The plastic forming is classified to the integral forming and the local forming, and the local is again classified to the point,the line,and the face forming.The comparisons of the advantages and disadvantages between the different forming techniques were made and the differences and features of the die forming and the flexible forming as well as the sectional forming and the continuous forming.Based on the above discussions,an effective,continuous,and flexible forming concept and a multipoint rolling technique by using the bendable,adjustable,and flexible rollers were proposed.A flexible rolling device for the freeform surface parts was developed and its experimental test was performed.The results show that the proposed concept is feasible and practicable.

Aluminumdoped zinc oxide (AZO) thin films have been deposited on quartz glass substrates by pulsed laser deposition from ablating metallic targets in oxygen atmosphere (11 Pa of oxygen pressure) and under a substrate temperature of 150 ℃. The structural, electrical and optical properties of these films were investigated as a function of Aldoping content in the film. It was observed that 1.37% of Al content in the film is the optimum doping content that makes the film to achieve the minimum film resisitivity and strong ultraviolet emission.

In the light of the poor antidive and antilift performances of a certain SUV,virtual prototype models of its front and rear suspension systems were built using the virtual prototype technique.Their kinematic and dynamic behaviors were revealed through simulations.The hardpoint of the guide mechanism of the front suspension system in the simulation model was modified under the condition of minimal possible modification of the components.The simulation results show that when the connection point of the lower control arm with the front body is pulled down by 20 mm,the vehicle antidive and antilift performances are improved by 80% and 30% respectively.In order to check the effect of the modification on the vehicle handling and stability,a complete virtual prototype model was built for the vehicle,the virtual prototype and real vehicle tests about the vehicle steadystate circular response were performed according to the China national standard,and both results were in a good agreement,validating the proposed virtual prototype model.Furthermore,the virtual tests of the steer release and step steer were performed on the complete vehicle virtual prototype model before and after the hardpoint modification of the front suspension system,showing that the modification has little effect on these performances.Therefore,the proposed modification on the front suspension hardpoint is acceptable.

To improve the steering easiness and road feel of the vehicle, an electric power steering（EPS） system with the steering torque as the control target was developed.A Simulink based simulation model was built for the vehicle EPS system on the basis of the analysis of the mathematics model of the EPS system. The simulation results show that the developed EPS system improves the steering easiness and road feel, at the same time, the control performance of the system does not change with the parameters of the system and has stable steering torque characteristic.

In order to investigate the engine combustion mode switchover between the spark ignition(SI) combustion and the homogeneous charge compression ignition(HCCI) combustion, a test engine was developed from refitting the ZS1105 series production diesel engine by modification of the compression ratio,the fuel supply system,and the intake system.By means of the adjustment of the intake charge preheating,the dual combustion mode switchover was realized smoothly on the test engine,which can be used as a platform for investigation of such combustion mode switchover.The criteria to distinguish the upper and lower boundaries of HCCI combustion were defined and the operation regions of the ethanol SI-HCCI-SI combustion modes were obtained.

Based on the available experimental results, a simulation study on the temperature field of a controllable active thermo-atmosphere combustor was presented with a non-reacting central jet issuing into a vitiated co-flow. The influences of the co-flow velocity, the co-flow temperature, and the jet issuing velocity on the thermal atmosphere have been evaluated. The results are useful for the optimization of experimental scheme design.

In the applied research of the distributed power station system, according to the principles of the total energy system and the cascade utilization of energy, a thermodynamic model of the endoreversible combined power cycle with the heat leak and the irreversible heat transfer process operating between the internal combustion engine and the Stirling engine was investigated based on the theory of the finite time thermodynamics. The expressions about power output and thermal efficiency were derived, which were denoted by the compression ratio and temperature ratio and other parameters. The relations between power output and efficiency and the effect of heat leak and other parameters on the performance of the combined cycle were analyzed by detailed numerical examples. The obtained results enlarge the scope of the theory of the finite time thermodynamics and provide a basis for the performance study of the practical combined power cycle.

The concept of road segment congestion quotiety was introduced based on the information of travel time and segment length,and the traffic state matrix of the region was built.The congestion quotiety of the whole coordination region was calculated by averaging the weighted quotieties of the segments and was used to judge the traffic state of the region.A simulation road network was built based on the simulation software VISSIM to check the feasibility of the proposed method.Simulation results show that the method based on the congestion quotiety is simple and practical,can judge the traffic state,provides a background environment for the development of the coordination strategy.

Combining with the road design norm,considering the dynamic performance of the vehicle and the climatic environment of the road,the road safety in the design drawing stage of the second class highway from Antu to Erdobaihe,Jilin Province,was evaluated by the simulation study of the different road alignments.The essential factor analysis about the effect of the road alignment on the traffic safety was studied.A concept of the mixing of the road alignment norm with the safety standard was suggested to provide a reference for the examination and verification of the road design safety.

An approach to estimate the average link travel time (ALTT) using taxi GPS data was developed. In this approach the GPS data is calibrated first. Then according to the amount of GPS data samples different estimation methods are employed. For large samples, the average value of the samples is used to estimate ALTT; while for small samples, the order statistics is used. The confidence interval and the degree of confidence of the estimated ALTT are also calculated. The estimation of ALTT using real taxi GPS data was compared with that using loop data, and result shows that the estimation error using GPS data is lower than that using loop data, indicating that the proposed approach can be applied in practice.

The test methods of VCADRF and VCAAC were introduced in the light of the application of gradation design for SAC13 in an actual project.The basic principle is the VCA formed by the coarse aggregate is filled with the fine aggregate,the filling,the asphalt and the preconcerted void content.In the paper the selected tentative gradation was tested with the VCADRF method and the results showed it met the requirement of the close skeletondense structure.Then the VCAAC method was applied to test the compacted specimen.The results indicate that the void content of the compacted specimen of this gradation is too high,having need of adjustment according to the VCAAC method to satisfy the requirement of close skeletondense structure.The demonstrated example proved the convenience and the flexibility of the abovementioned test method,and the adjusted design gradation was characterized by the good hightemperature stability,water resistance,skidresisting capability and water impermeability.

Based on the research of the finishing machining process of the surface,a new polishing technique,the hydrodynamic suspension polishing was proposed.A hydrodynamic suspension test system based on a threecoordinates NC milling machine was developed and the polishing experiment was performed on the optical glass(K9). The polishing process was analyzed and the removal mechanisms of the brittle material were discussed.It is found that the surface roughness is reduced by the plastic deformation and plastic removal under the impacting and shearing of the abrasive.The results show that the polishing by the hydrodynamic suspension with colloided SiC nanoabrasive is valid for the nano meter cutting and the nanometer supersmooth surface processing.

A novel inertial piezoelectric actuator was proposed which was characterized by controlling the friction between the moving mechanism and the supporting faces.The working principle of the actuator was described and a dynamic model of the inertial moving mechanism was built.The kinematic simulation of the mechanism was performed by means of the software Matlab and the simulated displacement curves were acquired.An inertial piezoelectric actuator to accomplish the linear reciprocating movement was developed and its performance test was carried out.The test results were in good agreement with the simulated ones.The results show that the scheme of the novel moving mechanism is valid and the developed actuator is characterized by stable moving performance,high resolution,step uniformity,and linearization of speed curves.

An equation for calculating the bearing stresses under different operating condition was derived based on the Andrews dynamic model for the bearing clearance.The equation was applied to the simulation of the impact test bench with pendulum and a simulate program was compiled.The simulated results were compared with those of the simulation without consideration of the bearing clearance and those of the experiments,showing the simulation result based on the bearing clearance model was closer to the experimental one than that without considering the clearance.The suggested bearing clearance model can be used to the dynamic simulation of the other mechanical systems.

An adaptive slicing algorithm for the rapid prototyping of the functionally gradient material components was suggested by considering the informations of both the geometry and the material distribution.The layer thickness was computed according to the area variation rate between the two adjacent slices,the material variation gradient and the material resolution.After slicing,the material informations of the component were stored discretely by means of the material composition function and the material resolution.An example of the slicing was given.

The equivalent method was combined with FEM in this paper, the equivalent dynamics model for RRR-RRP II degree bar-group was established on the basis of EFEM and the dynamics differential equation was given. Accounting the key point of building dynamic model to be setting up 2 steps tensor Tia and 3 steps tenser Tibc and assembling the equivalent mass matrix. The process to establish dynamics model was simple and rapid and the generality is well. Analyzed and solved the dynamics model to the RRRRRP 6 bars mechanism that has been given, the curves of dynamics responding of it was obtained. It is testified by dynamic analysis example that the method presented by this paper has a greater advantage and higher efficiency comparing with traditional method as describing the differential equation and calculating it by using computer and having redeemed the defect that it is easy to make as doing dynamic analysis because of the complexity of kinematics relation of system and the personal skill.

A fault tree model was built for the fault diagnosis of the wheel loader hydraulic system based on the fuzzy fault tree analysis method.An expert system rules database was compiled,an optimized search strategy was defined,and a fault diagnosis module consisting of a knowledge database and an inference unit was developed.The developed wheel loader hydraulic system fault diagnosis system was embedded into a wheel loader made by Shandong Lingong Construction Machinery Co.,Ltd.The test results show that the developed system is simple and easy to use,reliable and stable, and can be applied in other construction machineries after corresponding refit.

The bionic non-smooth technique was applied to the centrifugal water pump to modify the liquid flow structure in the pump to reduce the hydraulic drag in the liquidsolid interface,resulting in the enhancement of the pump efficiency.A test system was developed and an L9(34) orthogonal experiment design was used to test the pump efficiency under different flow rates,lifts,electric voltages and currents.The results show that some models with the bionic nonsmooth working surfaces are very effective,with the pump efficiency enhancement of 3.45%,that means its growth rate of 6.81%.The major and minor factors in two flow rate segments and their optimized levels were found,providing a preliminary basis for the application of the bionic nonsmooth technique to the contrifugal water pump.

The surface wettability of the butterfly wings of fortyeight species (thirtyfour genera, eight families) was qualitatively and quantitatively studied. Scanning electronic microscopy observation shows that the length of the wings' scales is in the range from 55 μm to 150 μm and the width is in the range from 35 μm to 105 μm. The gibbosities on some scales with height of 200 nm to 900 nm were also observed. The static contact angle of water droplet on the surfaces of the wings was measured using an optical contact angle measuring system. Results show that the contact angle is in the range from 134.0° to 159.2°, which indicates the surfaces of the wings are superhydrophobic. The rolling angle along the scale arrangement is smaller than 3° and against the arrangement is larger than 65°, indicating that the scale structure of the wings is evidently anisotropic. The wettability of the surface of the wings is attributed to the coeffects of the micro and nanostructures of the scales.

In order to manage the open ontologies effectively, coordinate the ontologies of different fields, even with the semantic difference of the ontologies of the same field, providing a based Multi-Agent Ontology Management Architecture(MOMA), it has the following characteristic: ①The concept included in Agent communication entity adopts ontology method to describe, divide the social management of one loose layer to ontologies. ②Setting up an open ontologies environment of stepping the platform dynamically, enable Agent to obtain, share and manage various kinds of ontology resources, and it can offer Agent service according to the ontology service needed too. It also can support other questions, such as the negotiation, coordination and communication of different Agents.

A region movement model is proposed on the basis of the conceptual neighborhood structure in Region Connection and Calculus (RCC) and the fuzzy sets theory. This model is able to deal with the space regions that can move or change with time in a dynamic environment. Also on the basis of the path consistency algorithm proposed by Allen, a fuzzy reasoning algorithm of path consistency is presented for this model.

A restricted BAN classifier learning algorithm－GBAN based on genetic algorithm is proposed. Genetic algorithm was used to study the network structure. The structure of TAN classifier was extended by restricting the complexity of the structure of BAN classifier. And then a restricted BAN classifier is obtained. As far as this classifier's structure studying, the fitness function based on logarithm likelihood was designed. The code scheme of network structure , and the corresponding genetic operators are designed. As a result, the algorithm converges on the overall optimal structure. The experimental result indicated that GBAN algorithm has good classifying effect and is more accurate than TAN classifier when the relationship between attributes of a data set is relatively complicated.

With the continuous increase in available web services, there is a growing demand to compose web service workflows, thus to perform complex business processes. This paper proposes a model for constructing semantic web service workflows, in which the OWLS/UDDI Matchmaker is applied. Procedures of the construction of semantic web service workflows are presented. Using the semantic descriptions of web services this model is capable to search and match web services automatically, and to realize the automatic construction of web service workflows.

A virtual instrument software automatic generation framework is developed. This framework eliminates the demand for users to possess the capability of software design in developing virtual instrument software. It is based on the modules of virtual instrument software functions and is driven by users' requirements. The functional module sets and the relationship among modules are established. By analyzing the user demand for virtual instrument software, the design process is dispelled in developing software under current graphic development environment. The feasibility of the framework is verified on a self developed software platform.

On the basis of the simple graph presentation of Redundant Arrays of Inexpensive Disks (RAID) doubleerasurecorrecting codes and the judgment theorem of doubleerasurecorrecting data layout, the optimal redundancy data layout theorem was proposed. This theorem proves the lower bound of the number of disks of doubleerasurecorrecting data layout of full-2 code (corresponding to complete graph), it also proves the structure features of the optimal redundancy data layout. A construction method for optimal redundancy doubleerasurecorrecting data layout based on perfect 1-factorization was also presented. Compared with other methods of doubleerasurecorrecting codes/data layouts, the reliability of this method is higher, the update penalty and the redundancy are optimal, and the complexity of encoding/decoding is lower. Moreover the applicability of this construction method is high.

Since common scheduling analysis algorithms in the model of distributed embedded systems constructed by Petri nets based on resources do not consider parallel problems, they can not get correct scheduling length. Aiming at this deficiency of the conventional scheduling analysis algorithms, a new algorithm adapting to the distributed embedded systems was presented by the introduction of time base and firing field relative to initial state class.

To address the problem of low accuracy of timing characterization caused by the large scale of VLSI circuits, a new timing characterization method to improve the accuracy of timing models in hierarchical circuits is proposed. The concepts of the local false path and global false path are introduced to the method. Experiment results show that, for some benchmark circuits with thousands of gates, this method is capable to generate timing models with very accurate path delay. Results also show that, without resorting intensive search-based or symbolic techniques to account for the circuit functionality, this method can provide correct calculation results and eliminate the vast majority of the false paths caused by mode connections in hierarchical design. This ensures the accuracy of circuit delay.

A hierarchical object detection method based on vision saliency was proposed for automatic remote sensing image interpretation. A bias function for each single feature was defined, and a compound bias function for synthesis of vision saliency image computed by each function was also defined. For the low resolution original image, each bias function was used to generate a vision saliency image corresponding to a certain feature. Then synthesized vision saliency image was computed using the compound bias function from these vision saliency images. Furthermore, the area where the compound bias function was maximum was located as the most promising object area. These three steps were repeated over the located areas in higher resolution level image until the objects were found or assuredly absent. Experiments show that this method can skip over the region where there is no object as early as possible, get better detection result compared with traditional methods, and cost much less computational expenditure, which proves to be a good basis for later object recognition.

A new method for medical image classification was proposed which is based on fuzzy support vector machine. In this method a fuzzy degree is defined for each training point, and the level of the fuzzy degree is determined by the amount of information in the subjecting function. Thus the traditional support vector machine algorithm is improved. Experiments on images with different noise levels were conducted and results show that the proposed method is able to classify the brain tissues in the MRI images with high precision. In application of this method the cost and time of computation can also be reduced.

To solve the problem of how to choose appropriate kernel bandwidths in mean shift filtering, the relationship between mean shift and scale space filtering was analyzed. The analysis showed that when normal kernel is adopted, mean shift is a clustering method based on Gaussian scale space filtering. On the basis of this analysis, a non-linear scale space filtering method relying on mean shift was presented. The proposed filter was controlled only by kernel bandwidth in spatial domain that was chosen at every sample point adaptively by local signal characteristic. Experiments show that his method overcomes the defect of blocky effect in mean shift filtering, and is superior to mean shift filtering, Gaussian scale space filtering and median filter.

An optimal control algorithm for T-S fuzzy systems with time domain hard constraints including control input constraints and state constraints was proposed, in which the output energy was chosen as performance function with the initial state under a bound. This algorithm was realized by minimizing the invariant ellipsoid of the closed loop system. Some sufficient conditions dependent on the invariant set for the satisfaction of the time domain hard constraints were also derived. The multi-objective design was finally deduced to solve an optimization problem with a set of nonlinear matrix inequalities. The solution for this problem was given, and the feasibility of this scheme was discussed. Simulation results for the application in inverted pendulum validate this control algorithm.

An self-correction PID chaos control scheme consisting of neural network identifier and neural network controller was presented. Davidon least square method was used to learning weights of two neural networks. Considering dynamic characters of multi-variable chaos system, the chaos system can implement fixed point control and synchronization by controlling single variable. The effectiveness of the proposed control method was verified by numerical simulations.

In order to improve LOG filter performance for stereo vision, a four step pretreatment method was proposed. This method was divided into 4 steps: wiping out random noise by mean filter, reducing Gaussian noise by Gaussian filter, balancing brightness difference between stereo image pair through histogram equalization, and enhancing image edges and details by Laplace sharpness. Experimental results show that this four step pretreatment method can eliminate image noise and reduce brightness difference effectively.

In the light of the features of the micro-manipulations,i.e.,the operation space is infinitesimal,the track is estimable,and both the required precisions of the motion and positioning are high,a vision servo control algorithm was proposed based on the motion estimation and the image match.The algorithm introduces the motion estimation into the vision servo system to perform the regional search instead of the match search in the whole video space by estimating the next possible location of the end actuator.Therefore,the time needed for image match is reduced significantly,the real-time feature of the algorithm is enhanced,and its practicability and robustness are improved.The proposed algorithm was used to solve the problems associated with the precise displacement and positioning in the cell injection micromanipulation,and the realtime tracking and motion positioning were realized with good precision and robustness.

Based on a magnetic field controlled electron spin quantum system, utilizing the mechanism of realizing the unitary evolution of the quantum state through the quantum gate, an optimal control strategy to prepare an arbitrary quantum state is proposed. The procedure is as follows: after the modeling of the electron quantum bit system and the control field, applying the Liealgebra, the control output is calculated by analogy with the classical optimal control method to achieve the target. The theoretical analysis and the simulation results showed the superiority of the proposed strategy.

A method to design a random clock error test bench in verilog was introduced which combines the simulating verification of the video mode detector. This method utilizes and improves the random number generation function provided by verilog to generate a set of random vectors, which approximately obey Gauss distribution. Simulation on the module can be carried out by imitating the real timing of the input signals. Simulation results indicate that the proposed method can effectively verify the impacts of the clock jitter and the random phase skew between signals on the design. The method can be used in other random tests.

According to the reliability theory about the mean time to failure and the system reliability evaluation,the reliability of the high capacity storage battery arrays with different connection schemes was evaluated.It is found that the reliability of the array with parallelseries connection is superior to that with seriesparallel connection or the combination of series and parallel connections.The result is significant for the design of power engineering and the reconstruction of PC system in the electrical power system.

A VLSI architecture for the implementation of image compression coding using pipeline and time division multiplexing technology in lifting wavelet transform and SPIHT was proposed. This structure uses less hardware resource, and increases operation speed and stability. SPIHT algorithm with multi-threshold quadruple parallel processing can enhance coding efficiency. The FPGA realization based on this structure can satisfy the requirement of image storage and transmission.

To withstand various attacks such as cropping and image mosaic in geography information copyright protection, a transform coefficient polarity based watermarking algorithm was proposed. In this algorithm binary were constructed to represent watermark bit 0 and 1 respectively. Since polarity is almost invariant, watermarking can be embedded by modulating the correlation value between polarity sequence and binary sequence. Special classification based polarity sequence distributing in whole DCT domain may increase robustness against cropping. Moreover, watermark retrieval scheme is resilient to image mosaic. Experiments show that the proposed method has good watermark imperceptibility, and resistant to light variety from -90 to 90, contrast variety from -90 to 80, scale, cropping and image mosaic, and it is robust for combined attack.

A semi-blind channel estimation scheme for MultipleInput Output MultipleOutput (MIMO) systems was proposed which is base on the timedomain precoding. This scheme is able to overcome the shortcoming of multiuser interference and to improve the utilization efficiency of high frequency band. In comparison with traditional ZF frequencydomain precoding, this timedomain precoding scheme shows better performance especially at low SignaltoNoisy Ratio (SNR). Also the complexity of this precoding scheme is lower and it is easier to implement. Theoretical analysis and performance simulation were conducted that validate the advantages of the proposed scheme.

On the basis of the optical navigation scheme with objective celestial body information, the encounter task is realized. The accuracies of orbit determination and ephemeris are improved by using navigation camera. On this basis, the Unscented Particle Filter (UPF) is introduced to the navigation scheme to overcome the shortcomings of large initial estimation errors and non-Gaussian distribution of the state or measurement errors that may exist in the determination of the deep space orbit. Taking the advantage of UPF in tackling nonlinear and non-Gaussian problems, the disadvantages in the approximate treatments of nonlinear and non-Gaussian state model and measurement model by traditional EKF, UKF and PF schemes are also gotten over. The feasibility of the proposed scheme is demonstrated by numerical simulations.

Based on the survey and measure of the three typical underground car parks in Changchun,Shenyang and Bejing, carbon monoxide concentrations did not exceed the standard limits in the condition of no mechanical ventilation. In the winter of severe cold or cold zones, natural ventilation rates of buoyancy driven flow can meet ventilation requirement basically in underground car park. In summer the natural ventilation rates of buoyancy driven flow decrease,in the meantime the cold start and idle emissions of cars are significantly lower in summer than in winter. Calculation of natural ventilation rates by tracer gas decay method was proposed, and definition of indoor criterion of carbon monoxide concentration limit and the carbon monoxide emission from cars were discussed.

A constant wavelength light absorption sensor was developed using the feature that the biologically active formazane solution extracted from the waste water has the maximal absorption peak for the light with wavelength 485 nm.This sensor transforms the biological activity of the extracted solution into the electric signal.Based on it, a portable intelligent waste water activity determinator was developed,with which the biological activity index data can be read directly,stored and printed out.The precise calibration of the instrument was realized by the appropriate design of its hardware and software.

A model was developed to estimate the traffic pollutant concentrations along roadside in one step with artificial neural network technique. First, factors affecting roadside pollutant concentrations were classified into three categories. And then a structure of the artificial neural network was designed to analyze the mathematical relationship between the pollutant concentrations and the factors such as traffic flow attributes, meteorological condition and road spatial configuration. At last great amount of data were collected for training the neural network to obtain a forecasting model which had a high accuracy.

Polysilicicferric (PSF) coagulant was prepared by copolymerization method. The oxidization of PSF on dissolved organic matters (DOMs) (humic acid HA as surrogate matter for DOMs) was studied. The capability to remove organic matters by PSF was investigated in comparison with that by polyferric aluminum (PFA). The mechanism of removing DOMs by PSF was analyzed primarily. Results show that PSF possesses oxidizing function; it changes the molecular structure and the surface nature of HA; it also alters the nature of the interface between HA and water solution, which leads to the improvement of the adsorption of HA. The pH value impacts the complexation mode between HA, PSF and kaolin. The oxidization/adsorptioncharge neutralization/destabilization is the precondition of the coagulation mechanism of PSF, bridging is the necessary condition, and sweeping is the supplement in coagulation process.

Displacement field function is built using meshless methods as trial and test function to seek partial differential equation, including base function and weight function selecting, shape function and its derivative computing. At the same time, the influence of the size of influence domain of weight function on computational precision of displacement field function is analyzed. And the rationality of displacement field function is validated by definite examples.

Based on the magneticelasticity nonlinear kinetic equations, physical equations, electrical kinetic equations and the expression of Lorentz forces, the magneticelasticity kinetic buckling equation of a currentcarrying plate applied mechanical load in a magnet field were derived. The equation was transformed into the standard Mathieu equation by using Galerkin method. Thus, the magneticelasticity kinetic problem was changed into a problem of solving the Mathieu equation. The criterion equation of the magneticelasticity kinetic problem was gotten by the eigenvalue relation of Mathieu equation's coefficient λ and η. The rectangular plates made of different materials were taken as calculating examples. The curves of the relations among the relative parameters are obtained when the plates are in the situation of critical buckling. The analysis and discussion of the calculating conclusion are given out.

A method to measure the corn moisture by acoustic analysis was proposed. The sound wave was acquired by a computer system equipped with a sound sensor. After the filtration of noise from the sound signals, the features, such as the acoustic intensity, spectrum energy and peak value were calculated to describe the sound waves. Multiple linear regression, binominal regression and neural network methods were applied in the analyses. Results show that there exist prominent relationships between these features and the moisture content, and among these features acoustic intensity is the most significant one. Using these relationships the corn moisture content can be accurately measured. The predicted data by the binominal regression model is the closest one to the real data.

The effects of Coix lachrymaljobi extract (EEC) on the mycelia growth, bioactive substance production and antioxide activity of Ganoderma lucidum in submerged fermentation were investigated. Addition of EEC at the level of 6 mg·L-1 led the biomass yields to increase from 8.29 ± 0.32 g·L^-1 to 10.19 ± 0.51 g·L^-1. The maximal yields, 0.72 ± 0.03 g·L^-1, of polysaccharides was obtained at EEC concentration of 10 mg·L^-1, and the maximal yields, 0.12±0.01 mg·L^-1, of ganoderic acid was obtained at EEC concentration of 14 mg·L^-1. Comparing with no EEC the increases in biomass, polysaccharides and ganoderic acid yields were 14.6%, 34.21% and 254.2%, respectively. The addition of EEC at the level of 14 mg·L^-1 also enhanced the free radical scavenging activity of Ganoderma lucidum, particularly the activity on superoxide anion radical, and the maximal inhibition rate reached 43.2±0.21%, which is about 61.6% higher than that with no EEC.So EEC could visibly enhance the growth,production and activity of scavenging free radical of Glucidum.