through (inter-) actions along the organisational processes within the framework of the organisational environment. Argyris and Schön conceptualised organisational learning processes in the sense of a theory of action (Argyris ). Organisational learning then is, to discover problems and mismatching situations, to correct them, and to change the organisational knowledge base in a way, that reflect new problem solving and action competences. Learning processes enhance the organisational knowledge, which itself may be represented in the minds of organisational members (normally only a certain part per employee) or stored in the organisational memory. Earlier in this paper we already stated the importance for knowledge management of making implicit knowledge explicit and available for the organisation and of possibly storing it in the organisational memory (system) as external knowledge. In this section we start by identifying basic organisational learning cycles. Through appropriate combinations of such learning cycles more complex learning scenarios in an organisation can be described. Especially, important known organisational learning types are covered by this approach, including single-looplearning and double-looplearning (Argyris , , Vlismas ).
The second difficulty encountered is the lack of scaffolding for the transfer between the single and doublelooplearning cycle. Within the singlelooplearning cycle, trainees are re- quired to ask: “this is what happened” -> “this is what might work to deal with it” -> “was my intervention effective?” However, the ‘working with meaning’ is more complex be- cause in order for trainees to transfer to the doubleloop learn- ing cycle, a second further removed past tense is required. Par- ticipants need to ask themselves what they have learned from the experience. This requires a stepping back from the experi- ence and rediscovering the learning event as vividly as possible so that a re-evaluation of that experience can take place. This is then followed by linking this “emergent knowing” to a “new paradigm” or new belief system constructed based on reflection on action. This “new paradigm” should seek to understand the ex- perience by considering much broader issues than those found in the classroom such as community factors and ethical, moral or political issues. It is evident that this is a complex process of reflection. To scaffold the procedure, a stepping stone was provided bridging the single and doubleloop learn-
Double-looplearning requires people to ask questions about the reasons and motives behind the policy (Argyris, 1994). Therefore, an education system that suppresses teachers’ feelings and complaints or that does not take the underlying causes of the problems into a reexamination and yet aims to achieve goals is a singleloop system. May be the recently introduced “Ethiopian Education Development Roadmap”, if approached from a doubleloop model of cation, will come up with a long term solution to the problem.
In this section, the authors argue how Argyris and Schön’s formalisms can be applied within the context of participatory design for subjective well-being. As already mentioned, we frame happiness as a function of human adaptation within self-organizing design activities. Conversely, a number of researchers and thinkers have argued that the ability to be happy and content with life is a central criterion of adaptation (e.g., Diener, 1984; Jahoda, 1958; Taylor & Brown, 1988). Findings from positive psychology illustrate the importance of intentional activities (Lyubomirsky, Sheldon & Schkade, 2005) that comprise a balance between skills and challenges (Csíkszentmihályi, 1990). Each participatory design activity (making, using or learning) can be perceived as a meaningful activity that challenges the skills of all engaging participants. Making products together can be framed as finding the edges of each other’s physical, mental and emotional potential through incremental ‘single-loop’ adaptations. As long as the governing variables stay within their limits, the same ingredients (attitudes, be-goals and standards) are challenged and optimized within a ‘single-loop’. However, once conflicts arise between the physiological edges and the governing variables, a ‘double-loop’ learning cycle is triggered. As a result, one or more ingredients change, which results in new adaptation strategies. In practice, the group adapts or changes its belief system and perceives its goals, skills or values from a whole new perspective through the interaction with the environment.
Paradoxes can be described as problems with two extreme solutions, where both solutions are true. Thus the conflict between innovative learning (double-looplearning), in which both the assumptions and the standards/strategies are modified, and routine learning (single-looplearning), which concerns only the action strategies (behaviours), constitutes a potential paradox. One can presume that in an efficient crisis management the most helpful cycle would consist of single-looplearning, while learning through the crisis, or learning to avoid crisis or fostering deep changes within the enterprise would require double-looplearning. Innovative learning is a radical change in methods of operation together with a change in the objectives (or standards) and the premises of the action. It can be assumed that single-looplearning, which is based on the ability to detect and correct errors with a given set of operating standards, is the most useful for effective crisis management. In the case of double- looplearning, both the standards and the basic assumptions are modified. Such learning is conducive to innovation, challenging goals, and it is more important for long-term survival [2; 63-69]. It entails the need to test the cognitive models.
Elements with linear phase response can be used to improve the antenna bandwidth. Linearization of phase response may be done in several ways including using a thick substrate, multiple stacked patches , and phase-delay lines , etc. However, these designs will lead to other shortcomings such as additional fabrication complexity, increased weight, and higher loss. Any attempt at broadening the bandwidth of the reflectarray should not solely rely on the thick substrate and multiple stacked patches but should also exploit the rich literature on the single-layer structure. In recent years, many single-layer elements have attracted much more attention due to their low-cost manufacturing and other advantages. Many novel designs of the single-layer structure have been proposed for enhancing the reflectarray bandwidth, including the multi cross loop elements , microstrip elements composed of rectangular patch and rectangular ring .
techniques, e.g. electron microscopy [13,14] and computer aided image analysis , loop models enabled a more precise specification of the loop shape and size, depending on the geometrical parameters of the loop. Contemporary computer graphics have been lately used to additionally illustrate the knitted loop geometry [16,17]. In addition to mechanical and energy loop models, which interpret the behavior of knitted structures under loading, the geometrical loop models have recently returned to focus. Geometrical loop models assist finding the effective parameters which cause dimensional changes during relaxation, enable planning the production of a piece of fabric before knitting, assist designing knitted structures for technical applications, obtain computer simulations of knitted structures for design or fault-assessing purposes, help creating a physical loop model, help evaluating curling etc .
The two most important design criteria for a transmitarray element are its phase range and transmission magnitude. Firstly, in order for an element design to be suitable for a large array where there is possibly phase shifting range in the aperture ﬁelds, a phase tuning range of 360 ◦ is required. Secondly, the transmission magnitude needs to be close to 1 (0 dB) to ensure a high aperture eﬃciency. Thus, the design process must involve both the transmission phase shift and magnitude of the element. In addition, the severe drawback of the microstrip transmitarray is its limited bandwidth performance, which is usually 7% or less, and intense eﬀorts have been made in recent years to overcome this shortcoming [1, 4, 5, 7] and . In this paper, we present a novel transmitarray operating at 9.5 GHz which uses a double-petal loop as the unit cell element and oﬀers a wider bandwidth than previously achieved ones.
The EIS data for BIM-NH at a concentration of 10 ppm is shown in Figure 10. From 3 to 11 h the Nyquist plot (Figure 10a) shows two time constants. At HF a capacitive loop that increases with time due to the inhibitor adsorption on the metal surface. This capacitive loop represents the active state of the metal surface. At LF the adsorption of the inhibitor molecules is manifested by the appearance of an inductive loop by the mechanism described in section 3.2.1. After 14 h of exposure the adsorption of the inhibitor molecules is no longer the controlling process and a charge transfer control process took over. A more capacitive behavior is observed by an increase in the phase angle shown in Figure 10b, suggesting the continuous formation of an inhibitor film. A non-well defined second time constant evolved at HF after 14 h, attributed to the inhibitor film. The Bode plot of phase angle vs. frequency shows a slight asymmetry that implies a second time constant . The presence of an inhibitor film was also supported by the two negative slopes observed in the Bode plot of modulus of impedance (|Z|) vs. frequency .
Multimedia learning systems have however focused largely on the visual and auditory senses to the exclusion of others like haptics and olfactory senses, thereby, reducing the additional advantages that focusing on these extra senses can deliver. Technology use in education had evolved over the years, and it is beginning to focus on Virtual Reality (VR) technology due to its ability to support immersive and collaborative learning (Christou, 2010; Monahan, McArdle, & Bertolotto, 2008) and teaching through simulation (Hamid, Aziz, & Azizi, 2014; Mujber, Szecsi, & Hashmi, 2004), and gaming (Choi, Jung, & Noh, 2015). VR is considered the next frontier of computer interaction (Fildes, 2015); it is employed in various types of training programmes (Farra, Miller, & Hodgson, 2015; Yiannakopoulou, Nikiteas, Perrea, & Tsigris, 2015), and it is specifically useful for supporting effective approaches to learner engagement and motivation (Buckley & Doyle, 2016; Kuo & Chuang, 2016). In fields like chemistry, where learners have to engage spatial skills (Dünser, Steinbügl, Kaufmann, & Glück, 2006; Hauptman, 2010), VR can deliver outstanding advantages as educational technology and/or learning environment. In particular, it can support the visualization of abstract concepts like atoms, molecules, bonds, and others within an immersive environment.
Abstract—Stochastic distribution control (SDC) systems are known to have the two-dimensional characteristics regarding time and probability space of a random variables, respectively. A double closed-loop structure, which includes iterative learning (IL) modeling (ILM) and iterative learning control (ILC), is proposed for non-Gaussian SDC systems. The ILM is arranged in the outer loop, which takes a longer period for each cycle termed as a BATCH. Each BATCH is divided into a modeling period and a number of control intervals, called batches, being arranged in the inner loop for ILC. The output probability density functions (PDFs) of the system are approximated by a radial basis function neural network (RBFNN) model, whose parameters are updated via ILM in each BATCH. Based on the RBFNN approximation of the output PDF, a state-space model is constructed by employing the subspace parameter estimation method. An IL optimal controller is then designed by decreasing the PDF tracking errors from batch to batch. Model simulations are carried out on an 4th-order numerical example to examine the effectiveness of the proposed algorithm. To further assess its application feasibility, a flame shape distribution control simulation platform for a combustion process in a coal-fired gate boiler system is constructed by integrating WinCC interface, Matlab simulation programs and OPC communication together. Simulation study over this industrial simulation platform shows that, the output PDF tracking performance can be efficiently achieved by this double closed-loop IL strategy.
Solar still, which converts available brackish or impure water into potable water, can be used to supply drinking water for the people living in arid and remote areas. But, this still is not popular because of its lower productivity. This research work presents the theoretical and actual performance of the single basin double slope solar still and explores the different methods to improve the efficiency. A single basin double slope still with overall size of 2.5 m x 1.5m x 1 m was fabricated and tested under laboratory conditions (still – laboratory) and in actual solar radiation conditions (still – solar) at Chidambaram (9o11’N, 77o52’E), a city in southern India. Experiments were conducted for different depths of water up to 0.2 cm and with different basin materials. Different wick materials like light cotton cloth, jute cloth and light sponge sheet were used. Aluminium rectangular fins arranged in length and breadth wise covered with cotton cloth and jute cloth were also used in the basin. The solid materials like quartzite rocks of different sizes, naturally washed stones, cement concrete pieces and, brick pieces and iron turnings were used in the basin. The above said solid materials were not used so far. Experiments were conducted using different glass thickness with different inclination and orientation and the variations in transmittance were studied. A regression equation was established, to calculate the transmittance of the given thickness glass plate at any place and at any time for given radiation conditions. The variations of energy losses at cover plates were studied.
Self-balancing robot  developed as early as 1986, originated in Japan, the initial idea is to design a machine that can stand automatically. In 1986, Professor K.Y. of Nippon Telekom University conceived an automatic standing robot , but the robot can only move forward on a fixed track. In 1996, Y.S.H. and S.Y. of Tsukuba University designed and implemented a two-wheeled robot. It uses a three-stage closed-loop control system to control the autonomous cruising of the robot in a plane . In 2002, F.G. developed the JOE . It can walk freely on slopes due to its centralized structure. In 2005, Professor R.H. at Carnegie Mellon Institute of Robotics and his team successfully developed the first ball-wheel robot, and named BALLBOT . In 2008, The Murata girl was introduced in the Japan Robot Exhibition . It maintains the lateral balance by turning the flywheel equipped in the robot body. In 2010, the REZERO was introduced by University of Zurich , which using the LQR control algorithm. This is a representative of the spherical self-balancing robot.
Due the large number of configurations used in these experiments, we will describe only the signal in the QPDSF configuration. The configuration in Figure 1(f) showed that each turn consistently gave a signal attenua- tion of 12 dB, this loss is due to three circulators, two TBF filters, and two WDMs. Thus, the amplified signal will propagate through the CIR1 from port1 to port2 then travel through EDF1; the signal will be affected by the first amplification from EDF1, through port2 into port3 of CIR2, passing through the first TBF1 filter into port1 and back to port2 to be amplified during the second pass by EDF1 into port2 of CIR1, and therefore will propa- gate again in the second stage through EDF2, CIR3, and TBF2 for the third and fourth passes. The output signal power was displayed through the OSA from port4 of CIR1. Traveling from port1 to port4 of CIR1, the signal will be affected by four amplifications during the four passes, or, as we mentioned, the quadruple pass double stage with filter configuration .
process that has been implemented by the interpreter is using conventional learning strategy. A simple learning strategy learner provides and distributes sub syllabus consisting of various subject matter to be discussed and discussed, both individually and in groups. The success and diversity of learning achievement of interpretation is closely related to several things, including the implementation of learning strategies implemented and also on the level of student learning independence. The conventional learning strategy emphasizes unidirectional lectures and discussions. A one-way discussion is that students ask lecturers and lecturers to respond to questions posed by learners. There is almost no intensive communication between learners and other learners on the subject being discussed. Thus, there is never an error of information between learners with one another. This is one of the triggers of low learning achievement of learners in the field of tafseer.
ture with the “double-banana” as an example (Cheng et al., 2009), and Rojas proposed the “double-banana” closure con- ditions as a paradigm for position analysis in robots (Rojas and Thomas, 2013). The cause for the “double-banana” para- dox is that there exist “isolated” DOFs in the mechanism (e.g. the bar spin between two spherical joints) which create an own level of transmission kinematics, and which thus need to be tracked in order to be able to detect which substructures are rigid and which are movable. We propose in this paper a novel procedure for tracking such isolated DOFs by us- ing an alternative method for describing kinematical interde- pendencies, called the “kinematical network” (Kecskeméthy, 1993), in which the mechanism is regarded not as a sys- tem of connected bodies, but as a system of interconnected kinematical loops. A kinematical loop can be regarded as a closed chain of bodies which corresponds to an elementary cycle in a graph theoretical sense. If the underlying graph of the mechanism is not at least 2-edge-connected, i.e. it con- tains bridges or is separated, the graph first is dissected into its 2-edge-connected components, called clusters, for which the loop decomposition algorithm is applied separately. Of course, rigidity needs to be detected only within each clus- ter, as non-connected or 1-edge-connected clusters can never reduce the DOF in comparison with the sum of their inter- nal DOFs. In the following, the method of “kinematical net- work” is briefly reviewed here for better understanding of the present paper, and then our novel rigidity-detection proce- dure for “double-banana” types of mechanisms is elucidated.
When 0 dbm input power is applied then there are different Q factors values observed at different transmission distance for single and cascaded modulators in fig 3.4. It is clear that double MZM modulator gives best performance among them. And travel maximum achievable distance with acceptable Q value. Among them single EAM give the worst performance it travel only upto 1800 Km with acceptable limit. Single MZM and Double EAM gives approximately same performance.
scientists. The task was to label sentences with Communication which implies some form of com- munication between the two parties involved in the contract. The training data consists of 28, 174 sen- tences extracted from 149 IBM procurement con- tracts and the held out, test data consists of 1259 sentences extracted from 11 non-IBM procure- ment contracts. The initial set of 188 weighted linguistic expressions learned using deep learning performed at 67% F1 (the harmonic mean of pre- cision and recall) on the test set. Note that, as mentioned in Section 3.2, weights can lead to lack of explainability. Thus the data scientist’s task is to use HEIDL to generate from this initial set of linguistic expressions, a smaller set expressed in pure first-order logic that achieves maximal per- formance on the out-of-domain test set. For each data scientist, we initialize HEIDL with the initial linguistic expressions, the training set sentences, and the corresponding predicates and dictionaries. Our baseline is a well established sentence clas- sification neural network, based on bi-directional long short-term memory (LSTM). More precisely, the LSTM replaces the tokens in the input sentence with their corresponding 300-dimensional GloVe embedding, computes an intermediate hidden state which are then max-pooled, fed into a fully con- nected layer and ReLU activation before passing it through sigmoid activation to get a probability of predicting the label. To improve training of the LSTM, we employed a variety of dropout regular- izations: variational dropout after the embedding lookup layer, weight dropout in the LSTM layer, and dropout in the fully connected layer.
All these results are proof of the versatility that HIL offers in the implementation and emulation of plants. Reason why, this alternative to control nonlinear systems like double inverted pendulum, whose physical mechanism is difficult to build, was chosen. Among the papers found with this system, publications from  and  can be highlighted, where they control the angular and spatial position of the bars and mobile, on which is supported, with a LQR controller. In the same way, classic control systems along with neural networks have been used for estimate the controller gains and achieve the stabilization of the system . Another example is the work performed in , where concepts related to artificial intelligence were used for controlling a similar system, achieving a real-time control, with more robustness and stability than using conventional methods.
The aim of this work is to study the evolution of the electrode surface during the single and double reactivation methods at microscopic scale. The tested material has been a duplex stainless steel (UNS 1.4462) in its as-received state and heated at 825 ºC for 1 hour in an inert atmosphere. Tests have been carried out in an electrochemical minicell that can be put in the stage of a confocal microscope. These devices allow the in-situ observation of the electrode surface at microscopic scale during the tests. Differences in the evolution of the electrode surface have been observed between the single and the doubleloop methods. Furthermore, the sensitised sample shows different evolution during the electrochemical tests than the as-received sample. These differences are due to the formation of new phases during the heat treatment. Therefore, the minicell permits obtaining additional information during the electrochemical characterisation of the degree of sensitisation in duplex stainless steels.