Elzaburu S.L.P

Electric pump comprising:- a gerotor comprising an outer rotor (4) and an inner rotor (3), the outer rotor (4) assembled around the inner rotor (3), the outer rotor (4) and the inner rotor (3) being configured to pump a fluid,- an electric outrunner motor comprising a stator (7, 8, 9) and a rotor (6) being configured to drive the gerotor, the stator (7, 8, 9) and the rotor (6) assembled around the gerotor.

Hold down and release device (1), suitable for switching from a hold down or closed position to a release or open position, that comprises: - a housing (2) comprising a through hole (3) and several guide means (4) configured to delimit radial channels (5) starting from the through hole (3), - a ring element (6) with a central cavity (15) and placed on the housing (2) concentrically with respect to the through hole (3) of the housing (2), the wall delimiting the central cavity (15) having several recesses (20) on it, such that the guide means (4) are placed inside the cavity (15) and surrounded by the ring element (6), the ring element (6) being configured to rotate on the housing (2) around an axis of the through hole (3), - at least three assemblies, each of them comprising a block (7) and at least an additional element (8, 8', 8"), such that each assembly is placed on a channel (5) delimited by two of the guide means (4), the additional element (8) being connected to the block (7) on the end of the block (7) which is more distant from the through hole (3), the additional element (8, 8', 8") being configured to couple in one of the recesses (20) of the ring element (6), and - activation means configured to actuate on the ring element (6) to be able to rotate it.

Embodiments are directed towards surface protective articles including a cast machine direction oriented multilayer polyethylene carrier and an adhesive composition.

The present disclosure provides a wind power generation apparatus, including: at least two support platforms, stacked along a first direction; a wind turbine and a support frame, arranged between two adjacent support platforms; where the support frame is connected between the two adjacent support platforms; the wind turbine is detachably connected to one of the two adjacent support platforms, and spaced apart from the support frame and the other of the two adjacent support platforms. In an embodiment of the present disclosure, a difficulty of disassembling the wind turbine in the wind power generation apparatus may be reduced.

Computer implemented method for training a feed-forward deep learning model intended to the detection and instance segmentation of objects, comprising the steps of: inserting (S1) a gradient reversal layer (GRL) after a hidden layer of the deep learning model, inserting (S2) a domain classifier layer (DCL) after the gradient reversal layer (GRL); introducing (S3) a set of input values (I) of training data into an input layer of the deep learning model; obtaining (S4) a set of output values (O) in response to said input values (I); calculating (S5) a loss function of the last hidden layer of the deep learning model, calculating (S6) the gradient of the loss function; propagating backwards the loss function and repeating steps (S5) and (S6) for each hidden layers and adjusting (S7) the weight of each hidden layer, so as to minimize the gradient of the loss function.

The present invention relates to the technical field of sealing materials, and discloses a high-temperature vermiculite sealing material for SOFC, and a preparation method thereof. The material is prepared from modified chemically expanded vermiculite and negative-ion high-temperature expanded vermiculite. The modified chemically expanded vermiculite carries positive charges, and the negative-ion high-temperature expanded vermiculite carries negative charges. After the modified chemically expanded vermiculite and the negative-ion high-temperature expanded vermiculite are mixed, electrostatic attraction is generated to achieve combination. The material does not use binders, so that the loss on ignition is significantly reduced without changing the combined strength of the sealing material. The stearate serves as an intercalation agent of chemically expanded vermiculite. The modified chemically expanded vermiculite prepared from the stearate has high-temperature resistant compressive strength at 700°C. The negative-ion high-temperature expanded vermiculite is formed by using negative ion air to treat high-temperature expanded vermiculite to enable the high-temperature expanded vermiculite to load and accumulate negative charges. Using the negative ion air to treat the high-temperature expanded vermiculite does not require the use of chemical agents, the process is simple, and large-scale production can be achieved.

The present invention relates to a masterbatch plastic resin-free composition comprising from 75 wt. % to 90 wt. % of mineral filler, pigment, fibre or any combination thereof, from 10 wt. % to 25 wt. % of binder additives, a method for preparing said masterbatch plastic resin-free composition, a plastic article comprising said masterbatch composition, a method for preparing said plastic article and the use of the masterbatch plastic resin-free composition.

The present invention refers to a multilayer plastic laminate recycling process and a delamination solution, wherein the process comprises the following steps: providing the multilayer plastic laminate and the delamination solution comprising an alkaline component and an tensioactive; conditioning the multilayer plastic laminate to obtain multilayer flakes comprising a mesh size under 20 mm; contacting the multilayer flakes with the delamination solution being a stirred step at 25 - 90 °C for 1 - 48 h, and the solution/solid ratio (w/w) is higher than 2, to obtain a mixture of a delamination sludge and monolayer plastic flakes; separating the delamination sludge of the mixture of monolayer plastic flakes to obtain a mixture of monolayer flakes; physically separating the mixture of monolayer plastic flakes to obtain monolayer first polymer flakes and drying of the monolayer first polymer flakes to obtain high-purity monolayer flakes.

Double floor (1) for a vehicle trunk comprising:- a rigid structure (2) defining a main surface delimited by a rear end (3) according to a longitudinal direction, facing a rear tailgate (12) of the trunk, a front end (4) opposite the rear end (3), and two side ends substantially perpendicular to the front end (4) and the rear end (3); and- a flexible covering (5) over at least the main surface of the rigid structure (2),wherein the flexible covering (5) comprises an excess part, such that said excess part of the flexible covering (5) defines a flexible spoiler (6) that projects beyond the front end (4) of the rigid structure (2) along the longitudinal direction. Said flexible spoiler (6) is created by folding the excess part of the flexible covering (5) onto itself, forming at least two flaps (7, 7'). The excess part is separated from the main surface of the rigid structure (2), allowing relative movement of the at least two flaps (7, 7') of said excess part relative to the main surface of the rigid structure (2).

The present invention provides a method and an apparatus for online prediction of a measurement error of an electronic current transformer. The method includes: obtaining environmental data of an environment where the electronic current transformer is located, and obtaining error data of the electronic current transformer in an automatic prediction manner by a neural network, thereby implementing online automatic prediction of an error of the electronic current transformer, improving efficiency for determining the error of the electronic current transformer, and further enhancing calibration efficiency of the electronic current transformer. Further, a target swimming penetration neural network includes a retention layer, and the retention layer is used to store an output from a hidden layer in a previous iteration. Compared with a conventional neural network, a structure of this retention layer can effectively enhance both the overall and local stability of the swimming penetration neural network, thereby improving accuracy of error prediction by the electronic current transformer.