This paper is concerned with the maximum power point tracking control of photovoltaic systems. In solar power generation using photovoltaic systems, DC-DC converters are usually inserted between photovoltaic systems and loads so that we can change operation points and thereby increase efficiency. However, the characteristic of the output power with respect to the duty ratio of DC-DC converters has multiple local maxima and drastically changes according to weather variation. In this paper, we propose an algorithm for online tuning of the duty ratio based on particle swarm optimization so that we can realize maximum power point tracking under such intractable circumstances. We finally verify the effectiveness of the algorithm by practical experiments.
Feedback Error Learning (FEL) has been actively studied in an adaptive control framework. When the plant is a time-variant or nonlinear system, however, conventional FEL schemes do not work any more because a single inverse model is not enough. As a remedy to this problem, Just-In-Time (JIT) modeling has been proposed, where we memorize regression data for all input/output behavior of the plant and construct local models as required in real time. However, a scheme for implementing JIT modeling, Locally Weighted Regression (LWR), has problems that it is computationally heavy to search for query point among huge amount of the past input/output data, and that regression calculation is difficult when the number of information vectors is insufficient. In view of these points, this paper proposes a new scheme. In the proposed scheme, we seek local models by a schedule parameter, instead of searching time series by input/output data. We call this scheme Scheduled LWR. By applying it to the FEL, we can reduce the search time without losing practicality and also solve the lack of sample number. We verify the effectiveness of the proposed scheme by simulation using a 2-link robot arm model.
We introduce tractable global optimization algorithms for small boolean quadratic programming problems using reflected Gray coding technique. In the algorithms, the space complexity is limited to O(n2). Furthermore, we can implement them without multiplications or floating point operations when the instances are integers. Numerical experiments revealed that the proposed algorithms run as fast as interior point algorithms for solving convex relaxation problems of the original QPs.