Samira Roshany-Yamchi

Kalman Filter-Based Distributed Predictive Control of Large-Scale Multi-Rate Systems: Application to Power Networks

In this paper, a novel distributed Kalman filter (KF) algorithm along with a distributed model predictive control (MPC) scheme for large-scale multi-rate systems is proposed. The decomposed multi-rate system consists of smaller subsystems with linear dynamics that are coupled via states. These subsystems are multi-rate systems in the sense that either output measurements or input updates are not available at certain sampling times. Such systems can arise, e.g., when the number of sensors is smaller than the number of variables to be controlled, or when measurements of outputs cannot be completed simultaneously because of practical limitations. The multi-rate nature gives rise to lack of information, which will cause uncertainty in the systems performance. To circumvent this problem, we propose a distributed KF-based MPC scheme, in which multiple control and estimation agents each determine actions for their own parts of the system. Via communication, the agents can in a cooperative way take one anothers actions into account. The main task of the proposed distributed KF is to compensate for the information loss due to the multi-rate nature of the systems by providing optimal estimation of the missing information. A demanding two-area power network example is used to demonstrate the effectiveness of the proposed method.

Distributed Model Predictive Control and Estimation of Large-Scale Multi-Rate Systems

Abstract In this paper, we propose a new method for control of large-scale multi-rate systems with linear dynamics that are coupled via inputs. These systems are multi-rate systems in the sense that either output measurements or input updates are not available at certain sampling times. Such systems can arise, e.g., when the number of sensors is less than the number of variables to be controlled, or when measurements of outputs cannot be completed simultaneously because of applicational limitations. The multi-rate nature gives rise to lack of information, which will cause uncertainty in the systems performance. A distributed model predictive control (MPC) approach based on Nash game theory is proposed to control multi-agent multi-rate systems in which multiple control agents each determine actions for their own parts of the system. Via communication, the agents can in a cooperative way take one anothers actions into account. To compensate for the information loss due to the multi-rate nature of the systems under study, a distributed Kalman Filter is proposed to provide the optimal estimation of the missing information. Using simulation studies on a distillation column the added value of the proposed distributed MPC and Kalman Filter method is illustrated in comparison with a centralized MPC with centralized Kalman Filter, and a distributed MPC method with a fully decentralized (i.e., no communication) Kalman Filter.

Nash-Based Distributed MPC for Multi-Rate Systems

In this chapter, a new Nash-based distributed MPC method is proposed to control large-scale multi-rate systems with linear dynamics that are coupled via inputs. These systems are multi-rate systems in the sense that either output measurements or input updates are not available at certain sampling times. Such systems can arise when the number of sensors is less than the number of variables to be controlled or when measurements of outputs cannot be completed simultaneously because of applicational limitations. The multi-rate nature gives rise to a lack of information which will cause uncertainty in the system’s performance. To compensate for the information loss due to the multi-rate nature of the systems under study, a distributed Kalman filter is proposed to provide an optimal estimate of the missing information.

Game theory-based distributed predictive control: Application to shell oil fractionator

In this paper we propose to study the underlying properties of our formerly proposed distributed multi-rate predictive control strategy based on Nash game [1]. In this proposed method a set of multi-rate constrained agents communicate to accomplish their goals. The problems of how to decide the multirate communication strategy, share the inputs, estimated states and observers gains are solved using tools from game theory. The proposed scheme is demonstrated through a simulation example. Eventually different multi-rate scenarios have been simulated to present the performance of the method under all possible scenarios and also comparison of these scenarios together.

Negotiated distributed estimation with guaranteed performance for bandwidth-limited situations

Abstract This paper proposes a guaranteed distributed estimator for large-scale, state-coupled plants. Each agent must find appropriate sets, described by zonotopes, to contain the actual state of the plant and reduce the estimation uncertainties. To do so, the agents collaborate by sending information through a network in which bandwidth restrictions need to be taken into account. An estimation algorithm is proposed, where the amount of information transmitted is negotiated through an iterative procedure that trades off communication burden and estimation performance. The negotiation process between the agents is similar to a non-cooperative game, which ends when the Nash equilibrium is reached. The algorithm can be tuned to weight estimation performance, communication costs, and discrepancies in the amount of information transmitted by neighboring agents. The estimation structure is tested by simulation on a plant consisting of a set of inverted pendulums coupled by springs.

Selective distributed model predictive control for comfort satisfaction in multi-zone buildings

Distributed model predictive control (DMPC) for thermal regulation in multi-zone buildings continues to gain attention over centralized approaches. Particularly, centralized control approaches have been shown to become impractical when applied to large-scale buildings due to for example, computation complexity, modeling complexity of large buildings and availability of required sensor and actuator infrastructure. In this paper a novel selective DMPC algorithm is developed in which, each agent optimizes a cost function to minimize the control effort in order to save energy and satisfying the comfort bound. The proposed method is useful especially in building thermal regulation when the objective is to keep the temperature of each zone in the building within the defined comfort bound.

Distributed set-membership observers for interconnected multi-rate systems

Abstract This paper presents a new distributed observer for interconnected multi-rate systems. The developed observer belongs to the family of set-membership estimators, and the use of zonotopes is proposed to mathematically describe the sets, a choice motivated by the available mathematical background in operations such as intersections, combinations and linear manipulations. The main features of the proposed distributed observer are (a) the actual state of any subsystem always belongs to the computed sets; (b) the volume of these sets is minimized in real time; (c) under equivalent assumptions, the performance of the observer approaches to that of an analogous distributed Kalman filter.

Input variables selection of fuzzy dynamic models by using genetic algorithm

In this paper, an input selection technique for fuzzy-based dynamic modelling is presented. When a system is composed of a large number of input-output variables with strong coupled dynamics, a systematic method is required in order to minimise the modelling error through the selection of proper variables. The selection approach is based on a Genetic algorithm optimisation to obtain solutions as binary strings that effectively determine a proper input field structure.