PdE-based model reference adaptive control of uncertain heterogeneous multiagent networks

Robust globally stable model reference adaptive control (MRAC) laws recently derived for systems described by parabolic and hyperbolic partial differential equations (PDEs) with spatially-varying coefficients under distributed sensing and actuation are extended to heterogeneous multiagent networks characterized by parameter uncertainty. The extension is carried out using partial difference equations (PdEs) on graphs that preserve parabolic- and hyperbolic-like cumulative network behavior. Unlike in the PDE case, only boundary input is specified for the reference model. The algorithms proposed directly incorporate this boundary reference input into the reference PdE to generate the distributed admissible reference evolution profile followed by the agents. The agent evolution thus depends only on the interaction with the adjacent agents, making the system fully decentralized. Numerical examples are presented as well, including the case of the switched topology associated with a sudden loss of an agent. The resulting PdE MRAC laws inherit the robust linear structure of their PDE counterparts.     

A preferential attachment model with Poisson growth for scale-free networks

We propose a scale-free network model with a tunable power-law exponent. The Poisson growth model, as we call it, is an offshoot of the celebrated model of Barabási and Albert where a network is generated iteratively from a small seed network; at each step a node is added together with a number of incident edges preferentially attached to nodes already in the network. A key feature of our model is that the number of edges added at each step is a random variable with Poisson distribution, and, unlike the Barabási–Albert model where this quantity is fixed, it can generate any network. Our model is motivated by an application in Bayesian inference implemented as Markov chain Monte Carlo to estimate a network; for this purpose, we also give a formula for the probability of a network under our model.     

Learning agents for the multi-mode project scheduling problem

Intelligent optimization refers to the promising technique of integrating learning mechanisms into (meta-)heuristic search. In this paper, we use multi-agent reinforcement learning for building high-quality solutions for the multi-mode resource-constrained project scheduling problem (MRCPSP). We use a network of distributed reinforcement learning agents that cooperate to jointly learn a well-performing constructive heuristic. Each agent, being responsible for one activity, uses two simple learning devices, called learning automata, that learn to select a successor activity order and a mode, respectively. By coupling the reward signals for both learning tasks, we can clearly show the advantage of using reinforcement learning in search. We present some comparative results, to show that our method can compete with the best performing algorithms for the MRCPSP, yet using only simple learning schemes without the burden of complex fine-tuning.     

RBF Neural Network Based Prediction on Blade Surface Pressure Fields in Compressors北大核心CSCD

The airflow characteristics of the internal flow path of an aero-engine compressor are complex, and the vortex flow field around the blade is characterized by high pressure, high speed, rotation, and unsteadiness. Therefore, there is an urgent need to calculate and predict the aerodynamic characteristics of the complex flow field around the compressor blade efficiently and accurately. The computational fluid dynamics (CFD) method was used to generate the aerodynamic load distribution on the blade surface under different operating conditions for the study of the complex flow fields around aero-engine blades. The radial based function (RBF) neural network was applied to establish the pressure surface aerodynamic load prediction model, and the neural network modeling method was combined with the flow field calculation. The neural network method can learn and train the CFD-based data set to properly compensate the errors from the CFD, which provides a reference for the effective prediction of the complex flow fields around aero-engine compressor blades. © 2023 Editorial Office of Applied Mathematics and Mechanics. All rights reserved.     

Efficient structure of noisy communication networks

In the canonical network model, the connections model, only three specific network structures are generically efficient: complete, empty, and star networks. This renders many plausible network structures inefficient. We show that requiring robustness with respect to stochastic information transmission failures rehabilitates incomplete, redundant network structures. Specifically, we show that star and complete networks are not generally robust to transmission failures, that circular and quasi-circular networks are efficient at intermediate costs in four-player networks, and that if either of them is efficient, then at least one of them is pairwise stable even without reallocation. Thus, incomplete, redundant networks are efficient and stable at intermediate costs.     

A damped bipedal inverted pendulum for human–structure interaction analysis

The bipedal inverted pendulum with damping has been adopted to simulate human–structure interaction recently. However, the lack of analysis and verification has provided motivation for further investigation. Leg damping and energy compensation strategy are required for the bipedal inverted pendulum to regulate gait patterns on vibrating structures. In this paper, the Hunt–Crossley model is adopted to get zeros contact force at touch down, while energy compensation is achieved by adjusting the stiffness and rest length of the legs. The damped bipedal inverted pendulum can achieve stable periodic gait with a lower energy input and flatter attack angle so that more gaits are available, compared to the template, referred to as spring-load inverted pendulum. The measured and simulated vertical ground reaction force-time histories are in good agreement. In addition, the dynamic load factors are also within a reasonable range. Parametric analysis shows that the damped bipedal inverted pendulum can achieve stable gaits of 1.6 to 2.4 Hz with a reasonable first harmonic dynamic load factor, which covers the normal walking step frequency. The proposed model in this paper can be applied to human–structure interaction analysis.     

Boosting cultural algorithms with a heterogeneous layered social fabric influence function

In this paper we investigate the emergence and power of a complex social system based upon principles of cultural evolution. Cultural Algorithms employ a basic set of knowledge sources, each related to knowledge observed in various social species. Here we extend the influence and integration function in Cultural Algorithms by adding a mechanism by which knowledge sources can spread their influence throughout a population in the presence of a heterogeneous layered social network. The interaction (overlapping) of the knowledge sources, represented as bounding boxes on the landscape, at the right level projects how efficient the cooperation is between the agents in the resultant ??Social Network??. The inter-related structures that emerge with this approach are critical to the effective functioning of the approach. We view these structures as constituting a ??normal form?? for Cultures within these real-valued optimization landscapes. Our goal will be to identify the minimum social structure needed to solve problems of certain complexities. If this can be accomplished, it means that there will be a correspondence between the social structure and the problem environment in which it emerged. An escalating sequence of complex benchmark problems to our system will be presented. We conclude by suggesting the emergent features are what give cultural systems their power to learn and adapt.     

Learning dynamics in second order networks

A model of a second order neural network incorporating a weight adjusting learning component and time delays in processing and transmission is formulated. Delay-independent sufficient conditions are derived for the existence of an asymptotically and exponentially stable equilibrium state. The learning dynamics is modelled with an unsupervised Hebbian-type learning algorithm together with a forgetting term. If there is nothing for the network to learn or if there are no second order synaptic interactions, then our analysis will correspond to one of the standard model of neural networks.     

Dynamical Properties of a Perceptron Learning Process: Structural Stability Under Numerics and?Shadowing

In this paper two aspects of numerical dynamics are used for an artificial neural network (ANN) analysis. It is shown that topological conjugacy of gradient dynamical systems and both the shadowing and inverse shadowing properties have nontrivial implications in the analysis of a perceptron learning process. The main result is that, generically, any such process is stable under numerics and robust. Implementation aspects are discussed as well. The analysis is based on the theorem concerning global topological conjugacy of cascades generated by a gradient flow on a compact manifold without a boundary.     

Equilibrium selection and altruistic behavior in noncooperative social networks

Given their importance in determining the outcome of many economic interactions, different models have been proposed to determine how social networks form and which structures are stable. In Bala and Goyal (Econometrica 68, 1181–1229, 2000), the one-sided link formation model has been considered, which is based on a noncooperative game of network formation. They found that the empty networks, the wheel in the one-way flow of benefits case and the center-sponsored star in the two-way flow case play a fundamental role since they are strict Nash equilibria of the corresponding games for a certain class of payoff functions. In this paper, we first prove that all these network structures are in weakly dominated strategies whenever there are no strict Nash equilibria. Then, we exhibit a more accurate selection device between these network architectures by considering “altruistic behavior” refinements. Such refinements that we investigate here in the framework of finite strategy sets games have been introduced by the authors in previous papers.     

Fixed-bed drying simulation of agricultural products using a new backward finite difference scheme

This work is concerned with the numerical simulation of fixed-bed corn drying using MSU (Michigan State University) drying model. The classical numerical procedure for MSU model relies on an explicit method of finite differences which requires certain stability conditions between the step sizes of the time and space variables. The objective of the present paper is to establish a stable implicit method based on backward finite differences, in both time and space variables, which takes into account some specific empirical aspects of the problem. Computational results illustrate the efficiency and the flexibility of method.     

Network capacity management under competition

We consider capacity management games between airlines who transport passengers over a joint airline network. Passengers are likely to purchase alternative tickets of the same class from competing airlines if they do not get tickets from their preferred airlines. We propose a Nash and a generalized Nash game model to address the competitive network revenue management problem. These two models are based on well-known deterministic linear programming and probabilistic nonlinear programming approximations for the non-competitive network capacity management problem. We prove the existence of a Nash equilibrium for both games and investigate the uniqueness of a Nash equilibrium for the Nash game. We provide some further uniqueness and comparative statics analysis when the network is reduced to a single-leg flight structure with two products. The comparative statics analysis reveals some useful insights on how Nash equilibrium booking limits change monotonically in the prices of products. Our numerical results indicate that airlines can generate higher and more stable revenues from a booking scheme that is based on the combination of the partitioned booking-limit policy and the generalized Nash game model. The results also show that this booking scheme is robust irrespective of which booking scheme the competitor takes.     

A two step algorithm for solving a large scale semi-definite logit model

This paper proposes a two step algorithm for solving a large scale semi-definite logit model, which is appreciated as a powerful model in failure discriminant analysis. This problem has been successfully solved by a cutting plane (outer approximation) algorithm. However, it requires much more computation time than the corresponding linear logit model. A two step algorithm to be proposed in this paper is intended to reduce the amount of computation time by eliminating a certain portion of the data based on the information obtained by solving an associated linear logit model. It will be shown that this algorithm can generate a solution with almost the same quality as the solution obtained by solving the original large scale semi-definite model within a fraction of computation time.     

Production planning and warehouse management in supply networks with inter-facility mold transfers

This paper, considers with the problem of production capacity and warehouse management in a supply network in which inter-plant mold transfers are enabled. The supply network has a limited number of very expensive molds which can be transferred from a plant to another making it possible for each plant to produce the entire product gamut. It is assumed that warehouses in this supply network can be activated and deactivated as required, and that material transfers from a warehouse to another are also possible. The objective is to develop a capacity and warehouse management plan that satisfies the expected market demands with the lowest possible cost. A mixed integer programming model for the problem is suggested and its properties are discussed. A linear programming-based heuristic that combines Lagrangian relaxation and linear programming duality to generate lower and upper bounds for the problem is proposed. Finally, based on a designed experiment the performance of the heuristic on a set of generated test problems is reported and discussed.     

Applications of queuing theory to some problems in neuronal circuitry*

This paper discusses mathematical models for neurons and neuron-pair networks. Models are developed in which the parameters are related to basic physiological properties. Single-neuron models are treated first. The membrane-potential decay is modeled as a linear function, making it analogous to the virtual waiting time in a queue. Both spatial and temporal summation are incorporated into the model. Networks consisting of two neurons are then analyzed. It is shown that even though each neuron generates a renewal process, the interaction of the spike trains produces a nonrenewal process. A feedback inhibitory network is found to generate a bursting pattern of spikes. Expressions for the interspike-interval density function and the serial correlogram are derived based on the points of regenerating in the process, and verified by computer simulation. Finally, the feedback neuron-pair model is applied to spike-train data from the hippocampus of a rabbit.     

Network DEA pitfalls: Divisional efficiency and frontier projection under general network structures

Data envelopment analysis (DEA) is a method for measuring the efficiency of peer decision making units (DMUs). Recently network DEA models been developed to examine the efficiency of DMUs with internal structures. The internal network structures range from a simple two-stage process to a complex system where multiple divisions are linked together with intermediate measures. In general, there are two types of network DEA models. One is developed under the standard multiplier DEA models based upon the DEA ratio efficiency, and the other under the envelopment DEA models based upon production possibility sets. While the multiplier and envelopment DEA models are dual models and equivalent under the standard DEA, such is not necessarily true for the two types of network DEA models. Pitfalls in network DEA are discussed with respect to the determination of divisional efficiency, frontier type, and projections. We point out that the envelopment-based network DEA model should be used for determining the frontier projection for inefficient DMUs while the multiplier-based network DEA model should be used for determining the divisional efficiency. Finally, we demonstrate that under general network structures, the multiplier and envelopment network DEA models are two different approaches. The divisional efficiency obtained from the multiplier network DEA model can be infeasible in the envelopment network DEA model. This indicates that these two types of network DEA models use different concepts of efficiency. We further demonstrate that the envelopment model’s divisional efficiency may actually be the overall efficiency.     

An interactive graphical method for community detection in network data

The detection of community structures within network data is a type of graph analysis with increasing interest across a broad range of disciplines. In a network, communities represent clusters of nodes that exhibit strong intra-connections or relationships among nodes in the cluster. Current methodology for community detection often involves an algorithmic approach, and commonly partitions a graph into node clusters in an iterative manner before some stopping criterion is met. Other statistical approaches for community detection often require model choices and prior selection in Bayesian analyses, which are difficult without some amount of data inspection and pre-processing. Because communities are often fuzzily-defined human concepts, an alternative approach is to leverage human vision to identify communities. The work presents a tool for community detection in form of a web application, called gravicom, which facilitates the detection of community structures through visualization and direct user interaction. In the process of detecting communities, the gravicom application can serve as a standalone tool or as a step to potentially initialize (and/or post-process) another community detection algorithm. In this paper we discuss the design of gravicom and demonstrate its use for community detection with several network data sets. An “Appendix” describes details in the technical formulation of this web application built on the R package Shiny and the JavaScript library D3.     

我国上市公司被特别处理的组合预测模型研究

针对组合预测比单项预测具有更高的预测精度，本提出了一种基于模糊神经网络的上市公司被ST的非线性组合建模与预测新方法，并给出了相应的混合学习算法。通过与多元线性回归模型、Fisher模型和Logistc回归模型的预测结果对比表明，该方法具有预测精度高，学习与泛化能力强，适应性广的优点。     

Optimal Design of Hole Reinforcements for Composite Structures

In many technical applications like aerospace and automotive structures, holes in thin-walled composite components are necessary for some reason. It easily happens that the presence of a hole results in a detrimental stress concentration in the vicinity of the hole with a strength degradation and premature failure of the structure, respectively. In order to avoid the aforementioned overloading and to achieve a sufficient strength, in practice, a local reinforcement is employed. In the present study, reinforcements by elliptic doublers,as well as doublers adapted to reinforcement requirements in a layerwise manner, are considered. The increasing demands of a low weight and high strength for modern structures lead to the problem of an optimal reinforcement design. For this purpose, an appropriate optimization model is set up, a structural model is developed to describe the mechanical behavior (displacements, stresses, etc.) of such structures, and the techniques of mathematical structural optimization are used to find an optimal design in a systematic manner. In this study, the finite-element method is applied to the structural analysis. Eventually, an appropriate mathematical optimization algorithm is used to approach the desired design optimum in an iterative way. The implemented procedure works with a good reliability and efficiency and yields optimal reinforcement designs which are very useful for direct engineering applications.     

Algorithms for a multi-level network optimization problem

In this paper we work on a multi-level network optimization problem that integrates into the same model important aspects of: (i) discrete facility location, (ii) topological network design, and (iii) network dimensioning. Potential applications for the model are discussed, stressing its growing importance. The multi-level network optimization problem treated is defined and a mathematical programming formulation is presented. We make use of a branch-and-bound algorithm based on Lagrangean relaxation lower bounds to introduce some new powerful auxiliary algorithms to exactly solve the problem. We conduct a set of computational experiments that indicate the quality of the proposed approach.