Reaction wheels have been extensively used to control and stabilize a wide range of applications due to the angular momentum exchange that they provide to mechanical systems. However, there exist several limitations associated with actuator saturation and the presence of singularities that lead towards the use of different controllers and devices. Recently, a new and unusual configuration using two reaction wheels was proposed and illustrated in an inverted pendulum to drive and control it using a simple and well-known PID controller. In this context, this paper improves the previous results on the same system by proposing two more sophisticated nonlinear controllers: a nonlinear proportional-derivative and a sliding mode controller. The proof of the stability of each controller is also provided, and the asymptotic stability is proven. A friction model is experimentally updated into the differential equations of the pendulum and also included for the controllers designing. A real-time application verifies the performance of the controllers using low-cost hardware. Results based on the analysis of different performance indices highlight the improvement of applying these two nonlinear control techniques comparing to the previous paper using this configuration of actuators.

This paper initially proposes an optimization problem and after presents its optimal solution. Then, this result is applied to obtain relaxed conditions to design controllers for nonlinear plants described by Takagi-Sugeno (TS) models, based on fuzzy Lyapunov function (FLF) and Linear Matrix Inequalities (LMI). The FLF is given by V(x(t)) = x(t) ^T P(α(x(t)))x(t), where x(t) is the plant state vector, P(α(x(t))) = α ₁ (x(t))P ₁ + α ₂ (x(t))P ₂ + ⋯+ α _r (x(t))P _r , P _i =P _i ^T > 0 and α _i (x(t)) is the weight related to the local model i in the representation of the plant by TS fuzzy models, for i=1,2,⋯,r. When one calculates the time derivative of this V(x(t)), it appears the term x(t) ^T · P(α(x(t)))x(t), that is usually handled using conservative upper bounds, supposing that the bounds of the time derivative of α _i (x(t)), i=1,2,⋯,r, are available. The main result of this paper is a procedure to obtain optimal upper bounds for the term x(t) ^T · P(α(x(t)))x(t), such that they contemplate the maximum value and are always smaller than or equal to the maximum value. It is a relevant result on this subject, because these optimal upper bounds do not add any constraint. With these optimal upper bounds, a relaxed design method for stabilization of TS fuzzy models is proposed. Two numerical examples illustrate the effectiveness of this procedure.

The attitude control using reaction wheels as actuators has been one of the most popular ways to stabilize and repel external disturbances in aerospace devices. From the controlled change of the angular momentum rate of change using reaction wheels, it is possible to control the oscillation and direction rates of change of rigid bodies in space. Thus, the main idea of this article is to present a case study with different configurations of the well-known reaction wheel pendulum. The first is based on the classical configuration, and the second, a new one, a pendulum with two reaction wheels. For both configurations, proportional–integral–derivative controllers were designed and experimental devices were built to perform real-time controllers using low-cost hardware. The simulated and experimental results have shown that the pendulums were controlled using a simple controller in the inverted position and the results were satisfactory. Four performance indices were calculated to evaluate the results for each configuration. They showed that the pendulum with two reaction wheels worked better than the pendulum with one reaction wheel. Two actuators made it easier to move and control the pendulum in the inverted position.

This paper analyzes, designs and implements a reconfigurable phase-shifted full-bridge (PSFB) converter. It adopts the topology of the traditional PSFB converter and incorporates clamping circuits to solve some fundamental problems of conventional topology. In addition, auxiliary switches are employed for output reconfiguration, which allows expanding the output voltage range without compromising the system efficiency. Single pole double throw (SPDT) mechanical switches are used to realize series and parallel connections. In this paper, the characterization of the PSFB converter with clamping circuit and its design considerations are discussed. A 10 kW prototype with a power density of 0.485 W/cm³, 900 V input voltage and 400/800 V nominal output voltage was manufactured. The experimental results validated the analysis and confirmed the high conversion efficiency for a wide load range; an efficiency of 96.69% was obtained for the full load condition.

In this work, the practical implementation of the control system for the reaction wheel inverted pendulum (RWIP) was carried out using low-cost hardware components. A switch control strategy was adopted to operate the swing up and stabilization controllers. This switch control method integrates two distinct controllers: one designed to raise the pendulum to the inverted region, and another to stabilize it at the unstable equilibrium point in the inverted position. A nonlinear swing up controller was designed to address the problem of moving the pendulum from the stable equilibrium position to a favorable position for the application of a stabilization controller. For the stabilization region, linear controllers, proportional– integral–derivative (PID) and eigenvalue-placement—were designed, as well as a nonlinear controller based on the conventional sliding mode control (SMC) technique. The control laws were programmed using scripts developed on the Arduino IDE with the ESP32 prototyping board, and data collection was performed through serial communication between MATLAB® and the ESP32. Performance indices were calculated to evaluate the response of the controlled variable in the stabilization region.