Lubricating oil quantity adjustment system and lubricating oil quantity adjustment method

Abstract

A lubricating oil quantity adjustment system and a lubricating oil quantity adjustment method. The lubricating oil quantity adjustment system includes a storage device and a processor, and is connected to a machine via a data acquisition device. The machine includes a motor. The data acquisition device acquires current information of the motor. The storage device stores a machine learning model, wherein the machine learning model is trained using a training data set including multiple current information of motor operation and multiple temperature values ​​measured when the machine is operating. The processor is configured to use the data acquisition device to acquire the current information of the motor's current operation, use the machine learning model to predict the temperature value of the machine when the motor is operating under the current information, and calculate and adjust the lubricating oil quantity suitable for use when the machine is operating according to the predicted temperature value.

Description

Technical Field

[0001] The present invention relates to a lubricating oil quantity adjustment system and a lubricating oil quantity adjustment method. Background Art

[0002] The traditional oil supply system is not linked to the machine data and mainly relies on fixed-time quantitative oiling. However, in fact, the crankshaft punching machine uses a motor to drive the crankshaft, and the physical slider stroke is used to determine the crankshaft's swing frequency. The shorter the stroke, the higher the swing frequency. As a result, the temperature of the machine rises faster and the amount of oil required also increases.

[0003] With the advancement of servo technology, some machines are now equipped with automatic lubrication systems, which can adjust various processing modes and automatically provide the right amount of lubricant to lubricate the various components of the machine. However, the amount of lubricant provided by the automatic lubrication system is usually designed with an over-design, that is, the automatic lubrication system usually uses the most stringent processing conditions to supply the amount of lubricant, which often results in too much lubricant, causing problems such as lubricant waste and recycling. Summary of the invention

[0004] The present invention provides a lubricating oil quantity adjustment system and a lubricating oil quantity adjustment method, which can more efficiently provide a lubricating oil quantity suitable for use during machine operation.

[0005] The lubricating oil quantity adjustment system disclosed in the present invention includes a storage device and a processor, and is connected to a machine via a data acquisition device. The machine includes a motor. The data acquisition device acquires the current information of the motor. The storage device stores a machine learning model. The machine learning model is trained using a training data set including multiple current information of motor operation and multiple temperature values ​​measured when the machine is operating. The processor is coupled to the data acquisition device and the storage device to use the data acquisition device to acquire the current information of the current operation of the motor, use the machine learning model to predict the temperature value of the machine when the motor is operating under the current information, and calculate and adjust the amount of lubricating oil suitable for use when the machine is operating according to the predicted temperature value.

[0006] The lubricating oil quantity adjustment method disclosed in the present invention is suitable for adjusting the lubricating oil quantity used during machine operation by an electronic device. The method includes the following steps: obtaining the current information of the current operation of the motor of the machine; using a machine learning model to predict the temperature value of the machine when the motor operates under the current information, wherein the machine learning model is trained using a training data set including multiple current information of the motor operation and multiple temperature values ​​measured during the machine operation; and calculating and adjusting the lubricating oil quantity suitable for use during machine operation based on the predicted temperature value.

[0007] Based on the above, the lubricating oil quantity adjustment system and lubricating oil quantity adjustment method disclosed in the present invention can use machine learning to predict the temperature value of the machine when the motor operates under specific current information, and calculate and adjust the lubricating oil quantity suitable for use during machine operation to provide a lubricating oil quantity suitable for use during machine operation.

[0008] In order to make the present disclosure more clearly understood, embodiments are given below with reference to the accompanying drawings for detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Figure 1 is a schematic diagram of a lubricating oil quantity adjustment system according to an embodiment of the present disclosure.

[0010] Figure 2 is a flow chart of a method for adjusting the amount of lubricating oil according to an embodiment of the present disclosure.

[0011] Figure 3 It is a schematic diagram of a machine learning model according to an embodiment of the present disclosure.

[0012] Figure 4 is a schematic diagram of a recursive neural network according to an embodiment of the present disclosure.

[0013] Figure 5 is a schematic diagram of a machine at different stages according to an embodiment of the present disclosure.

[0014] Figure 6 The present invention is a flow chart of calculating the amount of lubricating oil suitable for use during machine operation according to a predicted temperature value according to an embodiment of the present invention.

[0015] Figure 7 is a flow chart of a method for adjusting the amount of lubricating oil according to another embodiment of the present disclosure.

[0016] Reference numerals list

[0017] 100: Lubricating oil quantity adjustment system

[0018] 110: Machine

[0019] 111: Motor

[0020] 120: Data acquisition equipment

[0021] 130: Storage device

[0022] 131: Machine Learning Models

[0023] 131a: Recurrent Neural Networks

[0024] 140: Processor

[0025] X: Input layer

[0026] S: Hidden layer

[0027] O: Output layer

[0028] x t 、x t-1 、x t+1 : Output of the input layer

[0029] s t 、s t-1 、s t+1 : Output of the hidden layer

[0030] o t , o t-1 , o t+1 : The predicted temperature value output by the output layer

[0031] U, V, W: weights

[0032] t: time point

[0033] S201~S203, S601~S604, S701~S703: Steps DETAILED DESCRIPTION

[0034] Figure 1 is a schematic diagram of a lubricating oil quantity adjustment system 100 according to an embodiment of the present disclosure. Figure 1 The lubricating oil quantity adjustment system 100 mainly includes a storage device 130 and a processor 140, and the lubricating oil quantity adjustment system 100 is connected to the machine 100 via a data acquisition device 120. The lubricating oil quantity adjustment system 100 can be an electronic device such as an independent computer or a controller of the machine 100.

[0035] The machine 110 is, for example, a punching machine including a motor 111 or other machine tools or mechanical equipment that requires lubrication or cooling with lubricating oil, but is not limited here.

[0036] The data acquisition device 120 is, for example, a wired connection device such as a universal serial bus (USB), RS232, a universal asynchronous receiver / transmitter (UART), an inter-integrated circuit (I2C), a serial peripheral interface (SPI), a display port, thunderbolt, or a local area network (LAN) interface, or a wireless connection device supporting a communication protocol such as wireless fidelity (Wi-Fi), RFID, Bluetooth, infrared, near-field communication (NFC), or device-to-device (D2D). The data acquisition device 120 is coupled to the motor 111 to acquire current information of the motor 111.

[0037] The storage device 130 is, for example, any type of fixed or removable random access memory (RAM), read-only memory (ROM), flash memory, hard disk or other similar device or a combination of these devices, and is used to store programs that can be executed by the processor 140. In some embodiments, the storage device 130 can also store a machine learning model 131, and the machine learning model 131 is, for example, a convolutional neural network (CNN), a recurrent neural network (RNN) or a long short-term memory (LSTM) recurrent neural network, which is not limited by the present disclosure.

[0038] The processor 140 is coupled to the data acquisition device 120 and the storage device 130 to control the operation of the lubricating oil quantity adjustment system 100. In some embodiments, the processor 140 is, for example, a central processing unit (CPU), or other programmable general-purpose or special-purpose microprocessor, digital signal processor (DSP), programmable controller, application specific integrated circuit (ASIC), field programmable gate array (FPGA), programmable logic controller (PLC) or other similar devices or a combination of these devices, and can load and execute the program stored in the storage device 130 to perform the lubricating oil quantity adjustment method of the embodiment of the present disclosure.

[0039] Figure 2 This is a flow chart of a method for adjusting the amount of lubricating oil according to an embodiment of the present disclosure. Figure 1 and Figure 2 The method of this embodiment is applicable to Figure 1 The lubricating oil quantity adjustment system 100 is provided, and the detailed steps of the lubricating oil quantity adjustment method disclosed in the present invention are described below with reference to the various components of the lubricating oil quantity adjustment system 100.

[0040] In step S201 , the processor 140 may be configured to utilize the data acquisition device 120 to acquire the current information of the current operation of the motor 111 of the machine 110 .

[0041] In step S202, the processor 140 may be configured to use the machine learning model 131 to predict the temperature value of the machine 110 when the motor 111 operates under this current information, wherein the machine learning model 131 is trained using a training data set including multiple current information of the motor 111 and multiple temperature values ​​measured when the machine 110 operates.

[0042] For example, Table 1 is an example of a training data set. The training data content shown in Table 1 is only an example, and the present disclosure is not limited to this. At time point T1, the processor 140 can use the data acquisition device 120 to obtain the current information (e.g., 1.5 amps) of the motor 111 operation, and use this current information and the temperature value (e.g., 20°C) measured when the machine 110 is operated at time point T1 as training data 1. At time point T2, the processor 140 can use the data acquisition device 120 to obtain the current information (e.g., 2 amps) of the motor 111 operation, and use this current information and the temperature value (e.g., 22°C) measured when the machine 110 is operated at time point T2 as training data 2. By analogy, a training data set including multiple training data can be obtained. In some embodiments, the processor 140 can be configured to take the root mean square (RMS) of the current information obtained in a time segment before using it as training data.

[0043]

[0044]

[0045] Table 1

[0046] Figure 3 is a schematic diagram of a machine learning model 131 according to an embodiment of the present disclosure. Figure 3 , the machine learning model 131 of this embodiment includes operation layers such as an input layer X, a hidden layer S, and an output layer O, and each operation layer may include multiple neurons. In other embodiments, the machine learning model 131 may include multiple hidden layers, and each operation layer may include a different number of neurons, which is not limited here.

[0047] In some embodiments, the machine learning model 131 adopts a recursive neural network suitable for processing temporal data such as contextual data. Figure 4 is a schematic diagram of a recursive neural network according to an embodiment of the present disclosure. Figure 3 and Figure 4 , the recursive neural network 131a of this embodiment is Figure 3 In one embodiment of the machine learning model 131, the operations are sequential. In particular, in the recursive neural network 131a, the training data of the hidden layer S comes from two sources, one from the output data x of the input layer X, and the other from the previous operation result s of the hidden layer S. t-1 , where U, V, and W are the weights used in the hidden layer S operation. The detailed operation method will be described in detail later. Expanding the recursive neural network 131a shows that the final result at time (t+1) is t+1It is the result of the interaction of the input at that moment and all historical data. Based on this, the recursive neural network 131a can achieve the purpose of modeling the time series, which is in line with the situation of continuous processing of the punch press.

[0048] Please refer to the following formulas 1 and 2. When performing the operation of the hidden layer S, the processor 140 can be configured to convert the output x of the input layer X at that time (time point t) into t and the output s of the previous hidden layer S t-1 As the input of the current hidden layer S, multiple neurons in the hidden layer S use the activation function f1 to calculate the output s of the current hidden layer S t , and the output layer O converts the output s of the hidden layer S into t Use the activation function f2 to convert to output o t , where U, V, W are weights, b s 、b o is the offset value.

[0049] s t =f1(Ux t +Ws t-1 +b s )...(Formula 1)

[0050] 0 t =f2(Vs t +b o )...(Formula 2)

[0051] In some embodiments, x t is the current information at time point t, s t-1 is the output of the hidden layer S at time point t-1, s t is the output of the hidden layer S at time point t, o t is the predicted temperature value output by the output layer O at time point t, the activation function f1 is, for example, a sigmoid function or a tanh function, and the activation function f2 is, for example, a softmax function, but is not limited thereto.

[0052] In some embodiments, the processor 140 may, for example, input the current information of the training data in Table 1 into the machine learning model 131, and obtain the predicted temperature value as shown in Table 2. At this time, the processor 140 may be configured to compare the predicted temperature value with the measured temperature value, and update the weight of each neuron in the hidden layer S according to the comparison result. In some embodiments, the processor 140 may be configured to use the predicted temperature value output by the machine learning model 131 and the (actually measured) temperature value to calculate the loss function and use it to measure whether the prediction result of the machine learning model 131 is accurate enough, and update the weight of each neuron in the hidden layer S accordingly. In other embodiments, the processor 140 may be configured to use the gradient descent method (GD) or the back propagation method (BP) to update the weight of each neuron in the hidden layer S, and the present disclosure is not limited to this.

[0053] Training Data Time Current information Predicted temperature value Temperature value 1 <![CDATA[T1]]> 1.5 amps 18℃ 20℃ 2 <![CDATA[T2]]> 2 amps 20℃ 22℃ 3 <![CDATA[T3]]> 2.5 amps 23℃ 25℃ … … … … …

[0054] Table 2

[0055] After updating the weights of each neuron in the hidden layer S, the processor 140 may be configured to repeat the above steps (i.e., performing operations using formulas 1 and 2, and comparing the predicted temperature value with the measured temperature value to update the weights) to train the machine learning model 131.

[0056] After completing the training of the machine learning model 131 , the processor 140 can use the (trained) machine learning model 131 to predict the temperature value of the platform 110 when the motor 111 operates under the current information.

[0057] Please go back Figure 2 After obtaining the predicted temperature value (step S202), in step S203, the processor 140 may be configured to calculate and adjust the amount of lubricating oil suitable for use when the machine 110 is operated according to the predicted temperature value, and its implementation will be described by the following embodiment.

[0058] Figure 5 is a schematic diagram of different stages of operation of the machine 110 according to an embodiment of the present disclosure. Figure 5 As shown in Table 3, the machine 110 can be divided into three stages according to whether it is in operation and whether lubricating oil is provided. Among them, the cooling stage is a stage in which the machine 110 stops operating and the lubricating oil stops being provided, the standby stage is a stage in which the machine 110 stops operating and the lubricating oil is provided, and the heating stage is a stage in which the machine 110 is in operation and the lubricating oil is provided.

[0059] Operational Phase Machine status Lubrication status Purpose Heating stage operate Oil supply Input Power Standby stage stop Oil supply Lubricating cooling / heating capability Cooling stage stop Stop fuel supply Heat dissipation capacity

[0060] Table 3

[0061] Figure 6 This is a flow chart of calculating the amount of lubricating oil suitable for machine operation according to the predicted temperature value according to one embodiment of the present disclosure. Figure 1 and Figure 6 The method of this embodiment is applicable to Figure 1 The lubricating oil quantity adjustment system 100 is provided, and the detailed steps of the lubricating oil quantity adjustment method disclosed in the present invention are described below with reference to the various components of the lubricating oil quantity adjustment system 100.

[0062] In step S601 , during the cooling stage, the processor 140 may be configured to utilize at least one heat dissipation related parameter of the machine 110 and the measured temperature change of the machine 110 to calculate the heat dissipation characteristic parameters of the machine 110 for natural heat dissipation.

[0063] Please refer to the following formulas 3-5. In some embodiments, the heat dissipation related parameters of the machine 110 may include the thermal conductivity h, the machine surface area A, the machine density ρ, the machine specific heat C p Or the machine volume V. Since the machine 110 is stopped and the lubricating oil is stopped during the cooling stage, the friction heat W in Formula 3 and Formula 5 μ and cooling (lubrication) heat W c are all 0, so the total heat W in Formula 3 and Formula 5 is 0. The processor 140 can be configured to use the measured temperature changes of the machine 110 at different time points, as well as the thermal conductivity h, machine surface area A, machine density ρ, machine specific heat C of the machine 110 p and the machine volume V, and the heat dissipation characteristic parameter m of the natural heat dissipation of the computer table 110 according to Formula 4.

[0064] W=W μ +W c ...(Formula 3)

[0065]

[0066]

[0067] Where W is the total heat, W μ is the friction heat, W c is the cooling (lubrication) heat. In the embodiment of the present disclosure, W c Also called the heat impact parameter.

[0068] In step S602 , during the standby stage, the processor 140 may be configured to calculate a heat impact parameter of the amount of lubricating oil used on the heat dissipation of the machine 110 using at least one heat dissipation related parameter of the lubricating oil, the measured temperature change of the machine 110 , and the calculated heat dissipation characteristic parameter.

[0069] Please refer to the following formula 6. In some embodiments, the heat dissipation related parameters of the lubricating oil may include the oil volume M or the specific heat s. Since the machine 110 is stopped and lubricating oil is provided in the standby stage, the friction heat W in formula 3 is μ The cooling (lubrication) heat W is 0 c If it is not 0, the processor 140 may be configured to calculate the heat impact parameter W using Formula 6 c .

[0070] W c =MsΔT c ...(Formula 6)

[0071] Where M is the amount of lubricating oil, s is the specific heat, ΔT c is the temperature difference.

[0072] In step S603, during the temperature rise phase, the processor 140 may be configured to utilize at least one operating parameter of the machine 110 and the calculated heat dissipation characteristic parameter m and heat impact parameter W. c , a function showing the relationship between the amount of lubricating oil used when the computer table 110 is operating and the temperature change of the computer table 110.

[0073] Please refer to the following formula 7. In some embodiments, at least one operating parameter of the machine 110 may include a friction coefficient η, a punching pressure F, a diameter d, a friction surface area πdL, or a punching number n per minute. Since the machine 110 is in operation and lubricating oil is provided during the heating stage, the friction heat W in formula 3 is μ Not 0 and cooling (lubrication) heat W c is not 0, the processor 140 may be configured to calculate the friction heat W using Formula 7 μ Then, the relationship function between the complete lubricating oil quantity and the machine temperature change is derived (i.e., ).

[0074] W μ =πdLηFn...(Formula 7)

[0075] Among them, W μ is the friction heat, η is the friction coefficient, F is the punching pressure, d is the diameter, πdL is the friction surface area, and n is the number of punches per minute.

[0076] In step S604, after calculating the relationship function, the processor 140 may be configured to use the relationship function to calculate the amount of lubricant suitable for use when the machine 110 is operated at the predicted temperature value. In other words, the processor 140 may use the relationship function to calculate the amount of lubricant suitable for use when the machine 110 is operated according to the predicted temperature value.

[0077] Please go back Figure 2 In step S203 , after calculating the amount of lubricating oil suitable for use during the operation of the machine 110 , the processor 140 may be configured to adjust the amount of lubricating oil suitable for use during the operation of the machine 110 according to the calculated amount of lubricating oil.

[0078] In some embodiments, in addition to using the current information of the motor operation as the input of the machine learning model to train the machine learning model, the amount of lubricating oil used during the machine operation can also be used as an input to train the machine learning model, thereby learning the relationship between the current information, the amount of lubricating oil and the temperature change.

[0079] In detail, Figure 7 is a flow chart of a method for adjusting the amount of lubricating oil according to another embodiment of the present disclosure. Figure 1 and Figure 7 The method of this embodiment is applicable to Figure 1 The lubricating oil quantity adjustment system 100 is provided, and the detailed steps of the lubricating oil quantity adjustment method disclosed in the present invention are described below with reference to the various components of the lubricating oil quantity adjustment system 100.

[0080] In step S701 , the processor 140 may be configured to obtain the current information of the motor 111 of the machine 110 currently operating and the amount of lubricating oil used when the machine 110 is operating.

[0081] In step S702, the processor 140 may be configured to use the machine learning model 131 to predict the temperature value of the machine 110 when the motor 111 operates under current information, wherein the machine learning model 131 is trained using a training data set including multiple current information of the motor 111 operation, the amount of lubricating oil used when the machine 110 operates, and multiple temperature values ​​measured when the machine 110 operates.

[0082] In some embodiments, the data acquisition device 120 may acquire the amount of lubricating oil used when the machine 110 is operated. Table 4 is another example of a training data set. The training data content shown in Table 4 is only an example, and the present disclosure is not limited thereto. At time point T1, the processor 140 may use the data acquisition device 120 to acquire the current information (e.g., 1.5 amps) of the motor 111 and the amount of lubricating oil (e.g., 200 grams) used when the machine 110 is operated, and use the current information, the amount of lubricating oil, and the temperature value (e.g., 20° C.) measured when the machine 110 is operated at time point T1 as training data 1. At time point T2, the processor 140 may use the data acquisition device 120 to acquire the current information (e.g., 2 amps) of the motor 111 and the amount of lubricating oil (e.g., 250 grams) used when the machine 110 is operated, and use the current information, the amount of lubricating oil, and the temperature value (e.g., 22° C.) measured when the machine 110 is operated at time point T2 as training data 2. By analogy, a training data set including multiple training data may be obtained.

[0083] Training Data Time Current information Lubricating oil quantity Temperature value 1 <![CDATA[T1]]> 1.5 amps 200 g 20℃ 2 <![CDATA[T2]]> 2 amps 250 g 22℃ 3 <![CDATA[T3]]> 2.5 amps 300 g 25℃ … … … … …

[0084] Table 4

[0085] The processor 140 may be configured to use Figure 4 The machine learning model 131 is established by using the architecture of the recursive neural network 131a shown in FIG. 4, and the machine learning model 131 is trained using the training data set shown in Table 4. Different from the previous embodiment, in this embodiment, the processor 140 inputs the current information and the amount of lubricating oil into the input layer of the machine learning model 131 in a time series, including the output x of the input layer X at that time (time point t) t (including current information and lubricating oil volume) and the output s of the previous hidden layer S t-1 As the input of the current hidden layer S, multiple neurons in the hidden layer S use the activation function f1 to calculate the output s of the current hidden layer S t , and the output layer O converts the output s of the hidden layer S into t Use the activation function f2 to convert to the predicted temperature value o t The detailed implementation method is the same as above Figure 4 The embodiments are the same as those of the present invention and will not be described in detail here.

[0086] In some embodiments, the processor 140 may input the current information and lubricant amount of the training data in Table 4 into the machine learning model 131, and obtain the predicted temperature value as shown in Table 5. At this time, the processor 140 may be configured to compare the predicted temperature value with the measured temperature value, and update the weight of each neuron in the hidden layer S according to the comparison result. The processor 140 may update the weight of each neuron in the hidden layer S by calculating the loss function, or using the gradient descent method or the back propagation method. The detailed implementation method is the same as the above-mentioned Figure 4 The embodiments are the same as those of the present invention and will not be described in detail here.

[0087]

[0088]

[0089] Table 5

[0090] After the training of the machine learning model 131 is completed, the processor 140 may be configured to use the (trained) machine learning model 131 to predict the temperature value of the platform 110 when the motor 111 operates with the current information.

[0091] In step S703, the processor 140 may be configured to calculate and adjust the amount of lubricating oil suitable for use when the machine 110 is operating according to the predicted temperature value. Figure 6 After calculating the amount of lubricating oil suitable for use in the operation of the machine 110, the processor 140 may be configured to adjust the amount of lubricating oil suitable for use in the operation of the machine 110 according to the calculated amount of lubricating oil.

[0092] In summary, the lubricating oil quantity adjustment system and lubricating oil quantity adjustment method disclosed herein use machine learning to predict the temperature change of the machine when the motor is operating, and calculate the amount of lubricating oil suitable for use during machine operation, thereby achieving optimization of oil quantity prediction and intelligent temperature control and lubrication. In particular, the machine learning model disclosed herein can be trained using multiple current information, lubricating oil quantity, and temperature values, and also improves the accuracy of predicted temperature.

[0093] Although the present disclosure has been disclosed as above by way of embodiments, it is not intended to limit the present disclosure. Any person skilled in the art may make slight changes and modifications without departing from the spirit and scope of the present disclosure. Therefore, the protection scope of the present disclosure shall be determined by the appended claims and their equivalent scope.

Claims

1. A lubricating oil quantity adjustment system, adapted to be connected to a machine via a data acquisition device, wherein the machine comprises a motor, the data acquisition device is coupled to the motor to obtain current information of the motor, the lubricating oil quantity adjustment system comprising: A storage device storing a machine learning model, wherein the machine learning model is trained using a training data set including a plurality of current information of the motor operation and a plurality of temperature values ​​measured when the machine is operating; as well as a processor coupled to the data acquisition device and the storage device, using the data acquisition device to acquire the current information of the current operation of the motor, using the machine learning model to predict the temperature value of the machine when the motor operates at the current information, and calculating and adjusting the amount of lubricating oil suitable for use when the machine is operating according to the predicted temperature value, Wherein, the processor is further configured to: In a cooling stage when the machine stops operating and the lubricating oil stops being provided, a heat dissipation characteristic parameter of the machine's natural heat dissipation is calculated using at least one heat dissipation-related parameter of the machine and the measured temperature change of the machine; In a standby stage when the machine stops operating and the lubricating oil is provided, a heat impact parameter of the effect of the amount of the lubricating oil used on the heat dissipation of the machine is calculated using at least one heat dissipation-related parameter of the lubricating oil, the measured temperature value change of the machine, and the calculated heat dissipation characteristic parameter; In the operation of the machine and in a temperature rise phase in which the lubricating oil is provided, a relationship function between the amount of the lubricating oil used in the operation of the machine and the temperature change of the machine is calculated using at least one operating parameter of the operation of the machine and the calculated heat dissipation characteristic parameter and the heat impact parameter, wherein the operating parameter of the machine includes one or a combination of a friction coefficient, a punching pressure, a diameter, a friction surface area, and a number of punches per minute; and The relationship function is used to calculate the amount of lubricating oil suitable for use when the machine is operating at the predicted temperature value.

2. The lubricating oil quantity adjustment system according to claim 1, wherein the machine learning model comprises an input layer, a hidden layer and an output layer, and the processor is further configured to: Inputting the current information into the input layer in time series; The output of the input layer at this time and the output of the hidden layer at the previous time are used as the input of the hidden layer, the multiple neurons of the hidden layer calculate the output of this time by using an activation function, and the output layer converts the output of the hidden layer into a predicted temperature value; Compare the predicted temperature value with the temperature value measured at that time, and update the weight of each neuron in the hidden layer according to the comparison result; as well as Repeat the above steps to train the machine learning model.

3. The lubricating oil quantity adjustment system according to claim 1, wherein the machine learning model comprises an input layer, a hidden layer, and an output layer, the data acquisition device further acquires the lubricating oil quantity used when the machine is operated, and the processor is further configured to: Inputting the current information and the lubricating oil amount into the input layer in time series; The output of the input layer at this time and the output of the hidden layer at the previous time are used as the input of the hidden layer, the multiple neurons of the hidden layer calculate the output of this time by using an activation function, and the output layer converts the output of the hidden layer into a predicted temperature value; Compare the predicted temperature value with the temperature value measured at that time, and update the weight of each neuron in the hidden layer according to the comparison result; as well as Repeat the above steps to train the machine learning model.

4. The lubricating oil quantity adjustment system according to claim 1, wherein the machine learning model comprises a convolutional neural network, a recursive neural network or a long short-term memory recursive neural network. 5 . The lubricating oil quantity adjustment system according to claim 1 , wherein the heat dissipation related parameters of the machine include one or a combination of thermal conductivity, machine surface area, machine density, machine specific heat, and machine volume. 6 . The lubricating oil quantity adjustment system according to claim 1 , wherein the heat dissipation related parameters of the lubricating oil include oil quantity and specific heat.

7. A method for adjusting the amount of lubricating oil, suitable for adjusting the amount of lubricating oil used in the operation of a machine by an electronic device, the method comprising the following steps: Acquiring current information of a motor currently operating on the machine; Using a machine learning model to predict the temperature value of the machine when the motor is operating under the current information, wherein the machine learning model is trained using a training data set including a plurality of current information of the motor operation and a plurality of temperature values ​​measured when the machine is operating; as well as Calculating and adjusting the amount of lubricating oil suitable for use when the machine is operating according to the predicted temperature value, The step of calculating and adjusting the amount of lubricating oil suitable for use when the machine is operating according to the predicted temperature value includes: In a cooling stage when the machine stops operating and the lubricating oil stops being provided, a heat dissipation characteristic parameter of the machine's natural heat dissipation is calculated using at least one heat dissipation-related parameter of the machine and the measured temperature change of the machine; In a standby stage when the machine stops operating and the lubricating oil is provided, a heat impact parameter of the effect of the amount of the lubricating oil used on the heat dissipation of the machine is calculated using at least one heat dissipation-related parameter of the lubricating oil, the measured temperature value change of the machine, and the calculated heat dissipation characteristic parameter; In the operation of the machine and in a temperature rise phase in which the lubricating oil is provided, a relationship function between the amount of the lubricating oil used in the operation of the machine and the temperature change of the machine is calculated using at least one operating parameter of the operation of the machine and the calculated heat dissipation characteristic parameter and the heat impact parameter, wherein the operating parameter of the machine includes one or a combination of a friction coefficient, a punching pressure, a diameter, a friction surface area, and a number of punches per minute; and The relationship function is used to calculate the amount of lubricating oil suitable for use when the machine is operating at the predicted temperature value.

8. The method according to claim 7, wherein the machine learning model comprises an input layer, a hidden layer, and an output layer, and the method further comprises: Inputting the current information into the input layer in time series; The output of the input layer at this time and the output of the hidden layer at the previous time are used as the input of the hidden layer, the multiple neurons of the hidden layer calculate the output of this time by using an activation function, and the output layer converts the output of the hidden layer into a predicted temperature value; Compare the predicted temperature value with the temperature value measured at that time, and update the weight of each neuron in the hidden layer according to the comparison result; as well as Repeat the above steps to train the machine learning model.

9. The method according to claim 7, wherein the machine learning model comprises an input layer, a hidden layer, and an output layer, and the method further comprises: Obtaining the amount of lubricating oil used when the machine is operating; Inputting the current information and the lubricating oil amount into the input layer in time series; The output of the input layer at this time and the output of the hidden layer at the previous time are used as the input of the hidden layer, the multiple neurons of the hidden layer calculate the output of this time by using an activation function, and the output layer converts the output of the hidden layer into a predicted temperature value; Compare the predicted temperature value with the temperature value measured at that time, and update the weight of each neuron in the hidden layer according to the comparison result; as well as Repeat the above steps to train the machine learning model.

10. The method of claim 7, wherein the machine learning model comprises a convolutional neural network, a recurrent neural network, or a long short-term memory recurrent neural network. 11 . The method according to claim 7 , wherein the heat dissipation related parameters of the machine include one or a combination of thermal conductivity, machine surface area, machine density, machine specific heat, and machine volume.

12. The method according to claim 7, wherein the heat dissipation related parameters of the lubricating oil include oil volume and specific heat.

Images