Real-­time parametric estimation of periodic wake-­foil interactions using bioinspired pressure sensing and machine learning

Periodic wake-­foil interactions occur in the collective swimming of bio­inspired robots. Wake interaction pattern estimation (and control) is crucial to thrust enhancement and propulsive efficiency optimization. In this paper, we study the wake interaction pattern estimation of two flapping foils in tandem configurations. The experiments are conducted at a Reynolds number of 1.41×10^4 in a water channel. A modified wake-­foil phase parameter Φ, which unifies the influences of inter­foil distance Lx, motion phase difference ∆φ and wake convection velocity Uv, is introduced to describe the wake interaction patterns parametrically. We use a differential pressure sensor on the downstream foil to capture wake interaction characteristics. Data sets at different tandem configurations are collected. The wake-­foil phase Φ is used to label the pressure signals. A one ­dimensional convolutional neural networks (1D-CNN) model is used to learn an end­to­end mapping between the raw pressure measurements and the wake-­foil phase Φ. The trained 1D-­CNN model shows accurate estimations (average error 3.5%) on random wake interaction patterns and is fast enough (within 40 ms). Then the trained 1D ­CNN model is applied to online thrust enhancement control of a downstream foil swimming in a periodic wake. Synchronous force monitoring and flow visualization demonstrate the effectiveness of the 1D-­CNN model. The limitations of the model are discussed. The proposed approach can be applied to the online estimation and control of wake interactions in the collective swimming and flying of biomimetic robots.

Distributed estimation and control of a four-tank process

A multivariable process of four interconnected water tanks is considered for modeling and control. The objective of the current study is to design and implement a distributed control and estimation (DEC) for a multivariable four-tank process. Distributed model and inter-nodal communication structure are derived from global state–space matrices, thus combining the topology of plant flow sheet and the interaction dynamics across the plant subunits. Using experimental data, the process dynamics and disturbance effects are modeled. A typical lab-scale system was simulated and the obtained results demonstrated the potential of the DEC algorithm.

Model-based electron density profile estimation and control, applied to ITER

In contemporary magnetic confinement devices, the density distribution is sensed with interferometers and actuated with feedback controlled gas injection and open-loop pellet injection. This is at variance with the density control for ITER and DEMO, that will depend mainly on pellet injection as an actuator in feed-back control. This paper presents recent developments in state estimation and control of the electron density profile for ITER using relevant sensors and actuators. As a first step, Thomson scattering is included in an existing dynamic state observer. Second, model predictive control is developed as a strategy to regulate the density profile while avoiding limits associated with the total density (Greenwald limit) or gradients in the density distribution (e.g. neo-classical impurity transport). Simulations show that high quality density profile estimation can be achieved with Thomson Scattering and that the controller is capable of regulating the distribution as desired.