Avoidance control

We consider dynamical systems subject to control by two agents, one of whom desires that no trajectory of the system, emanating from outside a given set, intersects the set, no matter what the admissible actions of the other agent. Conditions are given whose satisfaction assures that a given control results in avoidance. Furthermore, these conditions are constructive in that they yield an avoidance feedback control. Some examples are presented.     

Black swans and white eagles: on mathematics and finance

It is often objected that we cannot use mathematical methods in finance because (1) finance is dominated by unpredictable unique events (the black swans), (2) there are qualitative effects that cannot be quantified, and (3) the laws themselves of finance keep on changing. In this paper we discuss these three objections, offering arguments to reject them. We begin by reviewing the development of the physical sciences, pointing out parallels that are relevant for our discussion. Modern science has abandoned the objective of describing reality and has adopted an operational point of view that regards physical laws as tools to connect observations. Modern science is no longer deterministic, but has accepted a fundamental uncertainty in physical laws both at micro and macroscopic levels. Unpredictable pivotal events exist in the physical sciences as well in finance but this does not lead us to question the use of mathematics in the physical sciences. On the contrary, using principles of safe design, we try to understand how to avoid and contain unpredictability. Financial markets are manmade artifacts with, as actors, a large number of interacting agents. If we so wish, we can reduce the level of uncertainty present in markets: But if we try to do so describing financial markets with simple mathematical laws, we find that these laws are not stable but change over time, eventually with sudden structural breaks. This makes the use of mathematical finance difficult but not impossible. We can forecast human decision-making processes, crucial in forecasting financial markets, at the statistical level in aggregate. From an operational point of view, we have the tools to understand and describe the behavior of large number of interacting agents. At the present stage of development of our science, we need to use the mathematics of adaptive systems, changing mathematical models in function of different market states. However, reductionism to a small number of basic laws remains a fundamental objective of financial economics as it is in the physical sciences.     

Transshipment games with identical newsvendors and cooperation costs

In a transshipment game, supply chain agents cooperate to transship surplus products. Although the game has been well studied in the OR literature, the fundamental question whether the agents can afford cooperation costs to set up and maintain the game in the first place has not been addressed thus far. This paper addresses this question for the cooperative transshipment games with identical agents having normally distributed independent demands. We provide characterization of equal allocations which are in the core of symmetric games, and prove that not all transshipment games are convex. In particular, we prove that though individual allocations grow with the coalition size, the growth diminishes according to two rules of diminishing individual allocations. These results are the basis for studying the games with cooperation costs. We model the cooperation costs by the cooperation network topology and the cooperation cost per network link. We consider two network topologies, the clique and the hub, and prove bounds for the cost per link that render coalitions stable. These bounds always limit coalition size for cliques. However, the opposite is shown for hubs, namely newsvendors can afford cooperation costs only if their coalition is sufficiently large.     

Can remembering history from predecessor promote cooperation in the next generation?

People usually think that helping the next generation to remember history can promote cooperation in dilemma games. We show that is not always the case when agents have memory. Agents play with each neighbor by game history and strategies (such as TFT and WSLS), and the next generation inherits good strategies from the predecessor. We analyze the system’s cooperation ratio by comparing the 2 sources of history at the beginning of each generation: (a) inherited from the predecessor; (b) randomly initialized with different cooperation ratio. We find that with unconditional imitation update rule, agents who remember history get lower cooperation ratio than those who randomly initialize the history; while with replicator rule, higher initial cooperation ratio promotes higher final cooperation. We also do additional experiments to investigate the R, ST, P reciprocity and strategies distribution of the systems.     

Avoidance Control on Time Scales

We consider dynamic systems on time scales under the control of two agents. One of the agents desires to keep the state of the system out of a given set regardless of the other agent’s actions. Leitmann’s avoidance conditions are proved to be valid for dynamic systems evolving on an arbitrary time scale.     

Cooperative dispatching – exploiting the flexibility of an FMS by means of incremental optimization

We present a new control scheme for releasing parts into a flexible manufacturing system (FMS) that is based on incremental optimization. Our objective is to exploit the available routing flexibility of parts in an enhanced manner by viewing part release as an assignment problem using system status information. In particular, we propose an “intelligent” part release mechanism with some look-ahead and optimization features in order to allow for optimization-based “cooperation” of workcenters. The cooperative dispatching concept is implemented in an object-oriented computer simulation model, and experiments with a varying degree of average routing flexibility are performed. The experimental results are used for a statistical analysis of the benefits of cooperative dispatching versus the common approach of standard dispatching. Finally, we investigate the robustness of the presented FMS control scheme in the case of random machine breakdowns.     

Bipartite flocking for multi-agent systems

This paper addresses the bipartite flock control problem where a multi-agent system splits into two clusters upon internal or external excitations. Using structurally balanced signed graph theory, LaSalle’s invariance principle and Barbalat’s Lemma, we prove that the proposed algorithm guarantees a bipartite flocking behavior. In each of the two disjoint clusters, all individuals move with the same direction. Meanwhile, every pair of agents in different clusters moves with opposite directions. Moreover, all agents in the two separated clusters approach a common velocity magnitude, and collision avoidance among all agents is ensured as well. Finally, the proposed bipartite flock control method is examined by numerical simulations. The bipartite flocking motion addressed by this paper has its references in both natural collective motions and human group behaviors such as predator–prey and panic escaping scenarios.     

Optimal Control of a Ship for Collision Avoidance Maneuvers

We consider a ship subject to kinematic, dynamic, and moment equations and steered via rudder under the assumptions that the rudder angle and rudder angle time rate are subject to upper and lower bounds. We formulate and solve four Chebyshev problems of optimal control, the optimization criterion being the maximization with respect to the state and control history of the minimum value with respect to time of the distance between two identical ships, one maneuvering and one moving in a predetermined way.Problems P1 and P2 deal with collision avoidance maneuvers without cooperation, while Problems P3 and P4 deal with collision avoidance maneuvers with cooperation. In Problems P1 and P3, the maneuvering ship must reach the final point with a given lateral distance, zero yaw angle, and zero yaw angle time rate. In Problems P2 and P4, the additional requirement of quasi-steady state is imposed at the final point.The above Chebyshev problems, transformed into Bolza problems via suitable transformations, are solved via the sequential gradient-restoration algorithm in conjunction with a new singularity avoiding transformation which accounts automatically for the bounds on rudder angle and rudder angle time rate.The optimal control histories involve multiple subarcs along which either the rudder angle is kept at one of the extreme positions or the rudder angle time rate is held at one of the extreme values. In problems where quasi-steady state is imposed at the final point, there is a higher number of subarcs than in problems where quasi-steady state is not imposed; the higher number of subarcs is due to the additional requirement that the lateral velocity and rudder angle vanish at the final point.     

Impact of Quantisation on Consistency of DMPC in Street Traffic with Dynamic Priority Rules

Distributed systems typically require communication to steer the overall system to a reference or target state and keep it there. We consider an intersection scenario as a spacial set with vehicles as multi-agents with individual initial and target conditions. Utilising a Distributed Model Predictive Control scheme (DMPC), in every time step each vehicle solves an finite horizon optimal control problem. The resulting optimal state trajectory is projected onto a quantisation of the spatial set and the quantised trajectories are broadcasted to the other vehicles to ensure collision avoidance. The quantisation is mainly motivated to reduce the necessary communication effort among the agents. Here, we introduce prediction coherence as difference of two predictions in two successive time steps. We numerically explore the idea to utilise prediction coherence to derive a lower bound for the communication requirements among the vehicles. (© 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim)     

The outbreak of cooperation

We study the ongoing collective action problem among intentional agents whose choices depend not only on the past but also on their expectations as to how their actions will affect those of others. In this model agents act on the basis of imperfect information. We show that under these conditions the onset of overall cooperation can take place in a sudden and unexpected way. Likewise, defection can appear out of nowhere in very large, previously cooperating groups. These outbreaks mark the end of long transient states in which defection or cooperation persists in groups that cannot sustain it indefinitely.     

Quadratic stabilization of a class of switched nonlinear systems via single Lyapunov function

This paper is concerned with the problem of globally quadratic stabilization for a class of switched cascade systems. The system under consideration is composed of two subsystems: a linear switched part and a nonlinear part, which are also switched systems. The feedback control law and the switching law are designed respectively when the first part is stabilized under some switching law and when both parts can be stabilized under some switching laws. We construct the single Lyapunov functions and design the switching laws based on the structure characteristics of the switched system. Also, the designed switching laws are of hysteresis switching form in order to avoid sliding models.     

Effect of local information within network layers on the evolution of cooperation in duplex public goods games

Traditional works of public goods game (PGG) are often studied in simplex networks where agents play games through the same type of social interactions. In order to promote cooperation against the defection in PGGs in simplex network environment, many mechanisms have been proposed from different perspectives, such as the volunteering mechanisms, and the punishment and reward approaches. However, due to diverse types of interactions between agents in reality, the study of PGG should also consider the characteristic of multiplexity of networks. Hence, we firstly model the public goods game in the duplex network (for simplification of analysis, the duplex network is considered), in which agents have two types of social interactions, and thus the network is modeled as two network layers. This type of PGG is naturally named as duplex public goods game (D-PGG), in which agents can select one of the network layers to allocate their limited resources. Then for the new game environment (D-PGG), we propose a novel perspective to promote cooperation: degrading the information integrity, i.e., agents get information just from one network layer (local information) rather than from the whole duplex network (global information) in the evolution process. Finally, through theoretical analyses and simulations, we find that if agents imitate based on the local information of the payoff in the evolution, cooperation can be generally promoted; and the extent of promotion depends on both the network structure and the similarity of the network layers.     

Individual choice and reputation distribution of cooperative behaviors among heterogeneous groups

Cooperation is vital for our society, but the temptation of cheating on cooperative partners undermines cooperation. The mechanism of reputation is raised to countervail this temptation and therefore promote cooperation. Reputation microcosmically records individual choices, while cooperation macrocosmically refers to the group or averaged cooperation level. Reputation should be preferred in order to investigate how individual choices evolve. In this work, we study the distribution of reputation to figure out how individuals make choices within cooperation and defection. We decompose reputation into its mean and standard deviation and inspect effects of their factors respectively. To achieve this goal, we construct a model where agents of three groups or classes play the prisoners’ dilemma game with neighbors on a square lattice. It indicates in outcomes that the distribution of reputation is distinct from that of cooperation and both the mean and standard deviation of reputation follow clear patterns. Some factors have negative quadratic effects on reputation's mean or standard deviation, and some have merely linear effects.     

COALITIONS IN INTERNATIONAL FISHERIES MANAGEMENT

Abstract We show that with symmetric agents, noncooperation is the only stable coalition structure in a fishery with more than two countries. In the case of asymmetric fishing nations, partial or full cooperation may be stable even if the number of countries exceeds two. These are important results for recent fisheries economics papers that have not allowed for coalition formation. As an example how of one can use the model, we study the problem of new entrants into Regional Fisheries Management Organizations (RFMOs). We show that depending on the economic structure of the fishery, new entrants may make cooperation more difficult or easier.     

Disturbance rejection in robust PdE-based MRAC laws for uncertain heterogeneous multiagent networks under boundary reference

Disturbance is a pervasive source of uncertainty in most applications. This paper presents model reference adaptive control (MRAC) laws for uncertain multiagent networks with a disturbance rejection capability. The algorithms proposed can also be viewed as the extension of the robust model reference adaptive control (MRAC) laws with disturbance rejection recently derived for systems described by parabolic and hyperbolic partial differential equations (PDEs) with spatially-varying parameters under distributed sensing and actuation to heterogeneous multiagent networks characterized by parameter uncertainty. The latter 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. The resulting PdE MRAC laws inherit the robust linear structure of their PDE counterparts.     

Dynamic linear programming games with risk-averse players

Motivated by situations in which independent agents wish to cooperate in some uncertain endeavor over time, we study dynamic linear programming games, which generalize classical linear production games to multi-period settings under uncertainty. We specifically consider that players may have risk-averse attitudes towards uncertainty, and model this risk aversion using coherent conditional risk measures. For this setting, we study the strong sequential core, a natural extension of the core to dynamic settings. We characterize the strong sequential core as the set of allocations that satisfy a particular finite set of inequalities that depend on an auxiliary optimization model, and then leverage this characterization to establish sufficient conditions for emptiness and non-emptiness. Qualitatively, whereas the strong sequential core is always non-empty when players are risk-neutral, our results indicate that cooperation in the presence of risk aversion is much more difficult. We illustrate this with an application to cooperative newsvendor games, where we find that cooperation is possible when it least benefits players, and may be impossible when it offers more benefit.     

Learning formation control for fractional‐order multiagent systems

In this paper, we use 2 iterative learning control schemes (P‐type and PI‐type) with an initial learning rule to achieve the formation control of linear fractional‐order multiagent systems. To realize the finite‐time consensus, we assume repeatable operation environments as well as a fixed but directed communication topology for the fractional‐order multiagent systems. Both P‐type and PI‐type update laws are applied to generate the control commands for each agent. It is strictly proved that all agents are driven to achieve an asymptotical consensus as the iteration number increases. Two examples are simulated to verify the effectiveness of the proposed algorithms.     

On the core of cost-revenue games: Minimum cost spanning tree games with revenues

In this paper, we analyze cost sharing problems arising from a general service by explicitly taking into account the generated revenues. To this cost-revenue sharing problem, we associate a cooperative game with transferable utility, called cost-revenue game. By considering cooperation among the agents using the general service, the value of a coalition is defined as the maximum net revenues that the coalition may obtain by means of cooperation. As a result, a coalition may profit from not allowing all its members to get the service that generates the revenues. We focus on the study of the core of cost-revenue games. Under the assumption that cooperation among the members of the grand coalition grants the use of the service under consideration to all its members, it is shown that a cost-revenue game has a nonempty core for any vector of revenues if, and only if, the dual game of the cost game has a large core. Using this result, we investigate minimum cost spanning tree games with revenues. We show that if every connection cost can take only two values (low or high cost), then, the corresponding minimum cost spanning tree game with revenues has a nonempty core. Furthermore, we provide an example of a minimum cost spanning tree game with revenues with an empty core where every connection cost can take only one of three values (low, medium, or high cost).     

A cooperative hyper-heuristic search framework

In this paper, we aim to investigate the role of cooperation between low level heuristics within a hyper-heuristic framework. Since different low level heuristics have different strengths and weaknesses, we believe that cooperation can allow the strengths of one low level heuristic to compensate for the weaknesses of another. We propose an agent-based cooperative hyper-heuristic framework composed of a population of heuristic agents and a cooperative hyper-heuristic agent. The heuristic agents perform a local search through the same solution space starting from the same or different initial solution, and using different low level heuristics. The heuristic agents cooperate synchronously or asynchronously through the cooperative hyper-heuristic agent by exchanging the solutions of the low level heuristics. The cooperative hyper-heuristic agent makes use of a pool of the solutions of the low level heuristics for the overall selection of the low level heuristics and the exchange of solutions. Computational experiments carried out on a set of permutation flow shop benchmark instances illustrated the superior performance of the cooperative hyper-heuristic framework over sequential hyper-heuristics. Also, the comparative study of synchronous and asynchronous cooperative hyper-heuristics showed that asynchronous cooperative hyper-heuristics outperformed the synchronous ones.     

Multiagent cooperation for solving global optimization problems: an extendible framework with example cooperation strategies

This paper proposes the use of multiagent cooperation for solving global optimization problems through the introduction of a new multiagent environment, MANGO. The strength of the environment lays in its flexible structure based on communicating software agents that attempt to solve a problem cooperatively. This structure allows the execution of a wide range of global optimization algorithms described as a set of interacting operations. At one extreme, MANGO welcomes an individual non-cooperating agent, which is basically the traditional way of solving a global optimization problem. At the other extreme, autonomous agents existing in the environment cooperate as they see fit during run time. We explain the development and communication tools provided in the environment as well as examples of agent realizations and cooperation scenarios. We also show how the multiagent structure is more effective than having a single nonlinear optimization algorithm with randomly selected initial points.