Papers by Keyword: Model Identification

Abstract: Based on kinematics characteristic of two-wheeled differential drive mobile robot (WMR) and response characteristic of fact motor drive system, this paper presents the analysis method of the equivalent rotation inertia, and the entire vehicle load is assigned to each wheel, and then the wheel load is converted into the corresponding equivalent rotation inertia of the motor shaft of each wheel, and motion model of WMR are obtained for combining with quasi-equivalent (QE) state space model of double-loop direct current motor systems under variable load and kinematics model of WMR under the load changes. By using speed response data of the actual system and combining with genetic algorithm to accurately identify the model parameters. Finally, through experiments results of the WMR motion model and the second order model respectively comparing with the actual system which demonstrates the effectiveness of the proposing method and model.

Abstract: For the need of multi-channel simultaneous sampling in the modeling and identification of power system, a portable data acquisition system is introduced. It can sample multi-channel data simultaneously, and also have a mass data storage. The system is composed of data acquisition card and PC processing software. The C8051F410 MCU is the core of the data acquisition unit card. It not only provides multi-channel synchronous sampling, large storage capacity, easy to carry and low cost, but also can run offline, and bidirectional interaction with PC software. The data processing software is programmed by Visual Basic6.0. Its main function is sending sample orders and the sample data processing. It can store and process the sample data, also it can set the sampling parameters of the card. Through actual verification, the system is easy to use and meets the precision of requirement.

Abstract: This paper was according to the early crack identification on metal deep drawing, combined with the characteristics of the AE singal of deep drawing forming, determined to analyze it by db4 wavelet basis function for 3 layers, extracted the energy value at every point of the eight AE signal samples from deep drawing forming which crack state was known to constituted characteristic matrix, fuzzy equivalence matrix was obtained by a series of matrix transformation,which was regard as pattern library of crack that used in cluster analysis. Then extracted energy value at every point of AE signal from deep drawing experiments with there states unknown, calculate similarity between samples and the model we established by using euclidean distance formula after normalization processing, deep drawing state was determined by the principle of selecting near, finally compared the recognition result with crack situation deviced by naked eye and microscope, results show that the accuracy of fuzzy identification of early crackn based on fuzzy clustering is up to 90%.

Abstract: Under the shear rheological conditions, debris filled with mud weak interlayer shows viscoelastic rheological laws. On the basis of non-linear regression analysis of rheological parameters, the relationship between shear stress and rheological parameters was researched. The study reveals that, with the increase of shear stress, weak interlayers instantaneous deformation, constant strain rate and the final elastic strain increase, while the delay time has not been found certain regularity. Under the shear rheological conditions, when the stress level is low, the constitutive model of weak interlayer can be described by Burgers model; when the stress level is high, Cvisc model should be used.

Abstract: In order to solve the problems in existing protection for high voltage direct current (HVDC) transmission lines, such as insufficient sensitivity for main protection and slow operation for backup protection, a novel pilot protection for HVDC transmission lines based on model error function is proposed in this paper. In the principle, through analyzing the fault component networks, external fault is equivalent to a capacitance model, while the internal fault is equivalent to a inductance model. Taking capacitance model as a benchmark, the model error function was defined to describe the degree of consistency between actual fault data and the benchmark. When an external fault occurs, fault data agree with the capacitance model and model error equals zero. When an internal fault occurs, fault data do not agree with the capacitance model and model error does not equal zero. According to this, a protection criterion is constructed. Theoretical analysis and simulation show that the new principle, not affected by transition resistance and line distributed capacitance current, can identify internal faults and external faults reliably and rapidly.

Abstract: Validation and optimization of technical systems are central activities in the product development process. One part of it is the calibration and validation on a level, which covers the whole vehicle. The aspect, that plays the most important role in both validation and optimization, is the driving condition. Especially in the case of hybrid vehicles, state variables like the state of charge (SOC) have great influence on the operating strategy and therefore on assessment criteria.The article’s objective is to present a procedure, which performs the conditioning and brings the planned maneuver into an order, which reduces the total needed conditioning duration. Thereby a lot of time can be saved, according to the type and amount of the possible maneuver and state values. In addition to optimizing the order of conventional maneuver, the procedure can be used to optimize the list of maneuver in a DOE-Plan. Thereby the maneuver of the individual criteria can be re-sorted as well as the designparametervariation. The IPEK-X-in-the-Loop framework (XiL) is the basis for the approach and will be used as a validation environment in an acoustic roller test bench with vehicle-in-the-loop technology.

Abstract: This paper proposes a robust adaptive neural network controller (RANNC) for electrode regulator system. An equivalent model in affine-like is derived for electrode regulator system. Then, the nonlinear control law is derived directly based on the affine-like equivalent model identified with neural networks, which avoids complex control development and intensive computation. Pretraining is not required and the weights of the neural networks used in adaptive control are directly updated online based on the input-output measurement. The proposed nonlinear controller is verified by computer simulations.

Abstract: An unmanned helicopter dynamic model identification method based on immune particle swarm optimization (PSO) algorithm is approved in this paper. In order to improve the search efficiency of PSO and avoid the premature convergence, the PSO algorithm is combined with the immune algorithm. The unmanned helicopter model parameters are coded as particle, the error of flight test and math simulation model is objective function, and the dynamic model of unmanned helicopter is identified. The simulation result shows that the method has high identification precision and can realistically reflect the dynamic characteristics.

Abstract: Extreme temperature events in China have been simulated by using the Beijing Climate Center Climate System Model (BCC_CSM1.0). Daily observed temperature data for 1961-2006 at stations in China are used for comparison. The result shows that the value of simulation is consistent with the observation. But the consequence has a lower value especially Qinghai-Tibet Plateau; there is a good simulation for BCC_CSM1.0 for temperature and extreme indices over China in a certain range of error. Trends of indices of extreme temperature are increased.

Abstract: Hot pressure welding (HPW) is one of the inner lead bonding process which is widely used in microelectronics manufacturing. This paper presents a practical digital temperature controller for HPW. Based on the step response, an identification method is developed to obtain a nonlinear time-delay model for heating-up control design. The input-output linearization problem is solved by means of second-order Pade approximation and static inverse transform, and an optimal controller is obtained for the welding machine. The experimental equipment on the temperature control is designed based on LabVIEW and data acquisition card. Simulation and experimental results show that the dynamic performance, steady stage error and overshoot satisfy the designed requirements.