This work is aimed at optimizing the wind turbine rotor speed setpoint algorithm. Several intelligent adjustment strategies have been investigated in order to improve a reward function that takes into account the power captured from the wind and the turbine speed error. After different approaches including Reinforcement Learning, the best results were obtained using a Particle Swarm Optimization (PSO)-based wind turbine speed setpoint algorithm. A reward improvement of up to 10.67% has been achieved using PSO compared to a constant approach and 0.48% compared to a conventional approach. We conclude that the pitch angle is the most adequate input variable for the turbine speed setpoint algorithm compared to others such as rotor speed, or rotor angular acceleration.

With the increase of wind power capacity worldwide, researchers are focusing their attention on the operation and maintenance of wind turbines. A proper pitch controller must be designed to extend the life cycle of a wind turbine's blades and tower. The pitch control system has two main, but conflicting, objectives: to maximize the wind energy captured and converted into electrical energy and to minimize fatigue and mechanical load. Four metrics have been proposed to balance these two objectives. Also, diverse pitch controller strategies are proposed in this paper to evaluate these objectives. This paper proposes a novel metrics approach to achieve the conflicting objectives with a maintenance focus. It uses a 100 kW wind turbine as a case study to simulate the proposed pitch control strategies and evaluate with the metrics proposed. The results are showed in two tables due to two different wind models are used.

Abstract. The introduction of energy efficiency as a new goal into already complex production plans is a difficult challenge. Decision support systems can help with this problem but these systems are often resisted by end users who ultimately bear the responsibility for production outputs. This paper describes the design of a decision support tool that aims to increase the interpretability of decision support outputs. The concept of 'grey box' optimisation is introduced, where aspects of the optimisation engine are communicated to, and configurable by, the end user. A multi-objective optimisation algorithm is combined with an interactive visualisation to improve system observability and increase trust.Keywords: visualisation, optimisation, energy efficiency, manufacturing. IntroductionEnergy efficient manufacturing is a key research challenge for both industry and academia. Systemic energy waste is closely tied to strategic production decisions and therefore poses a complex operations-research problem. An example of this involves switching idle machine into a low-power mode. While this strategy is an effective way to save energy, it is not a straightforward task in many industrial environments. Energy savings are often subservient to production targets and decisions about changing machine states involve weighing up a complex set of goals and constraints. These include hard metrics such as production capacity, predicted inventory and product priorities as well as soft constraints such as technician skill level, engineering requests and machine recovery risks. Operations managers currently apply human expertise to cope with this complexity. In high product mix factories this problem can become very challenging and even before energy-saving is considered. Optimisation algorithms can be applied to reduce the problem space associated with this decision and to highlight energy saving opportunities; however an algorithmic approach is challenged by soft constraints and unpredictable changes in goals. In addition operations managers tend to be wary of decision support tools due to their perceived brittleness and lack of transparency [1].

No abstract