Retrofit control is a new control concept towards the control of large-scale network systems that are subject to change due to extension and reconstruction of local subsystems by multiple independent subsystem operators. The proposed approach is a control method that, rather than an entire system model, requires a model of the subsystem of interest for controller design. The retrofit controller can guarantee robust stability in the sense that the entire closed-loop system is stable for any variations of neighboring subsystems (assembly of which is called an environment), other than the subsystem of interest, as long as the preexisting network system (the feedback system of the environment and the subsystem of interest) is originally stable. This enables the modular design of decentralized control systems, i.e., multiple subsystem operators can independently plug in, plug out, and reconstruct respective local controllers without concerning the instability of the entire network system.
As theoretical achievements, we have given the parameterization of all the retrofit controllers in terms of a constrained version of the Youla parametrization. Furthermore, providing a particular design method of retrofit controllers, we have analyzed resultant system properties when multiple retrofit controllers are designed and implemented in parallel. As further extension of theory, we have developed a design method of a generalized version of retrofit controllers where system identification (data-based learning) of the unknown environment is incorporated for better performance improvement. The major advantage of this extension is that the stability of the resultant control system is robustly assured regardless of not only the stability of approximate environment models, but also the magnitude of modeling errors. This has actually good compatibility with existing system identification methods because the accuracy of identified models may neither be reliable nor assurable in reality.
Smart use of renewable energy is one of the key tasks towards reductions in greenhouse gas emissions. To push up the penetration level of renewable energy resources, it is crucial to devise a practical control method to improve the degree of power system stability because renewable power generation, such as photovoltaic and wind power generation, is highly volatile due to weather change. With this background, we have developed a plug-in-type local control method for renewable-integrated power systems based on retrofit control theory. In particular, we have conducted detailed simulation analysis of a wind-integrated IEEE 68-bus test system to show that the proposed plug-in-type local control can properly enhance system stability in a modular fashion. Furthermore, using a modified Japanese bulk power system model, called PV-integrated EAST30 model, we have shown that the proposed method is practically effective also for the volatility of photovoltaic power generation.
Conventional thermal generators generally require several hours for warm-up and cool-down. Therefore, it is crucial to conduct the day-ahead scheduling of startup and shutdown of thermal generators to keep supply-demand balancing for volatile demand and renewable power generation, the amounts of which are also predicted the day before. Such a scheduling problem of conventional generators under uncertainty of demand and renewable power prediction is called a robust unit commitment problem, which is commonly described as two-stage robust optimization.
For robust unit commitment, we have proposed a novel approach that can realize stable power supply with explicit consideration of the feasibility of realtime economic dispatch in realtime operation. In the proposed robust unit commitment problem, we aim at deciding not only the startup and shutdown schedules of thermal generators, but also the admissible operation ranges of generators and batteries that are sufficient for realtime adaptation to the volatility of demand and renewable power generation. Such an ability to realize realtime adaptation is mathematically described as box-based temporal decomposition of feasible regions of multiperiod generator and battery operation.
Sharp fluctuations in photovoltaic (PV) power generation and idiosyncratic power consumption make the exact net-demand prediction difficult. As a relatively new approach to reliable renewable power prediction, there can be found interval-valued prediction methods that determine confidence intervals with a certain probability, e.g., 95%. Such interval-valued net-demand prediction can be depicted as the red region in the figure below. To comply with this type of prediction, we have formulated an interval-valued economic dispatch problem in terms of parametric optimization, which is described as a problem of finding the tightest box, i.e., interval hull, that encloses the image of an output function consisting of the minimizer of parametric optimization. The resultant interval hulls of the optimal generation and battery charge/discharge schedules are depicted as the blue and green regions in the figure, regarded as necessary regulating capacities to absorb the fluctuation of the net-demand.
In network science, real world networks have been extensively studied on the basis of statistics and graph theory, and some universal properties of real world networks, such as the small-world property and the power law of degree distribution, have been found out. Major examples of research topics in this area include community detection, which is formulated as a problem to divide a given network into some meaningful subnetworks in the sense of, e.g., personal human relationship.
In our work, we take an approach to the community detection problem from a viewpoint of control theory. More specifically, assigning a state variable to each node of networks, we construct a set of clusters according to the similarity of state behavior for input signals. As a result, as shown in the figure below, we can detect a set of similar nodes, and can visualize meaningful inter-cluster connections in the sense of input-to-state mapping. In addition, by aggregating the state of each cluster into a lower-dimensional variable, we can realize model reduction that preserves interconnection structure among clusters.
There are a number of systems having an internal state that does not escape from the nonnegative orthant. Example of such systems include spatially-discrete reaction-diffusion systems, electric circuit systems, Markovian processes, and so forth. In linear systems theory, this class of systems is called positive systems, and it is known that the positivity of systems is characterized by nonnegativity of system matrices. Based on this characterization, a model reduction problem preserving the system positivity can be formulated as a problem to approximate the input-to-output mapping of systems while preserving the nonnegativity of system matrices. In our work, in addition to the positivity-preserving model reduction problem, we formulate a dissipativity-preserving model reduction problem, and give a solution to it based on generalized singular perturbation.
A number of systematic control design methods, such as robust control, have been developed, and the efficiency of those methods has been verified from various viewpoints. However, it is known that, for large-scale systems, such a sophisticated design method inevitably makes a resultant controller and observer high-dimensional. In this sense, the implementation of control design methods for large-scale systems is not necessarily straightforward, and it is crucial to address the approximation problem of controllers and observers with a specified error precision.
The controller and observer reduction problem is formulated as a structured model reduction problem to find a lower-dimensional controller and observer preserving the signal transmission structure shown in the figures below. Even though this kind of structured model reduction problems is known as one of difficult problems in systems theory, we successfully give a solution to it by utilizing our dissipativity-preserving model reduction.
I am (was?) a tennis man.
Stephan Schneider, Guidi, ma+...
Belgian and German beers...