Search results for "LM"
showing 10 items of 19289 documents
Online fitted policy iteration based on extreme learning machines
2016
Reinforcement learning (RL) is a learning paradigm that can be useful in a wide variety of real-world applications. However, its applicability to complex problems remains problematic due to different causes. Particularly important among these are the high quantity of data required by the agent to learn useful policies and the poor scalability to high-dimensional problems due to the use of local approximators. This paper presents a novel RL algorithm, called online fitted policy iteration (OFPI), that steps forward in both directions. OFPI is based on a semi-batch scheme that increases the convergence speed by reusing data and enables the use of global approximators by reformulating the valu…
Predictive pumping based on sensor data and weather forecast
2019
In energy production, peat extraction has a significant role in Finland. However, protection of nature has become more and more important globally. How do we solve this conflict of interests respecting both views? In peat production, one important phase is to drain peat bog so that peat production becomes available. This means that we have control over how we can lead water away from peat bog to nature without water contamination with solid and other harmful substances. In this paper we describe a novel method how fouling of water bodies from peat bog can be controlled more efficiently by using weather forecast to predict rainfall and thus, minimize the effluents to nature. peerReviewed
Patented intelligence: Cloning human decision models for Industry 4.0
2018
Industry 4.0 is a trend related to smart factories, which are cyber-physical spaces populated and controlled by the collective intelligence for the autonomous and highly flexible manufacturing purposes. Artificial Intelligence (AI) embedded into various planning, production, and management processes in Industry 4.0 must take the initiative and responsibility for making necessary real-time decisions in many cases. In this paper, we suggest the Pi-Mind technology as a compromise between completely human-expert-driven decision-making and AI-driven decision-making. Pi-Mind enables capturing, cloning and patenting essential parameters of the decision models from a particular human expert making …
Comparison of electropolishing behaviours of TSC, ALM and cast 316L stainless steel in H 3 PO 4 /H 2 SO 4
2017
Abstract In recent decades, new manufacturing processes have been developed such as Thermal Spray Coating (TSC) and Additive Layer Manufacturing (ALM), which reduce or avoid machining of parts with complex geometries. This study aims to develop an Electropolishing (EP) process for TSC and ALM 316L Stainless Steel (SS). EP is an anodic dissolution process currently used in industry to reduce surface roughness and obtain a bright and smooth finish. The EP mechanism was studied, in a mixture of H3PO4 and H2SO4, for TSC, ALM and “cast” SS in order to determine the best conditions (time, temperature, potential). Special attention was paid to surface characterization by combining several techniqu…
A new approach to simulate coating thickness in cold spray
2020
Abstract In the process of cold spray on complex components, the coating thickness is an important indicator to monitor and control. Current methods such as destructive tests or direct mechanical measurements can only be performed after spraying. Besides, these methods lead to production shutdown and additional costs . This article presents a novel approach predicting coating thickness for components with complex curved surfaces, especially in the case of shadow effects. Firstly, a three-dimensional geometric model of the coating profile based on Gaussian distribution was developed. In addition, the relative deposition efficiency (RDE) resulting from the different robot kinematic parameters…
Optimisation of refractory coatings realised with cored wire addition using a high-power diode laser
2005
Laser; Cladding; Refractory alloys; Factorial experiments; International audience; The objective or our research was to obtain refractory alloys using the high-power diode laser (HPDL) coating technique. After optimisation using factorial experiments, two different cladding regimes were clearly distinguished. It was also shown that a very narrow transition zone exists between the two regimes, and, inside this zone, clad layers having a satisfactory compromise between the response functions (surface aspect and cavity presence) were obtained. The main objective of our study, namely, the control of the operating parameters (geometrical and kinematical) to realise adequate coatings, without cav…
JOINT TOPOLOGY LEARNING AND GRAPH SIGNAL RECOVERY VIA KALMAN FILTER IN CAUSAL DATA PROCESSES
2018
In this paper, a joint graph-signal recovery approach is investigated when we have a set of noisy graph signals generated based on a causal graph process. By leveraging the Kalman filter framework, a three steps iterative algorithm is utilized to predict and update signal estimation as well as graph topology learning, called Topological Kalman Filter or TKF. Similar to the regular Kalman filter, we first predict the a posterior signal state based on the prior available data and then this prediction is updated and corrected based on the recently arrived measurement. But contrary to the conventional Kalman filter algorithm, we have no information of the transition matrix and hence we relate t…
Time-varying Sampled-data Observer with Asynchronous Measurements
2019
International audience; In this paper a time-varying observer for a linear continuous-time plant with asynchronous sampled measurements is proposed. The observer is contextualized in the hybrid systems framework providing an elegant setting for the proposed solution. In particular some theoretical tools are provided, in terms of LMIs, certifying asymptotic stability of a certain compact set where the estimation error is zero. We consider sampled asynchronous measurements that occur at arbitrary times in a certain window with an upper and lower bound. The design procedure, that we propose for the selection of the time-varying gain, is based on a constructive algorithm that is guaranteed to f…
Vibration control strategy for large-scale structures with incomplete multi-actuator system and neighbouring state information
2016
The synthesis of optimal controllers for vibrational protection of large-scale structures with multiple actuation devices and partial state information is a challenging problem. In this study, the authors present a design strategy that allows computing this kind of controllers by using standard linear matrix inequality optimisation tools. To illustrate the main elements of the new approach, a five-story structure equipped with two interstory actuation devices and subjected to a seismic disturbance is considered. For this control setup, three different controllers are designed: an ideal state-feedback H 8 controller with full access to the complete state information and two static output-fee…
Evaluation Framework for Analyzing the Applicability of Criteria Lists for the Selection of Requirements Management Tools Supporting Distributed Coll…
2016
Effective requirements management and enabling tools are critical for successfully developing and maintaining services and products. The identification and selection of an appropriate requirements management tool can be a costly, time-consuming, and error-prone undertaking especially in the context of software product line requirements management, requiring the tools to support both product and platform development activities that often involve geographically distributed, collaborating, and competing stakeholders. Criteria lists have been developed to facilitate the selection. This research (1) creates an evaluation framework to review the applicability of the lists for the selection of req…