Hermínio Duarte-Ramos

Learn More
The problem of order estimation and global stability in affine three-layered state-space neural networks is here addressed. An upper bound for the number of neurons to be inserted in the hidden layer is computed using a subspace technique. Some sufficient conditions for the global asymptotic stability are presented using the Lyapunov stability theory and(More)
An extended model-based predictive control scheme is proposed and implemented on a bench three-tanks system. This structure is based on a constrained local instantaneous linear model-based predictive controller complemented with a static offset compensator for guaranteeing that tracking errors converge to zero in a finite time. A non-linear state-space(More)
This paper focuses on online non-linear system identification via state-space neural networks. The training algorithm is based on a generalisation of the Kalman filter to non-linear systems by means of the unscented transformation. Experimental results from a laboratory heating system confirm the feasibility and effectiveness of the proposed methodology.
The human operator is, no doubt, the most complex and variable element of a Mechatronics system. On simpler manual control tasks, a linear model may be used to capture the human dynamics, however experiences on human operator response during pursuit manual tracking tasks, show that the dynamics of the human operator appear to depend on the specific task(More)
Most of the mechatronics equipments and gadgets that we all nowadays rely on incorporate some kind of multidimensional human-machine systems. There is an increasing concern for improving the usability, performance, ergonomics and safety of such devices, and ultimately this will lead to the mass-production of next-generation intelligent machines, which will(More)
An approach to the control of a distributed solar collector field relying on a non-linear adaptive constrained model-based predictive control scheme with steady-state offset compensation is developed and implemented. This methodology is based on a non-linear state-space neural networks within a model-based predictive control framework. The neural network(More)
This paper describes the application of a non-linear adaptive constrained model-based predictive control scheme to the distributed collector field of a solar power plant at the Plataforma Solar de Almería (Spain). This methodology exploits the intrinsic non-linear modelling capabilities of nonlinear state-space neural networks and their online training by(More)
Traditionally Man-Machine Interfaces (MMI) are concerned with the ergonomic aspects of the operation, often disregarding other aspects on how humans learn and use machines. The explicit use of the operator dynamics characterization for the definition of the Human-in-the-Loop control system may allow an improved performance for manual control systems. The(More)
The problem of order evaluation for an affine state-space neural network or equivalently the estimation of the number of neurons to be inserted in the hidden layer in a recurrent neural network is here addressed. The proposed method is based on a singular value decomposition applied to an oblique subspace projection given as the projection of the row space(More)