一种改进的支持向量机参数寻优方法

误差惩罚参数和核参数是决定SVM泛化能力的主要因素,对这两种参数的优化是SVM应用中需要解决的关键问题之一。论文在研究支持向量机理论的基础上,利用一些标准的测试数据集研究了均匀设计方法在RBF核参数优化问题上的表现。通过对比寻优精度,发现和传统均方设计方法相比,采用论文的方法能进一步提高精度。

Error Analysis and Compensation of Strapdown Inertial Navigation System

Based on error analysis, the influence of error sources on strapdown inertial navigation systems is discussed. And the maximum permissible component tolerances are established. In order to achieve the desired accuracy (defined by circular error probability), the types of appropriate sensors are chosen. The inertial measurement unit (IMU) is composed of those sensors. It is necessary to calibrate the sensors to obtain their error model coefficients of IMU. After calibration tests, the accuracy is calculated by uniform design method and it is proved that the accuracy of IMU is satisfied for the desired goal.

Quantitative determination of PFC3D microscopic parameters

It is important to calibrate micro-parameters for applying partied flow code(PFC)to study mechanical characteristics and failure mechanism of rock materials.Uniform design method is firstly adopted to determine the microscopic parameters of parallel-bonded particle model for three-dimensional discrete element particle flow code(PFC3D).Variation ranges of microscopic of the microscopic parameters are created by analyzing the effects of microscopic parameters on macroscopic parameters(elastic modulus E,Poisson ratio v,uniaxial compressive strengthσc,and ratio of crack initial stress to uniaxial compressive strengthσci/σc)in order to obtain the actual uniform design talbe.The calculation equations of the microscopic and macroscopic parameters of rock materials can be established by the actual uniform design table and the regression analysis and thus the PFC3D microscopic parameters can be quantitatively determined.The PFC3D simulated results of the intact and pre-cracked rock specimens under uniaxial and triaxial compressions(including the macroscopic mechanical parameters,stress−strain curves and failure process)are in good agreement with experimental results,which can prove the validity of the calculation equations of microscopic and macroscopic parameters.

Simultaneous hybrid modeling of a nosiheptide fermentation process using particle swarm optimization

Hybrid modeling approaches have recently been investigated as an attractive alternative to model fermentation processes. Normally, these approaches require estimation data to train the empirical model part of a hybrid model. This may result in decreasing the generalization ability of the derived hybrid model. Therefore, a simultaneous hybrid modeling approach is presented in this paper. It transforms the training of the empirical model part into a dynamic system parameter identification problem, and thus allows training the empirical model part with only measured data. An adaptive escaping particle swarm optimization(AEPSO) algorithm with escaping and adaptive inertia weight adjustment strategies is constructed to solve the resulting parameter identification problem, and thereby accomplish the training of the empirical model part. The uniform design method is used to determine the empirical model structure. The proposed simultaneous hybrid modeling approach has been used in a lab-scale nosiheptide batch fermentation process. The results show that it is effective and leads to a more consistent model with better generalization ability when compared to existing ones. The performance of AEPSO is also demonstrated.

中频数字正交解调接收机的研究及实现

数字接收机在现代雷达、电子战、通信等领域的应用越来越广泛,数字接收机的性能是由正交解调信号的幅相一致性决定的。这里描述了数字正交解调的核心部分—数字滤波器的设计,通过运用均方误差最小化准则设计法和切比雪夫等波纹逼近设计法设计FIR数字滤波器,并利用MATLAB平台进行仿真,比较分析表明了用均方误差最小化准则设计法设计具有严格线性相位特性的FIR数字滤波器的最优性,即:可以获得较好的幅度误差和相位误差,从而保证了数字接收机的性能。

Single-jet Spray Mixing with a Confined Crossflow

In order to achieve uniform mixing between spray droplets and crossflow, cold-model experiment of a hollow-cone water spray in an air crossflow is investigated via a numerical simulation. The simulation cases are designed by using the orthogonal design method. The Eulerian-Lagrangian formulation is employed for modeling the droplets-crossflow two-phase flow while the realizable k-ε turbulence model is used to describe the turbulence. A new index, mixedness quality, is proposed to assess the overall mixing of the droplets in the crossflow. The simulation results demonstrate that the counter-rotating vortex pair (CVP) imposes a more significant impact on the spatial distribution than on the size distribution of the droplets. Pairs of CVP with smaller scales are preferable for achieving a better mixing. The influencing factors are listed in the following order in terms of the degree of their impact from the greatest to the least: the Sauter diameter of the initial droplets, the mixing tube diameter, the spray angle, the velocity of the inlet crossflow, and the vertical velocity of the initial droplets. A moderate droplet diameter, a smaller tube diameter, a moderate spray angle, a greater crossflow velocity and a moderate vertical velocity of the droplet are favorable for achieving a higher mixedness quality of the jet spray in a confined crossflow.

Joint Transceiver Design for OFDM Systems

Orthogonal Frequency Division Multiplexing(OFDM) is an effective technique to deal with a frequency selective channel since it can convert the channel into some flat fading subchannels.However,very different output SNR values of the subchannels will lead to poor bit error performance when a linear equalizer and Equal Bit Allocation(EBA) are adopted in OFDM systems.So,we proposed three novel nonlinear Joint Transceiver(JT) schemes based on Zero-Forcing(ZF) criterion and Minimum Mean Square Error(MMSE) criterion respectively,which can transform all subchannels of an OFDM system into subchannels with identical channel gain.Thus,EBA is equivalent to the Optimum Bit Allocation(OBA) for these subchannels.Numerical analysis helps us to obtain the theoretical approximate BER values of the JT scheme.Simulation results verify the numerical analysis and confirm that the performance of our proposed JT scheme greatly outperforms the traditional linear equalizer with EBA at moderate and high SNR values.

Orthogonal Array-based Uniform Latin Hypercube Design and Its Construction

Orthogonal array-based uniform Latin hypercube design（uniform OALHD） is a class of orthogonal array-based Latin hypercube designs to have the best uniformity. In this paper, we provide a less computational algorithm to construct uniform OALHD in 2-dimensional space from Bundschuh and Zhu（1993）. And some uniform OALHDs are constructed by using our method.

Application of Uniform Design and SVR for Culture Medium Optimization of Organophosphorus Degradation Bacteria

In order to improve the degradation activity of organophosphorus degradation bacterial strains, uniform design and support vector regression(SVR) are used to optimize the formula of mineral salt medium for the strains based on single factor experiment. The formula being used after SVR optimization is glucose 1.24 g/L,NH4 NO3 0.73 g, KH2 PO4 0.96 g, KCl 0.50 g, MgSO4·7 H2 O 0.50 g, MnSO4 0.02 g, CaCl2 0.04 g, distilled water 1 000 mL;the highest degradation rate is obtained at pH7.0(SVR, 76.68%), superior to that of the QSAR model(74.67%). The formula optimization method based on uniform design and SVR is very promising to apply in various multiple-factor and multiple-level formula optimization experiments.

Numerical Investigation of Tip Clearance Effects in an Axial Transonic Compressor

Numerical investigations of the Darmstadt transonic single stage compressor (DTC), in the Rotor1-Stator1 configuration, aimed at advancing the understanding of the effect of different rotor tip gaps and transition modelling on the blade surfaces are presented. Steady three dimensional Reynolds Averaged Navier Stokes (RANS) simulations were performed to obtain the flow fields for the different configurations at different operating conditions using the RANS-Solver TRACE. The stage geometry and the multi-block structured grid were generated by G3DMESH and a grid sensitivity analysis was conducted. For the clearance gap region, a fully gridded special H-grid was chosen. Comparisons were made between the flow characteristic at design speed, representative for a transonic flow regime, and at 65% speed, representative for a subsonic flow regime. The computations were used to analyse the flow phenomena through the tip clearance region for the different configurations and their impact on the performance of the compressor stage.

A design method and numerical study for a new type parabolic trough solar collector with uniform solar flux distribution

The non-uniform concentrated solar flux distribution on the outer surface of the absorber tube can lead to large circumferential temperature difference and high local temperature of the absorber tube wall,which is one of the primary causes of parabolic trough solar receiver(PTR)failures.In this paper,a secondary reflector used as a homogenizing reflector(HR)in a conventional parabolic trough solar collector(PTSC)was recommended to homogenize the solar flux distribution and thus increase the reliability of the PTR.The design method of this new type PTSC with a HR was also proposed.Meanwhile,the concentrated solar flux distribution was calculated by adopting the Monte Carlo ray-trace(MCRT)method.Then,the coupled heat transfer process within the PTR was simulated by treating the solar flux calculated by the MCRT method as the heat flux boundary condition for the finite volume method model.The solar flux distribution on the outer surface of the absorber tube,the temperature field of the absorber tube wall,and the collector efficiency were analyzed in detail.It was revealed that the absorber tube could almost be heated uniformly in the PTSC with a HR.As a result,the circumferential temperature difference and the maximum temperature could be reduced significantly,while the efficiency tended to decrease slightly due to the inevitably increased optical loss.Under the conditions studied in this paper,although the collector efficiency decreased by about 4%,the circumferential temperature difference was reduced from about 25 to 3 K and the maximum temperature was reduced from667 to 661 K.

Unbiased Condition of the Dispersion Effects Estimator in Unreplicated Two-Level Factorial Experiments

This article presents a general form of the estimator for identifying dispersion effects from unreplicated two-level factorial experiments,and shows that the widely used estimators such as the BH,MH,and AMH estimators are all special cases of the proposed one,designated as the G estimator.The unbiased condition of the G estimator is proved,and a lower bound of variance of the G estimator is provided.A simulation based on a realistic design illustrates the variation of the variance and MSE(mean square error) of the G estimator on different coefficients.This estimator may be more flexible and has better performance than other methods such as the BH and MH estimators by appropriately selecting the coefficients.

An Optimal Design Method Based on Small Amplitude Homogenization

An optimal design method for two materials based on small amplitude homogenization is presented. The method allows to use quite general objective functions at the price that the two materials should have small contrasts in their relevant physical parameters. The following two applications are shown: Stress constrained compliance minimization and defect location in elastic bodies.