Aldo Cipriano

Forecasting ozone daily maximum levels at Santiago, Chile

Abstract In major urban areas, air pollution impact on health is serious enough to include it in the group of meteorological variables that are forecast daily. This work focusses on the comparison of different forecasting systems for daily maximum ozone levels at Santiago, Chile. The modelling tools used for these systems were linear time series, artificial neural networks and fuzzy models. The structure of the forecasting model was derived from basic principles and it includes a combination of persistence and daily maximum air temperature as input variables. Assessment of the models is based on two indices: their ability to forecast well an episode, and their tendency to forecast an episode that did not occur at the end (a false positive). All the models tried in this work showed good forecasting performance, with 70–95% of successful forecasts at two monitor sites: Downtown (moderate impacts) and Eastern (downwind, highest impacts). The number of false positives was not negligible, but this may be improved by expressing the forecast in broad classes: low, average, high, very high impacts; the fuzzy model was the most reliable forecast, with the lowest number of false positives among the different models evaluated. The quality of the results and the dynamics of ozone formation suggest the use of a forecast to warn people about excessive exposure during episodic days at Santiago.

Real-Time Control of Buses in a Transit Corridor Based on Vehicle Holding and Boarding Limits

A real-time mathematical programming model of buses operating on a transit corridor that incorporates vehicle-capacity constraints is proposed. The objective for the model is to minimize the total times experienced by all passengers in the system, from the moment they arrive at a stop to the moment they reach their destination. Two control policies are considered: (a) vehicle holding, which is applicable at any stop, and (b) boarding limits that constrain the number of passengers entering a vehicle even when the vehicle is at less than physical capacity, to increase operating speed. The objective function is quadratic, but not convex with linear constraints. This problem is solved by using MINOS in a reasonable amount of computation time. A case study in a high-demand scenario shows that the proposed control achieves reductions in the objective function of more than 22% and 12% compared with no control and only holding strategies, respectively.

A REAL TIME VISUAL SENSOR FOR SUPERVISION OF FLOTATION CELLS

This paper describes an expert system for the supervision of flotation plants based on ACEFLOT, a real time analyzer of the characteristics of the froth that is formed on the surface of flotation cells. The ACEFLOT analyzer is based on image processing and measures several physical variables of the froth, including colorimetric, geometric and dynamic information. On the other hand, the expert system detects abnormal operation states and suggests corrective actions, supporting operators on the supervision and control of the flotation plant.

Fuzzy predictive control of a solar power plant

This work presents the application of fuzzy predictive control to a solar power plant. The proposed predictive controller uses fuzzy characterization of goals and constraints, based on the fuzzy optimization framework for multi-objective satisfaction problems. This approach enhances model based predictive control (MBPC) allowing the specification of more complex requirements. A brief description of the solar power plant and its simulator is given. Basic concepts of predictive control and fuzzy predictive control are introduced. Two fuzzy predictive controllers using different membership functions are designed for a solar power plant, and they are compared with a classical predictive controller. The simulation results show that the fuzzy MBPC formulation, based on a well proven successful algorithm, gives a greater flexibility to characterize the goals and constraints than classical control.

An Expert System for Monitor Alarm Integration

Objective. Intensive care and operating room monitors generate data that are not fully utilized. False alarms are so frequent that attending personnel tends to disconnect them. We developed an expert system that could select and validate alarms by integration of seven vital signs monitored on-line from cardiac surgical patients. Methods. The system uses fuzzy logic and is able to work under incomplete or noisy information conditions. Patient status is inferred every 2 seconds from the analysis and integration of the variables and a unified alarm message is displayed on the screen. The proposed structure was implemented on a personal computer for simultaneous automatic surveillance of up to 9 patients. The system was compared with standard monitors (SpaceLabsTM PC2), using their default alarm settings. Twenty patients undergoing cardiac surgery were studied, while we ran our system and the standard monitor simultaneously. The number of alarms triggered by each system and their accuracy and relevance were compared. Two expert observers (one physician, one engineer) ascertained each alarm reported by each system as true or false. Results. Seventy-five percent of the alarms reported by the standard monitors were false, while less than 1% of those reported by the expert system were false. Sensitivity of the standard monitors was 79% and sensitivity of the expert system was 92%. Positive predictive value was 31% for the standard monitors and 97% for the expert system. Conclusions. Integration of information from several sources improved the reliability of alarms and markedly decreased the frequency of false alarms. Fuzzy logic may become a powerful tool for integration of physiological data.

Dynamic modelling and advanced multivariable control of conventional flotation circuits

Expert and predictive multivariable control algorithms for a conventional copper flotation circuit were assessed through simulations. These simulations were carried out with a nonlinear dynamic model, derived from mass balances and empirical relationships, that qualitatively reproduced the dynamic behaviour of a real plant well. In order to make the simulations more realistic, they included noisy measurements, stochastic parameter variations and input disturbances. New expert algorithms were able to keep the plant operating within a pre-defined zone for long periods without complete control saturation, unlike previous expert controllers. In addition, the inclusion of constraints in a multivariable predictive algorithm verified improved control system regulation and flexibility.

An event-driven simulator for multi-line metro systems and its application to Santiago de Chile metropolitan rail network

Metros are the principal means of public transportation in many of the world’s cities, and continue to grow in the face of rising demand. Expanding metro infrastructure is costly, however, and at a certain point becomes unsustainable. When this occurs the only feasible solution is to improve the train’s management system by using either offline approaches, such as pre-programming schedules which use historic information, or online approaches which employ system status information obtained during operation. A new planning or control system, be it on or off line, requires prior testing that usually involves conducting simulations. This paper presents the design and implementation of an event-driven dynamic simulator for multi-line metro systems, and its practical application for studying different operating strategies. The simulator is based on object-oriented programming and is capable of interacting with Matlab programs written by the user to design and evaluate real-time control strategies. This article describes the model upon which the simulator is based, presents the user interface, and demonstrates how to use the simulator for operating strategies evaluation in the Santiago de Chile multi-line metropolitan rail network.

An integrated system for supervision and economic optimal control of mineral processing plants

This work tackles the problem of dynamically optimising the performance of a mineral concentration plant, taking into account economic profits and technical constraints. The paper proposes a two-level control strategy with regulatory control and the optimisation of an objective function. The regulatory control employs linear model based multivariable predictive controllers with constraints on controlled and manipulated variables, while the optimiser maximises the economic profits using non-linear dynamic models and linear constraints. The results, drawn from simulations, show the proposed strategy to lead to a significant improvement in economic profits when compared against an exclusively regulatory strategy.

Fault tolerant measurement system based on Takagi--Sugeno fuzzy models for a gas turbine in a combined cycle power plant

A fault tolerant measurement system for a gas turbine in a combined cycle power plant, based on dynamic models, principal component analysis (PCA) and Q test, is presented. The proposed scheme makes use of a model-based symptom generator, which delivers fault signals obtained by using direct identification of parity relations and structured residuals. Symptoms are then analyzed in a statistical module achieving fault diagnosis and reconstruction of the faulty signals. The scheme presents as main advantage the ability of detecting faults in both input and output sensors due to its particular structure. Tests carried out on the gas turbine of the San Isidro combined cycle power plant in the V Region, Chile, show that Takagi-Sugeno fuzzy models present the best fitting performance and an acceptable computational cost in comparison with autoregressive exogenous, state space, and neural models. Real time software based on this scheme has been developed and connected to Osisoft PI System^(TM). The software is running at Endesa Monitoring and Diagnosis Center in Santiago, Chile.

Design of fuzzy model based predictive controllers and its application to an inverted pendulum

The paper presents the design and evaluation of a predictive control algorithm based on a fuzzy model that is applied to the angular stabilization of an inverted pendulum. This algorithm is favourably compared with a conventional generalized predictive control, specially for initial conditions further from the desired final state.