Energy management in power systems is a thorny optimization problem. With the sizes of systems rising, centralized optimization methods are restricted by their complexities of communications, while distributed optimization methods have emerged as a powerful tool for dealing with increasingly complex systems. However, convergence rates of some widely used distributed optimization methods, such as the standard alternating direction method of multipliers (ADMM), still have room for improvement. In this paper, a parallel and distributed optimization method for energy management of microgrids (MGs) is proposed to boost the convergence rate without sacrificing the accuracy of the optima, in which agents calculate, exchange and update in parallel. At first, a decomposition method is presented, where the objective functions and constraints of an original optimization problem with separable variables are decomposed into local objective functions and constraints for agents, which is the key to our method. Further, agents solve their local optimization problems independently and then exchange determined optima with their neighbors. Finally, the method is evaluated to solve economic dispatch with demand response for microgrids. The simulation results show that compared to the standard ADMM, for a given accuracy, the number of iterations in our method is only one third or even less than that of ADMM. Furthermore, our method can minimize the cost functions of distributed generation on supply side and maximize the profit functions of flexible loads on the demand side.
The averaging algorithm for consensus is widely used as a distributed secondary method for the control and management of microgrids. However, during each iteration it may break the system balance obtained by the primary control. In this paper, a distributed and networked method for the control and management of islanded microgrids is proposed, in which there is an agent based communication network as the top layer over a microgrid as the bottom layer. Further, a systematic method is presented to derive a set of distributed control laws for agents from any given communication network, where only nearest neighbor information is needed. The control laws consist of two terms, dynamic and fixed weights, in which the term with dynamic weights reassigns outputs of distributed generators in order to reach different targets. Moreover, this method offers a convenient way to achieve different targets of control and management by substituting a parameter in the control laws with dynamic weights. More importantly, the control laws with dynamic weights never break the system balance during iterations. We formally show that if agents apply the control laws to regulate distributed generators, their outputs will iteratively satisfy the given targets. Finally, simulations are carried out to evaluate the performance of the control laws. The results show that equal outputs, proportional outputs and the optimal incremental cost are obtained. Moreover, the voltage and frequency are still stable, when fluctuations of load demand and environmental conditions are considered.
We study the coevolution of quantum and classical strategies on weighted and directed random networks in the realm of the prisoner’s dilemma game. During the evolution, agents can break and rewire their links with the aim of maximizing payoffs, and they can also adjust the weights to indicate preferences, either positive or negative, towards their neighbors. The network structure itself is thus also subject to evolution. Importantly, the directionality of links does not affect the accumulation of payoffs nor the strategy transfers, but serves only to designate the owner of each particular link and with it the right to adjust the link as needed. We show that quantum strategies outperform classical strategies, and that the critical temptation to defect at which cooperative behavior can be maintained rises, if the network structure is updated frequently. Punishing neighbors by reducing the weights of their links also plays an important role in maintaining cooperation under adverse conditions. We find that the self-organization of the initially random network structure, driven by the evolutionary competition between quantum and classical strategies, leads to the spontaneous emergence of small average path length and a large clustering coefficient.
We study the impact of adaptive degrees of trust on the evolution of cooperation in the quantum prisoner's dilemma game. In addition to the strategies, links between players are also subject to evolution. Starting with a scale-free interaction network, players adjust trust towards their neighbors based on received payoffs. The latter governs the strategy adoption process, while trust governs the rewiring of links. As soon as the degree of trust towards a neighbor drops to zero, the link is rewired to another randomly chosen player within the network. We find that for small temptations to defect cooperators always dominate, while for intermediate and strong temptations a single quantum strategy is able to outperform all other strategies. In general, reciprocal trust remains within close relationships and favors the dominance of a single strategy. Due to coevolution, the power-law degree distributions transform to Poisson distributions.