A passive controller is developed to solve the problem of instability of steering-by-wire system caused by the inaccuracy of sensor and signal noise. The passive controller is designed with the analysis of the steering-by-wire system from the aspect of energy based on passive theory, thus ensuring the control stability of the system. The passivity of the steering-by-wire system is analyzed based on passivity analysis of mechanical steering system. The part of the steering-by-wire system that is easy to show the active state is confirmed, and the passive controller is designed to ensure its passivity, thus achieving the stability control of the system. The method is verified through real vehicle test. The results show that the passive controller designed based on passive theory can guarantee the control stability of the steering-by-wire system.

In order to obtain good energy consumption economy for Plug-in Hybrid Electric Vehicle (PHEV) and reasonably distribute battery power under different driving condition, a fuzzy inference controller is used to identify the driver's intention. The working mode is partitioned according to the hybrid system dynamic coupling mode, the State of Charge (SOC) of the battery and the engine working characteristic curve. Then, instantaneous optimization method is used to distribute engine and motor torque, in which the energy consumption economy is taken as the objective function. Finally, the energy management strategy based on driving intention is proposed. A vehicle model is built on the Matlab/Simulink platform, and simulated under New European Driving Cycle (NEDC) condition. Simulation results indicate that the proposed strategy can realize the optimization control of the engine and ISG motor and keep the engine operation points within the peak efficiency region. The fuel consumption per 100 km is saved by 8.24% in comparison with the energy management strategy of the charge depleting and charge sustaining mode. In order to further verify the proposed energy management strategy, vehicle road test is conducted, and results show that the fuel consumption per 100 km is reduced by 32.93% compared with prototype gasoline vehicle.

An inverse sub-structuring transfer path analysis method is proposed. First, the dynamic stiffness of powertrain mounting system is computed using the frequency response functions of the vehicle system and substructure systems. Then, the vibration signals are processed by wavelet de-noising technology, and the loads are computed by combing the vibration data measured in actual condition. Finally, the vibration or noise transfer path is analyzed. A case study is carried out with a mini bus to demonstrate the effectiveness of the proposed method. Compared with conventional transfer path analysis method, the efficiency of the proposed method increases by 50% and the accuracy is equivalent.

In order to achieve precise control of the lateral movement of vehicles, a robust controller for vehicle motion is proposed. Ideal lateral acceleration of the vehicle mass center is achieved by changing the input angle of the steering controller, and robust stability and smaller control error are obtained. First, according to 2-degree-freedom vehicle model, the steering system and road perturbations, the generalized state equation is established. Then a loop transformation of the state equation is made to get the controller. Finally stability analysis and experiments are carried out. The controller parameters vary with the vehicle speed, and the design of this controller is much simpler than other variable parameter controller. Stability analysis results show that the whole control system has internal stability. Vehicle experiment results show that the controller can make the acceleration error of the lateral motion of the vehicle mass center within 0.1 m/s².

To study the influence of multi-modal guidance information on the travel mode of commuters from the viewpoint of trip chain, A Revealed Preference (RP) and Stated Preference (SP) survey was carried out to get the revealed and stated trip chain data. Then a mixed logit model, which takes scale parameter difference, heterogeneity and reference dependence effect into account, was established. A simulation method was proposed to complete estimation. The estimation results show that the introduction of scale parameter difference is helpful to explain the unobvious variance differences between RP data and SP data when they are combined together; the adoption of heterogeneity and reference dependence effects are not only conductive to measure the impacts of some variables, whose random preferences vary obviously across individuals, more accurately, but also benefits to reflect the preference distribution of random variables. The accuracy and predictive capacity of the mixed logit model considering the three aspects are better than the multinomial logit model and the ordinary mixed logit model. It is also manifested that multi-modal guidance information service contributes to encourage commuters to shift from car trip-chain to other modes, and helps to ease traffic congestion at source.

Spatially disaggregated panel data models using negative binomial process, including pooled model, fixed-effects model and random-effect model, were developed. The contributing factors, such as roadway geometrics, traffic characteristics, land usage and rainfall were collected. Incidence Rate Ratio (IRR) and homogeneous segments were employed for analysis and prediction of crash count, which contains Property-Damage-Only (PDO) crashes, injury crashes, fatal crashes, and total crashes. Meanwhile, F test and Hausman test, log-likelihood value and Deviance Information Criterion (DIC) were used to compare the performance of the three types of models. It is found that the effects and statistical significance of covariates on the four types of crashes are not the same. For example, if the speed limit is increased by 1.609 km/h, the PDO rate, injury crash rate and total crash rate are reduced by 3.89%, 2.24% and 2.79%, respectively; however, the fatal crash rate is increased by 6.38%. Results show that the fixed-effects panel negative binomial model surpasses the pooled model and random-effect model, and the modeling is even better for crash count associated with more severe injuries and fatalities.

A calculation method of the dynamic characteristics of continuous beam bridge with variable cross-section is proposed based staging concept. In this method, each cross of a girder bridge is segmented; the undetermined coefficients are used to express the vibration mode function of each segment; the transfer equations within the undetermined coefficients of each cross are established based on intersegmental continuity condition. The equations of the full-bridge undetermined coefficients are formed based on all the boundary conditions and transfer equations. The method is able to calculate the natural frequency and vibration mode of the bridge by solving the equations of undetermined coefficients. This method is applied to solve the dynamic characteristics of a continuous box girder bridge to demonstrate its validity and reliability.

The analytical solution of variational analysis for prestressed frame is investigated. Combined with experiment of typical prestressed frame structure, the effective flange width of prestressed concrete beam is studied in details. Results show that the equivalent width of prestressed concrete bean can describe the real stress state. In engineering practice, short-lived state of prestressed concrete stress calculation is required. Prestress can change the two-way concrete floor load transfer law. The results of this work can provide reference for the design and calculation of prestressed concrete slab system.

The clay collected from an engineering site is studied with GDS cyclic triaxial test to analyze the characteristics of dynamic modulus and damping ratio. It is shown that the 1/E_d-ε_d curve can be described by the approximate linear equation and the confining pressure and vibration frequency have little influence on the damping ratio. However, the change of damping ratio curve is saltatory with the increase in consolidation ratio. Through the establishment of empirical formula of the test parameters in a certain precision range, the (E_dmax/Pa·k_c)-(σ_m/Pa) curve could be expressed by linear equations after normalization processing, the curve (λ_d/λ_dmax) and (1-G_d/G_dmax) could be expressed by the power function after normalization processing. These functions can be easily used and provide reference for practical engineering.

The performance of transient operating condition is different under different working conditions for turbocharged engine. The effects of different initial loads under constant engine speed, different engine speed with the same increasing torque process on engine combustion process and emissions were investigated on a turbocharged diesel engine. The results indicate that, compared with high idle increasing torque, initial load as partial load has improving effect on the responsiveness of air intake and combustion process at constant speed and the smoke opacity varies slightly. However, the change of NO_x is significant, which is about 80% more than that of high idle initial load with constant speed and increasing torque. Results also show that for different speed conditions, compared with high constant speed in transient combustion process, the deterioration of smoke opacity is worse and the improvement of NO_x emission is enhanced at low constant speed.

The ammonia content, decomposition temperature and decomposition rate of ammonium bicarbonate, ammonium carbonate and ammonium carbamate are investigated by theoretical analysis, Thermogravimetric Analysis (TGA) experiment and equilibrium pressure experiments. Results show that ammonium carbamate has the highest ammonia content, followed by ammonium carbonate, and ammonium bicarbonate has the lowest ammonia content. Ammonium carbonate and ammonium carbamate can decompose from room temperature, but when heated to 100 ℃ ammonium bicarbonate only decomposes 5.6%. When the temperature is below 90 ℃, the decomposition rate of ammonium carbamate is higher than that of ammonium carbonate. Under the same period and same temperature, ammonium carbamate produces more ammonia. The decomposition equilibrium pressures of ammonium carbonate and ammonium carbamate are similar; both are excellent candidates of reductant of Selective Catalytic Reduction (SCR) system, but ammonium carbamate has more obvious advantages.

A model calculation method of heat pump using earth energy system is proposed. Testing of an actual operation condition of ground-source heat pump in severe cold area has proven the reliability of the calculation method. This method was used to forecast the operation process and energy consumption of the project for years. Calculation results show that without other supplementary heating modes or measures, the system can not ensure the normal operation for years. By analysis and calculation of the controllable factors, a scheme to improve the operation condition is proposed, which can ensure the sustainable use of the earth energy technology.

The structure and working principles of the load-sensing static signal priority vale are analyzed in this paper. By establishing the complete mathematical model and creating a block diagram of the system, the parameters, which have large impact on the steering flow response, are quantitatively analyzed, and the AMESim simulation model is established. The response of the steering flow under different input conditions is quantitatively investigated through dynamic simulation, and the key structure parameters are optimized. Comparison of simulation and experiment results shows that the model is accurate. Simulation results demonstrate that the sizes of the orifices of priority vale LS and damper S have huge influence on the response of the steering flow and increasing diameters appropriately can improve the system performance.

Single-acting double-stator motor has four operating states, including inner or outer motor working alone, inner and outer motor united working and differential operating states. Through the calculation of the hydraulic force acting on the rotor in different operating states, the formula of acting force is acquired, the effect of the vane number, which is odd or even, on the acting force is considered. The simulation according to the prototype was carried out using MATLAB. The results show that the rotor radial force varies periodically within a certain range under four strained conditions. Compared with that of the classical vane pumps, the strained conditions of the rotor of single-acting double-stator are better. This study may provide reference for the design of the variable displacement mechanism for single-acting double-stator hydraulic motors.

The relationship between bearing preload and system stiffness was investigated under the condition that the structure turntable was fixed. The relationships among bearing preload, bearing stiffness and shafting stiffness were analyzed. The computation formula between bearing preload and shafting natural frequency was deduced. Using finite element analysis the relationship curve between bearing preload and system natural frequency was obtained. Natural frequency tests of a two-dimensional turntable verified the validity of the proposed method, and an effective method was provided to determine the optimum bearing preload.

A control method for five-axle linkage CNC machine tools was proposed based on EtherCAT, and special CNC processing equipment for face gear was developed. The principles and methods of hopping and milling were analyzed on the basis of the principle of face gear cutting. The CNC processing coordinate conversion and tool path generation of hopping and milling were conducted. The CNC methods of processing were given based on the structure of the special face gear machine, and processing experiments of face gear were conducted. The coordinate values of tooth surface points were detected by pmm-c700 CMM, and the deviation ranges in the normal direction of face gear surface were obtained. The results verify that the structure of the machine and the processing methods are feasible, which lay the foundation of engineering applications of face gears.

A hollow-through Down the Hole (DTH) air hammer with large diameter and a reverse circulation drill bit are developed; then computational simulation and field test are carried out. The test results show that the maximum penetration rate of the DTH air hammer can reach 6 m/h, and the average penetration rate is 4.5 m/h. in addition, the removing cuttings are smooth with reverse circulation. Furthermore, the cuttings are rarely accumulated in the bottom of borehole, and the borehole quality is excellent. Therefore, using the DTH air hammer and drill bit could resolve the problems of rock breaking and removing cuttings, and meet the requirements of penetration rate and borehole quality in the operation of large-diameter emergency rescue well drilling.

In-plane tension-compression experiments of TRIP800 steel were conducted with a self-developed device. Based on the experiments, the Yoshida-Uemori (Y-U) hardening model was built and verified, which was then used to analyze the springback of the U-shaped part. The results show that more springback is achieved on the Y-U model than that based on isotropic model.

The forelimb of German Shepherd Dog (GSD), a land quadruped animal, was selected as the subject, and the kinematical characteristics of each joint in for-lower limb under various gaits in the foot-ground impact phase were analyzed using optical three-dimensional motion capture technology. The focus is to reveal how does each fore-lower limb joint adjust and adapt to the ground impact. The analyses of joint movement sequence and joint angle variations show that GSD employs the coordination of multi-joint flexion motion of the fore-lower limb to buffer ground reaction force. Compared with metacarpal-phalangeal joint, the humeral joint, cubital joint and carpal joint of GSD play more important roles in ground impact absorption; while, the flexion motion of these three joints function in different contact phases. The flexion of carpal joint happens first, then the humeral joint, and finally the cubital joint. This may show that the coordination of multi-joint flexion of GSD has spatio-temporal characteristics. This study also finds that the initial contact angle between ground and GSD finger increases with the ground impact force. This adaptive adjustment may help GSD to prolong the cushion time during foot-ground impact phase, thus, decreasing the strength of ground reaction force.

In order to improve the crashworthiness performance of thin-wall tube, the oxhorn is taken as a biological prototype to analyze the similarity between oxhorn and thin-wall tube. The feature structures of the oxhorn against collision are extracted. These feature structures are applied to the design of bionic thin-wall tube. The nonlinear finite element method is used to study the deformation mode and energy-absorption properties of the bionic tube. The results show that the deformation mode of the bionic tube is asymptotic telescoping deformation. The specific energy-absorption of the bionic tube is 46.2 kJ/kg, which is about 1.3 times and 1.8 times higher than the four cell conical tube and conical tube. The crashworthiness performance of the bionic tube improves as the thickness of the core increases. When the collision happens in transverse, the crashworthiness performance of the bionic tube is better than the four cell tube and general conical tube, and its specific energy absorption is 10.4 kJ/kg. The crashworthiness performance of the bionic tube may provide reference for the design of energy-absorption components of vehicle.

The change of the wettabilitys of lotus leaves after been soaked in different solutions, acid(H₂SO₄),alkali (NaOH), ethanol(C₂H₅OH) and acetone(C₃H₆O),with different pH values and soaking times was investigated. The results show that pH value and soaking time have little influence on the water contact angle on the lotus leaves. The lotus leaf samples before and after the experiments were observed using scanning electron microscope(SEM)and field emission scanning electron microscope(FESEM-FEG). We find that there are no obvious changes of the surface morphology and wax crystal of the lotus leaf. Chemical composition analysis of the lotus leaves demonstrates that the inert reaction between acid-base solution and cuticular wax is the main reason of un-wettability for the lotus leaves. The main reason for the rapid loss of the superhydrophobic property of the lotus leaf samples after being soaked in ethanol and acetone solutions is that the surface morphology and cuticular wax are changed due to the high combining ability of alcohols and ketones with cuticular wax.

According to Lyapunov stable theory and the modularity of reconfigurable manipulator, a decentralized backstepping neural network controller is designed for the fault free system, and the adaptive neural network system is employed to compensate the reduced control accuracy from the interconnection term. An augmented subsystem is obtained by introducing a filter, which maps the sensor fault into pseudo actuator fault, and then the multi-step time-delay technique is adopted to compensate the concurrent failures and realize the fault-tolerant. Simulations illustrated by two 2-DOF reconfigurable manipulators with different configurations demonstrate the effectiveness of the proposed scheme.

Based on the analysis of large number of measured data obtained from a free toll lane of highway detection system and experimental research, a vehicle length measurement method is proposed based on BP neural network, which can measure the length when the vehicle is traveling. This method can accelerate the convergent speed of BP neural network by fuzzy adaptively adjusting the learning rate of BP neural network. A large number of real vehicle data have been obtained after this method was applied to the detection system in the highway toll lane of Lalin River, an exit of Jilin Province to Heilongjiang Province. The results show that the proposed method of dynamic vehicle length measurement can effectively improve the measuring precision of traveling vehicles.

This paper proposes a method to compare the Received Signal Strength Indicator (RSSI) of target tag with the RSSI of reference tags while taking many reference tags as a whole. First, this method figures out the area of target tag; then adds virtual reference tags in this area; and finally determines the position of target tag via weighted estimate. The simulation and hardware experiments demonstrate that the improved algorithm has better stability and stronger anti-interference without increasing any physical cost. Besides, the location accuracy is enhanced obviously, thus the algorithm is more applicable to indoor location system.

First, a new model of uncertain nonlinear closed-loop fault Networked Control System(NCS) is established, which integrates the network time-varying delay, actuator saturation and fault, and the discrete event-triggered condition in a unified framework. Then, utilizing the Lyapunov stability theory and delay decomposition technology, combing the improved Jensen integral inequality and the reciprocal convex combination lemma, the delay- and event-dependent invariant set condition for ensuring closed-loop fault NCS stability is derived. The co-design method for both robust fault-tolerant controller and the weight matrix of discrete event-triggered conditions is given. Finally,a simulation case is employed to demonstrate that the proposed method, which guarantees faulty NCS based on Discrete Event-Triggered Communication Scheme(DETCS), possesses good fault-tolerant performance, and saves the limited network resources by reasonably choosing the triggering parameters.

A robust controller framework for flexible joint robot is presented, in which the effect of nonlinear friction on control performance is considered. The nonlinear friction is denoted as inverse additive output uncertainty relative to the nominal model. Based on this the describing function is analyzed in frequency domain, and the weighting function of nonlinear friction is further calculated. Combing the friction uncertainty, the Linear Matrix Inequality (LMI) optimization is proposed as the benchmark for controller design, which leads to good performance robustness. Simulation results show that the proposed controller can provide excellent command tracking and regulation performance.

For simultaneously achieving NOx conversion efficiency and low ammonia slip, a new control-oriented model of urea-SCR systems and novel nonlinear controller based on triple-step nonlinear method are presented. The proposed controller could drive the parameter-varying nonlinear system to track the desired ammonia coverage ratio. The robustness of the controller against the measurement noises and system uncertainties are analyzed. Based on validated enDYNA model, transient FTP75 driving cycle simulation is conducted to evaluate the effectiveness of the proposed strategy. Comparisons with Monte Carlo methods' optimal PID controller are presented to show the advantages of the proposed strategy.

The airborne opto-electric stable platform model was analyzed and simplified to K/s with current loop. The disturbance to the platform stability and the disturbance rejection method were illustrated. The traditional square lag-lead compensation method, its strengths and weaknesses were introduced. The repetitive controller and Active Disturbance Rejection Controller (ADRC) were described. A repetitive controller and a two stage ADRC were designed, which were integrated into a new controller system taking the advantages of both. Speed stability experiment and target tracking experiment were carried out on a vibration platform. Results show that compared with conventional controllers, the proposed controller improves the disturbance isolation by 6.28 dB at least. On all accounts, the proposed control system has strong robustness, observably improves the disturbance isolation of the opto-electric platform and can provide guide for engineering applications.

Using an example of Fractional Order Proportional Differential (FOPD) controller, the parameter calculation procedure of the robust FOPD controller based on phasor method is discussed in detail. It is proved that the calculated controller parameters based on phasor method are unique. Besides, the proposed algorithm is programmed in Matlab. According to Matlab programs, the FOPD controller parameters solution algorithm and the FOPD controller are respectively designed for the systems compensation following the different given phase margin, different given gain crossover frequency and different given controlled objects. This study results show that not only the parameter calculation procedure of fractional order controllers is simplified and the amount of calculation is reduced, but also the calculated controller parameters are unique and valid.

In this paper, first, it is deduced that the sensor uncertainty is minimum with Voronoi partition; a node coverage weight by priority function of event area is defined. Then a nodes deployment optimization algorithm is proposed based on node coverage weight and virtual force. It guarantees the connectivity of network while the sensing uncertainty is minimal. Finally, computer simulation is carried to compare the deployability of the proposed algorithm with other algorithms under different node densities and different event priorities. Simulation results show that the proposed deployment algorithm is quick and effective with smaller sensing uncertainty compared with other deployment algorithms. Also the proposed algorithm has strong robustness.

The weighted bipartite graph method of complex network is employed to model user behavior when accessing APPs. Based on the real communication logs collected from the international gateway entrance of the domestic mobile operator in 2014, the propagation characteristics of APPs are analyzed. These characteristics include APP propagation range, user access interest, access time distribution, click properties and access relevance. The analysis results provide the basis for using bipartite network and APP relevance graph to find specific information dissemination, characterize user groups and monitor intelligent terminal in a real-time way.

A video-based face texture program was proposed, which can extract binary patterns from three orthogonal planes in the frequency domain of the Gabor transform (GLBP-TOP). First, each normalized face was transformed by Gabor wavelet to get enhanced Gabor magnitude map. Then, the LBP-TOP operator was applied to the maps to extract video texture. Finally, weighted Chi square statistics based on Fisher Criterion were used to realize the face classification. The proposed algorithm achieved good results in the Honda-UCSD database, and was effective and robust under changes of illustration and expressions.

In this paper, an unequal cluster data gathering protocol with optimal cluster size (UCPOCS) is proposed in order to resolve energy hole problem in wireless sensor networks. First, UCPOCS adopts time broadcast to substitute traditional discussion mechanism to select cluster head. Then, the optimal cluster radius is theoretically obtained according to the location message of the candidate cluster head, which clusters the whole network. Finally, the multi-hops routing mechanism selects the relay-nodes according to three messages of neighbor cluster heads, e.g. residual energy etc, which is suitable for uniform and non-uniform node distribution. Simulation is conducted and results demonstrate that the proposed UCPOCS can balance energy consumption effectively and the network lifetime can be prolonged.

A software architecture of wireless sensor network system, which consisted of data acquisition layer, control and communication layer and data storage and analysis layer, was established for distributed stress monitoring in high-speed railway track. A concurrent data processing method for the information from multi-sensor nodes was proposed to improve the key characteristics such as the data integrity and rapid response. The data storage and management model was constructed based on relational databases and data analysis machine to manage and analyze the continuously generated data. In addition, the data flow of the interactive manipulation for client software, called "Metadata capture-Ontology mapping-Data acquisition-Data visualization", was designed to display the equipment distribution and sensing data. Finally, the software architecture was implemented on the wireless sensor network platform at the China Academy of Railway Science. Experiment results show that the packet loss rate was less than 5% and transmission delay was less than 1 min. The performance of the WSN-based software architecture can satisfy the requirement of the distributed monitoring of high-speed rail stress.

To realize remote monitoring and management of manufacturing equipments, a method of information integration of Numerical Control (NC) equipments was proposed, which included interconnection architecture, information model for devices and workflow of information interchange. Machine tools were connected to internet through agents. A container based hierarchical unified information model was proposed to organize the heterogeneous data from variety machine tools and sensors. The information was transmitted over an internet protocol. A remote monitoring and managing system in accordance with the method was designed and implemented. The system was tested through visualization of manufacturing process and statistics of the overall equipment efficiency. The results demonstrate that the proposed method is a reliable and extensible solution for manufacturing information integration.

In object surface detection using photometric stereo algorithm, two main issues need to be solved; unknown Generalized Bas-Relief (GBR) parameters and surface containing mirror point, shadows and other interference. To solve incalibrated photometric stereo, a novel closed-form algorithm based on local maxima points is proposed to solve GBR parameters. The algorithm, which only solves group of dual quadratic equation without optimization or iterative process, greatly improves the speed and accuracy compared with the existing GBR solutions. To solve non-Lambertian surface issue, image segment and PCA technology are used to segment and remove interference to ensure the algorithm satisfying the application conditions. Higher accuracy and efficiency are proved by experiments.

The current recognition accuracy of digital image is lower. To overcome this problem, the method of image feature extraction based on wavelet space feature spectrum entropy is proposed. According to the principle that the energy is equal before and after the wavelet transform, the matrix of wavelet energy model is constructed. Then, the singular value decomposition of the matrix is performed. Finally, the feature spectrum entropy of the singular value is taken as the image feature. Experimental results show that the proposed image feature extraction method can obtain higher image recognition accuracy, and the effectiveness and practicality are proved.

On the basis of Gigabit Ethernet, two cameras connected to the same computer can work simultaneously. A plug and play data acquisition module was adopted as a synchronous signal generator. By means of LabVIEW programming, two synchronous pulse signals with a specified period were generated, which were then used to trigger two cameras. Microsoft Visual C++ was used to develop the synchronous acquisition software. The two cameras can capture images synchronously and the images can be automatically saved. The maximum synchronous frame rate is up to 18 frames per second, laying a good foundation for the research of binocular vision pulse information measurement.

In order to adapt the characters in the data process of Lunar Exploration, the workflow technology is proposed to deal with the situation that research excessively depends on software development. Yet the data process of Lunar Exploration can not be represented by the current workflow process definition meta-models. An innovative extended workflow process definition meta-model is proposed with the data model and the process model. Base on that, the extended data model and process model are studied. The data relationship is established through product tree that data are encapsulated and layered in order to facilitate the modeling. To reduce the modeling complexity of the mathematical formulas using DAG, semantic modeling based on MathML is approached. The Data Processing on Chang'E-3 is used to verify the model. Results show that the flexibility in data pre-processing system of Lunar Exploration Project can be solved. Prototype software has already been used in Lunar Exploration.

In large-scale video stream service in video cloud, advanced reservation can cause high blocking probability and low bandwidth utilization. To overcome such problems, investigation on time-constrained malleable bandwidth reservation scheme is carried out. First, a finite malleable reservation model is defined. Then, based on this model, a time-constrained bandwidth fragment adjusting algorithm is proposed to improve bandwidth utilization. Finally, considering multi-path service of video streams in video cloud, the reservation algorithm is explored into multi-path mode, thus improving the success probability of bandwidth reservation. Simulation results show that, compared with other bandwidth reservation schemes, the proposed solution can effectively guarantee the QoS of video service while with lower blocking probability of bandwidth reservation.

To meet the requirements of the sensitivities during the airborne electromagnetic inversion iteration, a variable step size difference accumulation algorithm is proposed. Based on the idea of replacing the derivative by difference, and adjusting the accumulating time by difference step coefficient, a fast high-precision algorithm is realized. The sensitivity matrix of multi-group earth model is calculated. The results show that the error of the sensitivity matrix is within 0.1% when the coefficient of difference step is usually selected 10. Comparison with the routine center differential algorithm and the analytical method indicates that the variable step size difference accumulation algorithm is superior to center differential algorithm in calculation accuracy, and is better than analytical method in calculation speed. Finally, the proposed algorithm is applied to airborne inversion simulation to prove its effectiveness.

In nonlinear inversion, neural network algorithm has the shortcomings of easy to fall into local minimum, slow convergence and poor inversion accuracy. To overcome these shortcomings, the Particle Swarm Optimization (PSO) is combined with Back Propagation (BP) neural network to achieve inversion (referred as PSBP). The effectiveness of the PSBP is verified by the classical geoelectric model. Compared with the linear inversion and BP neural network inversion methods, the proposed PSBP method has obvious advantages with good inversion result.

An electromagnetic noise test system based on MSP430 and Field Programmable Gate Array (FPGA) dual controller in the larmor frequency range 1 kHz ~ 3 kHz is designed. The system reduces equivalent input noise by the parallel low noise operational amplifiers. The system realizes the controllable passband and adjustable 128 gain level to get more information about the current environment of electromagnetic noise. NIOS II soft core built in FPGA is used to set up data acquisition system, and the data are archived in the SD card by the SPI bus so as to realize effective data acquisition and storage. Analysis of the indoor and outdoor measured data verifies the effectiveness of the system.

A Frequency-domain Regularized Wiener Filtering (FDRWF) is proposed. A novel objective function is constructed with an additional noise constrain item and a regularized factor. An amplitude adapting factor is brought in to solve the signal losses resulted from filtering. Moreover, the noise spectrum is estimated by the first arrival noise and noisy record. Experimental results on both synthetic records and field data show that the novel method has superior performance to median filtering and traditional Wiener filtering. Meanwhile, signal losses and derivation are reduced obviously. The effectiveness of the regularized idea in the novel method is proved.

In order to solve resource allocation problem in heterogeneous system with Delay-Constrained (DC) users and No-Delay-Constrained (NDC) users, under the power constraints for base station and relay stations to guarantee data rates for DC users and with the maximization of the sum-rate of NCD users, an adaptive allocation algorithm is proposed. In the subcarrier allocation, both subcarrier matching and non-matching of two-hop relay links are considered. In the power allocation, optimal power allocation is obtained. Simulation results show that the proposed resource allocation algorithm has better performance with lower computation complexity.

A robust tracking method for multiple frequency-hopping signals is presented. First, the measurement noise is modeled using the t distribution. Then, a novel Bayesian probability model is proposed to describe the multiple Frequency-Hopping (FH) signals tracking problem, such that the influence of the impulsive noise on the tracking performance can be attenuated efficiently. Finally, a Multiple Variational Sparse Bayesian Learning (M-VSBL) algorithm is derived for the proposed Bayesian model to estimate the hopping frequencies, directions of arrival and detect the hopping times. Simulation results demonstrate the robustness and accuracy of the proposed method under impulsive noise.

In Single Carrier Frequency Domain Equalization (SC-FDE) system, a novel equalization algorithm based on Mean Square Error (MSE) criterion was proposed in order to predigest the complexity of Iterative Block Decision Feedback Equalization (IBDFE). The coefficients of the filter can be kept unchanged in each iteration based on the given channel frequency response. The Unique Word (UW) sequence, which was inserted into frame structure and the cyclic prefix, could reduce the overhead of the system. The simulation results show that the performance of the proposed algorithm is better than the conventional linear equalization algorithm in high Signal to Noise Ratio (SNR) and close to the IBDFE algorithm in the same iteration times. However, the complexity of the proposed algorithm is lower than that of the conventional IBDFE algorithm.

To solve the problem of particles degeneracy in the particle filter algorithms, a Classified Resampling (CR) algorithm is proposed. This algorithm adopts different duplication schemes according to the quantity of selected particles; furthermore, it replenishes new particles in the case that the number of effective particles is reduced. Simulation results demonstrate that, with smaller number of particles or longer simulation period, the proposed algorithm has minor Root Mean Square Error (RMSE) compared with Multinomial Resampling (MR) and Systematic Resampling (SR), and with multiple simulations the variance of RMSE is smaller, which indicates that the robustness, durability and stability of the proposed algorithm are improved.

To solve the difficult problem of expectations caused by interaction between hidden variables when using Expectation-maximization (EM) algorithm to estimate parameters for hierarchical Markov Random Fields (MRF) model, the mean-field theory is introduced into Gaussian-MRF (GMRF) model. Parameters can be estimated easily through simple linear equation in case of without window function. An interactive potential function based on Bayesian belief propagation algorithm is proposed to change the situation that the fixed or variable weighted potential function can not express the interaction of image regions. Experiments demonstrate that the proposed method not only has good regional classification but also smoothly internal region. In addition, the mixed and confused phenomenon of traditional hierarchical MRF is improved in wavelet domain.

Aiming at the high correlation of optical channel, an Adaptive Generalized Spatial Modulation (AGSM) algorithm was proposed based on the pair error probability of GSM. The antenna combinations and modulation levels were optimized jointly based on the Channel State Information (CSI) for better Bit Error Rate (BER) performance. The computational complexity of the AGSM was high, so a Simplified AGSM (S-AGSM) was then put forward. The S-AGSM utilizes QR decomposition to limit the exhaustive search space, thus to lower the computational complexity. Simulation results show that the S-AGSM performs better than GSM and GSM with only antenna selection. Moreover, S-AGSM has lower complexity than AGSM.