摘要：【中文】本公开提供一种验证训练数据的方法、训练系统以及计算机程序产品。方法包含：将多个原始数据提供给多个标记者；检索多个标记结果，其中标记结果包含多个标记数据，且标记数据通过标记者经由标记原始数据而产生；通过比较标记结果来确定多个一致性，且因此确定标记结果对于训练人工智能机器是否有效；响应于确定标记结果有效，而确定标记结果的至少一特定部分对于训练人工智能机器有效。 【EN】The present disclosure provides a method of validating training data, a training system, and a computer program product. The method comprises the following steps: providing a plurality of raw data to a plurality of markers; retrieving a plurality of marking results, wherein the marking results comprise a plurality of marking data, and the marking data is generated by a marker by marking original data; determining a plurality of consistencies by comparing the labeling results and thus determining whether the labeling results are valid for training the artificial intelligence machine; in response to determining that the tagging result is valid, determining that at least a particular portion of the tagging result is valid for training the artificial intelligence machine.

摘要：【中文】本发明公开了一种页岩水相圈闭损害评价的实验参数获取方法，该方法通过开展应力敏感、速敏实验，明确页岩水相圈闭损害实验的岩心预处理围压及渗透率测量、返排实验的驱替压差；通过开展孔隙水矿化度测定及压裂液返排液测试实验，确定模拟地层水矿化度；通过对比开展不同回压下的渗透率测量实验，明确水相圈闭损害评价过程中的渗透率测量所需施加的回压。本发明旨在避免应力敏感损害、水敏损害、盐敏损害、速敏损害、气体滑脱效应对页岩水相圈闭损害评价实验结果的干扰，保证了页岩水相圈闭损害评价结果的准确性。 【EN】The invention discloses an experimental parameter acquisition method for evaluating shale aqueous phase trapping damage, which is characterized in that by developing stress sensitivity and speed sensitivity experiments, the method determines the core pretreatment confining pressure and permeability measurement of the shale aqueous phase trapping damage experiment and the displacement differential pressure of a flowback experiment; determining the simulated formation water mineralization degree by carrying out pore water mineralization degree measurement and fracturing fluid flowback fluid test experiments; by carrying out permeability measurement experiments under different back pressures in a contrast manner, the back pressure required to be applied for permeability measurement in the process of evaluating the damage of the water phase trapping is determined. The method aims to avoid the interference of stress sensitive damage, water sensitive damage, salt sensitive damage, speed sensitive damage and gas slippage effect on the evaluation experiment result of the shale aqueous phase trapping damage and ensure the accuracy of the evaluation result of the shale aqueous phase trapping damage.

摘要：【中文】本发明公开了一种页岩水相圈闭损害综合评价方法，制备页岩气藏的岩心，并对岩心分别开展页岩水相自吸及返排实验、水相圈闭损害前后的页岩甲烷吸附/解吸实验、水相圈闭损害前后页岩甲烷扩散系数测量实验，计算各个参数，再根据上述计算得到的各个参数计算岩心的页岩水相圈闭损害指数，最后根据页岩水相圈闭损害指数对页岩水相圈闭损害进行综合评价。本发明区分了基质和裂缝的损害情况，考虑了吸附气、游离气的占比，确定了页岩水相圈闭损害关键评价指标，以及各评价指标对总损害程度的贡献率，评价结果能够客观真实反应页岩水相圈闭损害程度。 【EN】The invention discloses a shale aqueous phase trapping damage comprehensive evaluation method, which comprises the steps of preparing a rock core of a shale gas reservoir, respectively carrying out a shale aqueous phase self-absorption and backflow experiment, a shale methane adsorption/desorption experiment before and after aqueous phase trapping damage and a shale methane diffusion coefficient measurement experiment before and after aqueous phase trapping damage on the rock core, calculating each parameter, calculating a shale aqueous phase trapping damage index of the rock core according to each parameter obtained by calculation, and finally carrying out comprehensive evaluation on the shale aqueous phase trapping damage according to the shale aqueous phase trapping damage index. The method distinguishes damage conditions of the matrix and the cracks, considers the ratio of adsorbed gas to free gas, determines key evaluation indexes of the shale aqueous phase entrapment damage and the contribution rate of each evaluation index to the total damage degree, and the evaluation result can objectively and truly reflect the shale aqueous phase entrapment damage degree.

摘要：【中文】本发明公开了一种含吸附气条件下页岩压裂液强制渗吸及返排实验方法，通过将裂缝发育、裂缝欠发育和基块页岩串联构成页岩多尺度储渗空间，将样品放置于长岩心夹持器中，向夹持器中注入甲烷气体，并通过岩心出口端回压阀控制孔隙压力，甲烷吸附平衡后，在岩心进口端正压差作用下实现压裂液强制渗吸，出口端监测压裂液渗吸过程中置换出的甲烷气体，在返排阶段，通过岩心进口端回压阀设置压差模拟压裂液返排过程。过程中实时监测岩心电阻，通过电阻反映压裂液在页岩多尺度储渗空间的渗吸和滞留量。本发明可以真实模拟含吸附气条件下压裂液在页岩多尺度储渗空间的渗吸和返排过程，为页岩气井压裂后优化焖井时间和制定合理的返排制度提供参考。 【EN】The invention discloses a shale fracturing fluid forced imbibition and flowback experimental method under the condition of containing adsorbed gas, which comprises the steps of forming a shale multi-scale seepage storage space by connecting crack development, crack underdevelopment and base shale in series, placing a sample in a long rock core holder, injecting methane gas into the holder, controlling pore pressure through a back pressure valve at the outlet end of a rock core, realizing the forced imbibition of the fracturing fluid under the action of positive pressure difference at the inlet end of the rock core after the methane adsorption is balanced, monitoring the methane gas displaced in the imbibition process of the fracturing fluid at the outlet end, and simulating the flowback process of the fracturing fluid through the back pressure valve at the inlet end of the rock core at the flowback stage. And monitoring the resistance of the rock core in real time in the process, and reflecting the seepage and retention of the fracturing fluid in the shale multi-scale seepage storage space through the resistance. The method can truly simulate the seepage and flowback process of the fracturing fluid in the shale multi-scale seepage storage space under the condition of containing the adsorbed gas, and provides reference for optimizing the soaking time and formulating a reasonable flowback system after the fracturing of the shale gas well.

摘要：【中文】本发明公开了一种钢筋笼的起吊吊架装置及方法，包括十字型吊架主体，由四个吊架支臂组成，每个吊架支臂的外端部和四个吊架支臂交汇的中心位置设有吊点，吊点上连接拉杆，且五个拉杆交汇后固定一吊环；所述每个吊架支臂上均设有若干安装孔，每个吊架支臂的其中一安装孔上固定一吊钩。本发明利用十字型的吊架来起吊钢筋笼，起吊时，整个吊架受力平衡，可以保证稳定的移动钢筋笼，将其放入桩孔内。 【EN】The invention discloses a hoisting hanger device and a hoisting hanger method for a reinforcement cage, and the hoisting hanger device comprises a cross-shaped hanger main body, wherein the cross-shaped hanger main body consists of four hanger support arms, a hoisting point is arranged at the center position where the outer end part of each hanger support arm and the four hanger support arms are intersected, a pull rod is connected to the hoisting point, and a hoisting ring is fixed after the five pull rods are intersected; and each hanger support arm is provided with a plurality of mounting holes, and one mounting hole of each hanger support arm is fixedly provided with a lifting hook. The cross-shaped hanging bracket is used for hoisting the reinforcement cage, and when the reinforcement cage is hoisted, the stress of the whole hanging bracket is balanced, so that the reinforcement cage can be stably moved and placed in the pile hole.

摘要：【中文】本发明属于语义分析技术领域，尤其为一种基于机器学习的电力领域的语意分析方法，包括以下步骤：获取用户反馈信息；对反馈信息按照情感极性分为消极或积极；机器学习模块介入；在语料中获取特征，并对积极和消极的情感词分开执行；最终得到两个分支，分别表示文本的正向感情值和负向感情值；通过电脑获取用户反馈信息，并将用户的反馈信息输入数据库，根据数据库内，具有预先设定的有关于感情极性的特征信息，按照情感极性分为消极或积极，在对信息的处理过程中机器学习模块介入，减少信息筛选时间，加快筛选效率，可以强化机器对关键特征信息的判断速度，积累更多的特征信息，做出更多直观的可视化结论。 【EN】The invention belongs to the technical field of semantic analysis, in particular to a semantic analysis method in the electric power field based on machine learning, which comprises the following steps: acquiring user feedback information; dividing feedback information into negative or positive according to emotion polarity; a machine learning module intervenes; acquiring features from the corpus and separately executing the positive and negative emotion words; finally, two branches are obtained, and the positive emotion value and the negative emotion value of the text are respectively represented; the feedback information of the user is acquired through the computer, the feedback information of the user is input into the database, the information is divided into negative or positive according to the emotion polarity according to the preset characteristic information related to the emotion polarity in the database, the machine learning module intervenes in the information processing process, the information screening time is shortened, the screening efficiency is accelerated, the judgment speed of the machine on key characteristic information can be enhanced, more characteristic information is accumulated, and more visual conclusions can be made.