Automatic feeding method and system for feeding equipment

Abstract

The present invention relates to the field of data processing technology, and in particular to an automatic feeding method and system for a feeding device. The method comprises: dividing the internal temperature of an electrode induction gas atomization powder making device into a plurality of periodic segments, determining a first abnormality degree according to the variation law between the internal temperatures of the same sequence number in different periodic segments, determining a second abnormality degree according to the fitting error of the internal temperature under the fitting function of the periodic segment and the difference in the fitting functions of different periodic segments, determining a correction window according to the first abnormality degree and the second abnormality degree, correcting the internal temperature, and then obtaining a corrected value of the internal temperature, determining the descending speed of the master alloy bar according to the corrected value of the internal temperature and the external temperature, and realizing automatic feeding. The present invention eliminates the interference of noise on the internal temperature, makes the feeding more accurate, and helps to improve the powder making efficiency of the electrode induction gas atomization powder making device.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to an automatic feeding method and system for feeding equipment. Background Art

[0002] Electrode induction atomization powder making equipment is mainly used for the preparation of active metal, refractory metal and their alloy powders, such as pure titanium and titanium alloys, high-temperature alloys, platinum-rhodium-platinum, intermetallic compounds, etc. Powder products are widely used in additive manufacturing, powder metallurgy, hot isostatic pressing, injection molding and other manufacturing process fields.

[0003] The patent document with the announcement number "CN111014701B" discloses a method for preparing high-conductivity metal powder by electrode induction melting gas atomization, comprising: (1) placing a pre-prepared alumina funnel into an induction coil in an atomization device; (2) placing a graphite funnel inside the alumina funnel; (3) loading a high-conductivity metal rod of a specific size into the atomization device; (4) evacuating the device and then filling it with high-purity argon as a protective gas; (5) heating and melting the high-conductivity metal rod through the induction coil, and then atomizing and powdering; (6) after the powder is cooled, taking out the powder from the powder collecting tank.

[0004] The above methods can speed up the powder making speed, but the powder making efficiency of the electrode induction melting gas atomization equipment is also related to the feeding speed. The electrode induction melting gas atomization equipment rotates the master alloy rod at a certain speed and descends into the conical coil below for heating and melting. When the speed is too fast, it may cause insufficient melting of the master alloy rod, thereby affecting the quality of the powder product. When the speed is too slow, it may affect the powder making efficiency.

[0005] Therefore, how to control the descending speed of the master alloy rod to improve the powder making efficiency of the electrode induction melting gas atomization equipment and the quality of the powder product is a problem that needs to be solved urgently. Summary of the invention

[0006] In order to solve the problem that the feeding speed of the master alloy rod affects the powder making efficiency of the electrode induction melting gas atomization equipment and the quality of the powder product, the present invention provides an automatic feeding method and system for the feeding equipment.

[0007] In a first aspect, the present invention provides an automatic feeding method for a feeding device, which adopts the following technical solution:

[0008] An automatic feeding method for a feeding device comprises the following steps:

[0009] Collect the external temperature and internal temperature of the electrode induction gas atomization powder making equipment at each moment; divide the internal temperature into multiple periodic segments; form a target sequence with the internal temperatures in all periodic segments having the same serial number as the internal temperature at the current moment; determine a smoothing coefficient according to the difference between the external temperature at the current moment and the external temperature at the corresponding moment of each element in the target sequence, and the smoothing coefficient is positively correlated with the difference; predict the internal temperature at the current moment according to the smoothing coefficient, and determine the first abnormality degree according to the difference between the predicted value and the actual value; fit the periodic segment to which the internal temperature at the current moment belongs, and determine the second abnormality degree according to the fitting error of the internal temperature at the current moment and the difference between the fitting functions of different periodic segments, and the second abnormality degree is positively correlated with the fitting error and negatively correlated with the difference between the fitting functions; determine the correction window of the internal temperature at the current moment, and the size of the correction window is positively correlated with the first abnormality degree and the second abnormality degree; correct the internal temperature at the current moment according to all internal temperatures in the correction window to obtain the corrected value of the internal temperature at the current moment; determine the descending speed of the master alloy bar according to the corrected value of the internal temperature and the external temperature to realize automatic feeding.

[0010] The effects are as follows: the present invention determines the descending speed of the master alloy rod according to the corrected value of the internal temperature and the external temperature, realizes the automatic feeding of the electrode induction gas atomization powder making equipment, and improves the powder making efficiency and the quality of the powder product; the present invention eliminates the interference of noise on the internal temperature by correcting the internal temperature, makes the feeding more accurate, and helps to improve the powder making efficiency of the electrode induction gas atomization powder making equipment; the present invention determines the first abnormal degree of the internal temperature in combination with the change of the external temperature, eliminates the interference of the internal temperature change caused by the external temperature on the internal temperature change caused by noise, makes the noise removal more accurate, and improves the accuracy of the correction result of the internal temperature; the present invention determines the second abnormal degree of the internal temperature according to the change law of the internal temperature in the cycle segment and the difference of the fitting function between different cycle segments, eliminates the interference of the normal change of the internal temperature of the smelting furnace body on the noise recognition, further improves the accuracy of denoising, makes the corrected value of the internal temperature more accurate, and thus improves the accuracy of the feeding.

[0011] Preferably, the internal temperature is divided into a plurality of periodic segments, including: an internal temperature sequence formed by internal temperatures at different times; performing STL decomposition on the internal temperature sequence to obtain periodic terms, performing Fourier transform on the periodic terms, and taking the inverse of the frequency corresponding to the maximum amplitude after Fourier transform as the period length; and using the period length to divide the internal temperature sequence into a plurality of different periodic segments, the length of each periodic segment being equal to the period length.

[0012] The effect is that the present invention takes into account that the internal temperature of the smelting furnace body varies periodically, decomposes the internal temperature to obtain periodic terms, and is helpful to analyze the periodic variation of the internal temperature.

[0013] Preferably, the smoothing coefficient satisfies the expression: ;in, Indicates the smoothing coefficient at the current moment; Indicates the external temperature at the current moment; Indicates the target external temperature sequence. External temperature; represents the number of external temperatures contained in the target external temperature sequence; Indicates the extreme difference of all external temperatures; the method for obtaining the target external temperature sequence is: the external temperatures at the moments corresponding to each internal temperature in the target sequence constitute the target external temperature sequence.

[0014] Preferably, the first abnormality degree satisfies the expression: ;in, Indicates the first abnormality level of the internal temperature at the current moment; Indicates the actual value of the internal temperature at the current moment; Represents the predicted value at the current moment; represents the mean of all pairwise differences in the internal temperatures in the target sequence; represents the hyperbolic tangent function.

[0015] Preferably, the fitting of the cycle segment to which the internal temperature at the current moment belongs includes: taking the cycle segment to which the internal temperature at the current moment belongs as the target cycle segment, performing polynomial fitting on all internal temperatures in the target cycle segment using the least squares method, and obtaining a fitting function for the target cycle segment; during the fitting process, the loss value satisfies the expression: ;in, represents the loss value; Indicates the serial number of the internal temperature at the current moment in the cycle segment to which it belongs; Indicates the first The first abnormal degree of internal temperature; Indicates the first The fitted value of the internal temperature; Indicates the first Internal temperature; Represents an exponential function with a natural constant as its base.

[0016] The effect is that the present invention determines the loss value in the fitting process in combination with the first abnormal degree of each internal temperature, so that the fitting function is closer to the change of the internal temperature with a small first abnormal degree, eliminates the influence of the possible abnormal internal temperature on the fitting function, and makes the obtained fitting function more accurate.

[0017] Preferably, the second abnormality degree satisfies the expression: ;in, Indicates the second abnormality level of the internal temperature at the current moment; Indicates the serial number of the internal temperature at the current moment in the cycle segment to which it belongs; represents the fitted value of the internal temperature at the current moment; Indicates the internal temperature at the current moment; is a hyperparameter; , , Indicates The coefficients of the fitting function for the period segment; , , The coefficient of the fitting function representing the period to which the internal temperature at the current moment belongs; represents the variance of the coefficients of the fitting function for all periodic segments; represents an exponential function with a natural constant as base; represents the hyperbolic tangent function; Indicates the number of period segments.

[0018] The effect is that the present invention determines the second abnormal degree of the internal temperature based on the fitting error, and corrects the fitting error in combination with the difference in the fitting functions of different periodic segments, fully considering the periodic changes of the internal temperature, so that the obtained second abnormal degree is more accurate.

[0019] Preferably, the determining of the correction window of the internal temperature at the current moment includes: the correction window size of the internal temperature at the current moment satisfies the expression: ; The corrected window size representing the internal temperature at the current moment; Indicates the preset initial window size; Indicates the preset window size adjustment factor; Indicates the first abnormality level of the internal temperature at the current moment; Indicates the second abnormality level of the internal temperature at the current moment; Represents the rounding function; in the target sequence, construct The internal temperature at the current moment is the last data of the correction window.

[0020] Preferably, the correcting the internal temperature at the current moment according to all the internal temperatures in the correction window includes: taking the weighted average of all the internal temperatures in the correction window as the correction value of the internal temperature at the current moment, wherein the weight of each internal temperature in the correction window satisfies the expression: ;in, Indicates the correction window. The weight of the internal temperature; Indicates the correction window. The first abnormal degree of internal temperature; Indicates the correction window. a second abnormal degree of internal temperature; represents an exponential function with a natural constant as base; Indicates the correction window size.

[0021] The effect is that the present invention determines the weight of the internal temperature according to the first abnormal degree and the second abnormal degree of the internal temperature in the correction window, thereby avoiding the interference of the abnormal internal temperature on the correction value of the internal temperature at the current moment, making the correction value of the internal temperature at the current moment more accurate.

[0022] Preferably, it also includes: collecting the output power, frequency, material and diameter of the electromagnetic induction coil; determining the descent speed of the master alloy rod according to the correction value of the internal temperature and the external temperature to achieve automatic feeding, including: inputting the material of the master alloy rod, the diameter of the master alloy rod, the external temperature at the current moment, the correction value of the internal temperature, the output power and frequency of the electromagnetic induction coil into a trained neural network, and outputting the descent speed of the master alloy rod; controlling the descent of the master alloy rod in the feeding equipment according to the descent speed to achieve automatic feeding of the electrode induction atomization powder making equipment.

[0023] In a second aspect, the present invention provides an automatic feeding system for a feeding device, which adopts the following technical solution:

[0024] An automatic feeding system for a feeding device comprises: a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the automatic feeding method for a feeding device is implemented.

[0025] By adopting the above technical solution, the above-mentioned automatic feeding method of a feeding device is generated into a computer program and stored in a memory so as to be loaded and executed by a processor, thereby making a terminal device based on the memory and the processor for easy use.

[0026] The present invention has the following technical effects:

[0027] (1) The present invention determines the descending speed of the master alloy rod according to the correction value of the internal temperature and the external temperature, realizes the automatic feeding of the electrode induction gas atomization powder making equipment, and improves the powder making efficiency and the quality of the powder product;

[0028] (2) The present invention eliminates the interference of noise on the internal temperature, making the feeding more accurate and helping to improve the powder making efficiency of the electrode induction atomization powder making equipment;

[0029] (3) The present invention eliminates the interference of internal temperature changes caused by external temperature on denoising, making noise removal more accurate and improving the accuracy of the internal temperature correction result;

[0030] (4) The present invention eliminates the interference of normal changes in the internal temperature of the smelting furnace on noise identification, further improves the accuracy of the correction of the internal temperature, and thus improves the accuracy of the feeding. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] By reading the detailed description below with reference to the accompanying drawings, the above and other purposes, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-restrictive manner, and the same or corresponding reference numerals represent the same or corresponding parts.

[0032] Figure 1 The present invention is a flowchart of a method for automatically feeding a feeding device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0033] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0034] It should be understood that when the terms "first", "second", etc. are used in the claims, descriptions, and drawings of the present invention, they are only used to distinguish different objects, rather than to describe a specific order. The terms "include" and "comprise" used in the description and claims of the present invention indicate the presence of the described features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their collections.

[0035] The embodiment of the present invention discloses an automatic feeding method for a feeding device, referring to Figure 1 , comprising steps S1 to S7:

[0036] S1: Collect the external temperature and internal temperature of the electrode induction atomization powder making equipment at every moment.

[0037] Temperature sensors are deployed in the feeding device of the electrode induction gas atomization powder making equipment and the smelting furnace to collect the external ambient temperature around the feeding device of the electrode induction gas atomization powder making equipment at every moment, which is recorded as the external temperature. The temperature of the smelting furnace of the electrode induction gas atomization powder making equipment at every moment is collected and recorded as the internal temperature.

[0038] The collection frequency in the present invention is once per second. In other embodiments, the implementer can set the collection frequency according to the actual implementation situation.

[0039] At this point, the collection of external temperature and internal temperature has been achieved.

[0040] S2: Smoothing of external temperature.

[0041] It should be noted that the external temperature is the ambient temperature around the feeding device of the electrode induction aerosol powder making equipment. When the ambient temperature changes, the change is gradual and no sudden change occurs. When there is a sudden temperature change, it may be caused by noise in the process of collecting the ambient temperature. Therefore, the present invention removes the noise influence in the external temperature by filtering the external temperature.

[0042] Specifically, the external temperature at each moment is filtered by a mean filtering method to achieve smoothing of the external temperature. The external temperatures that appear subsequently in the present invention are all smoothed external temperatures.

[0043] At this point, smoothing of the external temperature is achieved.

[0044] S3: Obtain the cycle length of the internal temperature.

[0045] It should be noted that the principle of melting the master alloy rod by the electrode induction gas atomization powder making equipment is: the slowly rotating master alloy rod is lowered into the conical electromagnetic induction coil through which the high-frequency alternating current passes, and the alternating magnetic field generated by the changing current causes eddy current effects of the same size and opposite directions on the master alloy rod. The output power of the electromagnetic induction coil is consumed on the resistance of the master alloy rod, causing a thermal effect on the surface of the master alloy rod. After a certain period of time, the high-temperature alloy begins to melt to form an alloy liquid flow.

[0046] In the above process, the quantitative relationship of the thermal effect of the effective current is calculated by the Joule-Lenz law: ,in For thermal effect, is the vortex size; is the resistance of the master alloy bar; It is the time for the eddy current to pass through the mother alloy rod. When the electromagnetic power is constant and the mother alloy rod is unchanged, the size of the eddy current is related to the resistivity of the mother alloy rod. When the resistivity is larger, the eddy current size is smaller, and when the resistivity is smaller, the eddy current size is larger. Under the action of the thermal effect, the temperature of the mother alloy rod continues to rise. Since the temperature rise will cause the resistivity of the mother alloy rod to rise, the eddy current size will be reduced, the thermal effect will change, and the temperature of the mother alloy rod will change, and then the internal temperature of the smelting furnace will change. In the process of the mother alloy rod continuously rotating and descending, the part of the mother alloy rod outside the smelting furnace body continuously enters the inside of the smelting furnace body, so that the temperature of the mother alloy rod is reduced. When the temperature is reduced, the resistivity of the mother alloy rod will decrease, so that the eddy current size will increase, the thermal effect will change, and the temperature of the mother alloy rod will change, and then the internal temperature of the smelting furnace body will change. Therefore, the collected internal temperature is constantly changing and is a periodic change. Therefore, the present invention analyzes the change cycle of the internal temperature.

[0047] Specifically, the current internal temperature and the previous The internal temperature at each moment constitutes the internal temperature sequence, and the STL (Seasonal-Trend decomposition procedure based on Loess) decomposition algorithm is used to decompose the internal temperature sequence into trend terms, period terms, and residual terms. It is the preset time number, which can be set by the implementer according to the actual implementation situation, for example =1000.

[0048] The period length is obtained according to the period term. Specifically, Fourier transform is performed on the period term, and the inverse of the frequency corresponding to the maximum amplitude after Fourier transform is taken as the period length.

[0049] At this point, the cycle length of the internal temperature is obtained.

[0050] S4: Divide the internal temperature sequence into multiple periodic segments, and form a target sequence with the internal temperatures in all periodic segments having the same serial number as the internal temperature at the current moment, determine a smoothing coefficient according to the difference between the external temperature at the current moment and the external temperature at the corresponding moment of each element in the target sequence, predict the internal temperature at the current moment according to the smoothing coefficient, and determine the first abnormality degree of the internal temperature at the current moment according to the difference between the predicted value and the actual value.

[0051] It should be noted that since the internal temperature changes periodically, the internal temperature of the same serial number should be relatively consistent in different cycles, but there may be a situation where the internal temperature changes due to external temperature changes. Therefore, the present invention analyzes the changes in the internal temperature in combination with the changes in the external temperature to eliminate the influence of the external temperature changes on the recognition of noise in the internal temperature.

[0052] Specifically, the internal temperature sequence is divided into multiple different cycle segments using the cycle length, and the length of each cycle segment is equal to the cycle length. The sequence number of the internal temperature at the current moment in the cycle segment to which it belongs is used as the target sequence number. For example, if the internal temperature at the current moment is the 31st internal temperature value in the cycle segment to which it belongs, 31 is used as the target sequence number.

[0053] The internal temperatures corresponding to the target numbers in all period segments are formed into a sequence as a target sequence. The external temperatures at the moments corresponding to the internal temperatures in the target sequence are formed into a target external temperature sequence.

[0054] Get the smoothing coefficient of the current moment in the target sequence according to the target external temperature sequence:

[0055] ;

[0056] in, Indicates the smoothing coefficient at the current moment; Represents the external temperature at the current moment, that is, the last element in the target external temperature sequence; Indicates the target external temperature sequence. External temperature; represents the number of external temperatures contained in the target external temperature sequence; Represents the range of all external temperatures.

[0057] represents the average difference between the current external temperature and the remaining external temperatures in the target external temperature sequence, It is used to normalize the average difference. If the average difference is larger, it means that the external temperature at the current moment has changed significantly compared with the external temperature at other moments, which may bring about changes in the internal temperature at the current moment. At this time, a larger smoothing coefficient is set. When the exponential smoothing method is used to predict the internal temperature at the current moment in the future, more attention is paid to the changes in the internal temperature close to the current moment, so that the predicted value can reflect the changes brought about by the external temperature; if the average difference is smaller, it means that the external temperature at the current moment has changed less than the external temperature at other moments. At this time, the internal temperature at the current moment theoretically changes less than the other internal temperatures in the target sequence. At this time, a smaller smoothing coefficient is set so that when the exponential smoothing method is used to predict the internal temperature at the current moment in the future, more attention is paid to the overall change trend of the target sequence.

[0058] All internal temperatures before the internal temperature at the current moment in the target sequence are taken as data samples, and the predicted value of the internal temperature at the current moment is obtained using the exponential smoothing method according to the smoothing coefficient at the current moment.

[0059] Determine the first abnormality degree of the internal temperature at the current moment according to the difference between the actual value of the internal temperature at the current moment and the predicted value of the internal temperature at the current moment:

[0060] ;

[0061] in, Indicates the first abnormality level of the internal temperature at the current moment; Indicates the actual value of the internal temperature at the current moment; represents the predicted value of the internal temperature at the current moment; represents the mean of all pairwise differences in the internal temperatures in the target sequence; Represents a hyperbolic tangent function; if the difference between the actual value of the internal temperature at the current moment and the predicted value is greater, it means that the internal temperature at the current moment is more abnormal, and the first abnormal degree of the internal temperature at the current moment is greater; conversely, if the difference between the actual value of the internal temperature at the current moment and the predicted value is smaller, it means that the internal temperature at the current moment is more normal, and the first abnormal degree of the internal temperature at the current moment is smaller.

[0062] At this point, the first abnormality level of the internal temperature at the current moment is acquired.

[0063] S5: Fitting the period segment to which the internal temperature at the current moment belongs, and determining the second abnormality degree according to the fitting error of the internal temperature at the current moment and the difference between the fitting functions of different period segments.

[0064] Specifically, the cycle segment to which the internal temperature at the current moment belongs is taken as the target cycle segment, and all internal temperatures in the target cycle segment are fitted using the least squares method to obtain the fitting function of the target cycle segment. The expression of the loss value in the fitting process is:

[0065] ;

[0066] in, represents the loss value; Indicates the target sequence number, that is, the sequence number of the internal temperature at the current moment in the cycle segment to which it belongs; Indicates the first The first abnormal degree of internal temperature; Indicates the first The fitting value of the internal temperature, in the present invention, the fitting function adopts a cubic polynomial, and the fitting function expression is: , , , are coefficients of the polynomial. In other embodiments, the implementer may set the fitting function according to the actual implementation situation; Indicates the first Internal temperature; Represents an exponential function with a natural constant as its base.

[0067] If the target period The greater the first abnormal degree of the internal temperature, the greater the The less accurate the internal temperature is, the The smaller the reference value of the internal temperature for the fitting function, the smaller the reference value of the internal temperature for the fitting function. The smaller the first abnormal degree of the internal temperature, the The more accurate the internal temperature is, the The greater the reference value of the internal temperature for the fitting function, the greater the reference value of the internal temperature for the fitting function. As the first The weight of the internal temperature is used to obtain the loss value.

[0068] For loss value Respectively , , The partial derivative of , , , and the obtained result is substituted into the fitting function expression to obtain the fitting function of the target period. It should be noted that the partial derivative of the loss value is calculated, and the coefficients when the partial derivative is equal to 0 are solved to obtain the fitting function. This process is a well-known technique in the least squares method, and the specific principle will not be repeated here.

[0069] According to the fitting error of the internal temperature at the current moment and the difference of the fitting functions in different period segments, the second abnormal degree of the internal temperature at the current moment is determined:

[0070] ;

[0071] in, Indicates the second abnormality level of the internal temperature at the current moment; Indicates the target sequence number, that is, the sequence number of the internal temperature at the current moment in the cycle segment to which it belongs; represents the fitted value of the internal temperature at the current moment; Indicates the internal temperature at the current moment; is a hyperparameter used to prevent Larger, resulting in It is always close to 1. The implementer can set the value of the hyperparameter according to the actual implementation situation, for example ; , , Indicates The polynomial coefficients of the fitting function for the periodic segments; , , The polynomial coefficients representing the fitting function of the target period segment; The variance of the polynomial coefficients representing the fitting function for all periodic segments; represents an exponential function with a natural constant as base; represents the hyperbolic tangent function; Indicates the number of period segments.

[0072] It represents the fitting error of the internal temperature at the current moment under the fitting function of the target cycle segment. When the fitting error is larger, it means that the internal temperature at the current moment does not conform to the changing law of the internal temperature in the target cycle segment, and the second abnormal degree of the internal temperature at the current moment is larger; conversely, when the fitting error is smaller, it means that the internal temperature at the current moment conforms to the changing law of the internal temperature in the target cycle segment, and the second abnormal degree of the internal temperature at the current moment is smaller.

[0073] The change of internal temperature is a periodic change, and the change of each period segment is basically the same, so the fitting function of each period segment should also be relatively small. Indicates the difference between the fitting function of the target period segment and the fitting functions of the remaining period segments; when the difference is larger, it means that the fitting function of the target period segment is less accurate. At this time, the fitting error of the internal temperature at the current moment calculated based on the fitting function of the target period segment is less accurate. To conduct When the difference between the fitting functions is larger, The smaller the reduction of , the smaller the difference between the fitting functions, which means that the fitting function of the target period is more accurate. The closer it is to 1, the The smaller the correction degree is, the more the second abnormal degree of the internal temperature at the current moment depends on the fitting error of the internal temperature at the current moment under the fitting function of the target cycle segment.

[0074] At this point, the first abnormality level of the internal temperature at the current moment is acquired.

[0075] S6: determining a correction window of the internal temperature at the current moment according to the first abnormality degree and the second abnormality degree of the internal temperature at the current moment, and correcting the internal temperature at the current moment according to all internal temperatures within the correction window to obtain a correction value of the internal temperature at the current moment.

[0076] The correction window size of the internal temperature at the current moment is determined according to the first abnormality degree and the second abnormality degree:

[0077] ;

[0078] The corrected window size representing the internal temperature at the current moment; Indicates the initial window size, the default value is 1, and the implementer can set it according to the actual implementation situation; Indicates the window size adjustment coefficient, the empirical value is 5, and the implementer can set it according to the actual implementation situation; Indicates the first abnormality level of the internal temperature at the current moment; Indicates the second abnormality level of the internal temperature at the current moment; Represents a rounding function. If the first abnormal degree or the second abnormal degree of the internal temperature at the current moment is larger, it means that the internal temperature at the current moment does not conform to the change or periodic change law brought by the external temperature, and the change of the internal temperature at the current moment is more likely to be caused by noise, etc. At this time, a larger correction window is needed to correct the internal temperature at the current moment to remove the influence of noise on the internal temperature at the current moment; on the contrary, if the first abnormal degree and the second abnormal degree of the internal temperature at the current moment are both smaller, it means that the internal temperature at the current moment conforms to the change and periodic change law brought by the external temperature, and the change of the internal temperature at the current moment is more likely to be a normal temperature change in the smelting process, and a smaller correction window is needed to ensure that the characteristics of the internal temperature at the current moment are retained.

[0079] In the target sequence, construct The internal temperature at the current moment is the last data of the correction window.

[0080] In one embodiment, the average value of all internal temperatures within the correction window is used as the correction value of the internal temperature at the current moment.

[0081] In another embodiment, the weighted average of all internal temperatures within the correction window is used as the correction value of the internal temperature at the current moment, wherein the weight of each internal temperature satisfies the expression:

[0082] ;

[0083] in, Indicates the correction window. The weight of the internal temperature; Indicates the correction window. The first abnormal degree of internal temperature; Indicates the correction window. a second abnormal degree of internal temperature; represents an exponential function with a natural constant as base; Indicates the size of the correction window. The smaller the first abnormal degree and the second abnormal degree of the internal temperature are, the smaller the internal temperature at the current moment is corrected. The larger the reference value of the internal temperature, the greater the The greater the weight of the internal temperature, the greater the weight of the internal temperature; otherwise, The smaller the reference value of the internal temperature, the smaller the The smaller the weight of the internal temperature.

[0084] At this point, the corrected value of the internal temperature at the current time is acquired.

[0085] S7: Determine the descending speed of the master alloy bar according to the correction value of the internal temperature and the external temperature to realize automatic feeding.

[0086] Record the material and diameter of the master alloy rod currently being smelted by the electrode induction gas atomization powder making equipment. Record the output power and frequency of the electromagnetic induction coil in the melting furnace of the electrode induction gas atomization powder making equipment at every moment.

[0087] The present invention sets the descending speed of the master alloy rod in the electrode induction gas atomization powder making equipment through a neural network, and the specific content of the neural network is as follows:

[0088] The neural network adopts a fully connected neural network; the input of the neural network is the material of the master alloy rod, the diameter of the master alloy rod, the external temperature, the correction value of the internal temperature, the output power and frequency of the electromagnetic induction coil, and the output is the descent speed of the master alloy rod; the data set of the neural network is a data set composed of master alloy rods of different materials and diameters, different external temperatures, different internal temperatures, different output powers of the electromagnetic induction coil, and different frequencies. The labels are the optimal descent speeds of master alloy rods of different materials and diameters at different output powers of the electromagnetic induction coil, different frequencies, different external temperatures, and different internal temperatures. The optimal descent speed is obtained by the implementers through a large number of experiments; the loss function of the neural network is the mean square error loss.

[0089] The material and diameter of the master alloy bar, the current external temperature, the correction value of the internal temperature, the output power and frequency of the electromagnetic induction coil are input into the trained neural network, and the descending speed of the master alloy bar is output. The descending speed of the master alloy bar in the feeding equipment is controlled according to the descending speed, so as to realize the automatic feeding of the electrode induction gas atomization powder making equipment.

[0090] At this point, the automatic feeding of the electrode induction atomization powder making equipment has been realized.

[0091] The embodiment of the present invention further discloses an automatic feeding system for a feeding device, comprising a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, an automatic feeding method for a feeding device according to the present invention is implemented.

[0092] The above system also includes other components well known to those skilled in the art, such as a communication bus and a communication interface, and their configuration and functions are known in the art, so they will not be described in detail here.

[0093] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

[0094] The above are all preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. An automatic feeding method for a feeding device, characterized in that: include: Collect the external and internal temperatures of the electrode induction atomization powder making equipment at every moment; Dividing the internal temperature into multiple cycle segments; The internal temperatures in all period segments with the same serial number as the internal temperature at the current moment constitute a target sequence; The smoothing coefficient is determined based on the difference between the external temperature at the current moment and the external temperature at the corresponding moment of each element in the target sequence, including: ;in, Indicates the smoothing coefficient at the current moment; Indicates the external temperature at the current moment; Indicates the target external temperature sequence. External temperature; represents the number of external temperatures contained in the target external temperature sequence; represents the extreme difference of all external temperatures; the method for obtaining the target external temperature sequence is: the external temperatures at the moments corresponding to each internal temperature in the target sequence constitute the target external temperature sequence, and the smoothing coefficient is positively correlated with the difference; the internal temperature at the current moment is predicted according to the smoothing coefficient, and the first abnormality degree is determined according to the difference between the predicted value and the actual value, including: ;in, Indicates the first abnormality level of the internal temperature at the current moment; Indicates the actual value of the internal temperature at the current moment; Represents the predicted value at the current moment; represents the mean of all pairwise differences in the internal temperatures in the target sequence; represents the hyperbolic tangent function; Fitting the cycle segment to which the internal temperature at the current moment belongs, and determining the second abnormality degree according to the fitting error of the internal temperature at the current moment and the difference between the fitting functions of different cycle segments includes: ;in, Indicates the second abnormality level of the internal temperature at the current moment; Indicates the serial number of the internal temperature at the current moment in the cycle segment to which it belongs; represents the fitted value of the internal temperature at the current moment; Indicates the internal temperature at the current moment; is a hyperparameter; , , Indicates The coefficients of the fitting function for the period segment; , , The coefficient of the fitting function representing the period to which the internal temperature at the current moment belongs; represents the variance of the coefficients of the fitting function for all periodic segments; represents an exponential function with a natural constant as base; represents the number of periodic segments, the second abnormality degree is positively correlated with the fitting error and negatively correlated with the difference between the fitting functions; Determining the correction window of the internal temperature at the current moment includes: the correction window size of the internal temperature at the current moment satisfies the expression: ; The corrected window size representing the internal temperature at the current moment; Indicates the preset initial window size; Indicates the preset window size adjustment factor; Indicates the first abnormality level of the internal temperature at the current moment; Indicates the second abnormality level of the internal temperature at the current moment; Represents the rounding function; in the target sequence, construct A correction window of size, wherein the internal temperature at the current moment is the last data of the correction window, and the size of the correction window is positively correlated with the first abnormality degree and the second abnormality degree; the internal temperature at the current moment is corrected according to all the internal temperatures in the correction window, and the correction value of the internal temperature at the current moment is obtained, which includes: The weighted mean of all internal temperatures in the correction window is used as the correction value of the internal temperature at the current moment, where the weight of each internal temperature in the correction window satisfies the expression: ;in, Indicates the correction window The weight of the internal temperature; Indicates the correction window. The first abnormal degree of internal temperature; Indicates the correction window. a second abnormal degree of internal temperature; The descending speed of the master alloy bar is determined according to the correction value of the internal temperature and the external temperature to realize automatic feeding.

2. The automatic feeding method of a feeding device according to claim 1, characterized in that: The internal temperature is divided into a plurality of period segments, including: The internal temperature sequence is composed of the internal temperatures at different times; the internal temperature sequence is decomposed by STL to obtain the periodic terms, the periodic terms are Fourier transformed, and the inverse of the frequency corresponding to the maximum amplitude after Fourier transformation is taken as the period length; The internal temperature sequence is divided into a plurality of different cycle segments using the cycle length, and the length of each cycle segment is equal to the cycle length.

3. The automatic feeding method of a feeding device according to claim 1, characterized in that: The fitting of the period segment to which the internal temperature at the current moment belongs includes: The cycle segment to which the internal temperature at the current moment belongs is taken as the target cycle segment, and the polynomial fitting of all internal temperatures in the target cycle segment is performed using the least squares method to obtain the fitting function of the target cycle segment; the loss value in the fitting process satisfies the expression: ； in, represents the loss value; Indicates the serial number of the internal temperature at the current moment in the cycle segment to which it belongs; Indicates the first The first abnormal degree of internal temperature; Indicates the first The fitted value of the internal temperature; Indicates the first Internal temperature; Represents an exponential function with a natural constant as its base.

4. The automatic feeding method of a feeding device according to claim 1, characterized in that: Also includes: Collect the output power, frequency, material and diameter of the electromagnetic induction coil; The method of determining the descending speed of the master alloy bar according to the correction value of the internal temperature and the external temperature to realize automatic feeding includes: The material of the master alloy rod, the diameter of the master alloy rod, the current external temperature, the correction value of the internal temperature, the output power and frequency of the electromagnetic induction coil are input into the trained neural network, and the descending speed of the master alloy rod is output; the descent of the master alloy rod in the feeding equipment is controlled according to the descent speed, so as to realize the automatic feeding of the electrode induction atomization powder making equipment.

5. An automatic feeding system for feeding equipment, characterized in that: include: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, an automatic feeding method for a feeding device according to any one of claims 1 to 4 is implemented.

Images