4results about How to "Reduce non-uniformity" patented technology

The invention discloses injection-casting high-Zn magnesium alloy based on a two-stage corrosion-resistant barrier and a preparation method of the injection-casting high-Zn magnesium alloy, and belongs to the technical field of magnesium alloy materials and metallurgy. Aiming at the corrosion resistance bottleneck caused by looseness of a surface film of the high-Zn system injection casting magnesium alloy and strong cathode activity of a second phase of a grain boundary, a new two-stage corrosion resistance barrier design thought is provided, and through main and micro element optimization design, injection casting forming process optimization and cooperative regulation and control of a precise heat treatment system, the high-Zn system injection casting magnesium alloy is obtained. And a two-stage corrosion-resistant barrier with a compact surface passivation layer and a fine net-shaped weak cathode phase is constructed. After the magnesium alloy prepared through the method is soaked in a 3.5 wt% NaCl solution for 7 days, the weight loss corrosion rate is smaller than or equal to 0.70 mm / year, the hydrogen evolution rate is smaller than or equal to 0.40 mL / cm / day, the corrosion resistance is remarkably superior to that of an existing commercial high-Zn magnesium alloy, and the magnesium alloy is suitable for the fields, with requirements for corrosion resistance and light weight, such as 3C, unmanned aerial vehicles, robots and automobiles.

The utility model provides a kind of grouting anchor rod for water conservancy and hydropower engineering foundation reinforcement construction, grouting anchor rod body 1 is overall cylindrical, the inside of grouting anchor rod body 1 is interval and provided with pressurized cheek 2, grouting pressurized rod 4 is fixedly installed in the inside of pressurized cheek 2 in penetrating mode;The inside of grouting anchor rod body 1 is interval and provided with multiple guide columns 3, the inside of grouting anchor rod body 1 is provided with a grouting pressurized rod 4, grouting pressurized rod 4 is fixedly installed in the inside of multiple guide columns 3 in penetrating mode, grouting pressurized rod 4 is also fixedly installed in the inside of pressurized cheek 2 in penetrating mode.Numerical simulation result shows that, using the grouting anchor rod of the utility model can effectively grout and fill up fine crack, and the rated pressure of hydraulic cylinder is greater than the rated grouting pressure of grouting pump, can produce certain pulse effect, reduce the non-uniformity of slurry, greatly improve the effect of grouting.

This invention provides a copper-zinc-tin-sulfur-selenium thin film, its preparation method, and a semiconductor device thereof. The preparation method includes: fully dissolving at least one metal source of copper, zinc, tin, sulfur, and selenium in a ligand solvent; then adding a main solvent, other metal sources, and a sulfur or selenium source; and thoroughly mixing to obtain a copper-zinc-tin-sulfur-selenium precursor solution based on a mixed solvent system. The ligand solvent is a small-molecule solvent with carbonyl or amine coordinating groups that provide lone pairs of electrons to metal cations. The precursor solution is then coated onto a substrate surface to prepare a precursor thin film. The precursor thin film is further subjected to post-selenization or post-sulfurization treatment to obtain the copper-zinc-tin-sulfur-selenium thin film. By introducing a small-molecule solvent with specific coordination ability to form discrete metal-ligand structural units with the metal precursor, metal clusters or oxygen bridge network structures are effectively suppressed. High homogenization of the precursor solution is achieved at the molecular scale, significantly improving the thickness and compositional uniformity of the copper-zinc-tin-sulfur-selenium precursor thin film.

The application discloses a kind of deep explainable multimodal discriminative representation method.Deep explainable multimodal discriminative representation network model is proposed in the application, the method aligns heterogeneous modal in discriminative common space to improve classification accuracy, and provides a sensitive feature analysis idea.Specifically, the application uses multilayer perceptron as an encoder, combines self-representation constraints embedded with structural information and generalized canonical correlation analysis constraints, maps multimodal data into discriminative common space, and improves recognition performance.In addition, the application decouples sensitive features by compressing complex networks through knowledge distillation to enhance the explainability of the model.