Search

Greedy algorithm is an important class of sparse recovery algorithm of compressive sensing. Most existing greedy recovery algorithms need the prior knowledge of the signal’s sparsity, which is unknown or even time-varying in actual applications, to determine the stop threshold of the iterative recovery process. Recently, the power of the residual signal is used to compare with the noise power to determine whether there is some sparse signal remaining. In this paper, the statistical property of the residual signal power is proved to follow the chi square distribution with parameter L(M-k), where L and M and the size of the observation matrix, and k is the recovery iteration times. Then, a semi-blind greedy algorithm is proposed, in which the stop threshold is adjust according to the recovery times. Simulation results illustrate that the proposed algorithm is more effective than the existing energy-based support recovery algorithm.

The intelligent optimization of multidimensional process quality parameters are of great significance for improving the process quality and intelligence of manufacturing process. Manufacturing process is complex with numerous parameters. In view of the coupling, nonlinear and uncertainty of multidimensional process quality parameters, this paper uses ensemble learning method to build the parameter mapping model which on behalf of the non-linear mapping relationship between the multidimensional process quality parameters and process quality. The objective function is constructed and the intelligent optimization of multidimensional process quality parameters in manufacturing process is realized by using firefly optimization algorithm. Taking the gluing process as an example, the accuracy of the multidimensional process quality parameter mapping model and the effectiveness of the optimization method are verified, which provide support for improving the process quality and intelligent manufacturing process.

Drug abuse is a serious problem in south of China. Because of drug abuse, many people lose their family, cause, even their lives. Chinese government has tried its best to fight against drug abuse and trafficking all along. However, drug abuse is still a serious problem in Yunnan, Guizhou and some other places. This article proposes a dynamic prediction model which can predict the change in number of drug related people. We adjust some parameters that are adjustable and get prediction results under different parameters. Through the analysis of these data, we get some useful information, and these information can provide data base for government decision makers to develop anti-drug specific action.

The dynamic control of process quality is of great significance to improve the intellectualization of manufacturing process. The real-time monitoring and evaluation of machining state provides support for the intelligent control of process quality. In view of the timevarying, coupling and dynamic characteristics of monitoring parameters, as well as the real-time dynamic correlation and nonlinear relationship between the processing state and the product quality, this paper uses the Stacked Auto-encoder (SAE) to optimize the multidimensional real-time monitoring parameters in the machining process. By using the hybrid model of SAE-BP neural network, the nonlinear mapping relation between multidimensional monitoring parameters and the processing state is characterized adaptively and the dynamic intelligent evaluation of the machining state in the intelligent manufacturing process is realized. Taking an experimental platform as an example, the validity of the dynamic intelligent evaluation of the SAE-BP neural network used in processing state is verified. The proposed method provides support for the real-time dynamic evaluation of machining state.

Danjiangkou reservoir water source area is the source of the water source area of south-to-north water diversion. In order to confirm the specific conditions of the local geological environment, geological environment surveys occupy a very important position. During the survey, a large amount of geological environment data was generated. These data are extremely scattered and cannot be used in a comprehensive manner. For more efficiently using the geological environment survey data and realizing the function of polyphyletic, heterogeneous data management and real-time dynamic update, the paper explained how to use the SQL Server database, rely on the MapGIS geographic information system platform, and carry out secondary development based on the double services of the MapGIS IGServer, and establish a Geological Environment Survey Big Data Integration System based on WebGIS. The polyphyletic, heterogeneous spatial data and attribute data are uniformly processed, stored, searched, managed and analyzed. Data support will be provided for in-depth geological research, analysis and land planning for the Danjiangkou Reservoir water source area on this system platform.

The requirement of real-time processing indeed poses challenges on implementing spectrum sensing algorithms, especially for time-varying spectrum environments. The time-varying feature of the spectrum for wireless communication countermeasure system is analyzed. And the mathematical model of the time-varying spectrum is proposed. With the help of adaptive filtering theory, the performance criteria of optimal spectrum sensing algorithm is proposed, in which the total error was composed by sensing error and lag error. The objective of the real-time spectrum sensing algorithm is to minimize the summation of sensing error and lag error. Finally, the proposed performance criterion was used to guide the channel estimation of intelligent frequency hopping communication system.

This paper mainly aims at the aerial image obtained by UAV, and proposes a new solar panel recognition method based on machine vision. In this paper, OpenCV and VS2013 are used to build the experimental platform. Firstly, the image is preprocessed, then the image is segmented in HSV color space, and then the morphological operation is carried out. Finally, according to the prior knowledge, the system can display the number of solar panels and the coordinates of each solar panel in the image in real time. It has a higher efficiency and success rate. The experimental results show that the algorithm is effective, and the model based on the algorithm design can realize real time and accuracy.

The switching transients of electrical and electronic equipment (such as high-voltage isolation switch or circuit breaker) can generate various types of harassment transient immunity. This paper takes the electrical fast burst (EFT/B) as an example to specifically introduce the uncertainty of which couples pulse composed by numerous electrical fast transient to the power port of electrical and electronic equipment. Focusing mainly on the evaluation uncertainty of pulse rise time, pulse duration, peak voltage, repetition frequency, pulse train duration and pulse train period of transient mono-pulse, this paper will give a quantitative description of the quality of the measurement results and determine their reliability.

The maintenance interval is the basic elements for the preventive maintenance of an airborne product. It is generally determined by using qualitative methods presently. How to establish a scientific and practical method to calculate the maintenance interval has become a important issue in the aviation. In order to solve the problem, this paper analyzes firstly that the depth of product maintenance make the impact on its reliability and points out that the reliability for before and after the repair of the non-fully repair sample have correlation, so this paper updates the traditional method of disposal the sample information and establishes a calculation model based on the Weibull distribution for the preventive maintenance interval of the product. Then it proposes a set of methods to calculate the maintenance interval of each preventive maintenance type such as Hard time removal or discard, Functional test and Operational check. Finally, it introduces the calculation steps and the examples. The practical application proves that the model is scientific, reasonable, simple and low investment, which can meet the engineering need for determining product maintenance interval.

Automatic modulation classification is very important in cognitive radio and communication reconnaissance systems. Two novel approaches for identifying the modulation format of general M-ary PSK signal are proposed, which are based on the phase entropy. Phase entropy of the first one is estimated in time domain with probability space partitioned into fixed dimensions. And for the second one, the frequency transform is first applied to the phase of the received signal and the entropy of the measured signal is then estimated. Based on a general hypothesis test, the entropies of different modulation signals are compared to classify them. The simulation results illustrate that the proposed algorithm has smaller computational complexity than existing classifier and the second one has better classification performance in low signal-to-noise ratio domain.

In the presence of narrowband interference, non-Wiener effects have been observed for the normalized least-mean-square (NLMS) algorithm and the affine projection algorithm (APA). The convergence performance of the affine projection algorithm (APA) is analyzed in the presence of strong narrowband interference. Firstly, the relationship between the APA and the NLMS algorithm is derived, with the help of decorrelation of the input signal. Then, the statistical properties of the decorrelated input signals are derived, from which we can see that the convergence performance of the APA is almost not affected by the strong narrowband interference. The computer simulation results support the theoretical analysis.

Entropy-based signal detector has been proposed recently to counteract noise uncertainty in cognitive radio networks. The time-domain and frequency-domain entropy-based detector are introduced, and detailed simulations are produced to study the detection and classification performance of the two detectors over Rayleigh and Rice fading channel. The results of computer simulation show that the frequency-domain entropy-based detector is robust in fading channel, but it can not be used to classify different modulation signals. In contrast, the time-domain entropy-based detector deteriorates rapidly at low signal-to-noise ratios (SNR), but it can classify different modulation signals in medium SNR domain.

In this article, we propose an improved algorithm of blind signal separation that jointly exploits the selection of rational nonlinear functions and quasi-Newton method. The proposed algorithm uses rational nonlinear functions in constructing the cost function, which have less computational complexity than the usual nonlinear functions such as hyperbolic tangent and Gaussian functions. We use quasi-Newton method to solve the solution procedure of the cost function based on maximum likelihood criterion which has good asymptotic performance. The source data for simulation are taken from generalized Gaussian distribution series, as well as realistic voice signal. Simulation results show superior performance of the proposed algorithm compared with classical ones such as JadeR and fastica.

The water consumption and yield of rice were studied in a paddy field in cold region of North China under intermittent irrigation (IT), moisture irrigation (MT) and continuous flooding irrigation conditions (CSF). The results showed that the total water consumption of rice under IT and MT was lower than that under CSF by 13.9% and 20.2%., respectively, and the ANOVA analysis showed that at the 0.05 level, the difference was significant. The actual yield of rice under IT and MT was respectively 9.4% and 12.7% higher than that under CSF, while the theoretical yield of rice under IT and MT was respectively 10.9% and 13.6% higher than that under CSF, and the difference was significant (P

RCM(Reliability-Centered Maintenance) philosophy logic decision methods has great applied in developing aircraft maintenance program, but the dynamic management of its process has lagged behind relatively. Because of this, the maintenance program can't earn effective verification or continuing improvement, the validity of maintenance tasks is also declined. Reliability supervision includes the collection, identification, evaluation and disposal of the various meaningful information from signs of the failure or damage is about to take place. It can dynamically verify the validity of maintenance tasks, and improve the maintenance program continuously, which is the guarantee of reliability and operational safety of aircraft. This paper keeps focus on the main problem of aircraft reliability supervision and continuous improvement of maintenance program. Which contains: Firstly, make certain supervision objects. Select those items to supervise, which have safety, operational, significant economic consequences of functionally failure with a Hazard Risk Assessment Matrix. Secondly, data collection system is established. Failures, maintenance and operation information were collected. Thirdly, method of reliability evaluation and early-warning was developed. Through the supervision and statistics of relevant reliability parameters, a trends statistical chart is drawn. Warning information is sent out while the limit value is outrun. Finally, analyses the causation and replies of problems. The causation should be found when parameter is out of alarm value. Improvement proposition about design and manufacture should be given when the product has quality problem. For the problems of maintenance program settings, perform analysis on the applicability of maintenance type, the effectiveness and rationality of maintenance interval, validity of the operation method. And then put forward concrete proposition on improving maintenance program and maintenance operation cards.

Field experiment with three irrigation treatments, i.e., shallow-water irrigation (control irrigation I and control irrigation II), continuous flooding irrigation were installed to study the water requirement, yield and water use efficiency (WUE) of rice of Sanjiang Plain under different irrigation treatment. The results showed that in field experiment, the total water requirement was 557.4, 551.34 and 717.25mm for control irrigation I, control irrigation II and continuous flooding irrigation, respectively during the entire growing season. The total water requirement of control irrigation I and control irrigation II were lower than that of continuous flooding irrigation by 19.5% and 23.1%. The yield of control irrigation I, control irrigation II and continuous flooding irrigation was respectively 10150, 10650 and 9450kg.hm−2. The yield of control irrigation I and control irrigation II were 7.4% and 12.7% higher than that of continuous flooding irrigation. The WUE of control irrigation I, control irrigation II and continuous flooding irrigation was respectively 1.82, 1.93 and 1.32kg.m−3. The WUE of control irrigation I and control irrigation II were higher than that of continuous flooding irrigation by 37.9% and 46.2%. The results showed that shallow-water Irrigation was a comparatively water-saving and yield-increasing irrigation treatment.

A novel evaluation method is put forward. Firstly, it is a breakthrough for uniform weight evaluation patterns; it evaluates each evaluated object with the weight vector highlighting the relative advantage characteristic from individual viewpoint; Secondly, it puts forward a depict structure expressional form of one's individual advantage characteristic; thirdly, it evaluates each object according to each evaluated object's advantage value parameters, because different evaluation methods may result in different sort orders, evenness method is adopted to increase accuracy of the comprehensive evaluation. Lastly, relying on cluster analysis of individual advantage structure weights, the behavior pattern can be identified. The calculation example proves that the method is effective and feasible.

This paper we describe the on-going specification and development of Radio Block Center (RBC), a component of the new-generation Chinese Rail Train Control System (CTCS). The RBC will be responsible of managing the high speed (350 km/h) movement of trains equipped with radio communication. Although large-scale and complex RBC is safety and stability system, as a high level of safety system it must provide a safety services in any case. A new Hot Standby control system as real-time backup system when the CTCS system have any accident, in order to control the high speed of the train and keep it safe, thus providing a high level of confidence in their safety in a cost-effective way.

To improve the convergence rate, multiple input vectors are used to update the filter coefficients in the affine projection algorithm (APA) family, however, different input vector affects the convergence differently. The ideal selection criterion of the input vectors is derived by the largest decrease of the mean-square deviation. This selection condition fits for different algorithms of the APA family. Based on the ideal selection criterion, a family of APAs which dynamically select input vectors is presented. The experimental results show that the proposed algorithms have fast convergence speed, small steady-state error compared to the conventional APAs. In addition, the proposed algorithms retain low overall computational complexity.