Bonding silver wire full-automatic operation monitoring method and system based on deep learning

Abstract

The invention provides a bonding silver wire full-automatic operation monitoring method and system based on deep learning, the method and system are applied to wire drawing equipment of bonding silver wires, the wire drawing equipment comprises an alternating current servo motor, and the alternating current servo motor is connected with a take-up reel and used for conducting wire drawing on the bonding silver wires; the bonding silver wire full-automatic operation monitoring method based on deep learning comprises the steps that the real-time frequency of a power grid is acquired; inputting the real-time frequency into a frequency pulse adjustment model to obtain a frequency adjustment pulse instruction; obtaining a current-period frequency pulse instruction based on the frequency adjustment pulse instruction and a current-period original pulse instruction; and the current-period frequency pulse instruction is sent to a driver of the alternating current servo motor, so that the preset rotating speed of the alternating current servo motor is kept unchanged, and the problem that the forming quality of the bonded silver wire after wire drawing is inconsistent due to the fact that the working rotating speed of the motor is unstable due to the fact that the frequency of a power grid is unstable at present can be solved.

Description

Technical Field

[0001] The present invention relates to the field of semiconductor technology, and in particular to a method and system for fully automatic operation monitoring of silver bonding wires based on deep learning. Background Art

[0002] Bonding wires are fine metal wires used to electrically connect the input / output bonding points of the chip's internal circuits with the internal contact points of the lead frame (or substrate) when assembling semiconductor devices and integrated circuits. The quality of the bonding directly affects the performance of the integrated circuit.

[0003] Currently, the production process of bonding gold wire (or bonding silver wire) usually requires wire drawing, cleaning, annealing, etc. When drawing the bonding silver wire, the bonding silver wire is usually passed through a drawing die and connected to a take-up reel. A motor is used to drive the take-up reel to rotate for wire drawing. The motor is usually powered by a power grid. However, the instability of the power grid frequency will lead to instability in the motor's operating speed, which in turn causes inconsistent forming quality of the bonding silver wire after drawing. Summary of the Invention

[0004] In response to the above technical problems, the embodiments of the present application propose a fully automatic operation monitoring method and system for bonding silver wire based on deep learning, which can solve the current problem of unstable motor operating speed caused by unstable grid frequency, and thus inconsistent forming quality of bonding silver wire after drawing.

[0005] In a first aspect, an embodiment of the present application provides a fully automatic operation monitoring method for silver bonding wire based on deep learning, which is applied to a silver bonding wire drawing device, wherein the silver bonding wire drawing device includes an AC servo motor connected to a wire take-up reel and used to draw the silver bonding wire;

[0006] The deep learning-based fully automatic operation monitoring method for silver bonding wires comprises:

[0007] Get the real-time frequency of the power grid;

[0008] Inputting the real-time frequency into a frequency pulse adjustment model to obtain a frequency adjustment pulse instruction;

[0009] Obtaining a current period frequency pulse instruction based on the frequency adjustment pulse instruction and the current period original pulse instruction;

[0010] The current frequency pulse instruction is sent to the driver of the AC servo motor to keep the preset speed of the AC servo motor unchanged.

[0011] In some embodiments, obtaining the current period frequency pulse instruction based on the frequency adjustment pulse instruction and the current period original pulse instruction includes:

[0012] Record the current time of obtaining the real-time frequency;

[0013] Determine the arrival time of the grid current at the AC servo motor based on the current time and the distance to be transmitted;

[0014] determining an adjustment period for adjusting the AC servo motor based on the arrival time;

[0015] The original pulse instruction of the current period is obtained based on the adjustment period, and the frequency pulse instruction of the current period is obtained based on the frequency adjustment pulse instruction and the original pulse instruction of the current period.

[0016] In some embodiments, the deep learning-based fully automatic operation monitoring method for silver bonding wires further includes:

[0017] Obtaining a workload increase curve of the AC servo motor when it is operating based on the preset rotational speed and the physical parameters of the bonding silver wire;

[0018] Inputting the workload increase curve into a load pulse adjustment model to obtain a load adjustment pulse instruction;

[0019] Obtaining a current period load frequency pulse instruction based on the load adjustment pulse instruction and the current period frequency pulse instruction;

[0020] The current load frequency pulse instruction is sent to the driver of the AC servo motor.

[0021] In some embodiments, the wire drawing apparatus further comprises an ambient temperature sensor;

[0022] The method for fully automatic operation monitoring of silver bonding wire based on deep learning further includes:

[0023] Detecting the real-time ambient temperature by the ambient temperature sensor;

[0024] Obtaining an output power curve of the AC servo motor based on an initial load and a working load increase curve of the AC servo motor;

[0025] Obtaining a temperature rise curve of the AC servo motor based on the output power curve and the real-time ambient temperature;

[0026] Obtaining a temperature rise adjustment pulse instruction for a corresponding temperature node based on the temperature rise curve;

[0027] The temperature rise adjustment pulse instruction is sent to the driver of the AC servo motor.

[0028] In some embodiments, obtaining the temperature rise curve of the AC servo motor based on the output power curve and the real-time ambient temperature includes:

[0029] Obtaining an initial temperature rise curve of the AC servo motor based on the output power curve;

[0030] The temperature rise curve is obtained based on the initial temperature rise curve and the real-time ambient temperature.

[0031] In some embodiments, the wire drawing device further includes a device temperature sensor;

[0032] The method for fully automatic operation monitoring of silver bonding wire based on deep learning further includes:

[0033] Detecting the real-time device temperature of the AC servo motor by the device temperature sensor;

[0034] If the real-time device temperature is greater than the device temperature threshold, the cooling system of the AC servo motor is turned on to cool the AC servo motor.

[0035] In some embodiments, if the real-time device temperature is greater than a device temperature threshold, turning on a cooling system of the AC servo motor to cool the AC servo motor further includes:

[0036] Obtaining a cooling efficiency curve of the cooling system;

[0037] Obtaining an actual temperature drop curve for cooling the AC servo motor based on the cooling efficiency curve and the temperature rise curve;

[0038] Obtaining a temperature drop adjustment pulse instruction based on the actual temperature drop curve and the real-time ambient temperature;

[0039] The temperature drop adjustment pulse instruction is sent to the driver of the AC servo motor.

[0040] In some embodiments, the wire drawing apparatus further comprises a vibration sensor;

[0041] The method for fully automatic operation monitoring of silver bonding wire based on deep learning further includes:

[0042] Acquiring a real-time vibration level of the AC servo motor through a vibration sensor provided on the AC servo motor;

[0043] obtaining a vibration adjustment pulse instruction based on the real-time vibration level;

[0044] The vibration adjustment pulse instruction is sent to the driver of the AC servo motor.

[0045] In a second aspect, an embodiment of the present application provides a fully automatic operation monitoring system for silver bonding wires based on deep learning, which is applied to a silver bonding wire drawing device, wherein the silver bonding wire drawing device includes an AC servo motor connected to a take-up reel and used to draw the silver bonding wire;

[0046] The deep learning-based fully automatic operation monitoring system for silver bonding wires includes:

[0047] An acquisition module is used to obtain the real-time frequency of the power grid;

[0048] A calculation module, configured to input the real-time frequency into a frequency pulse adjustment model to obtain a frequency adjustment pulse instruction; and obtain a current period frequency pulse instruction based on the frequency adjustment pulse instruction and the current period original pulse instruction;

[0049] The adjustment module is used to send the current frequency pulse instruction to the driver of the AC servo motor so that the preset speed of the AC servo motor remains unchanged.

[0050] In a third aspect, an embodiment of the present application provides an electronic device, comprising a fully automatic operation monitoring system for bonding silver wires based on deep learning as described in the second aspect.

[0051] The present application provides a fully automatic operation monitoring method and system for silver bonding wire based on deep learning, which is applied to a silver bonding wire drawing device, wherein the silver bonding wire drawing device includes an AC servo motor, which is connected to a take-up reel and is used to draw the silver bonding wire; the fully automatic operation monitoring method for silver bonding wire based on deep learning includes: obtaining the real-time frequency of the power grid; inputting the real-time frequency into a frequency pulse adjustment model to obtain a frequency adjustment pulse instruction; obtaining a current frequency pulse instruction based on the frequency adjustment pulse instruction and the current original pulse instruction; sending the current frequency pulse instruction to the driver of the AC servo motor to keep the preset speed of the AC servo motor unchanged, and being able to adjust the driving pulse instruction of the AC servo motor according to the real-time frequency of the power grid, which can solve the current problem of instability of the motor working speed caused by the instability of the power grid frequency, thereby causing inconsistent forming quality of the silver bonding wire after drawing, and improve the consistency of the forming quality of the silver bonding wire. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Hereinafter, the present invention will be described in more detail based on embodiments with reference to the accompanying drawings.

[0053] Figure 1 This is a flow chart of a method for fully automatic operation monitoring of silver bonding wires based on deep learning provided by one embodiment of the present invention;

[0054] Figure 2This is a schematic diagram of a fully automatic operation monitoring system for bonding silver wires based on deep learning provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0055] The present invention will be further described below with reference to the accompanying drawings.

[0056] Wire bonding is a connection method that uses thin metal wires and utilizes heat, pressure, and ultrasonic energy to tightly weld the metal wires to the substrate pads to achieve electrical interconnection between the chip and the substrate and information exchange between chips.

[0057] Existing bonding wires (or bonding gold wires, bonding silver wires) are all obtained by drawing a metal wire with a relatively large outer diameter through a wire drawing machine to the required diameter, and then undergoing tempering (annealing) and winding operations. The bonding gold wire equipment includes a wire drawing machine for drawing the metal wire, an annealing furnace for annealing the metal wire, and a winding device for winding the metal wire. During the wire drawing operation, the bonding gold wire (or bonding silver wire) is forced through a die under the action of an external force, so that the metal cross-sectional area is compressed and the required cross-sectional area shape and size are obtained. In the process of drawing the bonding gold wire, the speed of the wire drawing will affect the size and resistivity of the bonding gold wire. When drawing, the bonding silver wire is usually passed through the drawing die and connected to the take-up reel. A motor is used to drive the take-up reel to rotate for drawing. The motor is usually powered by the power grid. Due to the instability of the power grid frequency, the working speed of the motor will be unstable, which in turn causes inconsistent forming quality of the bonding silver wire after drawing.

[0058] like Figure 1 As shown, in order to solve the above technical problems, the embodiment of the present application provides a fully automatic operation monitoring method for bonding silver wire based on deep learning, which is applied to a wire drawing device for bonding silver wire, wherein the wire drawing device includes an AC servo motor, which is connected to a wire take-up reel and is used to draw the bonding silver wire;

[0059] The deep learning-based fully automatic operation monitoring method for silver bonding wires comprises:

[0060] S101: Acquire the real-time frequency of the power grid;

[0061] It should be noted that the real-time frequency of the power grid can be obtained by the phase difference method, that is, the frequency of the power grid is calculated by comparing the phase difference between two nodes in the power grid; the voltage, current, frequency and other parameters of the power grid can also be detected by a power quality analyzer. The present application does not specifically limit the method for obtaining the real-time frequency of the power grid; wherein, the real-time frequency of the power grid can be a predicted power grid frequency (that is, a predicted real-time frequency), that is, after measuring the real-time frequency of the power grid, the predicted real-time frequency can be obtained in combination with the control method of the power grid frequency, such as mechanical load control, generator control, etc., so that the real-time frequency of the power grid can be obtained in advance, and the adjustment strategy can be calculated and prepared in advance.

[0062] S102: Inputting the real-time frequency into a frequency pulse adjustment model to obtain a frequency adjustment pulse instruction;

[0063] S103: Obtaining a current period frequency pulse instruction based on the frequency adjustment pulse instruction and the current period original pulse instruction;

[0064] It should be noted that the frequency pulse adjustment model can be a deep learning model, a neural network model, etc. The frequency pulse adjustment model is obtained through training, that is, the initial model is trained by a large amount of data corresponding to the grid frequency and the speed of the AC servo motor to obtain the frequency pulse adjustment model.

[0065] S104: Sending the current frequency pulse instruction to the driver of the AC servo motor to keep the preset speed of the AC servo motor unchanged.

[0066] It should be noted that AC servo motors are usually controlled by pulse signals, that is, the speed of the AC servo motor is controlled by the frequency of the sent pulse signal, and the number of rotations (or moving distance) of the AC servo motor is controlled by the number of sent pulse signals. Therefore, by obtaining the frequency adjustment pulse instruction, the number of pulses in the original pulse instruction of this period can be increased or decreased, thereby offsetting the impact of the grid frequency fluctuation on the speed of the AC servo motor, so that the preset speed of the AC servo motor remains unchanged.

[0067] It should be noted that the frequency pulse instruction of this period is obtained based on the frequency adjustment pulse instruction and the original pulse instruction of this period, usually by superimposing the frequency adjustment pulse instruction and the original pulse instruction of this period. The frequency pulse adjustment instruction and the original pulse instruction of this period can also be sent to the driver of the AC servo motor respectively, and the superposition operation is performed at the driver to obtain the frequency pulse instruction of this period, and generate a corresponding number of pulse signals to control the speed of the AC servo motor.

[0068] In some embodiments, obtaining the current period frequency pulse instruction based on the frequency adjustment pulse instruction and the current period original pulse instruction includes:

[0069] Record the current time of obtaining the real-time frequency;

[0070] Determine the arrival time of the grid current at the AC servo motor based on the current time and the distance to be transmitted;

[0071] determining an adjustment period for adjusting the AC servo motor based on the arrival time;

[0072] The original pulse instruction of the current period is obtained based on the adjustment period, and the frequency pulse instruction of the current period is obtained based on the frequency adjustment pulse instruction and the original pulse instruction of the current period.

[0073] It should be noted that the real-time frequency of the power grid is usually detected at the source end of the power grid branch (or the power supply head of the power grid) so that targeted detection can be carried out on the line where the AC servo motor is located. The distance to be transmitted is the distance between the source end of the power grid branch and the AC servo motor (or the distance between the power supply head of the power grid and the AC servo motor). The arrival time can be obtained by the distance to be transmitted and the power transmission speed. According to the arrival time, the adjustment period of the electric energy of the real-time frequency arriving at the AC servo motor can be determined within multiple pulse adjustment periods of the AC servo motor, so that the frequency adjustment pulse instruction can be superimposed on the original pulse instruction of this period for adjustment.

[0074] In some embodiments, the deep learning-based fully automatic operation monitoring method for silver bonding wires further includes:

[0075] Obtaining a workload increase curve of the AC servo motor when it is operating based on the preset rotational speed and the physical parameters of the bonding silver wire;

[0076] Inputting the workload increase curve into a load pulse adjustment model to obtain a load adjustment pulse instruction;

[0077] Obtaining a current period load frequency pulse instruction based on the load adjustment pulse instruction and the current period frequency pulse instruction;

[0078] The current load frequency pulse instruction is sent to the driver of the AC servo motor.

[0079] It should be noted that during the operation of the AC servo motor, its speed will be affected by the load. Generally, the greater the load, the lower the speed, and the smaller the load, the higher the speed. As the wire drawing process proceeds, the bonding silver wire is gradually wound on the take-up reel connected to the AC servo motor, causing the workload of the AC servo motor to continue to increase, thereby reducing the speed of the AC servo motor. By obtaining the load pulse adjustment instruction according to the workload increase curve, targeted adjustments can be made to the situation where the workload of the AC servo motor increases during the wire drawing process, thereby keeping the preset speed of the AC servo motor unchanged.

[0080] It should be noted that the physical parameters of the bonding silver wire include the density and diameter of the bonding silver wire. The load increase speed of the AC servo motor (i.e., the workload increase curve) can be determined based on the physical parameters of the bonding silver wire (i.e., density and diameter) and the preset speed. The load pulse adjustment model can be a deep learning model, a neural network model, etc. By obtaining the current load frequency pulse instruction based on the load adjustment pulse instruction and the current frequency pulse instruction, targeted compensation can be performed for the instability of the AC servo motor speed caused by frequency fluctuations and load increase, so that the preset speed of the AC servo motor remains unchanged. The load adjustment pulse instruction, the frequency pulse adjustment instruction, and the current original pulse instruction can also be sent to the driver of the AC servo motor respectively, and superposition operation can be performed at the driver to obtain the current load frequency pulse instruction, and a corresponding number of pulse signals are generated to control the speed of the AC servo motor.

[0081] In some embodiments, the wire drawing apparatus further comprises an ambient temperature sensor;

[0082] The method for fully automatic operation monitoring of silver bonding wire based on deep learning further includes:

[0083] Detecting the real-time ambient temperature by the ambient temperature sensor;

[0084] Obtaining an output power curve of the AC servo motor based on an initial load and a working load increase curve of the AC servo motor;

[0085] Obtaining a temperature rise curve of the AC servo motor based on the output power curve and the real-time ambient temperature;

[0086] Obtaining a temperature rise adjustment pulse instruction for a corresponding temperature node based on the temperature rise curve;

[0087] The temperature rise adjustment pulse instruction is sent to the driver of the AC servo motor.

[0088] In some embodiments, obtaining the temperature rise curve of the AC servo motor based on the output power curve and the real-time ambient temperature includes:

[0089] Obtaining an initial temperature rise curve of the AC servo motor based on the output power curve;

[0090] The temperature rise curve is obtained based on the initial temperature rise curve and the real-time ambient temperature.

[0091] It should be noted that since the speed of the AC servo motor is kept constant and the torque of the AC servo motor is usually unchanged or changes little, as the workload of the AC servo motor increases, its power continues to increase, so that the output power curve of the AC servo motor can be obtained. The output power curve of the AC servo motor can be obtained through a deep learning model, that is, the initial load and workload increase curve of the AC servo motor are input into the deep learning model to obtain the output power curve.

[0092] It should be noted that the ambient temperature in which the AC servo motor operates will affect its heat dissipation, that is, the lower the ambient temperature, the better the natural heat dissipation of the AC servo motor; the higher the ambient temperature, the worse the natural heat dissipation of the AC servo motor. The quality of the natural heat dissipation of the AC servo motor directly affects the operating temperature of the AC servo motor, and the operating temperature of the AC servo motor affects its speed. By obtaining the temperature rise curve of the AC servo motor and obtaining the temperature rise adjustment pulse instruction of the corresponding temperature node according to the temperature rise curve, the corresponding temperature rise adjustment pulse instruction can be generated at the corresponding temperature node, thereby compensating for the influence of the temperature change on the speed of the AC servo motor, thereby maintaining the preset speed of the AC servo motor unchanged.

[0093] It should be noted that the calculation of the output power curve and the temperature rise curve can be achieved through a deep learning model, that is, the initial load and workload increase curve of the AC servo motor are input into the output power model to obtain the output power curve; the output power curve and the real-time ambient temperature are input into the temperature rise model to obtain the temperature rise curve; the calculation of the temperature rise adjustment pulse instruction can also be achieved through a deep learning model, that is, the temperature rise curve is input into the temperature rise adjustment pulse model to obtain the temperature rise adjustment pulse instruction. After the temperature rise adjustment pulse instruction is sent to the driver of the AC servo motor, the temperature rise adjustment pulse instruction, the load adjustment pulse instruction, the frequency pulse adjustment instruction, and the original pulse instruction of this period can be superimposed at the driver, and a corresponding number of pulse signals can be generated to control the speed of the AC servo motor.

[0094] In some embodiments, the fully automatic operation monitoring method for bonding silver wire based on deep learning, the wire drawing equipment further includes an equipment temperature sensor;

[0095] The method for fully automatic operation monitoring of silver bonding wire based on deep learning further includes:

[0096] Detecting the real-time device temperature of the AC servo motor by the device temperature sensor;

[0097] If the real-time device temperature is greater than the device temperature threshold, the cooling system of the AC servo motor is turned on to cool the AC servo motor.

[0098] It should be noted that AC servo motors usually need to operate within a certain temperature range. If the operating temperature of the AC servo motor (i.e., the real-time device temperature) is too high, it will seriously affect the working performance of the AC servo motor (such as speed and torque). At this time, it is not appropriate to adjust the speed of the AC servo motor by adjusting the pulse. Therefore, it is necessary to cool the AC servo motor through the cooling system to ensure its normal operation. The implementation equipment temperature can be 60°, 80°, etc., and this application does not make specific limitations on this.

[0099] In some embodiments, if the real-time device temperature is greater than a device temperature threshold, turning on a cooling system of the AC servo motor to cool the AC servo motor further includes:

[0100] Obtaining a cooling efficiency curve of the cooling system;

[0101] Obtaining an actual temperature drop curve for cooling the AC servo motor based on the cooling efficiency curve and the temperature rise curve;

[0102] Obtaining a temperature drop adjustment pulse instruction based on the actual temperature drop curve and the real-time ambient temperature;

[0103] The temperature drop adjustment pulse instruction is sent to the driver of the AC servo motor.

[0104] It should be noted that when cooling the AC servo motor, the AC servo motor not only has the cooling effect of the cooling system, but also has natural heat dissipation to the environment. Therefore, the natural heat dissipation curve of the AC servo motor can be obtained through the implementation equipment temperature and the real-time ambient temperature. By combining (i.e., superimposing, or superimposing after calculation) the natural heat dissipation curve and the actual cooling curve, the temperature drop adjustment pulse instruction can be obtained, so that the preset speed of the AC servo motor remains unchanged.

[0105] It should be noted that the cooling efficiency curve can be obtained through actual measurement, and the actual cooling curve and the temperature drop adjustment pulse instruction can be calculated and realized through a deep learning model, that is, the cooling efficiency curve and the temperature rise curve are input into the actual cooling model to obtain the actual cooling curve; the actual cooling curve and the real-time ambient temperature are input into the temperature drop adjustment model to obtain the temperature drop adjustment pulse instruction. After the temperature drop adjustment pulse instruction is sent to the driver of the AC servo motor, the temperature drop adjustment pulse instruction, the load adjustment pulse instruction, the frequency pulse adjustment instruction, and the original pulse instruction of this period can be superimposed at the driver, and a corresponding number of pulse signals can be generated to control the speed of the AC servo motor.

[0106] In some embodiments, the fully automatic operation monitoring method for bonding silver wire based on deep learning, the wire drawing equipment further includes a vibration sensor;

[0107] The deep learning-based fully automatic operation monitoring method for silver bonding wires further includes:

[0108] Obtaining a real-time vibration level of the AC servo motor through a vibration sensor provided on the AC servo motor;

[0109] obtaining a vibration adjustment pulse instruction based on the real-time vibration level;

[0110] The vibration adjustment pulse instruction is sent to the driver of the AC servo motor.

[0111] It should be noted that the vibration of the AC servo motor will also affect its rotational speed. The vibration adjustment pulse instruction based on the real-time vibration level can be obtained through a deep learning model, that is, the real-time vibration level is input into the vibration adjustment model to obtain the vibration adjustment pulse instruction. After the vibration adjustment pulse instruction is sent to the driver of the AC servo motor, the vibration adjustment pulse instruction, the temperature rise adjustment pulse instruction, the temperature drop adjustment pulse instruction, the load adjustment pulse instruction, the frequency pulse adjustment instruction, and the original pulse instruction of this period can be superimposed at the driver to generate a corresponding number of pulse signals to control the rotational speed of the AC servo motor. If the real-time vibration level is greater than the vibration threshold, a vibration abnormality reminder is issued.

[0112] In summary, the embodiment of the present application provides a fully automatic operation monitoring method for bonding silver wire based on deep learning, which is applied to a wire drawing device for bonding silver wire, wherein the wire drawing device includes an AC servo motor, which is connected to a wire take-up reel and is used to draw the bonding silver wire; the fully automatic operation monitoring method for bonding silver wire based on deep learning includes: obtaining the real-time frequency of the power grid; inputting the real-time frequency into a frequency pulse adjustment model to obtain a frequency adjustment pulse instruction; obtaining a current frequency pulse instruction based on the frequency adjustment pulse instruction and the current original pulse instruction; sending the current frequency pulse instruction to the driver of the AC servo motor to keep the preset speed of the AC servo motor unchanged, and being able to adjust the driving pulse instruction of the AC servo motor according to the real-time frequency of the power grid, and being able to solve the current problem of instability of the motor working speed caused by the instability of the power grid frequency, thereby causing inconsistent forming quality of the bonding silver wire after drawing, thereby improving the consistency of the forming quality of the bonding silver wire.

[0113] Second, as Figure 2 As shown, the embodiment of the present application provides a fully automatic operation monitoring system for bonding silver wire based on deep learning, characterized in that it is applied to a wire drawing device for bonding silver wire, the wire drawing device includes an AC servo motor, the AC servo motor is connected to a wire take-up reel, and is used to draw the bonding silver wire;

[0114] The deep learning-based fully automatic operation monitoring system for silver bonding wires includes:

[0115] An acquisition module 210 is used to acquire the real-time frequency of the power grid;

[0116] The calculation module 220 is configured to input the real-time frequency into a frequency pulse adjustment model to obtain a frequency adjustment pulse instruction; and obtain a current period frequency pulse instruction based on the frequency adjustment pulse instruction and the current period original pulse instruction;

[0117] The adjustment module 230 is configured to send the current frequency pulse instruction to the driver of the AC servo motor so as to keep the preset speed of the AC servo motor unchanged.

[0118] In some embodiments, obtaining the current period frequency pulse instruction based on the frequency adjustment pulse instruction and the current period original pulse instruction includes:

[0119] Record the current time of obtaining the real-time frequency;

[0120] Determine the arrival time of the grid current at the AC servo motor based on the current time and the distance to be transmitted;

[0121] determining an adjustment period for adjusting the AC servo motor based on the arrival time;

[0122] The original pulse instruction of the current period is obtained based on the adjustment period, and the frequency pulse instruction of the current period is obtained based on the frequency adjustment pulse instruction and the original pulse instruction of the current period.

[0123] In some embodiments, the deep learning-based fully automatic operation monitoring method for silver bonding wires further includes:

[0124] Obtaining a workload increase curve of the AC servo motor when it is operating based on the preset rotational speed and the physical parameters of the bonding silver wire;

[0125] Inputting the workload increase curve into a load pulse adjustment model to obtain a load adjustment pulse instruction;

[0126] Obtaining a current period load frequency pulse instruction based on the load adjustment pulse instruction and the current period frequency pulse instruction;

[0127] The current load frequency pulse instruction is sent to the driver of the AC servo motor.

[0128] In some embodiments, the wire drawing apparatus further comprises an ambient temperature sensor;

[0129] The method for fully automatic operation monitoring of silver bonding wire based on deep learning further includes:

[0130] Detecting the real-time ambient temperature by the ambient temperature sensor;

[0131] Obtaining an output power curve of the AC servo motor based on an initial load and a working load increase curve of the AC servo motor;

[0132] Obtaining a temperature rise curve of the AC servo motor based on the output power curve and the real-time ambient temperature;

[0133] Obtaining a temperature rise adjustment pulse instruction for a corresponding temperature node based on the temperature rise curve;

[0134] The temperature rise adjustment pulse instruction is sent to the driver of the AC servo motor.

[0135] In some embodiments, obtaining the temperature rise curve of the AC servo motor based on the output power curve and the real-time ambient temperature includes:

[0136] Obtaining an initial temperature rise curve of the AC servo motor based on the output power curve;

[0137] The temperature rise curve is obtained based on the initial temperature rise curve and the real-time ambient temperature.

[0138] In some embodiments, the wire drawing device further includes a device temperature sensor;

[0139] The deep learning-based fully automatic operation monitoring method for silver bonding wires further includes:

[0140] Detecting the real-time device temperature of the AC servo motor by the device temperature sensor;

[0141] If the real-time device temperature is greater than the device temperature threshold, the cooling system of the AC servo motor is turned on to cool the AC servo motor.

[0142] In some embodiments, if the real-time device temperature is greater than a device temperature threshold, turning on a cooling system of the AC servo motor to cool the AC servo motor further includes:

[0143] Obtaining a cooling efficiency curve of the cooling system;

[0144] Obtaining an actual temperature drop curve for cooling the AC servo motor based on the cooling efficiency curve and the temperature rise curve;

[0145] Obtaining a temperature drop adjustment pulse instruction based on the actual temperature drop curve and the real-time ambient temperature;

[0146] The temperature drop adjustment pulse instruction is sent to the driver of the AC servo motor.

[0147] In some embodiments, the wire drawing apparatus further comprises a vibration sensor;

[0148] The deep learning-based fully automatic operation monitoring method for silver bonding wires further includes:

[0149] Obtaining a real-time vibration level of the AC servo motor through a vibration sensor provided on the AC servo motor;

[0150] obtaining a vibration adjustment pulse instruction based on the real-time vibration level;

[0151] The vibration adjustment pulse instruction is sent to the driver of the AC servo motor.

[0152] In a third aspect, an embodiment of the present application provides an electronic device, comprising a fully automatic operation monitoring system for silver bonding wire based on deep learning as described in any embodiment of the second aspect.

[0153] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (apparatus, systems), and / or computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the steps in the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0154] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0155] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0156] The specific embodiments described above further illustrate the purpose, technical solutions and beneficial effects of the present application. It should be understood that the above description is only a specific embodiment of the present application and is not intended to limit the scope of protection of the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should be included in the scope of protection of the present application.

Claims

1. A method for fully automatic operation monitoring of silver bonding wire based on deep learning, characterized in that: A wire drawing device for silver bonding wire, comprising an AC servo motor connected to a take-up reel and used for drawing the silver bonding wire; The deep learning-based fully automatic operation monitoring method for silver bonding wires comprises: Get the real-time frequency of the power grid; Inputting the real-time frequency into a frequency pulse adjustment model to obtain a frequency adjustment pulse instruction; Obtaining a current period frequency pulse instruction based on the frequency adjustment pulse instruction and the current period original pulse instruction; The current frequency pulse instruction is sent to the driver of the AC servo motor to keep the preset speed of the AC servo motor unchanged.

2. The method for fully automatic operation monitoring of silver bonding wire based on deep learning according to claim 1, characterized in that: The obtaining of the current period frequency pulse instruction based on the frequency adjustment pulse instruction and the current period original pulse instruction includes: Record the current time of obtaining the real-time frequency; Determine the arrival time of the grid current at the AC servo motor based on the current time and the distance to be transmitted; determining an adjustment period for adjusting the AC servo motor based on the arrival time; The original pulse instruction of the current period is obtained based on the adjustment period, and the frequency pulse instruction of the current period is obtained based on the frequency adjustment pulse instruction and the original pulse instruction of the current period.

3. The method for fully automatic operation monitoring of silver bonding wire based on deep learning according to claim 1, characterized in that: Also includes: Obtaining a workload increase curve of the AC servo motor when it is operating based on the preset rotational speed and the physical parameters of the bonding silver wire; Inputting the workload increase curve into a load pulse adjustment model to obtain a load adjustment pulse instruction; Obtaining a current period load frequency pulse instruction based on the load adjustment pulse instruction and the current period frequency pulse instruction; The current load frequency pulse instruction is sent to the driver of the AC servo motor.

4. The method for fully automatic operation monitoring of silver bonding wire based on deep learning according to claim 1, characterized in that: The wire drawing equipment further includes an ambient temperature sensor; The method for fully automatic operation monitoring of silver bonding wire based on deep learning further includes: Detecting the real-time ambient temperature by the ambient temperature sensor; Obtaining an output power curve of the AC servo motor based on an initial load and a working load increase curve of the AC servo motor; Obtaining a temperature rise curve of the AC servo motor based on the output power curve and the real-time ambient temperature; Obtaining a temperature rise adjustment pulse instruction for a corresponding temperature node based on the temperature rise curve; The temperature rise adjustment pulse instruction is sent to the driver of the AC servo motor.

5. The method for fully automatic operation monitoring of silver bonding wire based on deep learning according to claim 4, characterized in that: The step of obtaining a temperature rise curve of the AC servo motor based on the output power curve and the real-time ambient temperature includes: Obtaining an initial temperature rise curve of the AC servo motor based on the output power curve; The temperature rise curve is obtained based on the initial temperature rise curve and the real-time ambient temperature.

6. The method for fully automatic operation monitoring of silver bonding wire based on deep learning according to claim 4, characterized in that: The wire drawing equipment further includes an equipment temperature sensor; The method for fully automatic operation monitoring of silver bonding wire based on deep learning further includes: Detecting the real-time device temperature of the AC servo motor by the device temperature sensor; If the real-time device temperature is greater than the device temperature threshold, the cooling system of the AC servo motor is turned on to cool the AC servo motor.

7. The method for fully automatic operation monitoring of silver bonding wire based on deep learning according to claim 6, characterized in that: If the real-time device temperature is greater than the device temperature threshold, the cooling system of the AC servo motor is turned on to cool the AC servo motor, further comprising: Obtaining a cooling efficiency curve of the cooling system; Obtaining an actual temperature drop curve for cooling the AC servo motor based on the cooling efficiency curve and the temperature rise curve; Obtaining a temperature drop adjustment pulse instruction based on the actual temperature drop curve and the real-time ambient temperature; The temperature drop adjustment pulse instruction is sent to the driver of the AC servo motor.

8. The method for fully automatic operation monitoring of silver bonding wire based on deep learning according to any one of claims 1 to 7, characterized in that: The wire drawing equipment further includes a vibration sensor; The method for fully automatic operation monitoring of silver bonding wire based on deep learning further includes: Acquiring a real-time vibration level of the AC servo motor through a vibration sensor provided on the AC servo motor; obtaining a vibration adjustment pulse instruction based on the real-time vibration level; The vibration adjustment pulse instruction is sent to the driver of the AC servo motor.

9. A fully automatic operation monitoring system for silver bonding wire based on deep learning, characterized in that: A wire drawing device for silver bonding wire, comprising an AC servo motor connected to a take-up reel and used for drawing the silver bonding wire; The deep learning-based fully automatic operation monitoring system for silver bonding wires includes: An acquisition module is used to obtain the real-time frequency of the power grid; A calculation module, configured to input the real-time frequency into a frequency pulse adjustment model to obtain a frequency adjustment pulse instruction; and obtain a current period frequency pulse instruction based on the frequency adjustment pulse instruction and the current period original pulse instruction; The adjustment module is used to send the current frequency pulse instruction to the driver of the AC servo motor so that the preset speed of the AC servo motor remains unchanged.

10. An electronic device, characterized in that: Including the fully automatic operation monitoring system for bonding silver wire based on deep learning as described in claim 9.

Images