The low oxygen pressure self-fused Al-Cr coatings were made on the surface of the samples from the Ti-6Al-4V alloy. The effects of processing parameters such as fusing time, etc, on the formation of the coating and the effects of the Al-Cr coating on the cyclic oxidation behavior of the alloy in the air at the temperature of 873 K were investigated by observing and analyzing the surface morphology and the cross-section structures of the coating. The results show that the thickness of the inter-diffusion zone can be controlled by the fusing time, and processing for an appropriate fusing time may greatly improve the adherence of the coating to the substrate and enhance the cyclic oxidation resistance of the coating. The oxidation resistance of the coated sample is almost 300 times higher than that of the uncoated sample.

The mechanism of brittle fracture of hot-forged cast dies was discussed and the effective solutions were put forward. The brittle fracture of hot-forged cast dies was attributed to imperfection of materials and processes, which results in unstable lifespan. Through optimizing materials and processes, brittle fractures can be avoided. The early brittle fracture seldom occurs, which is resulted from abnormal heat treatment and operation. Impact fatigue fracture is the dominant pattern in brittle fractures of cast die which is caused by the fabrication defects, such as sand inclusion, micro-crack and thermal fatigue crack, which are taken as initiated crack to propagate to fracture.

The application of the artificial intelligence such as artificial neural network, fuzzy control, expert system and intelligent agent etc. In the resistance spot welding including its processing parameters design, process control, quality prediction and evaluation were reviewed. The artificial intelligence applied to resistance spot welding was more powerful on solving the nonlinear and ill-structured problems. Integrating various artificial intelligence technologies and making full use of their respective advantages have become the current trend.

By means of the quadratic regression combination design process, the regression equations relating the wettability, the mechanical property and the corrosion resistance of the solder to its element compositions were established. The influences of the elements on the wettability, the joint shearing strength and the corrosion resistance were analyzed. The results prove that adding more Zn improves the joint shearing strength, but at the same time the wettability and the corrosion resistance decline; adding appropriate amount of Bi can improve the wettability, joint shearing strength and the corrosion resistance distinctly. The interaction of Zn and Bi also affects the performance of the solder. After considering all the effects of solder ingredients to the melting point, the pasty range and so on, the contribution of Sn-Zn-Bi Pb-free solder used in electronic industry has been optimized. The result is 8%～9% for Zn and 3%～4% for Bi.

Based on the summarization of the die-sets for flashless-die-forging at home and abroad, double-action hydraulic die-sets for the flashless-die-forging were developed.The opening and closing of the mould is performed by an upper and a lower hydraulic cylinders.The movement of the upper and lower punches are controlled by a four-barlinkage.The developed die-sets were tested in an 8000 kN frame type hydraulic press for the cold forging of Jetta car differential half-axle gears. The results show that the design of the die sets is reasonable, its function can be realized safely,and the performance parameters meet the design specifications.

The merits of 4-wheel independent driving(4WID) for vehicles and the possibility of implementation of this driving system in electric vehicles(VE) were discussed. The differences of the vehicle dynamics control(VDC) systems between internal combustion engine vehicles and VE were described, and a novel VDC approach to 4WID VE has been brought forward. The approach combines the feedforward compensation for the front wheel steering angle and vehicle speed and the feedback compensation based on the side slip angle and gas rate deflection to control the vehicle movement state. The factors of feedforward and feedback compensation were determined by the optimal control technique. An integrated EV mathematical model was thus built, and a simulation model generated by MATLAB/Simulink software was used to perform a simulation study. Results show the developed VDC system ensure a satisfactory vehicle dynamic performance under different road and driving conditions.

A double wishbone suspension model of a pickup was built using ADAMS/CAR. The kinematics of the suspension was uncovered base on the general theory. The experimental results show that the anti-dive/anti-lift、ride rate and total roll rate performance are improved remarkable by putting 20 mm down the low control arm front point connected to the body.

The torque converter W305 was modeled in detail and the internal flow was computed using the program STAR-CD. To treat the interaction between various stages with different speeds, the mixing plane model was used in the calculation. Based on the numerical results, the internal flow field was analyzed with the emphasis on flow loss. The analysis is very valuable for torque converter design. The comparison between calculation and test data shows that the numerical analysis is sufficiently precise.

A calculating program based on Excel XP for calculating various parameters was brought forward, with which we can calculate cooler's each construction parameters quickly, repeatedly and accurately, and can analyze the effect each parameters had on performance of the cooler. An EGR cooler was designed for YC4112ZLQ turbo-charged and inter-cooled diesel engine based on EGR cooler's characteristic with experience formulas, and the flow resistance checking calculation was given. Moreover the paper includes plenty of practical calculating formulas and directive parameters of this diesel engine's EGR cooler, such as total heat exchange area S =0.196 m², heat exchange tube's length L=0.55 m, diameter r=7 mm and its arrangement form.

When many factors (e.g. more than 4 factors) are to be investigated in an experiment, the uniform design is an efficient experimental fractional factorial design. The design was used to implement the seeding performance test of an air-blowing vertical-rotor type precision soybean seed-metering device. The effects of the device structure parameters on the performance were studied at the atmospheric air pressure in a laboratory. The seed-filling include angle, the seed-cleaning angle and the operating speed,etc., were considered in the experiment, and 6 levels for each factor were taken. The experimentally optimized seed-filling include angle is 46?, and the seed-cleaning angle is 55? under these angles, approximately 100% seed drop rate can be acquired at operating speed less than 6 km/h. The results are of help to the development of air-blowing seed-metering device.

The diurnal variations of the cucumber stem sap flow(SSF) under different environmental conditions and their effects on SSF were discussed. The study revealed that the dirunal variation of SSF has a clear rhythm. SSF ascends promptly in the morning(6:00～10:00 am) and peaks at noon(11:30～14:00), then descends slowly towards evening and reaches down to the valley at night.SSF is greatly affected by the meteorological factors such as solar radiation intensily, air huminity and temperature,etc. Under the same environment, the correlation factor ρ of diurnal SSF curves between cucumber stem with and without sufficient water supply reflects the water deficit degree of peant. When ρ <0.80～0.77, the leaves of cucumber begin to wither. Based on the variation of SSF, a plant water deficiency diagnosis system using the neural network technique was developed. The experiments showed that the system prediction is in good agreement with the measured results.

The behavior of a high-yield mutant WZ-4 bacterium of the cell-surface phytase yeast was studied. The results indicate that the optimum pH value for the yeast is 5, the optimum temperature is 50 ℃, and the K_m value for the Na-phytate is 0.666 mmol/L. Besides, the phytase production of WZ-4 bacterium is controlled by the phosphorus concentration in the culture medium. The maximum production of phytase occurrs in the culture medium containing 0.5 mg phosphorus per 100mL. When the concentration of phosphorus in the medium is more than 10 mg phosphorus per 100mL, the phytase production is completely repressed.

A preview kinematic equation of vehicle was established at first, and a dynamic equation for vehicle steering system was obtained through system identification experiments. Then, a more integrated steering control model was established based on the preview kinematic equation. A steering controller was designed by means of the sliding variable structure control theory. A simulation was performed to analyze systematically the problems met in the system design and experiments such as the effect of the front-wheel steering angle on the controller output, the static deflection found in experiments, and the relation between the preview lateral deflection and vehicle actual lateral deflection,etc., and a method to eliminate the deflection in the control software was suggested. The steering control system was verified by field test and its favorable track capability has been demonstrated.

A new type of the PZT driving device to realize the planar movement by using the bimorph oscillator as the driver was presented. The performance characteristic of the driving device was introduced and its working mechanism was discussed. It was revealed that the device realized the directional movement by controlling the difference between the driving force of device leg and the friction force of interfaces. An experimental device of PZT locomotive carrying platform was developed and the tests were performed. Results proved that the device is really able to move directionally with allowable load under control of appropriately designed electric circuit.

A system made up of the engine, the variable delivery pump, the control valve and the load of a hydraulic excavator was studied to explore the principle and the technique of the power matching between its elements. An adaptive control of the power system against the operating conditions was implemented according to the pressure at the pump outlet. The power matching between the engine and the pump was achieved by the rotation speed feedback control. The power matching between the pump and the load was brought about by using a microcomputer to adjust the multipassage directional valve opening according to the flow rate at the pump outlet. The proposed technique on-line real-time adjusts the PID parameters by means of fuzzy algorithm to realize the energy-saving adaptive control over the hydraulic excavator without the shortcoming of pressure only or rotation only control.

On the basis of investigating the advantages and disadvantages of the existing pile hammers made in China and abroad, a new design scheme of hydro-pneumatic driving system for the pile hammer was put forward for the purpose of structure compactress manufacture simplicity, low cost and light driving efficiency. The system consist of a double-pole single-action hydraulic cylinder with nitrogen gas to accumulate energy and a transfering oil tank. A set of control valve was used to implement the control guideline "close before open". The developed prototype is characterized by high driving efficiency, compact structure, smooth and reliable operation, demonstrating the feasibility and outstanding character of the scheme. It may provide the technical foundation for developing the equivalent products in our country.

In order to improve performances of the working device of the loader, the dynamic simulation about the six-rod reversing mechanism was carried out using the software ADAMS, and the comprehensive optimum design with the theories and methods of the orthogonal experiment was also finished. The research showed that the dynamic siulation had many merits, such as the simple model, the easy control, the exact action, the strong visual effect, the comprehensive analysis, the few programs and so on. Using the theories and methods of the orthogonal experiment can consider the requests of the mechanism and the technics in a large range, so comprehensive optimum design fits the requests of the part renewal or the whole design of the products of the enterprises.

Kalman prediction method was applied to simulation of tracking mobile target of theodolite, and simulating study on tracking mobile target was conducted under different target beacon velocity. Input matrix of state equation G(k) for constant angular velocity model and constant angular acceleration model was presented separately. According to a principle of least standard deviation of Kalman prediction curve and practical curve, the deviation ratio of state noise and test noise Q/R was fixed, and the ratio Q/R for theodolite was optimized. The composed Kalman leap prediction method was presented. The simulating results calculated by composed Kalman leap prediction method for two models were compared. The adaptation ranges of composed Kalman leap prediction method of two models for simulating theodolite's tracking mobile target were conducted.

A new method was presented aiming at the three-dimensional parameter (bearing, elevation and range) estimation of near-field sources, Which uses the cross-array and four-order cumulant to estimate bearing-elevation-range parameter of sources. One subarray of the cross-array only has two sensors, that decreasing the number of the array sensors. it also has the merits of no needing spectral peak searching and the good estimate precision, and furthermore, the method can be applied to Gaussian color noise environments. The effectiveness are illustrated by the simulation results.

The problem of estimating the parameters of a Chirp signal was considered. A novel approach was proposed for the Chirp parameter estimation by exploiting the polynomial phase transform (PT) and the adaptive recursive least square (RLS) algorithm. The polynomial phase transform of the polynomial phase signal results in an exponential signal which satisfies the AR model. The estimation of the lower order polynomial phase coefficients can be thus adopted one by one. The main advantage of this algorithm is the improved speed of calculation, less expensive computation, and strong robustness to signal energy level. Simulations show good performance of the proposed algorithm at low signal to noise ratio (SNR).

A novel implementation scheme for optimizing throughput called eigen channel partition method was proposed using adaptive rate modulation for MIMO-OFDM systems.In this scheme sub-channels precoders and decoders jointly partition each frequency bin into a set of min(M,N) parallel eigen channels for M transmit and N receive antennas. All the parallel eigen channels are divided into groups with same assigned rate to all the channels within each group in order to maximize system throughput. Simulation results show the computation complexity is greatly reduced and the throughput and BER performance of the proposed scheme closely approaches to those of conventional method or the ideal case on the channel-by-channel basis.

After studying group space-time coding in mutipath and multiuser CDMA systems, a new joint group space-time coding and power control method was proposed. By partitioning antennas into small groups as the transmitter, using individual space-time codes to transmit information from each antenna group, optimizing transmit power, and decoding an individual space-time code with group interference suppression method and ML detecting at the receiver, this method can improve the diversity gain and error performance without increasing receiver complexity.

Motivated by superior performance of Root-MUSIC method and the advantages of using signal cyclostationarity in communication systems, a new Root-MUSIC-like polynomial algorithm used in cyclic MUSIC was introduced. Unlike other cyclic correlation based methods, both cyclic correlation matrix and cyclic conjugate correlation matrix are exploited in this algorithm in order to null out cyclic correlation of stationary additive noise and all other interferences with different cyclic frequencies. It is shown through simulations that the computation cost is significantly reduced and the resolution is largely increased when using this method. And the proposed method is also able to handle more sources with fewer sensors and efficiently solves uncorrelated signals.

A novel method to hide binary visual pattern in halftone images by stochastic error diffusion with the Jarvis kernel was proposed. When changing a continue tone image to a halftone image, a special technique to embed a binary image into the halftone image was used. To extract watermark in the halftone image, the image difference technology was applied. After the original halftone image and the halftone image with watermark were made difference, the hidden binary visual patterns will appear. Simulation results show that the proposed algorithm has good visual mode, and when the halftone image with the watermark is severely scratched, the detection algorithm can efficiently detect the watermark, which proves this proposed algorithm is robust to the scratching attack.

Echo and noise severely impacts communication quality in communications systems. By using spectral subtraction method with voice activity detection (VAD) technique and affine projection algorithm (APA) with double talk detector (DTD), an acoustic echo and noise canceller (AENC) was proposed. Simulation results show that the proposed AENC method can efficiently reduce noise, possesses high echo parameter estimation accuracy in fast convergent speed and accurate double talk detection, and distinguishes double-talk and echo path change with robustness.

Due to heavy computational complexity existing in the ESPRIT algorithm of rapidly time variant fading channel prediction, an algorithm for channels prediction based on the double smoothing technique and reassembling the covariance matrix is given。Through the theoretical analysis and simulation, it has been indicated that the modified algorithm can reduce dimensions of the covariance matrix and decrease the computational complexity。Meanwhile it can reduce the mean-root-square error of channel's prediction coefficient.

Aiming at the shortcomings of PID and RZN control algorithms, an I-RZN control scheme on the basis of intelligent control idea was proposed. This control scheme relies on observing process output variables and human universal control idea to make control decisions. The corresponding control decision rules were taken according to the different fields and properties of the process output variables, even in the case of taking the same control decision, the control parameters are changeable, which realized quick and steady response in dynamic state and stable and non-error in steady state.

Considering the disadvantages in the complex structure, a great deal of maintenance and tedious record effort of ordinary instrument used for detecting the boiler parameters, feasible schemes for detecting the boiler parameters using intelligent instruments were proposed. The design and the whole implemented course of the intelligent instruments with relevant technical treatment were explained in detail and the schemes were carried out in the Singlechip with MCS-48 series. The practical application shows that the intelligent instruments realized full-automatic display and record with multi-way parameter, and solved many checking problems using ordinary instrument, such as the record of parameters is unstandard, non-unified, etc., so the analysis is fast on time and high speed, and the efficiency is improved. It is more important that the testing and calculating with many kinds of different parameters can be completed which needs only to revise revising the software, and needn't to change the hardware.

The research status on intrusion detection was studied and the analysis was conducted for the component based and agent based distribution intrusion modules in distribution intrusion detection systems. Combined these two modules, a hybrid distributed intrusion detection system(HIDS) on the basis of network based intrusion system(NIDS) and mainframe intrusion detection system(MIDS) was designed. Functions in the different parts of the system, working procedure, message format, and communication method were discussed. Then the system deployment scheme and experiment process were given. The experimental results show that the system can accurately detect various intrusion actions and quickly take countermeasures to block intrusive connections.

Aiming at the key problem of synthetic Link travel time prediction in Urban Traffic Flow Guidance System (UTFGS). A Vehicle link travel time prediction algorithm based on Malcov Queuing model was presented. With a quantity of traffic measurement data, some model parameters were simplized and confirmed, thus getting a high precision and also making the model more become applicable.

A combined algorithm based on multi-sensor information fusion was brought forward. First, statistic estimation method was used to fuse the data from several sensors located at the same site on the link, thus the time average speed of this site could be gotten; then the adaptive weighting average method was used to get the space average speed by fusing these time average speed; finally, the effectiveness of this combined information fusion method was tested using the real data.

Beam test of carbon fiber reinforced concrete(CFRC) was done. vaviation rule of electric resistance of CFRC bending specimen under different loading conditions was discussed. The results show that the upper electric resistance of CFRC bending specimen decreases and then increases as loading became larger, but lower electric resistance decreases gradually. Under repeated loading, the electric resistance of tensile region concrete increases continually, but that of compressive region concrete firstly decreases and then increases. The residual electric resistance reflects inner damage and its accumulation of CFRC specimen. Under alternative loading, the electric resistance of CFRC increases as loading cycle prolongs, until the specimen is damaged. The variation rule of electric resistance depends on the electric pole mode.

Based on the property of sparse matrices in finite element method (FEM), a two-way cross list for the storage of FEM matrices was proposed, in which the large-scale sparse matrix is partitioned into small matrices, and only none-zero matrices were stored. With four pointers and two position indeces as the head of every small matrices, the row and column indeces are not needed. Thus, the effective storage is larger than 84.21%. The proposed data structure is able to use block algorithm, which is important in high performance computing and is more efficient in many computing environments than those organized at lower linear algebraic lever. In addition, it makes data motion very fast when the memory complexity is reduced to O(1).