摘要：【中文】本发明提供一种高速耐正负浪涌的模拟开关电路，包括第一、二、三晶体管电路。第一晶体管电路包括第一晶体管及第一闸极驱动电路，当第一晶体管的闸极于第一电压位准时，启动第一晶体管以第一级方式来降低正、负浪涌信号的电压位准峰值。第二晶体管电路包括第二晶体管及第二闸极驱动电路，当第二晶体管的闸极于第二电压位准时，启动第二晶体管以第二级方式来降低正、负浪涌信号的电压位准峰值。第三晶体管电路包括第三晶体管及第三闸极驱动电路，当第三晶体管的闸极于第三电压位准时，启动第三晶体管以第三级方式来降低正、负浪涌信号的电压位准峰值传送至低压输出端。本发明提出的模拟开关电路有效抵抗电路于上电时产生高压正负浪涌。 【EN】The invention provides a high-speed positive and negative surge resistant analog switch circuit which comprises a first transistor circuit, a second transistor circuit and a third transistor circuit. The first transistor circuit comprises a first transistor and a first gate driving circuit, and when the gate of the first transistor is at a first voltage level, the first transistor is started to reduce the voltage level peak value of the positive surge signal and the negative surge signal in a first stage mode. The second transistor circuit comprises a second transistor and a second gate driving circuit, and when the gate of the second transistor is at a second voltage level, the second transistor is started to reduce the voltage level peak value of the positive surge signal and the negative surge signal in a second-stage mode. The third transistor circuit comprises a third transistor and a third gate driving circuit, when the gate of the third transistor is at a third voltage level, the third transistor is started to reduce the voltage level peak value of the positive and negative surge signals in a third-level mode and transmit the voltage level peak value to the low-voltage output end. The analog switch circuit provided by the invention can effectively resist the high-voltage positive and negative surge generated when the circuit is powered on.

摘要：【中文】本发明公开了一种基于模糊集理论的Delta机器人控制方法，首先建立机器人的动力学模型，并选取表示机器人关节摩擦现象的非线性摩擦力模型；将机器人动力学模型中含有不确定性的项分离出来，根据机器人动力学模型中的标称项建立控制器中的标称补偿环节；取正定对角矩阵设计控制器中的控制环节对初始位置误差进行补偿；根据机器人动力学模型中与不确定性有关的项，构造代表系统不确定项上界信息的函数并验证；基于所描述的不确定性信息设计二次性能指标，利用D‑运算和模糊运算求解最优控制参数；将优化后的控制参数带入不确定性控制项，得到最终优化控制输出，通过优化二次性能指标，从而得到最优控制增益，适用于系统存在不确定性因素时的最优控制。 【EN】The invention discloses a Delta robot control method based on a fuzzy set theory, which comprises the steps of firstly establishing a kinetic model of a robot, and selecting a nonlinear friction model representing the friction phenomenon of joints of the robot; separating the uncertain items in the robot dynamic model, and establishing a nominal compensation link in the controller according to the nominal items in the robot dynamic model; a control link in the positive definite diagonal matrix design controller is taken to compensate the initial position error; according to the items related to uncertainty in the robot dynamics model, constructing and verifying a function representing upper bound information of the uncertainty items of the system; designing a secondary performance index based on the described uncertainty information, and solving an optimal control parameter by using D-operation and fuzzy operation; and (4) substituting the optimized control parameters into an uncertainty control item to obtain final optimization control output, and optimizing secondary performance indexes to obtain optimal control gain, so that the method is suitable for optimal control when uncertainty factors exist in the system.

摘要：【中文】本发明提供的一种单晶硅加料数据的确定方法、装置及设备，包括：获取单晶炉的工步状态；在工步状态为卸离硅棒工步的情况下，获取单晶炉的运行参数数据；在根据运行参数数据，确定开始进行加料操作的情况下，获取单晶炉中坩埚的信息、单晶炉所需原料的原料信息，并根据坩埚的信息、原料信息和运行参数数据，确定加料数据。本发明中，在工步状态为卸离硅棒工步时，就可以通过单晶炉的运行参数数据、坩埚的信息和原料信息，确定加料数据，使得制备单晶硅的加料操作更加快速和准确，缩短了确定单晶硅加料数据的时间，提高了单晶硅的生产效率。 【EN】The invention provides a method, a device and equipment for determining monocrystalline silicon charging data, which comprises the following steps: acquiring the process step state of the single crystal furnace; acquiring operation parameter data of the single crystal furnace under the condition that the process step state is the silicon rod unloading process step; and under the condition that the charging operation is determined to be started according to the operation parameter data, acquiring the information of a crucible in the single crystal furnace and the raw material information of raw materials required by the single crystal furnace, and determining the charging data according to the information of the crucible, the raw material information and the operation parameter data. According to the invention, when the silicon rod unloading process step is in the process step of silicon rod unloading, the feeding data can be determined through the operation parameter data of the single crystal furnace, the crucible information and the raw material information, so that the feeding operation for preparing the monocrystalline silicon is quicker and more accurate, the time for determining the monocrystalline silicon feeding data is shortened, and the production efficiency of the monocrystalline silicon is improved.

摘要：【中文】本公开提供一种单晶炉漏硅检测方法、设备及存储介质，涉及太阳能光伏领域，能够及时检测单晶炉是否漏硅，提高单晶炉的安全性，降低事故风险。具体技术方案为：获取同一时刻单晶炉坩埚内硅液液面的理论高度值和实际高度值；计算硅液液面的理论高度值和实际高度值的差值；在差值大于预设阈值时，确定单晶炉发生漏硅。本发明用于漏硅检测。 【EN】The invention provides a silicon leakage detection method and device for a single crystal furnace and a storage medium, relates to the field of solar photovoltaic, and can detect whether the single crystal furnace leaks silicon in time, improve the safety of the single crystal furnace and reduce the accident risk. The specific technical scheme is as follows: obtaining a theoretical height value and an actual height value of the liquid level of the silicon liquid in the crucible of the single crystal furnace at the same moment; calculating the difference value between the theoretical height value and the actual height value of the liquid level of the silicon liquid; and when the difference value is larger than a preset threshold value, determining that the silicon leakage occurs in the single crystal furnace. The invention is used for silicon leakage detection.

摘要：【中文】本发明提供一种单晶硅生长用掺杂剂的控制方法、装置、设备及存储介质，包括：在当前单晶硅制备完成后，获取当前的工艺参数，将当前的工艺参数输入预设模型中，确定掺杂剂的目标挥发比例；获取在当前单晶硅制备完成后的掺杂剂的理论剩余量；根据理论剩余量和目标挥发比例，确定掺杂剂的实际剩余量；获取下次单晶硅生长用掺杂剂的理论需求量，并根据掺杂剂的实际剩余量和下次单晶硅生长用掺杂剂的理论需求量，控制下次单晶硅生长用掺杂剂的实际需求量。本发明实施例通过采用预设模型，能够准确获取当前单晶硅在拉至完成后至下次填料过程中掺杂剂的挥发比例，进而更好的控制下次填料的掺杂剂的量，使得生长的下次单晶硅的电阻率符合预设的要求。 【EN】The invention provides a method, a device, equipment and a storage medium for controlling a dopant for monocrystalline silicon growth, wherein the method comprises the following steps: after the preparation of the current monocrystalline silicon is finished, obtaining current process parameters, inputting the current process parameters into a preset model, and determining the target volatilization proportion of the dopant; obtaining the theoretical residual amount of the dopant after the preparation of the current monocrystalline silicon is finished; determining the actual residual amount of the doping agent according to the theoretical residual amount and the target volatilization proportion; and obtaining the theoretical demand of the dopant for next monocrystalline silicon growth, and controlling the actual demand of the dopant for next monocrystalline silicon growth according to the actual residual amount of the dopant and the theoretical demand of the dopant for next monocrystalline silicon growth. According to the embodiment of the invention, the preset model is adopted, so that the volatilization proportion of the dopant in the process from the completion of pulling to the next filling of the current monocrystalline silicon can be accurately obtained, the dopant amount of the next filling is further better controlled, and the resistivity of the grown next monocrystalline silicon meets the preset requirement.

摘要：【中文】本发明公开了一种基于DTW算法和ResNet网络的转辙机故障诊断方法，包括：对每台转辙机设备的历史动作曲线分别进行预处理，再利用DTW算法选取对应的标准模板曲线；对曲线进行归一化与拼接操作，对得到的矩阵形式的曲线数据中部分曲线数据进行人工标记，作为训练用的模板曲线；利用模板曲线对预先构建的且引入了ResNet网络的卷积神经网络进行训练，使得ResNet网络能够提取出表征转辙机实时动作状态的特征参数；将需要进行故障诊断的曲线输入至训练好的卷积神经网络中，得到故障概率。该方法能够针对每台转辙机设备自适应选取标准模板曲线用于神经网络训练检测，能有效提高转辙机设备工作异常检出率，利于提前发现故障隐患，减少因设备故障对行车造成的影响。 【EN】The invention discloses a point switch fault diagnosis method based on a DTW algorithm and a ResNet network, which comprises the following steps: respectively preprocessing the historical action curve of each point switch device, and selecting a corresponding standard template curve by using a DTW algorithm; normalizing and splicing the curves, and manually marking part of curve data in the obtained curve data in a matrix form to be used as a template curve for training; training a convolution neural network which is constructed in advance and is introduced with a ResNet network by utilizing a template curve, so that the ResNet network can extract characteristic parameters representing the real-time action state of the point switch; and inputting the curve needing fault diagnosis into the trained convolutional neural network to obtain the fault probability. The method can adaptively select the standard template curve for neural network training detection aiming at each point switch device, can effectively improve the abnormal work detection rate of the point switch device, is beneficial to finding out the hidden trouble in advance, and reduces the influence on driving caused by the device trouble.