Table of Content

15 August 2023, Volume 40 Issue 4 Previous Issue    Next Issue

Select

Source Free Domain Adaptation Based on Feature Structure WANG Shipeng, SUN Jian, XU Zongben 2023, 40 (4):  511-522.  doi: 10.3969/j.issn.1005-3085.2023.04.001 Abstract ( 142 )   PDF (220KB) ( 154 )   Save The goal of domain adaptation is transferring knowledge learned from the labeled source domain to the unlabeled target domain, where the distributions of data from the source domain and target domain are different. Prior domain adaptation methods typically assume data from the source domain are available when learning to adapt to the target domain, which might be not possible due to privacy issues and data security. In order to transfer knowledge to the target domain without accessing data from the source domain, a new source-free domain adaptation method called FCDC is developed in this paper. SFDC splits data from the target domain into two parts according to their confidence and estimates pseudo labels with different strategies for the two parts. For data of high confidence, their pseudo labels are the predictions of the neural network. For data of low confidence, their pseudo labels are guided by data of high confidence, which is modeled as an optimization problem, and the solution of the problem gives the pseudo label of data of low confidence. To make the pseudo labels of data of low confidence more reliable, FCDC utilizes information maximum loss on these data to produce well-behavior clusters. Meanwhile, FCDC takes advantage of self-supervision loss on data of high confidence to make the features of these data more diverse and surround data of low confidence. Experimental results show that the proposed FCDC is an effective method for source-free domain adaptation. Related Articles | Metrics

Select

Modeling and Analysis of Operation Effectiveness Evaluation for TT{\&}C Equipment LIANG Jun, SONG Jianguo, CHEN Jiaming, LIU Jianye, LI Xiaochao, LEI Han 2023, 40 (4):  523-544.  doi: 10.3969/j.issn.1005-3085.2023.04.002 Abstract ( 75 )   PDF (1629KB) ( 240 )   Save In recent years, the number of spacecraft has rapidly increased, the types of telemetry track and command (TT{\&}C) equipment have become more diverse, and the variety of TT{\&}C requirements has become increasingly diverse, making TT{\&}C become a complex system engineering problem. At present, the evaluation of the daily operation effectiveness of TT{\&}C equipment is still in the stage of manual statistical data and one-sided calculation of a single index, which lacks a systematic and effective evaluation model and is difficult to comprehen-sively reflect the actual operation effectiveness of TT{\&}C equipment. In order to scientifically and accurately analyze the daily operation effectiveness for TT{\&}C equipment, the evaluation modeling is studied from different time scales of time period, day and month in this paper. Based on the evaluation indexes such as load rate, idle availability rate and idle fragment rate, the evaluation model of average operation effectiveness for TT{\&}C equipment in the geographical region is first constructed. Second, the evaluation model of daily operation effectiveness for the TT{\&}C equipment with the largest contribution of load duration in each region is established. Finally, a significance statistical test calculation formula for the operation effectiveness of TT{\&}C equipment is further constructed, which can effectively test whether there are significance in the same evaluation model between different TT{\&}C regions or different TT{\&}C equipment. In addition, the operation data of 21 TT{\&}C equipment in the west, east, south and north of the space TT{\&}C network are used for operation effectiveness evaluation and analysis in the simulation case, which provides scientific and effective evaluation means support for further improving the effectiveness of TT{\&}C equipment in space TT{\&}C system. Related Articles | Metrics

Select

Research on Optimal Reinsurance Strategy Based on Multi-dependent Insurance Business and Competition YANG Peng 2023, 40 (4):  545-558.  doi: 10.3969/j.issn.1005-3085.2023.04.003 Abstract ( 120 )   PDF (1196KB) ( 158 )   Save Under a competition between an insurance company and a reinsurance company, the formulation problem of the optimal reinsurance strategy is studied. There is an insurance company in the insurance market that operates $n$ kinds of dependent insurance businesses. For each insurance business, the insurance company reduces the claim risk by adopting proportional reinsurance. The insurance company and reinsurance company increase their wealth by investing in a common risk-free asset. Based on the relative performance, we quantify the competition between the insurance company and the reinsurance company, and derive the relative wealth process of the insurance company. Furthermore, we establish the stochastic optimization problem that the insurance company care about, i.e., find the optimal reinsurance strategy to maximize the expected terminal wealth while minimizing the variance of the terminal wealth. Using stochastic calculus and stochastic control theory, we establish the extended Hamilton-Jacobi-Bellman (HJB) equation and verification theorem that the value function satisfies, and then using the stochastic optimization theory, we obtain the explicit solution of the optimal time-consistent reinsurance strategy and the corresponding optimal value function, and theoretically discuss the insurance and economic significance of the optimal reinsurance strategy. Furthermore, for the two insurance business case, the dependence of claim amounts and claim distributions are discussed, respectively, and the corresponding optimal reinsurance strategy is given. Finally, we analyze the influence of insurance business dependence, competition degree and the number of insurance businesses on the optimal time-consistent reinsurance strategy through numerical experiments. The research results can effectively guide the reinsurance practice of insurance companies. Related Articles | Metrics

Select

Multi-channel Supply Chain Pricing Decisions Considering Heterogeneous Consumers WANG Rufeng, HOU Pengwen, ZHENG Wei 2023, 40 (4):  559-575.  doi: 10.3969/j.issn.1005-3085.2023.04.004 Abstract ( 63 )   PDF (478KB) ( 137 )   Save With the booming of e-commerce, more and more industries open multi-channels. A supply chain model is studied in which both the manufacturer and the retailer introduce the e-channel in addition to the traditional physical channel. This model incorporates heterogeneous consumers' lower valuation of the product sold through the e-channel and the discounted prices available in these channels. Using the traditional single-channel supply chain as a benchmark, Stackelberg games are applied to obtain the optimal equilibrium solutions of the benchmark and the proposed model. The results show that, compared with the benchmark, the wholesale price and the profits and the manufacturer's ability to control the market all will be improved; the retailer will reduce the retail price when the proportion of demand for the retailer's Internet channel is less and vice versa, but her profits will always be less than that of the benchmark. The manufacturer's profits will be better with respect to his proportion of demand of the Internet channel; numerical experiments show that the retailer's profits may increase or decrease with respect to her proportion of demand of the Internet channel, which is influenced by discount rate and the consumer acceptance of the Internet channel. Numerical experiments further show that we observe that the manufacturer's profit will decrease (increase) with respect to discount rate (the consumer acceptance of the Internet channel), but the retailer's profit will increase (decrease) with respect to discount rate (the consumer acceptance of the Internet channel). The research has a guiding significance for the multi-channel supply chain pricing considering the discounted prices online. Related Articles | Metrics

Select

Study on Joint Emission Reduction Decision-making of Low-carbon Supply Chain Considering Risk Aversion under Carbon Caps ZHANG Na, ZHANG Keyong 2023, 40 (4):  576-590.  doi: 10.3969/j.issn.1005-3085.2023.04.005 Abstract ( 62 )   PDF (3961KB) ( 101 )   Save The impact of Carbon Caps and the risk aversion characteristics of system members on system optimal decisions for a supply chain consisting of a single manufacturer and a single retailer was studied. Numerical examples are used for verification. The results show that the dominant firm is more profitable under a cap-and-trade policy, but can harmonize the profits of supply chain members by adjusting the price of carbon trading. The risk aversion characteristics of system members is not conducive to the joint carbon reduction of manufacturers and retailers, and it has a negative effect on the utility of system members. Retailers' sharing of abatement costs can weaken the negative effects of manufacturers' risk aversion characteristics on their own utility, but increases the damage to their own utility. Joint emission reductions are better under centralized decision-making, while supply chain benefits are better under decentralized decision-making. Related Articles | Metrics

Select

A Modified Spectral Conjugate Method ZHU Yixuan, SONG Enbin 2023, 40 (4):  591-604.  doi: 10.3969/j.issn.1005-3085.2023.04.006 Abstract ( 73 )   PDF (3852KB) ( 130 )   Save A modified spectral conjugate gradient method is proposed to solve unconstrained optimization problems. The search direction of the algorithm is the descent direction, and it has global convergence under the Wolfe-Powell line search. It is proved that the algorithm has a linear convergence rate under some suitable conditions. Some numerical experiments on some standard functions show that the algorithm is effective on the facts including number of iteration, number of function calculations and program running time, and has certain advantages compared with related algorithms. Finally, the algorithm is applied to image denoising problems. Some different noise effects are applied for the classical images Lena and Camera. The comparison with related algorithms in references shows that the algorithm has good image denoising effect about the fact of the PSNR. Related Articles | Metrics

Select

The Steady-state Queue Size for the $Geo^{X}/G/1$ Queue Controlled by Service Time Backlog LIU Renbin, TANG Yinghui 2023, 40 (4):  605-620.  doi: 10.3969/j.issn.1005-3085.2023.04.007 Abstract ( 77 )   PDF (257KB) ( 112 )   Save A discrete time batch arrival $Geo^{X}/G/1$ queue is considered, in which the startup of server is controlled by the sum of service times of all arrivals (called service time backlog). The customers enter the system in batch arrival. When the service time backlog of all arriving customers exceeds a given non-negative integer $D$, the server starts its service and lasts until a busy period ends (this policy is called the $D$ policy). The model can offer some theoretical basis for the practitioners of wireless sensor network. Firstly, in the preparation work, the queue size and service time backlog at the start of a busy period, and the busy and idle periods are discussed. Then, by the classifications of the customers who arrive during the idle and busy periods, and the method of probabilistic analysis, the steady-state queue sizes at a departure time and an arbitrary time $n^{+}$ are studied. As two special cases, the steady-state queue sizes for the $Geo^{X}/G/1$ and $D$-policy $Geo/G/1$ queueing systems are derived. Finally, a kind of wireless sensor node is modelled, and the minimum power consumption is numerically gotten. Related Articles | Metrics

Select

Stationary Analysis of Fluid Model Driven by Multi-server Working Vacation Queue with Optional Service LI Zikun, XU Xiuli 2023, 40 (4):  621-633.  doi: 10.3969/j.issn.1005-3085.2023.04.008 Abstract ( 73 )   PDF (313KB) ( 96 )   Save Fluid queue model has become a powerful tool for service system modeling and analysis. In the real-time service system, customers may have optional additional services in addition to the necessary basic services. On the other hand, the service system may operate at low speeds due to various factors such as performing other tasks, which reduce the service efficiency. Taking the above situation into consideration, a new fluid queue model is constructed based on a multi-server queuing system with optional services and working vacation as an external random environment. The steady-state queue length distribution of driving system is obtained by means of quasi birth-and-death process and matrix-geometric solution method, and the net input rate structure and steady state conditions of the fluid model are given. The matrix differential equation satisfied by the steady-state joint distribution of fluid model is derived by using probability analysis method and stochastic process theory. Several important function sequences are constructed according to the structural characteristics of the above matrix differential equation. With the Laplace transform (LT) and Laplace-Stieltjes transform (LST) methods, the steady-state distribution of fluid level in the buffer, mean fluid level and the probability of empty buffer are given. Finally, numerical analysis is presented to analyze the influence of system parameters on performance indices. The results show that the system performance can be optimized and the service efficiency can be improved by adjusting the model parameters in practical background. Related Articles | Metrics

Select

Dynamics of Stochastic Nicholson Model with Patch Structure and Multiple Delays LIU Rong, ZHANG Fengqin 2023, 40 (4):  634-646.  doi: 10.3969/j.issn.1005-3085.2023.04.009 Abstract ( 91 )   PDF (1194KB) ( 89 )   Save The stochastic Nicholson model with patch structure and multiple delays is considered. Firstly, the existence of a unique global positive solution is established by constructing a proper Lyapunov function. Next, by constructing suitable stochastic Lyapunov functions, and using Chebyshev inequality, Borel-Cantelli lemma and exponential martingale inequality, the properties of the solution, such as stochastic ultimate boundedness and non-positivity of sample Lyapunov exponents, etc., are studied. Then, under the condition that all time delays are equal, the sufficient conditions for the extinction of species in each patch are given by using the Burkholder-Davis-Gundy inequality and a strong law of large numbers. Finally, some numerical results are presented, which shows that the interaction between patches has the advantage of population survival, and the greater the time delay, the slower the extinction of species. The results generalize and improve the previous related results, such as removing the condition of the existence and uniqueness theorem of global positive solutions and narrowing the boundary of the sample Lyapunov exponents in related literature. Related Articles | Metrics

Select

Hopf Bifurcation Control Analysis of a Time-delay Fractional Computer Virus Model GAO Yanxin, SHI Jianping 2023, 40 (4):  647-660.  doi: 10.3969/j.issn.1005-3085.2023.04.010 Abstract ( 55 )   PDF (1085KB) ( 81 )   Save The control problem of Hopf bifurcation for a time-delayed fractional computer virus model is considered. A periodic pulse delay feedback controller is introduced into the first equation of the uncontrolled system. The semi-analytical and semi-numerical method is used to prove the equivalence between the controlled system and its linearized average system. From this theoretical basis, the critical value of Hopf bifurcation in the controlled system is derived by the linearization method and the fractional Laplace transform, and the range of the stability domain and the variation of the stability domain with the controller parameters are obtained. The results show that the Hopf bifurcation of the system can be effectively delayed by adjusting the gain parameter value. Finally, the correctness and feasibility of the theoretical analysis are verified by appropriate numerical simulation, and the influence of different control gain parameters on the system stability domain is compared. Related Articles | Metrics

Select

Stability and Hopf Bifurcation of a Diffusive Model with Prey Fear and Hunting Cooperation CHEN Qingwan, LIU Wenqing 2023, 40 (4):  661-671.  doi: 10.3969/j.issn.1005-3085.2023.04.011 Abstract ( 56 )   PDF (2297KB) ( 111 )   Save A predator-prey reaction-diffusion model with a fear factor and hunting cooperative is studied to explore the effects of two factors on the behavior of the predator-prey system dynamics. The local asymptotic stability of the positive equilibrium point is analyzed by the characteristic equation of the positive equilibrium point. The Hopf bifurcation point $\alpha^*$ is obtained by taking the hunting cooperation coefficient $\alpha$ as the bifurcation parameter without considering the fear factor. When the cooperation coefficient is greater than $\alpha^*$, the fear factor is used as the bifurcation parameter to obtain the Hopf bifurcation point $e^*$, the spatially homogeneous and non-homogeneous periodic solutions are generated near the Hopf bifurcation point. In addition, the conditions of Turing instability caused by diffusion are discussed. The results show that the system has a spatially inhomogeneous steady state when the ratio of predator and prey diffusivity is small. These conclusions can provide theoretical basis for how to maintain ecological balance. Finally, the conclusions are verified by numerical simulation. Related Articles | Metrics

Select

A Class of Nonlinear Functional Asymptotic Solution of Ecological System of Contagious Diseases XU Jianzhong, MO Jiaqi 2023, 40 (4):  672-680.  doi: 10.3969/j.issn.1005-3085.2023.04.012 Abstract ( 70 )   PDF (616KB) ( 99 )   Save A nonlinear population model of an infectious disease transmission is proposed, and the method of functional homotopic mapping is used to explore the law of the infectious disease transmission. A pair of functional homotopic mappings is constructed and the corresponding linear system of the model is discussed. Under some circumstances, the zero point is a stable node, that is, the original epidemic transmission system is a stable node at zero point. Therefore, when the time variable $t$ of the system model tends to infinity, it tends to zero solution. Better measures can be taken to control the epidemic disease. Finally, an example illustrates the correctness of the method used. The functional homotopic mapping method can take corresponding measures to control it. The expression of the solution obtained can be further analyzed. It is therefore possible to continue further discussions of the various properties of other related physical quantities. Related Articles | Metrics