In this work a tertiary horizontal-flow blast furnace slag (BFS) filter was tested for phosphorus removal under laboratory and field conditions. Laboratory experiments were conducted in a bench-scale BFS filter for phosphorus adsorption capacity at equilibrium conditions using P-spiked water. On-site, the performance of a pilot-scale BFS filter was monitored for phosphorus removal from the final effluent of a pilot-scale WSP system located in Esholt (Bradford, UK), which comprises a primary facultative pond and a secondary aerated rock filter (ARF) in series. Adsorption capacity (q) results showed that BFS has high affinity for inorganic P species and it can remove up to 30 g P/kg BFS; however, q values are strongly dependent on the initial P concentration. The pilot-scale BFS filter tested on-site performed better for longer (<2 mg P/L) than the laboratory-scale filter, even though despite the latter received the same hydraulic and P loadings (1.8 m3/m3 d and 18 g P/m3 d, respectively). Both the laboratory and field results showed that BFS filters are an appropriate low-cost technology to upgrading small wastewater treatment systems for phosphorus removal.

Tracer experiments are of concern to wastewater treatment engineers and researchers because of the importance of determining hydraulic regimes and retention times in wastewater treatment units. In this work, a pilot-scale maturation waste stabilisation pond (WSP) was spiked with Rhodamine WT, in order to determine how suspended organic matter would interfere with its performance as a tracer in a domestic wastewater treatment unit which had a high content of suspended algal biomass. A primary maturation pond was spiked in three separate runs with different levels of algae (high, medium and low), with a known amount of Rhodamine WT (20% w/v); the tracer was measured in the pond effluent in real time every 20 min for 3? (the theoretical retention time, ? = 17 days). Algal biomass was monitored weekly from influent, column and effluent water samples by chlorophyll-a determination. The results show that algal biomass has a strong influence on the behaviour of Rhodamine WT as a tracer and therefore the hydraulic characteristics calculated from tracer curves may be affected by tracer adsorption on suspended organic matter.

IEEE 802.15.4e is emerging as a robust standard for industrial Internet of Things. This standard is particularly effective for low power, multi-hop networks, and typically supports applications with low throughput. However, with the advances in energy harvesting, the need for supporting better performing solutions is also gaining momentum. This paper presents and analyses a modified link scheduling algorithm for IEEE 802.15.4e based solutions. Three approaches are considered. With the first approach, only the gateway node is equipped with multiple radios. With the second approach, non-gateway nodes are equipped with multiple radios as well. With the third approach, multiple single radio gateways are considered. Results from simulations have been presented to demonstrate the improvement in performance compared to the typical single radio, single gateway deployments.

In this work, an hybrid, self-configurable, multilayered and evolutionary subsumption architecture for cognitive agents is developed. Each layer of the multilayered architecture is modeled by one different Reinforcement Machine Learning System (RMLS) based on bio-inspired techniques. In this research an evolutionary mechanism based on Gene Expression Programming to self-configure the behaviour arbitration between layers is suggested. In addition, a co-evolutionary mechanism to evolve behaviours in an independent and parallel fashion is used too. The proposed approach was tested in an animat environment (artificial life) using a multi-agent platform and it exhibited several learning capabilities and emergent properties for self-configuring internal agent’s architecture.

In this work, a hybrid, self-configurable, multilayered and evolutio-nary architecture for cognitive agents is developed. Each layer of the subsump-tion architecture is modeled by one different Machine Learning System MLS based on bio-inspired techniques. In this research an evolutionary mechanism supported on Gene Expression Programming to self-configure the behaviour arbitration between layers is suggested. In addition, a co-evolutionary mechan-ism to evolve behaviours in an independent and parallel fashion is used. The proposed approach was tested in an animat environment using a multi-agent platform and it exhibited several learning capabilities and emergent properties for self-configuring internal agent’s architecture.

Flip chip interconnection design and process capability has been extended with the use of copper pillar, specially for finer pitches, beyond what was possible with area array and standard C4 reflow and solder bumps. Many trends in device packaging are fueling this trend, specially the needs for thin packaging, increased function integration, challenging thermal, mechanical and physical specifications. In terms of processing, devices using Cu Pillar are most commonly assembled using thermo-compression bonding due to sizable challenges in extending the conventional mass reflow solution, typical for solder bumps. TC is specially recommended for handling very thin die in sparse or peripheral bump layouts, which are a predominant share of all devices converting to Cu pillar, where a bonding head is used to both hold the die flat and in true alignment with the substrate while supplying the thermal energy necessary to complete the interconnection. This process most often also requires the use of a non-conductive paste/film for stress absorption in recognition to the fact that underfill materials have great limitations in filling the underdie cavity. In this paper, we explore avenues to extend the current processing envelope for Cu pillar bumped devices using C4 reflow. Benefits of this approach are shown as decreased cost of ownership for materials, lower fixed costs due to using mostly depreciated equipment. Yet, challenges in defining new design rules, reformulating materials and extending process flows are among some of the barriers for further adoption of this solution in production.

Miel I is a 188 m high RCC dam in Colombia, at elevation 454 m a.s.l. The RCC mix has a cement content of 85 to 160 kg/m3. To meet contractual schedule, the original design of an upstream face made of slip formed reinforced concrete was changed to a Carpi drained exposed composite Polyvinylchloride geomembrane (= PVC geocomposite) system, placed on a 0.4 m thick zone of RCC material enriched with cement grout and vibrated. Placement of the geocomposite system in horizontal stages allowed starting of dam impoundment while RCC placement was still ongoing at higher elevation. The geocomposite system was completed in 2002, before finishing of complementary concrete works. ISAGEN, the owner of the dam, constantly monitors the dam’s and geocomposite system’s performance. The average leakage of the dam registered since the first impoundment of the reservoir has been of 2 l/s for the geocomposite’s drainage system, and of 25 l/s from the abutments. Such levels are below the design parameters that defined the maximum acceptable leakage through the geocomposite in 9.7 l/s and the maximum acceptable leakage through the abutments in 30 l/s. The measured leakage has never surpassed the historical values (with stable lecture through 11 years), or overcome the maximum allowed drainage design values defined during the design stage. After 11 years of good performance and operation, in summer 2013 ISAGEN decided to have a comprehensive monitoring of the general conditions and the weathering behaviour of the waterproofing geocomposite at Miel I. The objectives of the investigation were: assessing the reason of small cuts that had been detected and repaired (in total 15 in 11 years on a 31,000 square meters surface), assessing the conditions of the PVC geocomposite, ascertaining at which extent its characteristics had changed over service, and evaluating if the weathering behaviour was in accordance with the expected durability. Investigation started with inspection at the upstream face and in the drainage galleries, followed by sampling of the exposed PVC geocomposites. Testing in an independent laboratory allowed comparing the physical and chemical properties of the aged geocomposite sample collected at the dam with the physical and chemical properties of the same geocomposite as manufactured. Evaluation of test results, based also on most recent research on geomembrane weathering mechanisms, completed the investigation. Results confirmed that the behaviour of the PVC geocomposite is extremely good and fully in line with expectations. The paper, after a brief outline of the geocomposite system installed at the dam, will describe all steps of the investigation, and detail the test results and the conclusions drawn.

In this work, an hybrid, self-configurable, multilayered and evolutionary subsumption architecture for cognitive agents is developed. Each layer of the multilayered architecture is modeled by one different Machine Learning System (MLS) based on bio-inspired techniques such as Extended Classifier Systems (XCS), Artificial Immune Systems (AIS), Neuro Connectionist Q-Learning (NQL) and Learning Classifier Systems (LCS) among others. In this research an evolutionary mechanism based on Gene Expression Programming (GEP) to self-configure the behaviour arbitration between layers is suggested. In addition, a co-evolutionary mechanism to evolve behaviours in an independent and parallel fashion is used. The proposed approach was tested in an animat environment using a multi-agent platform and it exhibited several learning capabilities and emergent properties for self-configuring internal agent’s architecture.

Power conversion systems are becoming fundamental with the increasing power generation from renewable energy sources due to global warming and energy security concerns. In particular, wind energy systems require advance power converter devices to be able to supply power to the electric network or remote isolated areas in an effective way. A practical approach to these devices is required to fully comprehend their principles of operation and how they are integrated with the expanding renewable electricity production. This thesis describes the design and construction process of an inverter and a DC-DC converter, part of the AC-DC-AC link required to connect a laboratory model of a variable speed wind power system to the grid or to an isolated load.

Actualmente los sistemas de control de riego SCADA son diseñados con modelos y protocolos que dependen exclusivamente de cada fabricante. Si se requiere una interoperabilidad a nivel de programación de riegos y estado real de actuadores y sensores, con algún sistema de gestión que no sea del mismo fabricante, significa una alta inversión económica y tecnológica, que en la mayoría de casos no es posible realizar. Adicionalmente los mecanismos de diagnóstico de fallos actuales, no permiten determinar los fallos que se presentan en los sensores, lo cual se vuelve crítico cuando se presenta una anomalía en el sistema y no se puede cumplir con una demanda de riego exigida. La finalidad principal de esta tesis es contribuir a la estandarización de los sistemas de control de riego y a la integración de mecanismos de diagnóstico de fallos avanzados. Para ello y partiendo de un caso de estudio de un sistema de control de riego piloto, la memoria se encuentra constituida por las siguientes tres partes: En la primera, se propone un modelo de procesos que pretende estandarizar los sistemas de control de riego y su interoperabilidad con sistemas de gestión. Todo esto bajo las normativas ISA-88 e ISA-95 y la especificación XML del ISA-88: Batch-ML. En este punto la tesis trata la descripción de los modelos físicos, de procedimientos y de procesos de sistemas de control de riego, por medio del lenguaje de marcas extensibles (XML), centrándose principalmente en la programación de órdenes de riego y en la determinación en tiempo real del estado de los dispositivos del sistema. Para validar dicha descripción, en la segunda parte se presenta la implementación de un software prototipo que supervisa y controla, por medio del protocolo industrial OPC, un sistema de control de riego piloto que actualmente pertenece al grupo de investigación CEPIMA (Centre d´Enginyeria de Processos i Medi Ambient) de la UPC, el cual colabora como socio académico del grupo de empresas públicas TRAGSA. Finalmente en la tercera parte, se realiza un estudio del diagnóstico de fallos del sistema - caso de estudio, partiendo de la problemática que este tipo de sistemas conlleva, al sólo poder determinar automáticamente los fallos que se presentan en las válvulas más no en los sensores de caudal existentes en el sistema. Por lo tanto se plantea utilizar una técnica de diagnóstico de fallos a nivel estructural, que permite detectar y aislar los fallos desde el análisis del comportamiento del mismo sistema. Todo esto desde el enfoque FDI.