ISSN 
2004 Vol. 33, No. 1
2004, 33(1): 1-4.
Abstract:
This paper is based on the gray-level co-occurrence matrix method, and particularly study some texture features used for the classification of SAR images, including difference variance、difference average、difference entropy、contrast、energy、variance、sum variance、inverse difference moment and correlation etc. Furthermore we have abstracted features of SAR image and studied classification characteristic. Using criterion called distance of inside classes and between classes, we can get a few features of a SAR image which is good at image classification. At last, making use of BP neural network of three layers, we proceed image classification and get the satisfied results.
This paper is based on the gray-level co-occurrence matrix method, and particularly study some texture features used for the classification of SAR images, including difference variance、difference average、difference entropy、contrast、energy、variance、sum variance、inverse difference moment and correlation etc. Furthermore we have abstracted features of SAR image and studied classification characteristic. Using criterion called distance of inside classes and between classes, we can get a few features of a SAR image which is good at image classification. At last, making use of BP neural network of three layers, we proceed image classification and get the satisfied results.
2004, 33(1): 5-8.
Abstract:
A new subspace algorithm(LSM-ALSM) is discussed in this paper. By rotating the subspaces of ALSM, LSM changes their independent relations and improves the ALSM classifier's recognition capacity in alike characters. To realize the best balance between recognized rate, rejected rate and error rate, we presents performance function instead of experience function to determine the refuse rules. The experiment with limited character set shows that LSM-ALSM method improves the recognized rate and credibility effectively.
A new subspace algorithm(LSM-ALSM) is discussed in this paper. By rotating the subspaces of ALSM, LSM changes their independent relations and improves the ALSM classifier's recognition capacity in alike characters. To realize the best balance between recognized rate, rejected rate and error rate, we presents performance function instead of experience function to determine the refuse rules. The experiment with limited character set shows that LSM-ALSM method improves the recognized rate and credibility effectively.
2004, 33(1): 9-11,34.
Abstract:
Wigner ville distribation and short time Fourier transform are the main time-frequency analysis methods. However, both of them have their advantages and limitations. On the base of the definition of the two transform, this paper analyzes the connection between them, then proposes an improved method through the construction of window function. Computer simulation shows the time-frequency character of signal is improved and the method is applicable to the analysis of multicomponent nonstationary signal.
Wigner ville distribation and short time Fourier transform are the main time-frequency analysis methods. However, both of them have their advantages and limitations. On the base of the definition of the two transform, this paper analyzes the connection between them, then proposes an improved method through the construction of window function. Computer simulation shows the time-frequency character of signal is improved and the method is applicable to the analysis of multicomponent nonstationary signal.
2004, 33(1): 12-15.
Abstract:
Direct Digital Synthesis is a new kind of frequency synthesizing technology that is widely applied in arbitrary waveform generator. This paper introduces a realization of DDS, which is based on programmable device. It also presented a method of realizing frequency modulation by real-time changing frequency tuning word of high-speed phase accumulator, Frequency sweep, FM and FSK modulation are analyzed and designed. The data can be built stably and the frequency be synthesized is correct.
Direct Digital Synthesis is a new kind of frequency synthesizing technology that is widely applied in arbitrary waveform generator. This paper introduces a realization of DDS, which is based on programmable device. It also presented a method of realizing frequency modulation by real-time changing frequency tuning word of high-speed phase accumulator, Frequency sweep, FM and FSK modulation are analyzed and designed. The data can be built stably and the frequency be synthesized is correct.
2004, 33(1): 16-18,62.
Abstract:
The fundmental of designing hairpin microstrip filter and the application of ADS in the design of hairpin microstrip filter with frequency in the L-band is introduced. The approach of obtaining the relationship between microstrip resonator's spacing and microstrip resonator's coupling coefficient is given. The results of model simulation and EM simulation are shown. At last,the design example and measured result for a hairpin bandpass filter of fifth order with frequency in the L-band are given.
The fundmental of designing hairpin microstrip filter and the application of ADS in the design of hairpin microstrip filter with frequency in the L-band is introduced. The approach of obtaining the relationship between microstrip resonator's spacing and microstrip resonator's coupling coefficient is given. The results of model simulation and EM simulation are shown. At last,the design example and measured result for a hairpin bandpass filter of fifth order with frequency in the L-band are given.
2004, 33(1): 19-22.
Abstract:
To realize a fast and accurate parameter estimation of multiple linear frequency modulated signals, using the property of ambiguity function and Radon transform, the paper presents a fast algorithm for estimation chirp rates of multiple LFM signals by ambiguity function calculated by fractional autocorrelation, and by Radon transform. The result of simulation on the evaluation module board of TMS320C6701 shows that the algorithm has a better parameter performance and high computing efficiency under the low SNR.
To realize a fast and accurate parameter estimation of multiple linear frequency modulated signals, using the property of ambiguity function and Radon transform, the paper presents a fast algorithm for estimation chirp rates of multiple LFM signals by ambiguity function calculated by fractional autocorrelation, and by Radon transform. The result of simulation on the evaluation module board of TMS320C6701 shows that the algorithm has a better parameter performance and high computing efficiency under the low SNR.
2004, 33(1): 23-26.
Abstract:
The consideration of Enhanced Covariance Matrix Technique(ECMT) in DOA estimation for coherent signal sources using large-scale array is discussed and introduced in the paper. This technique is to reduce the noise power contribution to the covariance matrix so as to enhance the signal to noise ratio at reception. It's found in our simulation and analysis that the technique can achieve high resolution in resolving coherent signal sources and closely spaced signals. The simulation results are given to validate the consideration.
The consideration of Enhanced Covariance Matrix Technique(ECMT) in DOA estimation for coherent signal sources using large-scale array is discussed and introduced in the paper. This technique is to reduce the noise power contribution to the covariance matrix so as to enhance the signal to noise ratio at reception. It's found in our simulation and analysis that the technique can achieve high resolution in resolving coherent signal sources and closely spaced signals. The simulation results are given to validate the consideration.
2004, 33(1): 27-30.
Abstract:
The radar used Linear frequency modulated continuous wave(LFMCW) signal can detect the information of range and Doppler of target. This paper analyzes the LFMCW signal used in HF radar, then deduces the expression of its beat frequency, and the theoretical for range/Doppler processing used two-dimensional fast Fourier transform(FFT) technique. The range/Doppler processing and the adaptive digital beamforming(ADBF) technique combined to extract the information of target echoes while control the effect of interrupt and noise. The simulation results are provided in this paper.
The radar used Linear frequency modulated continuous wave(LFMCW) signal can detect the information of range and Doppler of target. This paper analyzes the LFMCW signal used in HF radar, then deduces the expression of its beat frequency, and the theoretical for range/Doppler processing used two-dimensional fast Fourier transform(FFT) technique. The range/Doppler processing and the adaptive digital beamforming(ADBF) technique combined to extract the information of target echoes while control the effect of interrupt and noise. The simulation results are provided in this paper.
2004, 33(1): 31-34.
Abstract:
In this paper, we construct MW R-S code based on Reed-Solomon code, this code belong to 2-D codes, its auto-relativity is zero and the cross-relativity is one. Analyzed the capacity of 2-D R-S code with λa=λa=λ=1 based on the Johnson boundary condition, compared with the code we constructed, it shows that the code we constructed is nearly optimal. And we carried on the datum analysis on the performance of the code in 2-D OCDMA system with only considering the multi-user interference, got the curves of BER via the number of simultaneous users in 2-D OCDMA system with different prime p. This result supplies important reference for the further theory investigation on 2-D OCDMA system.
In this paper, we construct MW R-S code based on Reed-Solomon code, this code belong to 2-D codes, its auto-relativity is zero and the cross-relativity is one. Analyzed the capacity of 2-D R-S code with λa=λa=λ=1 based on the Johnson boundary condition, compared with the code we constructed, it shows that the code we constructed is nearly optimal. And we carried on the datum analysis on the performance of the code in 2-D OCDMA system with only considering the multi-user interference, got the curves of BER via the number of simultaneous users in 2-D OCDMA system with different prime p. This result supplies important reference for the further theory investigation on 2-D OCDMA system.
2004, 33(1): 35-38.
Abstract:
In this paper, the noise of receiver and its probability distribution are analysed, and the spectrum of outer space is calculated. Then, a new way to calculate minimal detectable power of laser radar is given, which is deduced on the base of signal detection and two statistic concepts:false alarm rates and detection probabilities. System simulation proves the minimal detectable power calculated by this method can optimize the laser radar system.
In this paper, the noise of receiver and its probability distribution are analysed, and the spectrum of outer space is calculated. Then, a new way to calculate minimal detectable power of laser radar is given, which is deduced on the base of signal detection and two statistic concepts:false alarm rates and detection probabilities. System simulation proves the minimal detectable power calculated by this method can optimize the laser radar system.
2004, 33(1): 39-42.
Abstract:
The interfering mechanism on the photoelectric detector is analyzed in damage mechanism and jamming threshold. The model of laser interfering probability is given. The self-adaptive interference and even time-domain interference are provided and the theoretic probability of two different types of interference on the IR scanning imaging systems is calculated. The effectiveness of surveillance and interference mode is proved in the experimental results. It is also proved that the laser self-adaptive interference is the first choice in different types of optoelectronic countermeasures.
The interfering mechanism on the photoelectric detector is analyzed in damage mechanism and jamming threshold. The model of laser interfering probability is given. The self-adaptive interference and even time-domain interference are provided and the theoretic probability of two different types of interference on the IR scanning imaging systems is calculated. The effectiveness of surveillance and interference mode is proved in the experimental results. It is also proved that the laser self-adaptive interference is the first choice in different types of optoelectronic countermeasures.
2004, 33(1): 43-45.
Abstract:
Two phase mask plate with independent chirp signals are used to encode image in the joint transform correlator,so as to make encrypt the original image.The characteristic of this method is when decoding,the rotation of encrypted image can not affected decoding effect,and moreover,owing to the fact that phase distribution is a decisive function only grasped by owner of secure key and the secure key can be regenerated.It makes the practical use convenient.The results of computer simulation shows the feasibility of this method.
Two phase mask plate with independent chirp signals are used to encode image in the joint transform correlator,so as to make encrypt the original image.The characteristic of this method is when decoding,the rotation of encrypted image can not affected decoding effect,and moreover,owing to the fact that phase distribution is a decisive function only grasped by owner of secure key and the secure key can be regenerated.It makes the practical use convenient.The results of computer simulation shows the feasibility of this method.
2004, 33(1): 46-49.
Abstract:
The time-frequency distribution series(TFDS) has been tested effective in the balance of the cross-term suppression and the time-frequency resolution. It is promising in the application of the target recognition. But the computational load is too large to bear with the length of the signal increasing. By getting ride of the redundant computing about the time-frequency atom, we overcome the difficulty. The simulation of the multi-component non-stational signal shows that the improvement of the TFDS method accelerates the computing. It can be consulted under the use of the real-time processing of the time-frequency characteristic of the target echo.
The time-frequency distribution series(TFDS) has been tested effective in the balance of the cross-term suppression and the time-frequency resolution. It is promising in the application of the target recognition. But the computational load is too large to bear with the length of the signal increasing. By getting ride of the redundant computing about the time-frequency atom, we overcome the difficulty. The simulation of the multi-component non-stational signal shows that the improvement of the TFDS method accelerates the computing. It can be consulted under the use of the real-time processing of the time-frequency characteristic of the target echo.
2004, 33(1): 50-52.
Abstract:
The plasma-assisted slow-wave oscillator is a class of unique high-power microwave source, A class of fibre-optic controlled gases-puff system for plasma-cathode electron gun of the plasma assisted slow-wave oscillator is developed in this paper. Gases were injected into the hollow cathode, and the modulator is triggered at arbitrary delay to generate low-pressure glow discharge at the same time. And the discharge experiment of the electron gun in this system was presented. Operating on pulse-gas-puff-pressure at 0.28~0.4 Pa, representative discharge current up to 120~167 A, and transporting e-beam current up to 80 A.
The plasma-assisted slow-wave oscillator is a class of unique high-power microwave source, A class of fibre-optic controlled gases-puff system for plasma-cathode electron gun of the plasma assisted slow-wave oscillator is developed in this paper. Gases were injected into the hollow cathode, and the modulator is triggered at arbitrary delay to generate low-pressure glow discharge at the same time. And the discharge experiment of the electron gun in this system was presented. Operating on pulse-gas-puff-pressure at 0.28~0.4 Pa, representative discharge current up to 120~167 A, and transporting e-beam current up to 80 A.
2004, 33(1): 53-55.
Abstract:
On the basis of having studied the characteristic of JN43' coke oven, Generalized predicative control algorithm has been designed for coke oven heating control system, by using of direct algorithm, calculate work on line is reduced. the control algorithm preferably settled down the fluctuation of flue temperature due to some factors,like as the amount of coal, coking period, the moisture in compound coal and so on. the results show that the fluctuation of flue temperature reduced from ±10℃ to ±5℃, and annual average economized BFG capacity for 1 032 m3, the energy saving is 3.1%.
On the basis of having studied the characteristic of JN43' coke oven, Generalized predicative control algorithm has been designed for coke oven heating control system, by using of direct algorithm, calculate work on line is reduced. the control algorithm preferably settled down the fluctuation of flue temperature due to some factors,like as the amount of coal, coking period, the moisture in compound coal and so on. the results show that the fluctuation of flue temperature reduced from ±10℃ to ±5℃, and annual average economized BFG capacity for 1 032 m3, the energy saving is 3.1%.
2004, 33(1): 56-58,70.
Abstract:
The new low-loss wide-angle Y-branch is designed and optimized in this paper, in detail. Device performance is examined numerically by finite-difference beam propagation method. The simulated results show the radiation loss of these devices is very small with relatively wide branching angle. At last a new 1×4 optical splitter composed of three new Y-branches is calculated. The simulation shows the new splitter has the characteristics of low-loss and simple-structure, and its length is decreased greatly.
The new low-loss wide-angle Y-branch is designed and optimized in this paper, in detail. Device performance is examined numerically by finite-difference beam propagation method. The simulated results show the radiation loss of these devices is very small with relatively wide branching angle. At last a new 1×4 optical splitter composed of three new Y-branches is calculated. The simulation shows the new splitter has the characteristics of low-loss and simple-structure, and its length is decreased greatly.
2004, 33(1): 59-62.
Abstract:
Deformation structure induced by indentation in GaAs single crystal has been studied using transmission electron microscopy. The result shows that resettle-like asymmetric dislocations, which consist of elongated-arm dislocations and short-arm ones, and twofold symmetrical twins occurres around indentation. The dislocation configuration formed by the nucleation, movement and interaction of the Frank-Read sources on six slipping planes of the GaAs crystal, and is closely related to the various mobility of the α and β dislocations in GaAs crystal. The formation of the twins is closely related to the stacking_faults.
Deformation structure induced by indentation in GaAs single crystal has been studied using transmission electron microscopy. The result shows that resettle-like asymmetric dislocations, which consist of elongated-arm dislocations and short-arm ones, and twofold symmetrical twins occurres around indentation. The dislocation configuration formed by the nucleation, movement and interaction of the Frank-Read sources on six slipping planes of the GaAs crystal, and is closely related to the various mobility of the α and β dislocations in GaAs crystal. The formation of the twins is closely related to the stacking_faults.
2004, 33(1): 63-66.
Abstract:
This paper focuses on the independent component analysis presenting a review on the basic model, the main method, the mathematical principle of blind source separation, analyzing the possibility of separation. The paper puts forward the method that based on the neural network unsupervised learning, also, improves the index on separation effects. Under the biology information processing background, combining artificial neural network with information theory to resolve this kind of problem can get the separation matrix adaptively by itself. It can separate mixtures which have both positive and negative kurtosis, and promotes the research of blind source separation greatly.
This paper focuses on the independent component analysis presenting a review on the basic model, the main method, the mathematical principle of blind source separation, analyzing the possibility of separation. The paper puts forward the method that based on the neural network unsupervised learning, also, improves the index on separation effects. Under the biology information processing background, combining artificial neural network with information theory to resolve this kind of problem can get the separation matrix adaptively by itself. It can separate mixtures which have both positive and negative kurtosis, and promotes the research of blind source separation greatly.
2004, 33(1): 67-70.
Abstract:
Raise the concept of distributed VPN and create the system model, which extend the concept of classical VPN. From the as aspects of security, it overcome the classical VPN's disadvantages of failing to protect the internal communication of LANs. The result of DVPN's performance analyzing indicates that it can enhance the parallelity of the system,reduce the storage requirement from complexity O(n2) to O(n) and make the database grow linearly. As well,give a system implementation, which proves that it can improve the security and efficiency of the system.
Raise the concept of distributed VPN and create the system model, which extend the concept of classical VPN. From the as aspects of security, it overcome the classical VPN's disadvantages of failing to protect the internal communication of LANs. The result of DVPN's performance analyzing indicates that it can enhance the parallelity of the system,reduce the storage requirement from complexity O(n2) to O(n) and make the database grow linearly. As well,give a system implementation, which proves that it can improve the security and efficiency of the system.
2004, 33(1): 71-74.
Abstract:
This paper presents a novel IDS mechanism based on the theory of data fusion-DFIDS, which is designed, according to intrusion patterns with complex features, to heighten the accuracy of the system while detecting complex intrusion acts. The performance of the whole detecting system is enhanced because DFYDS effectively overcomes the limitations of conventional single sensor detecting system regarding aspects of data collecting and analyzing by using optimized model of the fusion of parallel-distributed detection and decision. In this paper, a theoretical model of DFIDS is established and compared with traditional IDS, which proves that DFIDS is of better performance as to detecting accuracy.
This paper presents a novel IDS mechanism based on the theory of data fusion-DFIDS, which is designed, according to intrusion patterns with complex features, to heighten the accuracy of the system while detecting complex intrusion acts. The performance of the whole detecting system is enhanced because DFYDS effectively overcomes the limitations of conventional single sensor detecting system regarding aspects of data collecting and analyzing by using optimized model of the fusion of parallel-distributed detection and decision. In this paper, a theoretical model of DFIDS is established and compared with traditional IDS, which proves that DFIDS is of better performance as to detecting accuracy.
2004, 33(1): 75-78.
Abstract:
Mainly on the principle of Ipchains embedded in Linux kernel has been discussed. A series of rules can be set up with the input chains or output chains or forward chains or user defined chains, these rules can filter the packet of input or output or forward by host computer and can also filter packet from one network card to another network card by computer. Therefore, the access control can be implemented base on data of source address and destination address and protocol number. This technology is suitable for the users such as small enterprise or department to set up a firewall without professional firewall production. An example is presented for demonstrating how to use the proposed technology as firewall in LAN.
Mainly on the principle of Ipchains embedded in Linux kernel has been discussed. A series of rules can be set up with the input chains or output chains or forward chains or user defined chains, these rules can filter the packet of input or output or forward by host computer and can also filter packet from one network card to another network card by computer. Therefore, the access control can be implemented base on data of source address and destination address and protocol number. This technology is suitable for the users such as small enterprise or department to set up a firewall without professional firewall production. An example is presented for demonstrating how to use the proposed technology as firewall in LAN.
2004, 33(1): 79-82.
Abstract:
In order to save energy in operation of pumps and ensure water supply with constant pressure and use for fire control, the principle of saving energy of pump in variable frequency variable speed drive system was analyzed. The system structure was designed with fuzzy controller, variable frequency variable speed drive system and three main pumps as core. Some debugging experiences were introduced. This system are obvious effect in saving energy and its idea for design, interface technique may be used for reference for technique reform in analogy water supply system..
In order to save energy in operation of pumps and ensure water supply with constant pressure and use for fire control, the principle of saving energy of pump in variable frequency variable speed drive system was analyzed. The system structure was designed with fuzzy controller, variable frequency variable speed drive system and three main pumps as core. Some debugging experiences were introduced. This system are obvious effect in saving energy and its idea for design, interface technique may be used for reference for technique reform in analogy water supply system..
2004, 33(1): 83-86.
Abstract:
It is very key to build the steady-state models of the industrial processes. Using the step signals of set-points changes in the course of optimizing control as the input signals to identifying the steady-state models is good identification technique and reduces the disturbed to the processes. In this paper, a new identification technique to obtain the steady-state models is presented for the bilinear industrial processes. which uses the step signals of set-points changes in the course of optimizing control as the input step signals to identifying the steady-state models of industrial processes described by bilinear models,is presented.The novel technique is simple,high identification accurate and less disturb to the industrial processes.Finally,the simulation results show that the new approach is very efficient and practical.
It is very key to build the steady-state models of the industrial processes. Using the step signals of set-points changes in the course of optimizing control as the input signals to identifying the steady-state models is good identification technique and reduces the disturbed to the processes. In this paper, a new identification technique to obtain the steady-state models is presented for the bilinear industrial processes. which uses the step signals of set-points changes in the course of optimizing control as the input step signals to identifying the steady-state models of industrial processes described by bilinear models,is presented.The novel technique is simple,high identification accurate and less disturb to the industrial processes.Finally,the simulation results show that the new approach is very efficient and practical.
2004, 33(1): 87-90.
Abstract:
The main difficult in the application of inverse system method to nonlinear control system is to determine the invertibility of the system. In order to investigate new method for analyzing invertibility of system, a rank criterion method is introduced to multivariable affine nonlinear system, by which the problem of invertibility is converted to determine the rank condition of the Jacobi matrix about output and its derivation of system. The rank criterion theorem such that affine nonlinear system be invertible is proved and an algorithm is obtained to determine invertibility of the system. Finally, the algorithm mentioned is verified by using an example. The effectiveness of rank criterion method is validated by comparison with differential geometry method and inverse system method.
The main difficult in the application of inverse system method to nonlinear control system is to determine the invertibility of the system. In order to investigate new method for analyzing invertibility of system, a rank criterion method is introduced to multivariable affine nonlinear system, by which the problem of invertibility is converted to determine the rank condition of the Jacobi matrix about output and its derivation of system. The rank criterion theorem such that affine nonlinear system be invertible is proved and an algorithm is obtained to determine invertibility of the system. Finally, the algorithm mentioned is verified by using an example. The effectiveness of rank criterion method is validated by comparison with differential geometry method and inverse system method.
2004, 33(1): 91-93.
Abstract:
This note presents the asymptotic behavior of the solution and stability of zero solution. Continuous dependence for initial value of the solutions and some translation algorithms are employed. There is no need to solve the equation. Using the sign of eigenvalues decide the asymptotic behavior of the solutions and the stability of zero solutions. These process is presented in the note explicitly.
This note presents the asymptotic behavior of the solution and stability of zero solution. Continuous dependence for initial value of the solutions and some translation algorithms are employed. There is no need to solve the equation. Using the sign of eigenvalues decide the asymptotic behavior of the solutions and the stability of zero solutions. These process is presented in the note explicitly.
2004, 33(1): 94-97.
Abstract:
Aim at the chart solution of two-dimensional linear programming questions, a property theorem about optimum solution of two-dimensional linear programming questions was given, and a alglebraic algorithm of two-dimensional linear programming questions was gained. Then generalize the algorithm, a new algorithm of linear programming questions can gain.The new algorithm's rapidity of convergence is faster than simplex method.
Aim at the chart solution of two-dimensional linear programming questions, a property theorem about optimum solution of two-dimensional linear programming questions was given, and a alglebraic algorithm of two-dimensional linear programming questions was gained. Then generalize the algorithm, a new algorithm of linear programming questions can gain.The new algorithm's rapidity of convergence is faster than simplex method.
2004, 33(1): 98-101.
Abstract:
The diameter and W-wide diameter of a graph measure delay time in a collateral and distribute net, in this paper, two kinds of W-wide diameter are studied.The first, the minimum number of 2-wide diameter of c(n,t) is studied, where graph c(n,t) is formed by adding t edges beyond a given n-cycal Cn, An upped and down bounds of h(n,t) are obtained. Fathermore, the guess accuracy number of h(n,t) is offerd At the same time, based on the analysis of the connectivity of the join graph,an upped bound of W-wide diameter of join graph is also obtained.
The diameter and W-wide diameter of a graph measure delay time in a collateral and distribute net, in this paper, two kinds of W-wide diameter are studied.The first, the minimum number of 2-wide diameter of c(n,t) is studied, where graph c(n,t) is formed by adding t edges beyond a given n-cycal Cn, An upped and down bounds of h(n,t) are obtained. Fathermore, the guess accuracy number of h(n,t) is offerd At the same time, based on the analysis of the connectivity of the join graph,an upped bound of W-wide diameter of join graph is also obtained.
2004, 33(1): 102-104.
Abstract:
In this paper, we first apply properties of irreducible matrices to study nonsingular H-matrices, then conditions for irreducible matrices to be nonsingular H-matrices are obtained. Applying the properties of non-zero elements chain matrices, conditions for them to be nonsingular H-matrices are also obtained. Results obtained do not require that matrices should be diagonal dominant, and are applicable to more matrices. These results are sufficient conditions of nonsingular H-matrices.
In this paper, we first apply properties of irreducible matrices to study nonsingular H-matrices, then conditions for irreducible matrices to be nonsingular H-matrices are obtained. Applying the properties of non-zero elements chain matrices, conditions for them to be nonsingular H-matrices are also obtained. Results obtained do not require that matrices should be diagonal dominant, and are applicable to more matrices. These results are sufficient conditions of nonsingular H-matrices.
2004, 33(1): 105-108.
Abstract:
The combination forecasting model based on Theil coefficient is a new kind of correlation one. In this paper its properties are studied. On basis of new concepts, such as superior combination forecasting and dominant forecasting method etc, the existence of noninferior combination forecasting, superior combination forecasting and redundant forecasting method are discussed. Finally redundant information is given in the determining theorem.
The combination forecasting model based on Theil coefficient is a new kind of correlation one. In this paper its properties are studied. On basis of new concepts, such as superior combination forecasting and dominant forecasting method etc, the existence of noninferior combination forecasting, superior combination forecasting and redundant forecasting method are discussed. Finally redundant information is given in the determining theorem.
2004, 33(1): 109-112.
Abstract:
In this paper, we analyzed the change in the value of rolling hedger's position and studied the basis of rolling hedge and basis risk. The rolling hedge model is estabilished based on the change in the value of rolling hedger's position and drove out the optimal hedging ratio and the minimum rolling basis risk. The dynamic adjustment was discussed in the paper on the basis of the sharing idea from actuarial science.
In this paper, we analyzed the change in the value of rolling hedger's position and studied the basis of rolling hedge and basis risk. The rolling hedge model is estabilished based on the change in the value of rolling hedger's position and drove out the optimal hedging ratio and the minimum rolling basis risk. The dynamic adjustment was discussed in the paper on the basis of the sharing idea from actuarial science.
