线性、非线性与广义线性回归模型

本文研究了线性、非线性与广义线性回归模型之间的关系，澄清了一些重要的统计概念，并对广义线性模型作了简要介绍。广义线性模型是当前统计科学发展的一个方向，统计研究与应用的许多高、新领域都在很大程度上以广义线性模型为其理论基础。将其介绍给我国广大读者对于促进我国统计事业的发展，建设大统计学科有重要意义。

Modeling and Simulation of the Characteristics of Pneumatic Cushion Cylinders

The analysis of the characteristics of the cushion process of the pneumatic cushion cylinder is presented, and the nonlinear model of pneumatic cushion cylinders is built in the form of nonlinear differential equations. Besides, through the simulation of the pressure in the cushion chamber, the characteristics of the pneumatic cushion cylinder are obtained, which helps to understand the performance of the pneumatic cushion cylinder and improve or design the better cushion structure.

Determination of Melamine by Infrared Spectroscopy Based on Nonlinear Modeling

[Objective] The aim was to develop a nonlinear model of quantitative analysis of melamine content by infrared spectroscopy and provide theoretical basis for the nondestructive detection of melamine. [Method] According to dynamics,mathematical modeling and optimization theory,linear and nonlinear models were respectively set up by taking an absorption peak of 1 550 cm-1 as characteristic absorption peak. [Result] The correlation coefficient of nonlinear model was 0.922 7 and the recovery was 96%,which showed that the nonlinear model was more accurate than linearity model with correlation coefficient of 0.904 9 and recovery of 557%. [Conclusion] It is feasible to determine melamine content by using the nonlinear model quantitatively.

Direct adaptive fuzzy control based on integral-type Lyapunov function

A new scheme of direct adaptive fuzzy controller for a class of nonlinear systems with unknown triangular control gain structure is proposed. The design is based on the principle of sliding mode control and the approximation capability of the first type fuzzy systems. By introducing integral-type Lyapunov function and adopting the adaptive compensation term of optimal approximation error, the closed-loop control system is proved to be globally stable, with tracking error converging to zero. Simulation results demonstrate the effectiveness of the approach.

Study on Growth and Development Rules and Growth Curve Fitting of Minghua Black Minks

Growth curves of Minghua black minks at 0-180 days old were fitted and analyzed by using two growth models Logistic and Gompertz. The results showed that the growth curves of Minghua black minks could be fitted very well by Logistic model and Gompertz model （the degree of fitting FF≥0.99）, but Gompertz model was better at fitting and predicting their weight.

Effect of Copper and Zinc on Accumulation of Vitamin E in Wheat Embryo-derived Callus

[Objective] The aim was to study the effect of the content of copper and zinc on in medium the vitamin E accumulation in wheat embryo-dreived callus.[Method] The mathematical models were established to describe the growth kinetics and the vitamin E accumulation in wheat embryo callus cells.With the aim of getting the highest accumulation of the secondary metabolite Vitamin E,the optimal combination of copper and zinc in medium was confirmed by testing.[Result] The results showed that the production of vitamin E in B5 medium reached the highest value with 2.0 mg/mL ZnSO4·7H2O and 0.1 mg/mL CuSO4·5H2O.The fitting degrees of kinetic models of vitamin E accumulation and cell growth were 97.53% and 95.60%,respectively,which indicated good nonlinear relationships.[Conclusion] Both copper and zinc could affect the accumulation of vitamin E in wheat germ callus,and Copper showed more prominent effect than Zn.Synergism existed in low copper and zinc concentration,and the inhibitive effect enhanced with the increase of the concentrations.

NONLINEAR DYNAMIC ANALYSIS OF THE PLATFORM STRUCTURES SUBJECTED TO ICE EXCITATION(PART Ⅰ): EQUATION OF MOTION

In this paper,we established the equation of motion of a platform structure with four legs by way of Hamilton' s Principle and obtained the nonlinear dynamical model of platform structures. By using Laplace transformation we obtain the mode shapes of the system. The nonlinear model of the platform structure in this paper provded an accurate way of analysis for the engineering prediction of the dynamical characteristics of the platform and the understanding of its vibrational mechanism.

Adaptive Volterra series model for nonlinear sensor compensation

A novel performance enhancement method of nonlinear sensor based on the Volterra series model is proposed. The Volterra series model, which is considered a nonlinear filter that can reduce sensor noise, presents an effective way for modeling and compensating a nonlinear sensor. In the experiment, the low accuracy pressure sensor MPX10 was used as the actual object, and higher accuracy sensor MPX2010 was used as the reference to provide the necessary teaching data for training the Volterra model. The simulation shows that the accuracy of MPX10 changes from 0.354-0.420 to 0.041-0.053 after the Volterra filter is applied. Obviously this scheme can effectively improve the sensor performance. Moreover, the scheme provides greater accuracy and environmental suitability for a nonlinear sensor.

OTA MACROMODEL AND QUARTER-SQUARE MULTIPLIER

This paper presents a new kind of macromodel of OTA,which can be used to solve the problem in which the two port macromodel couldnt reflect some functions of the OTA.The new model also opens up a new way for the simulation of the OTA circuit.This paper discusses the way of designing this model and simulating it in SPICE.The result proves its reasonable design and its simplicity in structure.In the application of this model,we design a complete symmetric double differential quarter square OTA multiplying unit by using four three port OTA macromodels.It successfully solved the problem of the unsymmetry of two input ports in an OTA multiplying unit.This result fully agrees with the experiment.

STUDY ON NON LINEAR VIBRATION RESPONSE OF PLATE GATES *

This paper deals with the vertical vibration, the horizontal translatory and flexural vibrations of plate gate with a flat bottom under the action of unstable submerged underflow. Based on a non linear resonant oscillator model by which the coupled excitation mechanism between the unstable vortices and gate motion can be simulated, the differential equations for the three types of gate vibrations are established. The parameters in the equations are determined by model experiments. Then, the steady state non linear responses of the three types of gate vibrations are calculated. The calculation results are in good agreement with the experimental data of gate vibrations with different reduced velocity and gate parameters by previous researchers.

Nonlinear elastic deformation of magnesium and cobalt by Preisach-Mayergoyz model

A classic hysteretic model, Preisach-Mayergoyz model （P-M model）, was used to calculate the nonlinear elastic deformation of magnesium （Mg） and cobalt （Co）. Mg and Co samples in cylinder shape were compressively tested by uniaxial test machine to obtain their stress—strain curves with hysteretic loops. The hysteretic loops do have two properties of P-M hysteretic systems： wiping out and congruency. It is proved that P-M model is applicable for the analysis of these two metals’ hysteresis. This model was applied on Mg at room temperature and Co at 300 ℃. By the P-M model, Co and Mg nonlinear elastic deformation can be calculated based on the stress history. The simulated stress—strain curves agree well with the experimental results. Therefore, the mechanical hysteresis of these two metals can be easily predicted by the classic P-M hysteretic model.

THE MODEL VALIDATION OF DYNAMIC NEURAL NETWORKS

This paper investigates the problem of the model validation in identifying discrete-time-nonlinear dynamic systems by using neural networks with a single hidden layer.Based on the estimation theory,a synthetic error-index(SEI)criterion for the neural network models has been developed.By using the powerful training algorithm of recursive prediction error (RPE),two simulated non-linear systems are studied,and the results show that the synthetic error-index criterion can be used to verify the dynamic neural network models.Furthermore,the proposed technique is much simple in calculation than that of the effective correlation tests.Finally,some problems required by further study are discussed.

Identification of Hammerstein Model Using Hybrid Neural Networks

The identification problem of Hammerstein model with extension to the multi input multi output (MIMO) case is studied. The proposed identification method uses a hybrid neural network (HNN) which consists of a multi layer feed forward neural network (MFNN) in cascade with a linear neural network (LNN). A unified back propagation (BP) algorithm is proposed to estimate the weights and the biases of the MFNN and the LNN simultaneously. Numerical examples are provided to show the efficiency of the proposed method.

DESIGN OF A NOVEL NONLINEAR DISORETE-TIME CONTROLMETHOD FOR SWITOHING OONVERTERS

Based on the high frequency and large signal discrete-time model for switching converters,a nonlinear discrete-time control method is designed on the principle of dynamic programming. In one switch period the optimal duty ratio is got in accordance with a minimum error of the output voltage,then the duty ratio is checked and regulated with the energy function in order to make the global performance best. In thisway the complete objective function is obtained. Based on the objective function,the fast and accurate control effect is accomplished.

ENHANCED UNSTEADY AND NONLINEAR ROTOR WAKE MODEL FOR REAL-TIME FLIG HT SIMULATION

WT5 'BZThis paper presents an unsteady and nonlinear wake model based on th e widely used Peters He finite state dynamic wake model with improvements. The swirl component in the tip trace plane (TTP) can be predicted, nonlinear items are added into the linear theory, and the old small angle assumption use d in matrix prediction is removed. All of these enha ncements are aimed at the low speed flight phase and formulations for the induce d velocity field just in the TTP frame are derived. The corresponding FORTRAN pr ogram is tested and optimized for the real time applications on PCs.

SOLVABILITY RESULTS OF A CONDITIONAL INPUT-OUTPUT EQUATION BASED ON A TYPE OF NONLINEAR LEONTIEF MODEL

A class of nonlinear and continuous type Leontief model and its corresponding conditional input-output equation are introduced, and two basic problems under the so called positive or negative boundary assumption are presented. By approaches of nonlinear analysis some solvability results of this equation and continuous perturbation properties of the relative solution sets are obtained, and some economic significance are illustrated by the remark.

Genetic Algorithm-Based Estimation of Nonlinear Transducer

This paper describes an innovative, genetic algorithm based inverse model of nonlinear transducer. In the inverse modeling, using a genetic algorithm, the unknown coefficients of the model are estimated accurately. The simulation results indicate that this technique provides greater flexibility and suitability than the existing methods. It is very easy to modify the nonlinear transducer on line. Thus the method improves the transducer's accuracy. With the help of genetic algorithm (GA), the model coefficients' training are less likely to be trapped in local minima than traditional gradient based search algorithms.

Velocity of a Liquid Drop with a Sorption-Controlled Surfactant

The terminal velocity of a liquid droplet settling in a sulfactant solution has been studied for the non-linear adsorption Langmuir frameworks accounting for monolayer saturation and non-ideal surfactant interactions. Most prior research uses a linear adsorption model which cannot capture these effects, The Maragoni migration of a liquid drop settling through a surfactant solution is examined by using Langmuir framework. The solution concentration Ceq is assumed large enough for the surfactant mass transfer to be adsorption-controlled. Langmuir model generates non-linear Marangoni stresses which diverge in the limit of approaching ∝, strongly retarding U'.

Diagnostics in generalized nonlinear models based on maximum L_q-likelihood estimation

In order to detect whether the data conforms to the given model, it is necessary to diagnose the data in the statistical way. The diagnostic problem in generalized nonlinear models based on the maximum Lq-likelihood estimation is considered. Three diagnostic statistics are used to detect whether the outliers exist in the data set. Simulation results show that when the sample size is small, the values of diagnostic statistics based on the maximum Lq-likelihood estimation are greater than the values based on the maximum likelihood estimation. As the sample size increases, the difference between the values of the diagnostic statistics based on two estimation methods diminishes gradually. It means that the outliers can be distinguished easier through the maximum Lq-likelihood method than those through the maximum likelihood estimation method.

Nonlinear Model Predictive Control Based on Support Vector Machine with Multi-kernel

Multi-kernel-based support vector machine （SVM） model structure of nonlinear systems and its specific identification method is proposed, which is composed of a SVM with linear kernel function followed in series by a SVM with spline kernel function. With the help of this model, nonlinear model predictive control can be transformed to linear model predictive control, and consequently a unified analytical solution of optimal input of multi-step-ahead predictive control is possible to derive. This algorithm does not require online iterative optimization in order to be suitable for real-time control with less calculation. The simulation results of pH neutralization process and CSTR reactor show the effectiveness and advantages of the presented algorithm.