Adama Traoré

Control of sludge height in a secondary settler using fuzzy algorithms

During the last years, the efficiency of pollution removal and sludge treatment processes are major concerns because of the more restrictive legislation related to nutrient discharge. A key point in the field of activated sludge wastewater treatment plants (WWTPs) is the secondary settler efficiency. In this paper, the COST 624 Simulation Benchmark (model and data) is used to develop a fuzzy controller of the sludge height in the secondary settler. The presented control strategy is based on simple on-line data (influent, removal and recycle flows) and daily analytical values of the sludge volume index (SVI) allowing the fuzzy algorithm to reduce sludge height variations and thus to increase the settling process efficiency. The developed controller has then been adapted and applied to the Cassa de la Selva activated sludge WWTP (Spain) model, with satisfying results.

Fault detection in a wastewater treatment plant

The wastewater treatment plants are very unstable and the waters to be treated are ill defined. The command of those processes needs to be done with advanced methods of control and supervision. Those methods have to take into account the bad knowledge of the processes. The most important situations that can imply problems for the plant are listed. The parameters that allow the comparison between the crisis situation and normal one are measured. The fuzzy logic method is used to distinguish the different situations in order to take a decision for the working of the plant. The method is tested in simulation and on a real plant.

Modeling of a geothermal and thermal solar process

Abstract This paper is based on a project combining solar and geothermal energies used for heating an individual home located in a mountainous region of southern France (Saint-Pierre Dels Forcats town). A solar field consists of 10 thermal solar collectors, with a total area of 21.25 m 2 which recharge thermally a borehole of 100 m of depth. A heat pump extracts the heat energy to restitute it to the home with a heating floor of 100 m 2 . The long-term objective is to develop a predictive control on the global system to realize financial and energy savings. In order to do this, a modeling of all the subsystems has to be done. This paper focuses of two of these sub-systems, the solar field and the heat pump. Depending on the level of knowledge of each subsystem, a physical model (thermal solar collectors) or a artificial intelligence methodology (heat pump) is used.

Classification of Abnormal Situations in a Waste Water Treatment Plant

The waste water treatment plants are very unstable and the waters to be treated are ill defined. The command of those processes needs to be done with advanced methods of control and supervision. They have to take into account the bad knowledge of the processes. The most important situations that imply problems for the plant are listed. Parameters for comparison between crisis and normal situation are measured. Fuzzy logic and support vector machines methods are used and compared to distinguish the different situations in order to take a decision. The method is tested in simulation and on a real plant.