Ichiro Maruta

Robust PID controller tuning based on the constrained particle swarm optimization

This paper proposes a novel tuning strategy for robust proportional-integral-derivative (PID) controllers based on the augmented Lagrangian particle swarm optimization (ALPSO). First, the problem of PID controller tuning satisfying multiple H ∞ performance criteria is considered, which is known to suffer from computational intractability and conservatism when any existing method is adopted. In order to give some remedy to such a design problem without using any complicated manipulations, the ALPSO based robust gain tuning scheme for PID controllers is introduced. It does not need any conservative assumption unlike the conventional methods, and often enables us to find the desired PID gains just by solving the constrained optimization problem in a straightforward way. However, it is difficult to guarantee its effectiveness in a theoretical way, because PSO is essentially a stochastic approach. Therefore, it is evaluated by several simulation examples, which demonstrate that the proposed approach works well to obtain PID controller parameters satisfying the multiple H ∞ performance criteria.

A simple and efficient constrained particle swarm optimization and its application to engineering design problems

Abstract Engineering optimization problems usually contain various constraints and mixed integer-discrete-continuous type of design variables. This article proposes an efficient particle swarm optimization (PSO) algorithm for such problems. First, the constrained optimization problem is transformed into an unconstrained problem without introducing any problem-dependent or user-defined parameters such as penalty factors or Lagrange multipliers, though such parameters are usually required in general optimization algorithms. Then, the above PSO method is extended to handle integer, discrete, and continuous design variables in a simple manner, yet with a high degree of precision. The proposed PSO scheme is fairly simple and thus it is easy to implement. In order to demonstrate the effectiveness of our method, several mechanical design optimization problems are solved, and the numerical results are compared with those reported in the literature.

Practical modeling and system identification of R/C servo motors

An R/C servo motor is a compact package of DC geared-motor associated with position servo controller. They are widely used in small-sized robotics and mechatronics by virtue of their compactness, easiness-to-use and high power/weight ratio. However, in order to improve control performance of mechatronic systems using R/C servo motors, such as biped robots or under-actuated systems, it is crucial to clarify their mathematical model. In this paper, we propose a simple and realistic internal model of R/C servo motors including the embedded servo controller, and estimate their physical parameters using continuous-time system identification method. We also provide a transfer function model of their referenceto-torque characteristics so that we can estimate the internal torque acting on the load.

Synthesis of fixed-structure H∞ controllers via Constrained Particle Swarm Optimization

Abstract This paper provides a design method of fixed-structure robust controllers satisfying multiple H∞ norm specifications by using a sort of randomized algorithms. First, a new tool to perform general constrained optimization is developed which does not need any gradient or derivative of the objective function. This tool is based on PSO (particle swarm optimization), which attracts a lot of attention recently in the evolutionary computation area due to its empirical evidence of its superiority in solving various non-convex problems. Second, it is shown how to design a fixed-structure controller satisfying given multiple H∞ specifications by using the developed optimization tool. Third, its effectiveness is evaluated through various numerical examples, because it is difficult to guarantee the performance of the proposed method theoretically due to a probabilistic nature of the PSO. The simulation results demonstrate its effectiveness clearly.

Particle swarm optimization based robust PID controller tuning scheme

A novel optimal tuning technique for robust PID controllers based on the augmented Lagrangian particle swarm optimization (ALPSO) is presented. To this aim, we first consider the problem on PID controller tuning satisfying Hinfin performance criteria. It suffers from computational intractability and conservatism when the existing methods are introduced. In the proposed approach, the ALPSO based optimal robust gain tuning technique for PID controllers is introduced in order to solve simply and directly such a design problem without using any complicated manipulations. The proposed scheme works without any conservative assumption required in the conventional methods, and also enables one to find a set of optimal gains by just solving the constrained optimization problem in a straightforward way. The numerical results demonstrate that our approach is not only computationally efficient but also considerably convenient to obtain a set of optimal PID controller parameters.

Identification of PWA models via data compression based on l 1 optimization

In this paper, a new identification method for piecewise affine (PWA) models is introduced. The method is based on the data-based representation of PWA maps and the data compression with l1 optimization technique, which enable the method to deal with large data sets. This method can be applied to a wide range of modeling problems, and an example with a DC motor system is shown in this paper to show the usability of the method.

Projection-based identification algorithm for grey-box continuous-time models

Abstract In this paper, a new identification method for continuous-time models, which can handle various grey-box structures and has strong robustness, is presented. The proposed method is based on an incremental model update scheme and the projection onto the subspace which reflects the model structure. By utilising these schemes, robustness of other continuous-time system identification methods and versatility of generic optimisation algorithms can be integrated into the proposed method. The effectiveness of the proposed method is demonstrated through numerical examples related to a grey-box model in closed-loop system and systems with unknown time-delay.

Modeling of dynamics in demand response for real-time pricing

To avoid the excessive suppression of electricity demand caused by conservative planning of electricity price, real-time minute-to-minute control of the electricity price, which is called real-time pricing, is required. However, the electricity demand response has the dynamics on a timescale shorter than an hour, which can cause catastrophic incidents if the real-time pricing is done without ample consideration, and the model for such dynamics is demanded. For this background, a new dynamic model for the electricity demand response is proposed in this paper. Since the main source of the price elasticity utilized in realtime pricing is considered to be the temperature setting of HVAC equipment, an experiment with an air conditioner is performed to get a quick sketch of the dynamics. Then, from the experimental results and knowledge from existing literatures, a kind of dynamic linear model with Markov-switching is introduced and shown to be a good model for the electricity demand response on a timescale shorter than an hour.

Identification of PWA models via optimal data compression

Abstract In this paper, a new identification method for non-parametric piecewise affine (PWA) models is introduced. The method is based on the non-parametric data-based representation of PWA maps and the data compression with l 1 optimization technique, which enable the method to deal with large data sets. In the proposed scheme, the prior knowledge about partitioning of the PWA map is not required, and the trade-off between complexity and accuracy of the model is easily adjusted by one parameter, which can be determined by holdout validation technique in practical situations. These features of the proposed method provide the high usability in practical problems, which is demonstrated through numerical and experimental examples.

A new approach for modeling hybrid systems based on the minimization of parameter transition in linear time-varying models

Since the methods for modeling piece-wise affine systems from its I/O data often require massive computation power and could be infeasible in realistic situations, we propose a new approach for modeling hybrid systems which are similar to piece-wise affine systems. In the approach, time-domain consistency of system parameters is utilized for constructing models, and linear time-varying models which resemble piece-wise affine systems are constructed via convex optimization techniques. Although the model construction procedure is simple and computationally inexpensive, the obtained models are informative and can be a basis for constructing hybrid models suitable for analysis and controller design. The effectiveness of the proposed approach is shown through an experiment, where a hybrid model for a DC motor system with non-linear friction is constructed based on the proposed approach.