Search

Traditional manufacturing processes of 3D force platform (three-dimensional force platform) are purchasing of three dimensional force sensor, calibrating of sensor in specialized agencies and building up with platform table and base. However, due to processing accuracy and installation error, the method of calibration before building up cannot guarantee the global accuracy of 3D force platform, and it is difficult to meet the requirements in practice. An improved method is proposed in this article that calibrating force sensors first, then dynamic monitoring the installation process, and global calibrating the 3D platform, verifying the calibration results at last. This article design a global calibrating system based on the global calibrating method which achieve the function of data extraction, classification, data evaluation and parameter optimization. The 3D force platform is calibrated again using the optimized parameters, and the result shows that the optimized calibration results are in coincide with the calculated results. At the same time, the method of sensitivity optimization based on global calibrating avoids the low efficiency caused by trial-and-error method and repeatedly loaded, and then improves the calibration efficiency and accuracy.

Link adaptation is an important technique to achieve high throughput in wireless communication systems. Effective SNR mapping (ESM) algorithm is widely used to provide equivalent SNR, which helps to predict the BLER performance. In this paper, we propose a new ESM - average effective SNR mapping (AESM) algorithm, which is applicable and efficient in LTE-A uplink. Compared with the well-known exponential ESM (EESM) algorithm, AESM not only provides good approximation, but also has lower complexity by getting rid of the calibration procedure, and replacing exponential and logarithmic calculation with simply averaging algorithm. We study the relationship between AESM and EESM, and demonstrate that the value of AESM is greater than that of EESM, and the difference between them is monotonically decreasing and non-decreasing function of the calibration factor and the variance of the mapped SNRs or SINRs, respectively. Finally, we apply AESM to link adaptation in LTE-A uplink, and the results illustrate that effective SNR obtained by AESM is almost the same with EESM and the performance of link adaptation is as good as expected.

Compensation structure, which should match with the overall enterprise system and be vital to the development of enterprises, has been the focus of theory and practice circles. Through the Delphi method by industry expert experiences, we attribute the decision-making factors of compensation structure as industry competition situation, labor factor, compensation function, business strategy, enterprise development stage, enterprise culture and enterprise organization structure. Compensation structure decision-making model is constructed in the paper employing fuzzy evaluation method that utilizes the relation between the size of degree of membership and compensation structure. Actual case is provided to illustrate the whole decision-making process, through setting up the aggregate of factor, the aggregate of evaluation, single-factor fuzzy evaluation, the aggregate of weight. Empirical results show that the power enterprise should adhere to post-oriented primarily and performance-oriented and ability-oriented peripherally. The decision-making result is more scientific, quantitative and objective, as a result that the enterprise finds out a suitable actual compensation structure.

In order to establish the training system which is suitable to the characteristics of enterprises and to improve the effectiveness of staff training, this paper discusses the competency theory and the application of correlative model in the entire training process. Comparing the staff actual competencies with the job competencies by competency assessment, this paper determines accurate training needs and sets up targeted training content according to the training needs, and then selects the appropriate training methods. To show the implementation of this model, a case on company human resources management is used. Empirical results show that, by using competency model theory, training programs combining with the actual situation of enterprises and employees can be set, and can also ensure that trainees learn the skills they really need. Training system design based on competency model can ensure the training more scientific, targeted and efficient.

A novel approach -adaptive wavelet neural network (AWNN) is proposed in this paper for energy consumption forecasting. The commonly used Morlet wavelet has been chosen as the activation function for hidden-layer neurons of feed-forward neural network (FFNN) which has been trained by hierarchical genetic algorithm combined with gradient descent algorithm. To demonstrate the effectiveness of the proposed approach, month-ahead prediction of energy consumption in Zhu Jiang Iron and Steel Groupcorporation is considered. The forecasted results clearly show that AWNN has better prediction properties and needs less learning time compared with WNN.

As the main noise source of HVDC, converter transformer converter's noise prediction precision and the choice of control methods will directly influence the level of noise prediction and management effect for converter station. Through the systematic study on the noise-generation mechanism, frequency spectrum, the type of sound source, noise control measures and so on, the paper highlights the BOX-IN technology and verifies the effect compared with sound barriers technology by SoundPlan. According to the field test, the result indicates that the noise reduction quantities of BOX-IN equipment achieves about 20dB(A) corresponding to the theoretical calculation, which will provide theoretical basis for improving the noise prediction precision of HVDC furtherly.

Combining with the characteristic of SOC test planning and test power, the optimization for SOC test based on quantum algorithm is introduced in this paper, in the paper, the corresponding mathematical model is established and the design of algorithm is given. After establishing the model with appropriate parameters, the power limit is ascertained, then the optimal individual can be distinguished. During the process of algorithm realization, genetic factors are changed according to a discriminant function, and the shortest test times with power constraints can be find out. Validation of the algorithm is made with ITC'02 SOC test benchmarks finally. Experimental results demonstrate the advantages of quantum algorithm compared with other algorithms[3] [6].

In order to meet the personalized demand of users and help users to be more effective on selecting potentially useful search engines, this paper presented a metasearch engine algorithm that is based on personalized demands of users. According to practical demands of users, different select strategies of metasearch engine algorithm are designed. The experimental results show that selection strategy of the algorithm can reflect some effects by the certain characteristics of the member search engines. So, this algorithm will be widely used in the future.

There are wide backgrounds of applications of a thin elastic rod. For an exact model of a Cosserat elastic rod which are not only bending and torsion but also shearable and extensible, constitutive relation for principal vector and principal moment on a cross section versus displacements of the section with arc coordinate are derived from Hooke's low. It shows that derivative of generalized coordinates in bending components of curvature-twisting vector and axial components of strain vector of center of a cross section which are all composed of linear strain is about arc coordinate before deflection of the rod. And the one in torsion components of curvature-twisting vector and shear components of strain vector of center of a cross section which are all composed of shearing strain is about arc coordinate after deflection of the rod. In the senses of first-order approximation, the principal vector is proportional to the strain vector of a point in the center line, and the principal moment is also proportional to the curvature-twisting vector about intrinsic arc coordinate.

The multiple attractors induced by a non-resonant double Hopf bifurcation in a two-neuron network with self/neighbor-delayed-connections are investigated in this paper. Various dynamical behaviors are classified in the neighborhood of the bifurcating point and are expressed approximately in a closed form by the center manifold reduction. The solution induced by Hopf bifurcation is expressed approximately by the method of multiple scales. It follows that the delay induces the coexistences of silencing and periodic spiking, two periodic spiking, periodic and quasi-periodic spiking. The basins of attraction are classified numerically. These results have some potential applications in designing the neural network according to the memory pattern and storage.

In wireless sensor networks (WSNs), extending networks lifetime is one of the fundamental and important issues. One of the ways is to choose the fewest number of sensors and maintains the coverage of domain. In this paper a new way of dividing domain into regular hexagon is presented, which is the dividing domain algorithms (DDA). And the simple procedure runs in a short time and the coverage sensor number is not larger than the procedure. At last, the simulation verifies the procedure is better than the other procedures.

The operation state of the urban road network units (road sections or junctions) under disasters and emergencies has strong failure (jam) dependence. While analyzing and computing the reliability of urban road network, road units jam dependence must be considered and the method of extreme probability network estimating technique (PENT) is utilized to ascribe each section and intersection of jam dependence with strong relativity to a fully-related group, and find out the related jam units of road network. On this basis, jam dependence is considered to calculate and assess unblocked reliability index of path, Origin-Destination (OD) and traffic systems so as to make the index accord with the characteristics of urban traffic flow, facilitate traffic management department and decision makers to apply the index to traffic practice, and provide an effective evaluation for road network layout, road network planning, traffic-building timing and anti-disaster planning.

The enterprise is an open system, so according to the second law of thermodynamics and the dissipative structure theory, the enterprise has to require an influx of negative entropy from external sources by exchanging substances, energy and information with its external environment in order to become a system with low entropy. Negative entropy principle indicates information is negative entropy what suggests the enterprise should obtain more and more information from outside. But the fact does not support this principle, for example, information overburden has been a serious issue for the enterprise to deal with. This paper demonstrates the irrationality of negative entropy principle base on the basic definition of information entropy, and proves the possibility of information is positive entropy. The purpose of this paper is to make the enterprise pay much more attention to the management of information resources.

In this paper the ferromagnetic properties of Co-doped SnO2 system under room temperature conditions are investigated using X-ray diffraction and vibrating sample magnetometers. The effects of substrates on the magnetic properties of doped films are also studied.

This paper proposes a reuse-oriented process component representation framework for the description and classification of process components (PCs). The key features of our approach are the division of process component descriptions into a goal-oriented PC's general information description (PGD), an implementation-oriented PC's specification description (PSD) and an improvement-oriented PC's data description (PDD). A facet-based PC classification scheme and an XML-based process description method is presented. A retrieval model based on tree-containment matching is given, which can improve the recall while maintaining a high level precision. Finally, a reuse based process and project management system (P2MS) to support process improvement and control is also introduced in this paper.

This paper proposes a new fast calculating method for the transient stability limits for the output power of generators, using the fast higher order Taylor series expansion and the equal area criterion for multimachine power systems. The results of calculation experiments show that this method not only has better accuracy but also possesses fast execution speed, and therefore has practical value. >