Jean Buisson

Comparison of Hybrid Control Techniques for Buck and Boost DC-DC Converters

Five recent techniques from hybrid and optimal control are evaluated on two power electronics benchmark problems. The benchmarks involve a number of practically interesting operating scenarios for fixed-frequency synchronous dc-dc converters. The specifications are defined such that good performance can only be obtained if the switched and nonlinear nature of the problem is accounted for during the design phase. A nonlinear action is featured in all methods either intrinsically or as external logic. The designs are evaluated and compared on the same experimental platform. Experiments show that the proposed methods display high performances, while respecting circuit constraints, thus protecting the semiconductor devices. Moreover, the complexity of the controllers is compatible with the high-frequency requirements of the considered application.

On the stabilisation of switching electrical power converters

This paper considers the control of switching power converters which are a particular class of hybrid systems. Such systems, which are controlled by switches, can be modeled using physical principles. Taking advantage of the energetical properties of their models, a Lyapunov function is proposed. This function, which has not to be computed but is systematically deduced from the physical model, allows to derive different stabilizing switching sequences. From a theoretical point of view, asymptotic stability can be obtained, but it requires null intervals between switching times. In order to ensure a minimum time between switchings, this Lyapunov function has to be increasing for a small duration by using a delay or a dead zone. A control law principle that guarantees the invariance of a specified domain with respect to state trajectories is proposed. Two examples are provided at the end of this paper that demonstrate the efficiency of the proposed approach.

Analysis of the bond graph model of hybrid physical systems with ideal switches

Abstract This paper deals with the modelling of hybrid physical systems. The bond graph technique is used to establish their knowledge model, based upon an ideal representation of the switches. These components are modelled either by flow or by effort sources according to their state and therefore modify the circuit topology at switching times. The paper shows the usefulness of the implicit representation to derive a unique implicit state equation with jumping parameters, to analyse the model properties, to derive an implicit state equation with nilpotency index one for each configuration and to compute the discontinuities. Also, a comparison between the chosen ideal modelling approach and the more common non-ideal approach is carried out using singular perturbations theory. After a presentation of the whole study in the most general context, its results are applied to power converters, which constitute a particular class of hybrid physical systems where switches only commutate in pairs. Finally, an example is developed.

Analysis of switching devices with bond graphs

Abstract Modelling and simulation of switching devices with bond graphs is a subject which has no totally satisfying solutions. In this paper, switching devices are represented by flow or effort sources, with a variable circuit topology at switching time. For the simulation, the usual method uses causality resistors to insure integral causality to energy storage elements. The choice of those resistors is quite arbitrary and can lead to stiff systems. We propose, for linear systems, an approach without such resistors. When components lose the integral causality, the order of the state vector changes, provoking the use of pulse variables. A solution is proposed to compute the amplitude of the pulse and the value of the new variables.

Hybrid Control Techniques for Switched-Mode DC-DC Converters Part I: The Step-Down Topology

Several recent techniques from hybrid and optimal control are evaluated on a power electronics benchmark problem. The benchmark involves a number of practically interesting operating scenarios for a fixed-frequency synchronous step-down DC-DC converter. The specifications are defined such that good performance only can be obtained if the switched and nonlinear nature of the problem is respected during the design phase.

Stabilization of switched affine systems: An application to the buck-boost converter

In this paper we extend a technique developed to design a feedback stabilizing control law for autonomous switched systems all modes of which are unstable. More specifically, we extend the switching table procedure to the class of affine switched systems, the dynamics of which either do not have an equilibrium point or, if they do, it is not common. This method is then applied to the DC-DC buck-boost converter. The design of the control law is based on dynamic programming and it results in a partition of the state space into switching look-up tables. A comparison with a Lyapunov based technique is also discussed.

Distributed model predictive control for building temperature regulation

This paper presents a predictive control structure for thermal regulation in buildings. The proposed method considers a dynamic cost function trying to exploit the intermittently operating mode of almost all types of buildings. One of the key idea is to use the knowledge about the occupation profile. For that purpose, the predictive control strategy is first presented for a single zone building then extended to a multizone building example. Two opposite control strategies commonly exists. The decentralized control structure, which does not offer good performances especially when the thermal coupling among adjacent rooms is not negligible, and on the other hand, the centralized control for which the computational demand grows exponentially with the size of the system, being very expensive for large scale buildings. Our solution is based on a distributed approach which takes the advantages of the both methods mentioned above. A distributed MPC algorithm with one information exchange per time step is proposed with good control performances and low computational requirements.

Hybrid control of a three-level three-cell dc-dc converter

This paper compares three synthesis methods for controlling a three-level three-cell dc-dc converter. The main contribution of this paper is to analyse different strategies: i) The Passivity Based Control that uses the notion of average model, ii) A stabilizing method in which a unique Lyapunov function is introduced and iii) A new predictive control approach, which relies on the use of optimization procedures.

MODEL PREDICTIVE CONTROL FOR HYBRID SYSTEMS UNDER A STATE PARTITION BASED MLD APPROACH (SPMLD)

This paper presents the State Partition based Mixed Logical Dynamical (SPMLD) formalism as a new modeling technique for a class of discrete-time hybrid systems, where the system is defined by different modes with continuous and logical control inputs and state variables, each model subject to linear constraints. The reformulation of the predictive strategy for hybrid systems under the SPMLD approach is then developed. This technique enables to considerably reduce the computation time (with respect to the classical MPC approaches for PWA and MLD models), as a positive feature for real time implementation. This strategy is applied in simulation to the control of a three tanks benchmark.

A dynamic horizon distributed predictive control approach for temperature regulation in multi-zone buildings

This paper proposes a distributed predictive control strategy for building temperature regulation. The originality of the approach consists of using a dynamic prediction horizon MPC, allowing a more effective disturbances rejection. Then this idea is extended, for multizone temperature control in buildings, in a distributed manner, where each controller solves a local optimization problem using information about the expected behavior of the other local regulators. The effectiveness of the proposed approach is showed through different simulations.