards, obtain experimental signals, and extract the statistical features of the signals to obtain a data set. Use the data set to train and test improved models, and compare them with traditional models. The experimental results show that the improved ANN model can converge faster and quantify the defect area accurately. Compared with the ANN quantization model, the accuracy in the training set has been improved by 17. 9%, reached 98. 3%. and increased by 16. 6% on the test set, reached 92. 2%.

Abstract:In order to effectively monitor the rolling bearing performance degradation trend and its abnormal fluctuations, a dynamic early warning method of rolling bearing performance degradation trend based on the relative energy entropy of variational mode decomposition (VMD) and the adaptive ARMA model is proposed. Methods VMD was used to decompose the life data of rolling bearing to obtain bandlimited intrinsic mode functions (BLIMFs). The energy of the BLIMFs component is analyzed by relative entropy, and the characteristics of rolling bearing performance degradation are extracted to obtain the bearing performance degradation evaluation index of VMD relative energy entropy. The energy entropy value extracted by VMD decomposition is used as an input for ARMA model for dynamic regression prediction. The test results show that this method can effectively monitor the degradation trend of rolling bearing performance and the abnormal fluctuation of indexes, and verify the effectiveness of the proposed method.

Abstract:In order to improve the fault diagnosis rate of rolling bearings, make full use of the difference in the recognition ability of the bearing operating state by the time domain, frequency domain and frequency domain features, and take into account that the features are prone to irrelevance, redundant interference and other issues, as well as the simple, fast and effective feature evaluation of the actual project method needs. Position optimized Fisher distance measure ( POFDM) method is proposed and applied to bearing fault characteristic select. The method is based on Fisher’s criterion, and the positional relationship between multi-class samples is used to correct the evaluation coefficient by the median method, which could reflect sensitivity of the state separation and aggregation. Thus, the features that can suppress the degree of state coincidence are selected. In addition, aiming at the problem that the intelligent diagnosis model is inefficient in seeking optimal feature set, feature set evaluation method based on multi-dimensional spatial measure-Fisher is proposed. The optimal feature set is selected based on the maximum value principle by calculating the distance measure index of different dimension candidate feature sets in multidimensional space. Finally, the proposed algorithm is verified by the bearing fault experiment. The experimental results show that the optimal low-dimensional feature set obtained by the proposed method achieves 99. 17% diagnostic accuracy of the SVM classifier when the number of feature combinations is 3, which can effectively diagnose bearing faults.

Abstract:To overcome the slow convergence and low accuracy of traditional particle swarm optimization (PSO) for slow convergence and low accuracy of multi-objective task scheduling in cloud computing, an optimized multi-objective task scheduling particle swarm optimization algorithm (MOTS-PSO) is proposed. Firstly, the nonlinear adaptive inertial weight is introduced to change the particle’ s optimization ability to avoid the algorithm from running into local optimum. Secondly, the flower pollination algorithm probability update mechanism is introduced to balance the global search and local optimization of the particles. In addition, we improve the global search position update formula. Finally, the firefly algorithm ( FA) is introduced to generate the elite solution to improve the local search position update formula. At the same time, we utilize the elite solution to perturb the particle position and to jump out of the local optimal state. Experiments show that the MOTS-PSO algorithm has 27. 1% and 19. 9% higher convergence speed and precision than the PSO algorithm, and 22. 09% and 5. 2% higher than the FA algorithm. Further experiments show that the MOTS-PSO algorithm is more effective than the PSO and FA algorithms in solving tasks of different sizes and numbers.

Abstract:The pulse wave signal collected by optical capacitance product method is easy to be interfered and the baseline drift noise occurs. An improved variational mode decomposition (VMD) algorithm is proposed to eliminate baseline drift noise. Firstly, the PPG signal is resolved into multi-modal components by means of variational mode decomposition, and the residual items of empirical mode decomposition (EMD) are eliminated. Finally, all modes are reconstructed. The experimental results show that the signal with drift effectivity could be reduced by the algorithm. Compared with the EMD algorithm, the noise of signal reduce ratio and mean square error of the improved VMD algorithm are 0. 26 and 1. 73, which can improve the quality of signals effectively.

Abstract:In order to solve the problem of current measurement in vehicle microgrid with rich current frequency components and complex electromagnetic environment, a new closed current measurement method is proposed, which is based on the advantages of low frequency characteristics of tunnel reluctance sensor and high frequency characteristics of rogowski coil. Ansoft Maxwell is used to simulate and analyze the shielding effectiveness of the shielding shell. The power frequency measurement experiment is carried out with Rogowski coil and tunnel magnetoresistance sensor, and the calibration is completed with the clamp current probe as the benchmark. The calibration parameters are applied to the peak pulse current measurement in the process of micro grid parallel and off grid, and the data composite processing is carried out. The experimental results show that the composite current measurement signal is in good agreement with the clamp current probe signal, especially performs as the reduction of the peak error, and the relative error of the composite peak pulse current is reduced to 3. 67%. The current measurement method based on the data combination of rogowski coil and tunnel magnetoresistance is of great significance for the electrical measurement of equipment.

Abstract:The recognition accuracy of dot matrix characters is low due to error segmentation, this paper proposes a dot matrix character detection method based on convolutional neural network (CNNs) recognition feedback. Firstly, multi-scale windows are used to acquire multiple candidate regions and CNNs are established to identify them. The voting mechanism is used to make comprehensive decisions on multiple recognition results, and then the lattice characters are reversed according to the decision result and the character segmentation is completed. Finally, a sliding flip window is proposed to segment and identify all characters. The experimental results show that the proposed method outperforms the traditional character recognition method in the segmentation accuracy and recognition rate of dot matrix characters, reaching 97. 53% and 97. 50% respectively.

Abstract:In order to study the dynamic characteristics of electric vehicle body-in-white, a structural modal parameter identification method based on natural excitation is proposed. By this method, the structural modal parameters of an electric vehicle body-in-white is identified. The first three-order structural modal parameters of the electric vehicle body-in-white are obtained, then the results of identification by this method are compared with those by traditional methods. It is found that the maximum error of natural frequency is 1. 8%, the maximum error of damping ratio is 13%, the modal shape is consistent, the correctness of this method is verified. Then, the dynamic characteristics of the Body-in-white of the electric vehicle are evaluated by the identified structural modal parameters and the working characteristics of the electric vehicle. The proposed natural excitation method simplifies the identification process of structural modal parameters, and can be applied to identify modal parameters of large and heavy structures which are not easy to excite. The results of modal parameter identification have certain guiding significance for the dynamic characteristic design of electric vehicle body in white.

Abstract:The instrument landing system provides an important guarantee for the safe landing of aircraft. In view of the increasing number of flights, how to ensure the efficient and safe operation of ILS has become an important issue for the civil aviation department. Through theoretical analysis and research on modulation depth difference (DDM), it is concluded that the main factor affecting DDM is modulation factor m90 . On this basis, the mathematical models of channel and modulation factor and carrier negative value are established by using control variable method and least square method respectively. Based on NM7000 equipment, a simulation system was built to dynamically simulate the channel adjustment process. Finally, the accuracy of the model was verified by comparing the established system with the actual situation. The data showed that the relative error of simulation was less than 1%, indicating that the proposed model was of high accuracy and could provide theoretical support for the reference of air traffic controllers.

Abstract:Static coupling is one of the most important factors affecting the measuring accuracy of three-axis force sensor. For the phenomenon of coupling, this paper introduces the principle of three-axis force sensor’ s couple. Analysis of two different linear decoupling algorithms, which were based on the Cramer theorem and the least square method. The static decoupling algorithm based on coupling error modeling is improved. Processing the data in the coupling direction using Newton’ s cubic interpolation method. The experimental results on the self-developed three-axis wrist force sensor show that the novel algorithm has high precision.

Abstract:The detection of infrared multi-target Image and video in complex background is the hotspot and difficulty of target detection. In order to detect infrared target in complex background more accurately, the algorithm of YOLOv3 is improved. Firstly, by increasing the feature scale on the basis of the original algorithm, the recognition accuracy of remote and complex background image is improved, and the BN network layer and convolution neural network are combined. The final detection results are obtained by layer fusion calculation. The analysis and comparison between the original algorithm and the improved algorithm show that the improved algorithm can better the average recognition accuracy from 64% to 88%, and the mAP from 51. 73 to 59. 28, which is verified that the improved YOLOv3 algorithm has better performance and more obvious advantages in infrared target detection.

Abstract:In the field of Beidou navigation and positioning, in order to obtain accurate carrier phase observation data, the cycle slip phenomenon in carrier phase observation must be effectively detected and repaired. This paper first constructs single-difference and double-difference detection quantities, then using wavelet transform to three-scale decomposition of single-difference and doubledifference detection sequences,and extracting high-frequency coefficients and low-frequency coefficients at the same time. Obvious cycle slips can be observed from singular values of high-frequency coefficients. Then, the singular values of high frequency coefficients are replaced by Lagrange linear interpolation, and the high frequency coefficients and the low frequency coefficients are reconstructed. Finally, the reconstructed signal is subtracted from the original signal to obtain the cycle slip difference value, which is used to repair the cycle slip layer by layer. In the experiment, 200 cycle-free epochs were selected, and different cycle slips were added at 100 epochs. Simulation results show that the method can effectively detect and repair cycle slips of more than 0. 5 weeks.