Mathematical analysis based on SAIQRS modelJi, Meng
doi: 10.1088/1742-6596/2898/1/012015pmid: N/A
Given the mobility of the population during epidemic transmission, the lethality of certain epidemics, and the fact that recovered individuals do not possess lifelong immunity, it is crucial to consider vaccination strategies to control epidemic spread. Vaccinating the population stimulates the production of antibodies, thereby reducing the disease transmission rate. Furthermore, individuals who have been infected with the same pathogen as the vaccine will generate a certain amount of antibodies, enhancing their resistance to future epidemics. In this paper, we develop a susceptible-antibodies-infected-isolated-removed-individuals (SAIQRS) model and derive the basic regeneration number R0 for the deterministic SAIQRS model. Next, the effect of the stochastic factor is taken into account by changing the transmission rate β1 → β1 + σB(t), which extends the model to a stochastic SVIQRS model and gives the conditions for the extinction and persistence of the disease. Finally, the theorems of the deterministic and stochastic models are verified by numerical simulations.
An adaptive event-based state estimation for coupled network with multiple fading measurement via partial nodes informationLin, Na; Chen, Dongyan; Zhao, Hairui
doi: 10.1088/1742-6596/2898/1/012008pmid: N/A
The state estimation (SE) problem is explored for coupled complex networks (CNs) subjected to general nonlinearity and coupled nonlinearity, as well as multiple fading measurement phenomena. The phenomenon of multiple fading measurement means that the measurement fading occurs in a random manner and each sensor has a separate fading probability in the information transmission. An adaptive event triggering mechanism (ETM) with a time-varying threshold is introduced for adjusting the frequency of information transmission. A two-stage partial-nodes-based (PNB) state estimator is designed using partial nodes’ information to estimate all true states. The PNB estimator gain matrix is parameterized by optimizing the trace of the upper bound matrix of the estimation error covariance (UBMEEC). The monotonicity of the trace of the UBMEEC with respect to the fading probability is clarified. Finally, a numerical example demonstrates the effectiveness of the proposed two-stage PNBSE method.
Effect of the septal dissection on the distribution of the residual oilLiu, Zhenyu; Tang, Yong; Yang, Yan; Yang, Zhao; Qu, Changyue
doi: 10.1088/1742-6596/2898/1/012007pmid: N/A
In order to explore the influence of the interlayer on residual oil, the LZ block in Songliao Basin was selected as the target, and the influence of macroscopic and interlayer location on the distribution of residual oil was studied through a streamlined method and numerical simulation. Among them, the distribution models of 9 types were established according to the position of the interlayer. The study shows that the interlayer greatly influences the enrichment degree of residual oil. Large interlayer plays an obvious role in layered water injection in the medium and low water content stages. In the high water content stage, the higher the position is, the more obvious the effect is. At the same reservoir thickness, the larger the interlayer size is, the stronger the retention effect of the 2PV interlayer on the remaining oil is. At different reservoir thickness, the greater the reservoir thickness is, the lower the integrated water content of the interlayer under 2PV is, and the higher the degree of extraction is, the weaker the residual oil retention effect is.
Light culling based on the limitations of human visual perceptionChen, Huiwen; Chen, Chunyi
doi: 10.1088/1742-6596/2898/1/012022pmid: N/A
We propose a method to remove virtual point light sources (VPLs) in the scene to reduce the time overhead of lighting calculation for shading points in non-visually sensitive areas. We quantify the degree of light stimulation to the human eye and use it to calculate the lighting contribution of VPLs in the scene to visually sensitive areas. Based on this contribution, we remove VPLs that contribute less to visually sensitive areas to improve rendering speed. Experimental results show that the FLIP evaluation index difference between this method and FLC [1] is only 0.00077, while the average number of frames is 1.15 times that of the method in FLC. This proves that this paper can effectively improve rendering speed.
Multi-agent cooperative encirclement based on improved MADDPG algorithmAi, Ling; Tang, Shaozhen; Yu, Jie
doi: 10.1088/1742-6596/2898/1/012033pmid: N/A
In this paper, we propose an improved Multi-agent Deep Deterministic Policy Gradient algorithm with a Priority Experience Replay mechanism (PER-MADDPG) to address the baseline algorithm’s high-dimensional state space challenges in multi-agent encirclement scenarios. The PER mechanism effectively mitigates the issue of non-stationary experience data distribution. By incorporating the Apollonian circle theory, we design an effective encirclement reward function that enables the multi-agent system to complete encirclement tasks in environments with static obstacles. Comparative simulation results show that the improved algorithm achieves faster reward value growth and higher average rewards.
Improved black-winged kite algorithm based on chaotic mapping and adversarial learningZhao, Mingjing; Su, Zhongji; Zhao, Chenyang; Hua, Zexi
doi: 10.1088/1742-6596/2898/1/012040pmid: N/A
Aiming at the shortcomings of the black-winged kite algorithm, such as poor uniformity of initial population distribution, subjective search and single direction, this paper proposes an improved black-winged kite algorithm based on chaotic mapping and adversarial learning. It then progresses to adopting Tent chaotic mapping to achieve a uniform initial solution distribution, introducing Beta random distribution and optimizing the nonlinear factor in the attack phase to make the search trend more in line with the demand, and introducing an adversarial learning mechanism to extend the search direction. Twelve benchmark test functions are tested and compared with five other algorithms, and the result shows that the algorithm significantly outperforms the other algorithms in terms of searching ability and optimization-finding accuracy.
Conformal prediction with censored data using Kaplan-Meier methodSun, Xiaolin; Wang, Yanhua
doi: 10.1088/1742-6596/2898/1/012030pmid: N/A
In this paper, we introduce a prediction algorithm founded on conformal prediction, tailored for constructing prediction intervals in the context of censored survival data. Conformal prediction frameworks distinguish themselves from other prediction paradigms by their non-empirical evaluation, reliance on user-defined confidence intervals for modeling errors, and widespread adoption across regression and classification methodologies, inclusive of survival analysis, in recent years. Herein, we present a novel application wherein the Kaplan-Meier method is employed to compute empirical quantiles of nonconformal scores, specifically tailored for censored schematic variables. This novel approach facilitates the generation of well-calibrated prediction intervals for survival times, augmenting any existing survival prediction algorithm. Validation of its efficacy and computational efficiency is performed on both the real-world dataset ‘SUPPORT’ and the synthetic dataset ‘RRNLNPH.’
A variable step and multi-constraint vehicle’s trajectory generation algorithm based on deep deterministic policy gradient networkHu, Xingzhi; Li, Mingjie; Zhao, Bendong; Wang, Zhiren; Lai, Jianqi
doi: 10.1088/1742-6596/2898/1/012042pmid: N/A
Aiming at the problem that the traditional vehicle’s trajectory generation method takes a long time and is difficult to calculate in real time, a variable step and multi-constraint trajectory generation algorithm based on a deep deterministic policy gradient (DDPG) network is proposed. Firstly, the dynamic model and constraint conditions of the vehicle are analyzed. On this basis, the reinforcement learning training model is constructed based on DDPG, and the state, action, and reward design of the training model are defined. At the same time, the variable step is introduced into the action model to improve the generated trajectory’s navigation precision. Finally, the proposed method is verified under different cases, and according to the results, the proposed method can realize quick generation of multi-constraint guidance trajectories.
Analysis of optimal planting scheme based on the North China regionLiang, Kun; Lei, Jiaying; Wu, Feiyang; Hu, Zhenning; Zhou, Yusong; Li, Dandan; Gong, Shiqin
doi: 10.1088/1742-6596/2898/1/012004pmid: N/A
This paper constructs a linear programming model for optimizing crop planting strategies in rural mountainous areas of North China by simulating data and predicting the planting scheme for each sample plot in the next six years. It sets decision variables as plot, year, season, and crop planting area, considering crop rotation and three-year legume planting restrictions. Applying an improved genetic algorithm for solving, which improves efficiency through multi-matrix chromosome coding and optimized crossover mutation operation, and deals with infeasible solutions through a rule-based repair mechanism to propose different optimal planting strategies. They are considering the fluctuating uncertainty of crop sales, yield, and planting cost based on the expected fluctuation range and reducing computational resources through orthogonal experimental design. A bi-objective robust optimization model was used to balance average profit maximization and profit minimization under different scenarios. This was solved by the improved NSGA-II algorithm. The final optimal planting scheme suggests increasing the planting proportion of vegetable crops and controlling the planting scale of crops with large cost fluctuations while guaranteeing grain crops like wheat and corn.