The programming language GCLA — A definitional approach to logic programming

We present a logic programming language, GCLA^*** (Generalized horn Clause LAnguage), that is based on a generalization of Prolog. This generalization is unusual in that it takes a quite different view of the meaning of a logic program—a “definitional” view rather than the traditional logical view. GCLA has a number of noteworthy properties, for instance hypothetical and non-monotonic reasoning. This makes implementation of reasoning in knowledge-based systems more direct in GCLA than in Prolog. GCLA is also general enough to incorporate functional programming as a special case. GCLA and its syntax and semantics are described. The use of various language constructs are illustrated with several examples.     

TeMAS–a multi-agent system for temporally rich domains

In this paper, we present the model and simulator of a multi-agent system (MAS) for temporally rich domains. The theoretical foundations of the model include a knowledge representation scheme based on an original modification of Petri nets, called Petri nets with time tokens (PNTTs), as well as temporal reasoning based on the extension of Allen's temporal logic. The proposed MAS, called TeMAS, has a hierarchical structure, consisting of different levels, where each level contains clusters of agents. A paradigm of hierarchically organized blackboards is used for the communication among agents, clusters, as well as levels. We describe an object-oriented implementation of a program simulator of TeMAS and give an example of the use of the simulator for interpretation of events in a dynamic scene. Slobodan Ribarić received the B.S. degree in electronics, the M.S. degree in automatics, and the Ph.D. degree in electrical engineering from the Faculty of Electrical Engineering, Ljubljana, Slovenia, in 1974, 1976, and 1982, respectively. He is currently a Full Professor at the Department of Electronics, Microelectronics, Computer and Intelligent Systems, Faculty of Electrical Engineering and Computing, University of Zagreb, Croatia. His research interests include pattern recognition, artificial intelligence, biometrics, computer architecture and robot vision. He has published more than 150 papers on these topics and authored four books (Microprocessor Architecture, The Fifth Computer Generation Architecture, Advanced Microprocessor Architectures, CISC and RISC Computer Architecture) and co-authored one book (An Introduction to Pattern Recognition). Dr. Ribarić is a Member of the IEEE, ISAI and IAPR. Tomislav Hrkać received the B.S. degree in computer science from the Faculty of Electrical Engineering and Computing at the University of Zagreb, Croatia, in 1999. Since October 2000, he has been a Researcher with the Department of Electronics, Microelectronics, Computer and Intelligent Systems at the same faculty. He received the M.S. degree in 2004. As a co-author, he published several papers in international conference proceedings and a paper in a reviewed scientific journal. He is a Student Member of IEEE.     

Logic-based modeling of information transfer in cyber–physical multi-agent systems

In modeling multi-agent systems, the structure of their communication is typically one of the most important aspects, especially for systems that strive toward self-organization or collaborative adaptation. Traditionally, such structures have often been described using logic-based approaches as they provide a formal foundation for many verification methods. However, these formalisms are typically not well suited to reflect the stochastic nature of communication in the cyber–physical setting. In particular, their level of abstraction is either too high to provide sufficient accuracy or too low to be practicable in more complex models. Therefore, we propose an extension of the logic-based modeling language SALMA, which we have introduced recently, that provides adequate high-level constructs for communication and data propagation, explicitly taking into account stochastic delays and errors. In combination with SALMA’s tool support for simulation and statistical model checking, this creates a pragmatic approach for verification and validation of cyber–physical multi-agent systems.     

Distributed cooperative H∞ optimal tracking control of MIMO nonlinear multi-agent systems in strict-feedback form via adaptive dynamic programming

The design of distributed cooperative H_∞ optimal controllers for multi-agent systems is a major challenge when the agents’ models are uncertain multi-input and multi-output nonlinear systems in strict-feedback form in the presence of external disturbances. In this paper, first, the distributed cooperative H_∞ optimal tracking problem is transformed into controlling the cooperative tracking error dynamics in affine form. Second, control schemes and online algorithms are proposed via adaptive dynamic programming (ADP) and the theory of zero-sum differential graphical games. The schemes use only one neural network (NN) for each agent instead of three from ADP to reduce computational complexity as well as avoid choosing initial NN weights for stabilising controllers. It is shown that despite not using knowledge of cooperative internal dynamics, the proposed algorithms not only approximate values to Nash equilibrium but also guarantee all signals, such as the NN weight approximation errors and the cooperative tracking errors in the closed-loop system, to be uniformly ultimately bounded. Finally, the effectiveness of the proposed method is shown by simulation results of an application to wheeled mobile multi-robot systems.     

Leader–follower synchronization of networked multi-agent systems via hybrid protocols

This paper studies the leader–follower synchronization problem of complex-valued networked multi-agent systems with time-delay. A new hybrid protocol including a continuous-time protocol which is based on the interaction topology of follower agents and a pinning delayed impulsive control protocol is proposed. By employing the Lyapunov functional method in complex domains and the mathematical analysis technique, several delay-dependent leader–follower synchronization criteria are established that take into account various sizes of delays. Particularly, our result shows that leader–follower synchronization of delayed complex-valued networked multi-agent systems can be achieved even if the proposed hybrid protocol is being subject to relatively large impulse delays. A numerical example is provided to illustrate the effectiveness of the theoretical results.     

Min–max control using parametric approximate dynamic programming

This study presents a computationally efficient approximate dynamic programming approach to control uncertain linear systems based on a min–max control formulation. The optimal cost-to-go function, which prescribes an optimal control policy, is estimated using piecewise parametric quadratic approximation. The approach requires simulation or operational data only at the bounds of additive disturbances or polyhedral uncertain parameters. This strategy significantly reduces the computational burden associated with dynamic programming and is not limited to a particular form of performance criterion as in previous approaches.     

Leader–follower fixed-time consensus of multi-agent systems with high-order integrator dynamics

The leader–follower fixed-time consensus of high-order multi-agent systems with external disturbances is investigated in this paper. A novel sliding manifold is designed to ensure that the tracking errors converge to zero in a fixed-time during the sliding motion. Then, a distributed control law is designed based on Lyapunov technique to drive the system states to the sliding manifold in finite-time independent of initial conditions. Finally, the efficiency of the proposed method is illustrated by numerical simulations.     

A distributed model predictive control scheme for leader–follower multi-agent systems

In this paper, we present a novel receding horizon control scheme for solving the formation problem of leader–follower configurations. The algorithm is based on set-theoretic ideas and is tuned for agents described by linear time-invariant (LTI) systems subject to input and state constraints. The novelty of the proposed framework relies on the capability to jointly use sequences of one-step controllable sets and polyhedral piecewise state-space partitions in order to online apply the ‘better’ control action in a distributed receding horizon fashion. Moreover, we prove that the design of both robust positively invariant sets and one-step-ahead controllable regions is achieved in a distributed sense. Simulations and numerical comparisons with respect to centralised and local-based strategies are finally performed on a group of mobile robots to demonstrate the effectiveness of the proposed control strategy.     

Leader–follower consensus of linear multi-agent systems with unknown external disturbances

This paper addresses the leader–follower consensus tracking problem for multi-agent systems with identical general linear dynamics and unknown external disturbances. First, a distributed extended state observer is proposed, where both the local states and disturbance of each agent are estimated simultaneously by using the relative output information between neighbors. Then a consensus algorithm is proposed for each agent based on the relative estimated states between neighbors and its own disturbance estimate. It is shown that, with the proposed observer-based consensus algorithm, the leader–follower consensus problem can be solved. Finally, we present a simulation example to demonstrate the effectiveness of the proposed algorithm.     

Dual–primal proximal point algorithms for extended convex programming

This paper presents a class of dual–primal proximal point algorithms (PPAs) for extended convex programming with linear constraints. By choosing appropriate proximal regularization matrices, the application of the general PPA to the equivalent variational inequality of the extended convex programming with linear constraints can result in easy proximal subproblems. In theory, the sequence generated by the general PPA may fail to converge since the proximal regularization matrix is asymmetric sometimes. So we construct descent directions derived from the solution obtained by the general PPA. Different step lengths and descent directions are chosen with the negligible additional computational load. The global convergence of the new algorithms is proved easily based on the fact that the sequences generated are Fejér monotone. Furthermore, we provide a simple proof for the O(1/t) convergence rate of these algorithms.     

Event-based privacy-preserving security consensus of multi-agent systems with encryption–decryption mechanism

The article concentrates on exploring the issue of privacy-preserving sliding mode consensus of multi-agent systems (MASs) with disturbance. An encryption and decryption algorithm has been proposed to address data security and privacy issues during data transmission. To optimize network resource allocation, a dynamic event-triggering mechanism has been introduced, which reduces the number of encrypted data while saving the computation cost. The consensus performance based on the sliding mode control strategy is achieved when the reachability of the slide-mode surface is guaranteed, and then the slide-mode controller is developed. Finally, an empirical demonstration through a numerical example validates the efficacy of the proposed strategy.     

A particle swarm–BFGS algorithm for nonlinear programming problems

This article proposes a hybrid optimization algorithm based on a modified BFGS and particle swarm optimization to solve medium scale nonlinear programs. The hybrid algorithm integrates the modified BFGS into particle swarm optimization to solve augmented Lagrangian penalty function. In doing so, the algorithm launches into a global search over the solution space while keeping a detailed exploration into the neighborhoods. To shed light on the merit of the algorithm, we provide a test bed consisting of 30 test problems to compare our algorithm against two of its variations along with two state-of-the-art nonlinear optimization algorithms. The numerical experiments illustrate that the proposed algorithm makes an effective use of hybrid framework when dealing with nonlinear equality constraints although its convergence cannot be guaranteed.     

The thermal conductivity of alumina–water nanofluids from the viewpoint of micromechanics

The thermal conductivity of alumina–water nanofluids is analysed from the viewpoint of micromechanics. Wiener and Hashin–Shtrikman bounds are recalled, and it is shown how these model-independent bounds can be used to check experimental results and literature data. Further, it is shown that in the infinite-phase-contrast case (i.e. for superconducting particles), a useful approximation is obtained for the lower Hashin–Shtrikman bound (= Maxwell model = Hamilton–Crosser model for spherical particles). It is also shown that the second-order approximation of this infinite-phase-contrast relation is identical to the Choy–Torquato cluster expansion for impenetrable spheres with a “well-separated” distribution, whose dilute limit approximation is identical to that of other popular models, such as the self-consistent and differential scheme approximation. It is found that a considerable amount of published data violates the Hashin–Shtrikman lower bound, i.e. lies below the Maxwell model; the possible reasons are discussed. On the other hand, it is shown that the strong temperature dependence reported for the thermal conductivity of nanofluids cannot be explained within the framework of micromechanics and, if confirmed by future measurement where macroscopic convection can safely be excluded, indeed requires other mechanisms of heat transfer, such as forced nano- or micro-convection caused by Brownian motion of nanoparticles.     

Coordination of multi-agent Euler–Lagrange systems via energy-shaping: Networking improves robustness

In this paper, the robust coordination of multi-agent systems via energy-shaping is studied. The agents are nonidentical, Euler–Lagrange systems with uncertain parameters which are regulated (with and without exchange of information between the agents) by the classical energy-based controller where the potential energy function is shaped such that, if the parameters are known, all agents converge globally to the same desired constant equilibrium. Under parameter uncertainty, the globally asymptotically stable (GAS) equilibrium point is shifted away from its desired value and this paper shows that adding information exchange between the agents to the decentralized control policy improves the steady-state performance. More precisely, it proves that if the undirected communication graph is connected, the equilibrium of the networked controller is always closer (in a suitable metric) to the desired one than that of the decentralized controller. The result holds for all interconnection gains if the potential energy functions are quadratic, else, it is true for sufficiently large gains. An additional advantage of networking is that the asymptotic stabilization objective can be achieved by using lower gains into the loop. Some experimental results (using two nonlinear manipulators) given support to the main results of the paper.     

The zeros of complex polynomials,matrix inequalities and non-linear programming†

The complex zeros of a general complex polynomial are localized by constructing the intersection of areas in the complex plane defined by various inequality bounds on the eigenvalues of the companion matrix and also, possibly, by other inequalities on the zeros of polynomials. This localization then provides an efficient starting point for determining the zeros by applying a non-linear optimizer, such as the Fletcher-Powell method, to the square of the modulus of the polynomial, |p(x+iy)|², in order to determine its minimums. The minimums of | p |² are zero and occur at the zeros of p(z). Experimentation indicates that Gershgorin's discs and similar results for Cassini's ovals supply rather sharp bounds for this purpose     

Hybrid CPU–GPU execution support in the skeleton programming framework SkePU

In this paper, we present a hybrid execution backend for the skeleton programming framework SkePU. The backend is capable of automatically dividing the workload and simultaneously executing the computation on a multi-core CPU and any number of accelerators, such as GPUs. We show how to efficiently partition the workload of skeletons such as Map, MapReduce, and Scan to allow hybrid execution on heterogeneous computer systems. We also show a unified way of predicting how the workload should be partitioned based on performance modeling. With experiments on typical skeleton instances, we show the speedup for all skeletons when using the new hybrid backend. We also evaluate the performance on some real-world applications. Finally, we show that the new implementation gives higher and more reliable performance compared to an old hybrid execution implementation based on dynamic scheduling.