scispace - formally typeset
Search or ask a question

Showing papers in "Systems engineering and electronics in 2006"


Journal Article
TL;DR: The paper defines the preference vector and coherence indexes of the entire group and provides a new theory and method of group decision in CHGDS and the new idea of group coherence decision mechanism basing on attributes weighted is provided.
Abstract: To answer the questions of the existing FCM involving local limit value and bad scalability,the paper puts forward an improved clustering algorithm MFCM(minimum fuzzy C-means) basing on MCDSA(minimum connected donating set algorithm).To help CHGDS(complex huge group-decision) using improved MFCM algorithm,the paper defines the preference vector and coherence indexes of the entire group and provides a new theory and method of group decision in CHGDS,it is validated by a test.Finally,the new idea of group coherence decision mechanism basing on attributes weighted is provided.

47 citations


Journal Article
TL;DR: A new evidence combination method is presented that enhances the rationality and reliability of the evidence combination through introducing weight coefficients and allotting the conflicted probability again and better results can be acquired.
Abstract: The evidence combination method is the key part of the evidence theory.It can combine the independent evidences from different information sources to acquire more reliable information.However,its application is limited by the shortcomings of Dempster-Shafer evidence combination method.This method may be useless when the conflict among evidences is rather great or even complete.A new evidence combination method is presented.In this improved method,the importance of the evidences is calculated.It enhances the rationality and reliability of the evidence combination through introducing weight coefficients and allotting the conflicted probability again and better results can be acquired.

18 citations


Journal Article
TL;DR: It is proved that the formula of dependent function based on "distance" and "location values" in extenics is error, and indicated that it cannot be applied to multi-sensor data fusion and its correlative fields.
Abstract: A variable fuzzy method based on the definition of relative difference function is proposed,which can identify the degree that object attracts and repels a property according to characteristic values of the object with regard to the property,so the variable fuzzy method pioneers a new way for recognition.The fuzzy variable recognition model of multi-sensor data fusion system is established,and then the model is applied to multi-sensor data fusion system for parts recognition.Finally an example is given.The case study proves that the model and method are reasonable and practicable.In the end,from theory and application,it is proved that the formula of dependent function based on "distance" and "location values" in extenics is error,and indicated that it cannot be applied to multi-sensor data fusion and its correlative fields.

18 citations


Journal Article
TL;DR: By applying the method to the target identification, the simulation experiment shows that it can identify the target accurately and is an effective and feasible multi-sensor data fusion method.
Abstract: Focused on the problems that it is difficult to determine the reliability of each sensor and how the data measured by different sensors are fused,a multi-sensor data fusion method based on correlation function and fuzzy integration function is proposed.The mutual supportability of multiple sensors is obtained from the correlation function.Then by the membership function,the reliability of information provided by each sensor is gained.Finally,the supposed fusion result of the object attribute from the multiple sensors could be produced on the basis of fuzzy integration function.This method is simple computationally and can objectively reflect the reliability of each sensor and interrelationship between these sensors.By applying the method to the target identification,the simulation experiment shows that it can identify the target accurately and is an effective and feasible multi-sensor data fusion method.

18 citations


Journal Article
TL;DR: A new learning algorithm based on the feedforward neural networks is proposed for solving the problems of interpolation using two systems of linear equations for getting the forms of weights' cubic spline functions without problems such as local minima, slow convergence, and dependent on initialized values arising from the steepest descent-like algorithms.
Abstract: A new learning algorithm based on the feedforward neural networks is proposed for solving the problems of interpolation.The network's topology is so simple that the number of neurons needed for solving the problem under investigation is independent of the number of patterns.The number of neurons needed for the training is simply expressed as the multiplication of the input patterns' vector dimension by the output patterns' vector dimension.Only one layer weights are trained.The trained weights consist of cubic spline functions,not constants like the popular learning algorithm such as backpropagation(BP) or radial basis function(RBF) networks.The new algorithm developed is the learning procedure to solve two systems of linear equations for getting the forms of weights' cubic spline functions,without problems such as local minima,slow convergence,and dependent on initialized values arising from the steepest descent-like algorithms.Finally,to illustrate the power of the new learning algorithm,some simulation examples are presented to show good performance in interpolation precision and training speed compared with the popular learning algorithm such as backpropagation and radial basis function networks.

17 citations


Journal Article
TL;DR: In this article, the problem of dynamic hybrid multi-attribute decision making under risk, which integrates with the precision number, interval number and fuzzy number, and an unknown weight information on indexes, is studied.
Abstract: The problem of dynamic hybrid multi-attribute decision making under risk,which integrates with the precision number,interval number and fuzzy number,and an unknown weight information on indexes,is studied.Based on the grey matrix relative degree,a new decision making method is presented,and a decision making example is given to demonstrate the feasibility and rationality of this new method.Therefore,a new and effective way is obtained to solve the problem of dynamic hybrid multi-attribute decision making under risk.

16 citations


Journal Article
TL;DR: Based on the definitions of entropy, a method of getting weight is proposed, and corresponding algorithm of grey fixed weight clustering decision-making is given, which uses the state data of system for calculating entropy and getting decision weight.
Abstract: The weight is given before clustering process in traditional grey fixed weight clustering decision,while it is not an objective method.And study on how to get the weight of different indexes scientifically has not thought much.Based on the definitions of entropy,a method of getting weight is proposed,and then corresponding algorithm of grey fixed weight clustering decision-making is given.This method uses the state data of system for calculating entropy and getting decision weight.Taken a practical problem as a sample,an empirical study is proposed.The results show that the method is superior to traditional one.It is easy to calculating,and its way to get weight is objective.So it is an effective development to grey clustering decisionmaking theory.

16 citations


Journal Article
TL;DR: An improved crossover operation is proposed, different crossover strategies are selected according to the diversity of population and the relativity of chromosome, andffective crossover operations are decreased greatly so the convergence of the algorithm is improved.
Abstract: Crossover operation is the most important operation of genetic algorithms,it is the key to the convergence of genetic algorithms.An improved crossover operation is proposed,diversity of population and relativity of chromosome are defined,different crossover strategies are selected according to the diversity of population and the relativity of chromosome,ineffective crossover operations are decreased greatly,so the convergence of the algorithm is improved.The simulation result of complicated function optimization shows that this improved crossover operation is much more effective than the standard crossover operation.

15 citations


Journal Article
TL;DR: Aiming at the problem of radar radiating source recognition system, a scheme applying wavelet transform to feature extraction of emitter is put forward, which can conquer the disadvantages of conventional methods, and increase the accuracy of envelope extraction.
Abstract: With the rapidly developing of electronic technology,the system and modulation manner of radar signals became more and more complicated and various,and circumstance of signals became increasing densenessIt results in that the routine method and theory of recognition can hardly satisfy practical requirement and can't effectively recognize for radar signalsSo,the strict demand has been presented for study on recognition of radar signalsAiming at the problem of radar radiating source recognition system,we put forward a scheme applying wavelet transform to feature extraction of emitterThe method can conquer the disadvantages of conventional methods,increase the accuracy of envelope extractionAt last,the results of test show that the method is effective

13 citations


Journal Article
TL;DR: A new method of multi-sensor data fusion based on relative distance based on fuzzy set theory that avoids the design of relation threshold and reduces the influence of prior information, and avoids the absoluteness of correlated support degree among measurement data.
Abstract: The cause of multi-sensor measurement data's uncertainty is analyzed,and a new method of multi-sensor data fusion based on relative distance is presented.The new method utilizes setting membership function of fuzzy set theory to dig the redundancy and complementary information, and adjusts the support degree to reasonably distribute the weight value of measure data.At the same time,the method also avoids the design of relation threshold and reduces the influence of prior information,then avoids the absoluteness of correlated support degree among measurement data.Finally,the simulation proves the validity of the method.

12 citations


Journal Article
TL;DR: Two feature extraction methods based on differential power spectrum(DPS) and differential cepstrum,originally used in the research area of speech signal processing and homomorphic signal processing are respectively introduced to the radar target recognition community.
Abstract: The problem of target recognition using the high resolution radar range profiles is discussedTwo feature extraction methods based on differential power spectrum(DPS) and differential cepstrum,originally used in the research area of speech signal processing and homomorphic signal processing are respectively introduced to the radar target recognition communityTwo differential power spectrum based features are applied to target classificationA multi-layered feed forward neural network with SARPROP(simulated annealing resilient propagation) algorithm is selected as classifierThe range profiles are obtained with step-frequency technique and the two-dimension backscatter distribution data of four different scaled aircraft modelsSimulations are presented to evaluate the classification performance with the above featuresThe results show that the differential power spectrum based feature is effective and robust for the radar target recognition

Journal Article
TL;DR: In this paper, a modified Grey Relational Analysis (GRA) method is proposed to solve the problem of multiple attribute decision-making with incomplete information on attribute weights to which the attribute values are given in terms of interval numbers.
Abstract: For the problem of multiple attribute decision-making with incomplete information on attribute weights to which the attribute values are given in terms of interval numbers,a modified Grey Relational Analysis(GRA) method is proposed.Then the calculation steps to solve it are given.As the key step,a single objective programming model is developed to determine the relation degree between every alternative and positive ideal point or negative ideal point.Then,according to the relation degree,the relative relation degree to the positive point is calculated to rank all alternatives.At last,a numerical example is provided to illustrate the proposed method.The result shows the approach is simple,effective and easy to calculate.

Journal Article
TL;DR: The application in the scheme selection for a certain V/STOL aircraft proves the validity of this decision-making method based fuzzy analytic hierarchy process.
Abstract: Aimed at the difficulty to make optimal selection from design scheme of complex systems,a decision-making method based fuzzy analytic hierarchy process(FAHP) is brought forward.A brief introduction to the basic theory of FAHP is presented and a mathematical model of fuzzy analytic hierarchy process is built,then the building method of fuzzy complementary judgment matrix,the formula of weight and the consistency check method of judgment matrix are given.The application in the scheme selection for a certain V/STOL aircraft proves the validity of this method.

Journal Article
TL;DR: This paper summarized the main results on fault diagnosis of networked control systems and investigated the fault detection and fault-tolerant control theory and technology in recent years.
Abstract: Networked control systems have many advantages compared with traditional control systems,but at the same time,the network-induced delay,packet dropout,asynchronous and other peculiarities of networked control systems may degrade the performance of closed-loop systemsWe have to develop new theory and techniques for control systems that operate through data networkFault diagnosis and fault-tolerant control are very important issues for practical control systems,particularly for safety-critical systemsWe investigated the fault detection and fault-tolerant control theory and technology for networked control system in recent yearsThis paper summarized our main results on fault diagnosis of networked control systems

Journal Article
TL;DR: Aimed at the fuzziness of the scheme evaluation, a new method based on entropy weight of fuzzy information is presented and an example proves the validity of this method.
Abstract: Aimed at the fuzziness of the scheme evaluation,a new method based on entropy weight of fuzzy information is presented.According to fuzzy rating matrix,approximation degree and distance of triangular fuzzy number to the ideal points,reasonable schemes are evaluated and the most optimal scheme is selected with only fuzzy judging matrix but without experts' weights.At last,an example proves the validity of this method.

Journal Article
TL;DR: According to the color spatial feature of an image, a new algorithm based on entropy is proposed, which introduces color spatial distribution entropy (SDE) as the spatial descriptor of animage which describes the distribution of pixels with the same color in the image.
Abstract: According to the color spatial feature of an image,a new algorithm based on entropy is proposedIt introduces color spatial distribution entropy(SDE) as the spatial descriptor of an image,which describes the distribution of pixels with the same color in the imageSDE is also rotation,translation and scale invariantIntegrated with the human vision and the characters of entropy,improvements are made on the new partition method and the weight functionBased on SDE and color histogram,two retrieval methods are proposed to compute the similarity of two imagesExperiments show that better results can be achieved than color histogram

Journal Article
TL;DR: A subaperture processing method is proposed for stepped frequency chirp signal(SFCS) compression by matching filtering for the echo of each transmitted subpulse andIFFT is performed to get range profile.
Abstract: A subaperture processing method is proposed for stepped frequency chirp signal(SFCS) compression.In this method,matched filtering for the echo of each transmitted subpulse is firstly performed,the filtered spectrum is then up moved,and coherently combined in frequency domain,and finally,IFFT is performed to get range profile.Another approach, by which the echo of each subpulse is firstly matched filtered,then digitally up mixed,and finally combined in time domain,is also introduced.Different methods of windowing are applied and compared with each(other.) Simulations show that different windowing methods affect the compression result dramatically.The method is suitable for both linearly stepped case and nonlinearly stepped case.

Journal Article
TL;DR: In this article, a new method of grey relational analysis (GRA) is proposed for multiple attribute decision-making problem in which both the attribute weights and attribute values are interval numbers, and the priority of alternative is determined by comparing the grey relational degree intervals of alternatives based on a ranking approach of interval numbers.
Abstract: With respect to multiple attribute decision-making problem in which both the attribute weights and attribute values are interval numbers,a new method of grey relational analysis(GRA) is proposed.Firstly,the definition of positive ideal solution of multiple attribute decision-making problem with interval numbers is given,and by which the grey relational coefficient of alternatives from the positive ideal solution is drived.Secondly,the grey relational degrees of alternatives are obtained by using the scale-multiplication and addition operations of interval numbers.The priority of alternative is determined by comparing the grey relational degree intervals of alternatives based on a ranking approach of the interval numbers.At last,a numerical example is provided to illustrate the proposed method.The result shows the approach is simple,effective and easy to implement.

Journal Article
TL;DR: In this article, the relation between difference equation and differential equation is studied by analyzing the mechanism in building grey models, and a "bridge" is set up between Difference Equation and differential Equation by sampling theorem and state transition matrix.
Abstract: Parameters of grey model GM(1,N) are often estimated based on related difference equation,but time response function is acquired by the solution of differential equation.No reasonable scientific or theoretical basis is given to explain the jump from difference equation to differential equation till now.The relation between difference equation and differential equation is studied by analyzing the mechanism in building grey models,and a "bridge" is set up between difference equation and differential equation by sampling theorem and state transition matrix.It proves this method has validity by analysis of simulation result.

Journal Article
TL;DR: A model solving approach based on heuristic method is presented, the framework of model solving and the detailed solving algorithm are also given and practically applications indicate that the model is valid and the obtained result is feasible.
Abstract: In order to measure area air defense deployment comprehensively and rationally,to solve the problem of optimization deployment under the circumstance of military strength that is known,a concept of protect value is put forward,which is an integrated measurement of protect ability and important degree of object defended;then a new kind of index system of evaluating an area air defense deployment,is built and the determining method of each factor is given.In this base,an area air defense optimization deployment model using protect value as objective function is built;in the mean while,a model solving approach based on heuristic method is presented,the framework of model solving and the detailed solving algorithm are also given.Practically applications indicate that the model is valid and the obtained result is feasible.

Journal Article
TL;DR: The super-efficiency DEA model is used to sufficiently compare and rank the efficient evaluation object in the case having more than one efficient evaluation objects and the problem of unrealistic weight is considered.
Abstract: To make multi-index evaluation containing cost-type index,income-type index and fixed-type index,DEA method is used to make evaluation,the DEA model used to make multi-index evaluation is established.To the problem of unrealistic weight,the weight restriction is considered to make it yield more reasonable input and output weights,and the super-efficiency DEA model is used to sufficiently compare and rank the efficient evaluation object in the case having more than one efficient evaluation objects.In the end,an empirical example is illustrated.

Journal Article
TL;DR: A set-pair-analysis-based method for multiple attribute decision making problems in which both the attribute weights and the elements of decision matrix are interval numbers is proposed, and its ranking rule has more intuitional meanings.
Abstract: A set-pair-analysis-based method is proposed for multiple attribute decision making problems in which both the attribute weights and the elements of decision matrix are interval numbers.By referring to the thought that the universe is divided into three parts in the set-pair-analysis theory,the interval evaluations are transformed into connection numbers.Two ranking criteria are presented and their rationality is analyzed.Based on the fact that the freedom degree of the interval attribute weights vector is different from that of the evaluation vector of every alternative in decision matrix,different methods to deal with the vectors are needed.Then an example is given to illustrate the proposed method.The result shows that the method is effective for ranking.Compared with popular approaches such as mathematical programming,the proposed method is more simply and conveniently,and its ranking rule has more intuitional meanings.

Journal Article
TL;DR: Aimed at the processing of unknown radar signals, a new clustering approach for deinterleaving the emitters is put forward by applying Kohonen net model and weighted values are determined based on entropy and statistical theory.
Abstract: Aimed at the processing of unknown radar signals,a new clustering approach for deinterleaving the emitters is put forward.By applying Kohonen net model,new conception of weighted value is introduced,weighted values are determined based on entropy and statistical theory,weighted and normalized Minkowsky distance is computed,minimum distance is veried,and similar clustering centers are merged according to hypothesis testing theory.In this way, the interleaved radar signals are sorted into individual cells.Computer simulation results show that the proposed approach can successfully deinterleave unknown radar signals.

Journal Article
TL;DR: A global and parallel hillclimbing searching strategy is presented, which realized non-ergodic searching and can be used to find the best matching point very quickly, and introduced a searching path-table to marker the point which had been searched.
Abstract: Image matching based on grey cross-correlation,has been widely applied to various field because of its high matching probability.But the searching strategy of conventional grey cross-correlation algorithm is ergodic,the speed of matching is slow.To improve the speed of image matching,a global and parallel hillclimbing searching strategy are presented,which realized non-ergodic searching and can be used to find the best matching point very quickly,also introduced a searching path-table to marker the point which had been searched.Experiment show that this approach has high speed and accuracy in image matching.

Journal Article
TL;DR: A new dynamic weight self-adjusting algorithm is presented for ensemble learning method and it is shown that ensemble classification performance is improved by use of this algorithm.
Abstract: As soon as a classifier is trained by AdaBoost ensemble learning algorithm,it has a constant weight for all test instances.A few of classifiers which have better classification performance for some instances hard to classified have usually small weights.A new dynamic weight self-adjusting algorithm is presented for ensemble learning method.The effective determining area of the test instance is computed automatically based on the classification behavior of classifiers.Some combine classifiers are selected and their weights are adjusted based on the effective determining area of the test instance.An integration decision is made by using of the statistics information of sets of instances.The experiment result shows that ensemble classification performance is improved by use of this algorithm.

Journal Article
Gao Zhi-wei1
TL;DR: In this article, the authors studied the problem of fault-tolerant control for singular bilinear systems and developed state-feedback controllers to ensure exponential stability of the closed-loop system and the required H∞ norm performance index.
Abstract: To study the problem of fault-tolerant control for singular bilinear systems,the concept of H∞ control for nonlinear systems is used to and develope state-feedback controllers.The existence conditions and design methods of the state-feedback controllers are proposed.The controllers,whether the actuators are normal or abnormal,could ensure exponential stability of the closed-loop system and the required H∞ norm performance index.And then the fault-tolerant control is achieved.Finally,a simulated example is given.


Journal Article
TL;DR: An authentication protocol against DoS attacks that is improved by asking the client to commit its system resources to the execution of the protocol before the server allocates its memory and processing time.
Abstract: Denial of service(DoS) by exhausting the server resources has become a major security threat in open networks.Particularly,wireless networks are vulnerable to DoS attacks when they have much fewer system resources than the wired counterparts.The basic strategy against DoS attacks is to impose an adjustable cost on the attackers while the attacks are being launched.An authentication protocol against DoS attacks that is improved by asking the client to commit its system resources to the execution of the protocol before the server allocates its memory and processing time.The server sends the client a puzzle whose solution requires a brute force search for some bits of the inverse of an one way Hash function.The difficulty of the puzzle is parameterized according to the server available resources.The server stores the protocol state and computes expensive public key operations only after it has verified the client's solution.The puzzle protects the server that authenticates the clients against resource exhaustion attacks during the first messages of the connection opening before the clients are reliably authenticated.Then a four pass public key authentication and key establishment protocol is proposed.

Journal Article
Shi Kai-quan1
TL;DR: The new metal material discovery-recognition is discussed by use of attribute value generation, and the attribute value analysis of newMetal material is made and the results presented coincide with the practical results.
Abstract: S-rough sets(singular rough sets)have two forms: one direction S-rough sets,and two direction S-rough sets,and they have heredity characteristics and memory characteristics.The S-rough sets are penetrated and shared with materials.The new metal material discovery-recognition is discussed.By use of attribute value generation,the attribute value analysis of new metal material is made and the results presented coincide with the practical results.The research on S-rough sets is a new direction,and S-rough sets is a mathematical tool for new material discovery-recognition.

Journal Article
TL;DR: The experimental results show that the compound optimum model PSO proposed can improve the convergence speed and global searching ability of the algorithm.
Abstract: The global optimum model(Gbest) of particle swarm optimization(PSO) has characteristics of fast convergence and good local searching ability but poor robustness,while the local optimum model(Lbest) has characteristics of good global searching ability and robustness but slow convergence.Incorporating the characteristics of these two models,a compound optimum model PSO is proposed.Tests are done using four Benchmarks functions.The experimental results show that,compared with Gbest model and Lbest model,the strategy proposed can improve the convergence speed and global searching ability of the algorithm.Lastly the new algorithm is applied to identify a nonlinear system model,and the result proves the effectiveness of the algorithm.