Search

as everyone knows, conditioning F2N2(Feed Forward Neural Network) in many ways, but ways, but has so far failed to completely solve the "convergence not quickly" and "convergence is not robust"of these two problems, thus affecting the application of F2N2. Aiming at this problem, pro this paper puts forward a new method of processing and application of F2N2, the method of flexible BPNN (Back Propagation Neural Network) algorithm and hierarchical optimization algo-rithm as a whole, each layeris independent rights attached to the conditioning process, carefully con-structed F2N2objective function is non linear series based on the description, not only can the optimization of each layer of the weight problem is simplified to a linear problem, but also can effec-tively controlthe linearity error. This paper provides several examples of application of F2N2automa-tic detection, the results of application show that, this method is superior to other methods.

This paper explains and demonstrates how an atomic force microscope (AFM) is working and how can be improved the measurements realized with this system. At the begining of the paper is a complex presentation of AFM and its improvement with capacitive sensors. Starting from the obtained information, we studied the influence of quality and linearity error of sensors to characterize surfaces using atomic force microscopy. We used the device that had sensors with a linearity error of 0.1%. For another set of measurements we used sensors with linearity error of 0.01%. Experiments realized on CoCrMo and Ti6Al4V materials surfaces demonstrated that the accuracy of sensors, respectively, the linearity error, from the measurement system has an important influence.

The bending sheet metal inevitably exist bending errors including angle error, linearity error and length error of sideline for reasons that the structure of press brake, the manufacturing precision of press brake and mould and the inhomogeneous characteristics of processed sheet metal. The processing errors of sheet metal can affect the assemblage, increasing subsequent repair to the mold and forming calibration, extending the product development cycle, restricting the further promotion and application of bending forming, especially on the forming high strength and high spring-back sheet metal. The PBH110-3100 CNC press brake in Jiangsu Yawei Co.Ltd. is studied to increase the bending precision of press brake in this study. The bending errors in press break are analyzed in depth according to the elastic mechanics theory, and the rule that the manufacturing precision of press brake affect the press precision is educed. The analysis results have real significance on improving the press precision and reliability of press brakes.

A new kind of straightness error measurement system based on laser interference is developed. High stability He-Ne laser beam which is collimated and broadened is cast on wedge-shaped glass plate, on which back and front the light reflects and interfere. The angle of the guide and target would be changed when the motion of the target is along the guide, if there is linearity error existing in the guide. So interference stripes would be moved by the changed angle of the guide and target. In this way the straightness error of the guide is transferred to the displacement of the interference stripes. Then interference stripes are tested by photoelectricity sensor and deviation of the laser beam is acquired. The comparison tests and repeated tests show that the straightness error is 623.103μm when the tested long guide is 50 meters long.

A long rail straightness measurement method is given for the portable laser alignment measurement system which is based on laser shearing interference fringes. Through theoretical derivation, mathematical model is set up between the laser phase differential linearity error and shearing interferometry. Large numbers of experiments have verified correctness of long rail straightness measurement method, analyzed the cause of error and provided ways for error reduction.

For the dispersion problem of the widespread sensors in the R & D process of measuring instruments, this paper proposes a practical method for the automatic calibration of linearity error. Meanwhile, it takes the pressure sensor as an example to describe different calibration methods such as using the hardware, the combination of software and hardware, using the software.

The operating performance of micro-gyroscope may be easily influenced by various factors, which changes the dynamic response accordingly. Generally, the influences can’t be quantified because of the complexity, and the accurate sensitivity and linearity measuring of micro-gyroscope meets enormous difficulties. In order to accomplish corresponding measuring and analysis accurately and efficiently, input excitation waves are designed based on mathematical analysis of the measuring principles and vibration response, and each measuring of sensitivity and linearity can be analyzed in a single test. In this way, the influences from manual operation and variation of measuring environment are well eliminated, and some key parameters of the dynamic response are obtained, such as sensitive voltage, response function, linearity error, etc. In addition, problems about imbalance in double drive vibration, design of the minimum excitation voltage and evaluation of the measuring error are also described. Therefore, some important references are provided for subsequent design, experiment and engineering application.

Crack and strain detecting is an important part of Structure Health Monitoring for aircrafts in fatigue environment. A signal conditioning circuit of a new strain-crack detecting system is developed in this paper. This circuit can deal with both intelligent coating sensor and strain gage, which have a great different original resistance between the two kinds of sensors. Comparing to the Wheatstone Bridge, when the strain is being detected, the new circuit gets rid of the influences of the lead-wires and decreases the wires’ number. The principle and the constitution of the circuit are introduced. And the circuit’s performances are demonstrated by a series of experiments. The linearity error is small and the stability is good.

Linearity is a very important parameter for seamless steel pipes. A real-time and on-line visual measurement system for seamless steel pipe linearity is presented. The system consists of several structured-light visual sensors. Each sensor can achieve the coordinate of the center of partial steel pipe in its local coordinate system. Through global calibration, all coordinates measured can be transformed into an integrated coordinate system. The linearity error of steel pipe can be assessed. This method can fulfill 100% on-line and real-time linearity measurement. A pair structure-light sensor is designed to improve accuracy and a suspension-wires method for sensor calibration and global calibration is used. Through experiments, it shows that the method not only meets the need of precise calibration but also significantly improves the efficiency and feasibility.