摘要：【中文】本发明提供一种抗疲劳的固体饮料的制备方法，包括以下步骤：(1)按重量份称取黄芪1‑2份、党参1份、大枣1‑3份；(2)将药材混合，加水煎煮，煎液100目筛网滤过，收集滤液；(3)滤液减压浓缩至相对密度为1.055‑1.12，得到清膏；(4)设置进样口温度120‑160℃，将清膏喷雾干燥成粉，包装。本发明制备的固体饮料组成成分搭配合理，可有效改善能量代谢，延缓疲劳发生，抗疲劳作用显著；而且在冲泡后能快速溶解，色泽均匀、口感好、无涩味或辛味。 【EN】The invention provides a preparation method of an anti-fatigue solid beverage, which comprises the following steps: (1) weighing 1-2 parts of astragalus, 1 part of codonopsis pilosula and 1-3 parts of Chinese date according to parts by weight; (2) mixing the medicinal materials, decocting in water, filtering decoction with 100 mesh sieve, and collecting filtrate; (3) concentrating the filtrate under reduced pressure to relative density of 1.055-1.12 to obtain fluid extract; (4) setting the injection port temperature at 120-. The solid beverage prepared by the invention has reasonable component matching, can effectively improve energy metabolism and delay fatigue occurrence, and has remarkable anti-fatigue effect; but also can be dissolved quickly after being brewed, has uniform color and good taste, and has no astringent taste or pungent taste.

摘要：【中文】本发明提供一种基于LSTM的电网一次调频数据预测方法及装置。本发明电网调度运行安全数据预测方法，包括：获取原始数据，所述原始数据为第一格式；将所述原始数据进行格式转换获得第二数据，所述第二数据为第二格式，所述第二格式不同于所述第一格式；对第二数据进行数据清洗，删除无用数据，获得第三数据，所述第三数据包括：时间、有功功率、直流外送、水电有功功率、火电有功功率；使用第一预测模型根据第三数据预测电网频率。本发明克服了现有技术中系统风险感知能力和预警能力缺失的问题，进一步提高了电网调度安全。 【EN】The invention provides a power grid primary frequency modulation data prediction method and device based on LSTM. The invention discloses a power grid dispatching operation safety data prediction method, which comprises the following steps: acquiring original data, wherein the original data is in a first format; carrying out format conversion on the original data to obtain second data, wherein the second data is in a second format, and the second format is different from the first format; performing data cleaning on the second data, deleting useless data, and obtaining third data, wherein the third data comprises: time, active power, direct current outgoing, hydroelectric active power and thermal power active power; and predicting the grid frequency from the third data using the first prediction model. The invention overcomes the problem of the lack of system risk perception capability and early warning capability in the prior art, and further improves the dispatching safety of the power grid.