Fredy Ruiz

A study on the use of virtual sensors in vehicle control

The design of linear virtual sensors to estimate yaw rate for vehicle stability control systems is investigated. Standard model-based virtual sensor design techniques are compared to novel direct virtual sensor (DVS) design methodologies. The obtained DVS is stable and it can be used in a large range of operating conditions. It is shown how the use of virtual sensors derived directly from data and a suitable choice of the measured variables in sensor design improves the estimation and control accuracy. The effectiveness of the proposed DVS design is demonstrated by its employment in an existing yaw rate feedback loop, based on an active front steering actuator and designed using internal model control techniques. Robust stability is guaranteed in the presence of model uncertainty and of the DVS. In particular, the presented study shows that the DVS technology can be conveniently taken into account to replace physical sensors to obtain low cost stability control solutions for application on A and B segment cars.

Direct design from data of optimal filters for LPV systems

In real-world applications, the traditional approach to filter design is based on a two-step procedure: (1) model identification from data; (2) filter design from the identified model. However, the two-step procedure is in general not optimal and evaluating the effects of the modeling error (occurring in the identification step) on the estimation error of the designed filter is a largely open problem. In this paper, a new approach to filter design for LPV systems is proposed, based on the direct design of the filter from data, avoiding the intermediate step of model identification. This approach, developed within a Set Membership framework, allows the design of optimal filters and the evaluation of non-conservative bounds on the estimation error. An applicative example is presented, related to the estimation of vehicle yaw rate, a variable used by safety control systems to improve the vehicle stability.

A new approach to optimal filter design for nonlinear systems

Optimal filters for nonlinear systems are in general difficult to derive/implement. The common approach is to obtain approximate solutions, e.g. based on linearizations along the performed path, as done in Extended Kalman Filters. However, no optimality properties can be guaranteed using these approximations, not even the stability of the estimation error is ensured. In this paper, a new method is presented, able to overcome this problem. The method is based on the direct identification of the filter from a set of data. Under the standard assumptions of the filtering literature, i.e. known system and noise properties, the data can be generated by simulations of system equations. A further feature of the method arises from the fact that, in most practical situations, the system to filter is not known, but it is possible to perform measurements on it. In these situations, a two-step approach is typically adopted: 1) a model of the system is identified from the available measurements; 2) a filter is designed from the identified model. A relevant problem of the two-step design is that, in presence of modeling errors, the two-step filter may display large performance deteriorations. The direct method allows the direct design of the filter from the measurements, avoiding this problem. The method is developed within a Set Membership framework. Optimal filters for nonlinear systems are obtained, where optimality refers to the minimization of the induced norm from the noise to the estimation error.

On the design of linear virtual sensors for low cost vehicle stability control

The design of linear direct virtual sensors (DVS) to estimate yaw rate for vehicle stability control systems is investigated. The obtained DVS is stable and it can be used in a large range of operating conditions. It is shown how the use of closed-loop collected data and a suitable choice of the measured variables in sensor design improves the estimation accuracy. The effectiveness of the proposed DVS design is demonstrated by its employment in an existing yaw rate feedback loop, based on an active front steering actuator and designed using internal model control techniques. Robust stability is guaranteed in the presence of model uncertainty and of the DVS. In particular, the presented study shows that the DVS technology can be conveniently taken into account to replace physical sensors to obtain low cost stability control solutions for application on A and B segment cars.

A predictive control approach for DC-DC power converters and cyclic switched systems

Classical control laws for power converters are based on the average model. Usually, they have a good performance in the transient. Nevertheless, the steady state behavior is not well-controlled (waveform, subharmonics, etc.). This article shows a predictive approach that reaches an optimal periodic cycle from a set-point of the average model. The method uses the sensitivity functions and a Newton algorithm which allows to track an optimal trajectory based on a cost function.

Multivariable estimation of a web winding system

This article presents the identification of a web winding system and the estimation of its reel radius as time-varying parameter. This system is non-linear and time-varying, then, a set of Linear, Time Invariant models are estimated for local operation regions. The reel radius estimator is a direct virtual sensor (DVS) directly designed from experimental data. The estimated models are compared to a single model, identified for the complete operation range and better results are obtained with the local models on fresh data not used in the identification phase. The DVS is able to estimate the reel radius from voltage, speed and tension measurements, with good precision.

Demand response: Understanding the rational behavior of consumers in a Peak Time Rebate Program

The optimal behavior of the demand side in an electricity market is studied when a consumer participates in a Peak Time Rebate Program (PTR). The main motivation is the growth of demand bidding programs around the world in order to reduce peak power consumption events. The situation is posed as a stochastic programming problem where the user chooses the optimal consumption profile to maximize his economic benefits. The consumer preferences are modeled as a risk-averse function under additive uncertainty. The case considered uses the previous consumption as the household-specific baseline for PTR program. As result, a rational user alters the baseline in order to increase his well-being. When the incentive is greater than the energy retail price, the best behavior for the user is to consume the maximum possible energy during the baseline setting period in order to get the highest profits during the PTR time. Thus, PTR program is not a favorable mechanism for a System Operator (SO) from an economic view whether the SO does not control properly the user participation.

Nonlinear model predictive control for a Ball&Beam

This article presents the development of a modelbased predictive control (MPC) for Ball & Beam system and results achieved in the implementation of this controller on a real plant. The purpose of this work is to show that the MPC is one of the optimal control strategies more employed in the research community over the last years due to good management of the nonlinearities and the fulfillment of mechanical and electrical constraints. The validation of the results is supported by comparing experimental data with theoretical results.

A Fast Approximation Algorithm for Set-Membership System Identification

Abstract The Set Membership nonlinear identification method is a flexible technique, suitable for the identification of non-linear systems where no information on the system structure is available. This method generates a non-parametric model, embedded on the identification data set, with optimality properties and bounds on the possible values the variable can assume. However, the complexity of the model grows with the product of the input dimension and the size of the data set. This is problematic when a big data set is employed. In this paper, the nearest point approximation to the exact Set Membership model, found in literature, is analyzed and a novel approximation is proposed, whose complexity does not depend on the size of the data set. Guaranteed bounds on the worst-case approximation error are given. Two examples, considering simulated and experimental data, illustrate the validity and applicability of the obtained results.

Production and characterization of a human lysosomal recombinant iduronate‐2‐sulfatase produced in Pichia pastoris

Hunter syndrome (Mucopolysaccharidosis II, MPS II) is an X‐linked lysosomal storage disease produced by the deficiency of the lysosomal enzyme iduronate‐2‐sulfatase (IDS). Currently, MPS II patients are mainly treated with enzyme replacement therapy (ERT) using recombinant enzymes produced in mammalian cells. As an alternative, several studies have shown the production of active and therapeutic forms of lysosomal proteins in microorganisms. In this paper, we report the production and characterization of a recombinant IDS produced in the yeast Pichia pastoris (prIDS). We evaluated the effect of culture conditions and gene sequence optimization on prIDS production. The results showed that the highest production of prIDS was obtained at oxygen‐limited conditions using a codon‐optimized IDS cDNA. The purified enzyme showed a final activity of 12.45 nmol mg−1 H−1 and an apparent molecular mass of about 90 kDa. The highest stability was achieved at pH 6.0, and prIDS also showed high stability in human serum. Noteworthy, the enzyme was taken up by culture cells in a dose‐dependent manner through mannose receptors, which allowed the delivery of the enzyme to the lysosome. In summary, these results show the potential of Pichia pastoris as a host to produce an IDS intended for a MPS II ERT.