Noise control with Kalman filter for active headrest

A control strategy with Kalman filter (KF) is proposed for active noise control of virtual error signal for active headset. Comparing with the gradient based algorithm, KF algorithm has faster convergence speed and better convergence performance. In this paper, the state equation of the system is established on the basis of virtual error sensing, and only the weight coefficients of the control filter are considered in the state variables. In order to ensure the convergence performance of the algorithm, an online updating strategy of KF parameters is proposed. The fast-array method is also introduced into the algorithm to reduce the computation. The simulation results show that the present strategy can improve the convergence speed and effectively reduce the noise signal at the virtual error point.

Modeling and analysis of compaction roller deformation for fiber placement

The fiber reinforced resin-based composite materials have the characteristics of high specific modulus and strength, impact resistance, creep resistance, and seismic resistance, which are widely used in the aviation, automotive, and marine industries. As one of the advanced composite material forming technologies, the fiber placement can manufacture the large-curvature composite parts with the high efficiency, high quality, high repeatability and low cost. The rubber compaction roller will deform under the pressure, in which the deformation will increase the contact area between the pressure roller and the substrate, and improve the interlaminar bonding degree. Due to the different deformation of the compaction roller caused by the different laying pressures, the deformation process of the fiber compaction roller was analyzed. By establishing the pressure stress distribution model for the compaction roller, the contact pressure, the downward deformation of the compaction roller and the deformation contact area between the compaction roller and the substrate are obtained. The mapping relationship among the contact curves of the substrate is verified by using the experiments.

Electromagnetic design method of linear induction motor considering lateral end effect

Due to the complex end effect, it is difficult for the linear induction motor to design and analyze with the "road" calculation and two-dimensional field calculation. Considering the end effect, skin effect and secondary leakage reactance of the eddy current of the secondary induction plate, a linear induction motor electromagnetic design and calculation program was programmed via VB language. A two-dimensional finite element analysis method was proposed to approximate the lateral end effect on linear induction motor performance by multiplying the two lateral end effect coefficients into secondary induction plate resistivity and air-gap relative permeability. The accuracy for both calculation of end effect coefficients by "road" calculated program and considering lateral end effect in two-dimensional finite element analysis was validated by using the experiment and three-dimensional full model finite element, respectively. This work proposes the calculation method of the lateral end effect, which is faster and more accurate for the electromagnetic design as well as the finite element simulation of linear induction motor.

The memory forensic research oriented to segment heap in Windows 10 system

The current forensic research on heaps mainly extracts information from the heap of Linux and the NT heap of Windows. However, the study of how to extract the information on the segment heap in the Windows 10 from dump files is not sufficient. To reproduce the internal information on the segment heap, this paper proposes a method for locating and extracting the internal information on the segment heap in the Windows 10 according to the field offset in the vtype description information of memory object. The method uses the pool scanning technology to locate the process object, obtains the starting position of the process heap and scans the process heap according to the structural information on the process object and the process environment block object. Then it locates the position of the segment heap with its feature values, thereby extracting its internal information. Based on the analysis results, five forensic plugins for extracting the information on the segment heap were developed on the Volatility framework. The experimental results show that this method can effectively extract the information on the address of each segment heap and its internal components in the memory and on the size of committed memory, etc. The information can help investigators to analyze the digital traces left in the memory by cyber criminals or cyber attackers.

Robust dynamic inversion control of flight control system based on L1 adaptive structure

Nonlinear Dynamic Inversion(NDI) control has excellent rapidity and decoupling ability, unfortunately it lacks the essential robustness to disturbance. From the perspective of enhancing the robustness, an adaptive NDI method based on L1 adaptive structure is proposed. The L1 adaptive structure is introduced into the NDI control to enhance its robustness, which also guarantees the stability and expected dynamic performance of the system suffering from the disturbance influence. Secondly, the flight control law of the advanced aircraft is designed based on the present method to improve the robustness and fault tolerance of the flight control system. Finally, the effectiveness of the flight control law based on the present approach is verified under the fault disturbance. The results showed that the flight control law based on L1 adaptive NDI has excellent dynamic performance and strong robustness to parameter uncertainties and disturbances.

DOA estimation method of spatial spectrum product for distributed coprime linear array

Coprime array isAsparse array composed of two uniform linear arrays with different spacing. When the two subarrays are inAnon-coherent distributed configuration, the direction of arrival (DOA) method based on the covariance analysis of the complete coprime array is no longer effective. According to the essential attribute that the distance between the elements of two subarrays can eliminate the angle ambiguity, based on the mathematical derivation, Aspatial spectral product DOA estimation method for incoherent distributed coprime arrays is proposed. Firstly, the spatial spectrum of each subarray is calculated by using the snapshot data of each subarray, and then the DOA estimation is realized by multiplying the spatial spectrum of each subarray. The simulation results show that the estimation accuracy and angle resolution of the present method are better than those of the traditional ambiguity resolution methods, and the estimation performance is good in the mutual coupling and low SNR environment, with the good adaptability and stability. Moreover, by using the flexibility of distributed array, the matching error is effectively solved through the rotation angle.

Research on simulation model validation based on probability relational analysis

The most basic and direct method for simulation model validation is to compare the consistency of missile flight data and simulation data under the same input conditions. However, the existing dynamic data consistency analysis methods are mainly suitable for the case between 1-D missile flight data and 1-D simulation data, and do not conform to the consistency test of single sample flight data and multi-sample simulation data in equipment qualification/finalization test. To solve this problem, a simulation model validation method based on probabilistic relational analysis is proposed. The consistency of output data is measured from the two scales of probability relational coefficient and probability relational degree. The probability relational coefficient is determined by calculating the cumulative distribution probability value of real missile flight samples in the distribution function constructed by simulation data. The probability correlation degree is calculated by judging whether the probability relational coefficient satisfies the uniform distribution of[0 1]. The consistency analysis problem of a kind of dynamic data association is solved accordingly. The correlation theorem that the probability relational degree must satisfy and its property are proved. Meanwhile the operation steps of simulation model verification based on probability correlation analysis are given. This method can process all multi-dimensional simulation data at the same time, and integrate the random factors in the test process, so it can make full use of the test information under the condition of small sample flight test, and improve precision and the reliability of simulation model verification. The rationality and validity of this method are further verified by numerical tests and application examples.

Research on interactive color design method of cultural creative product driven by group consensus

Aiming at assisting the color design of cultural creative product and outputting color schemes that correspond to multi-user image preference effectively, an interactive color design method driven by group consensus was proposed in this paper. Firstly, the triangular fuzzy number was used to quantitatively descript users' image preference and a consensus degree model of group users' image preference was constructed. By studying the process and basic principles of interactive genetic algorithms, the interactive evolution operation for cultural creative product color design was then implemented based on the group consistency of decision-making combined with the triangular fuzzy number, which can help generate cultural creative product color schemes that satisfy group consensus and satisfaction. Next, the color aesthetic measure model was used to rank and select the final product color design schemes. Finally, taking the color design of terracotta warriors' statue as an example, it is verified that the proposed method can effectively integrate the image preference of group users, and assist designers to better conduct cultural creative product color design with the consistent results of multi-user decisions.

A low-complexity orthogonal time frequency space modulation method for underwater acoustic communication

Compared with the orthogonal frequency division multiplexing (OFDM) modulation, the orthogonal time frequency space(OTFS) modulation has a lower peak-to-average power ratio. It can effectively resist the time selective fading caused by the Doppler effect and has significant performance advantages over doubly dispersive channels. However, the conventional OTFS linear minimum mean square error (LMMSE) method has a high complexity and is not easy to process in real time. In order to solve this problem, we propose a low-complexity equalization algorithm with infinite norm constraints based on the optimal coordinate reduction. The equalization algorithm not only obtains the optimal solution through a certain number of iterations and avoids direct matrix inversion but also equalizes infinite norm constraints to improve the symbol detection performance gains. At the same time, the OTFS delay-Doppler channel matrix we utilize is sparse and the two-norm squares of each column vector equally reduces the complexity of optimal coordinate descent. Finally, the simulation in the underwater acoustic communication scenario we designed verify the effectiveness of the proposed equalization algorithm. The simulation results show that the performance of the proposed equalization algorithm is close to that of the LMMSE method, while its low complexity is ensured.

Intelligent recommendation algorithm of mobile application crowdsourcing test based on deep learning

As the functions of mobile applications become more and more complex, the crowdsourcing testing puts higher demands on the professional skills of testers. Therefore, it is an important factor to ensure test quality how to effectively match test task requirements with test personnel's skill level and achieve accurate crowdsourcing test task recommendation. This paper proposes a crowdsourcing test task recommendation algorithm for mobile applications based on deep learning. Firstly, feature analysis is carried out for testing tasks and testers, and feature systems are designed respectively. Second, the resulting characteristic data is used as input data for the Stacked Marginalized Denoising Autoencoder (SMDA). The deep feature data learned from SMDA are combined as the input of Deep Neural Networks (DNN). Finally, the learning ability of DNN is used for prediction. Experimental results show that the proposed algorithm has obvious advantages in both performance and training time compared with CDL and AUTOSVD ++, which verifies the effectiveness of the proposed algorithm. The proposed algorithm can recommend testing tasks to appropriate testers and improve the precision of the algorithm.