A fault warning method and system for metal powder production equipment

Abstract

The present invention relates to the field of data processing technology, and in particular to a fault warning method and system for metal powder production equipment. The method includes the following steps: presetting the length of a historical pressure data segment, obtaining the actual value and predicted value of each data point in the historical pressure data segment of the data point to be monitored in the pressure time series of the production equipment, and the predicted value is calculated by using the prediction error factor through the HTFE algorithm; obtaining the slope of each data point to determine the rate of change of the historical pressure data segment; determining the prediction error index of the historical pressure data segment of the data point to be monitored; calculating the prediction error factor of the data point to be monitored; determining the fault monitoring result of the production equipment based on the prediction error factor of the data point to be monitored, and giving a warning for the abnormal result, which can effectively improve the accuracy of the fault warning of the obtained metal powder production equipment.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a fault early warning method and system for metal powder production equipment. Background Art

[0002] Metal powder is a key raw material for equipment in high-tech industries such as new energy, aerospace, and biomedicine. During the production process, fault warning of metal powder production equipment plays an important role in improving production efficiency, reducing maintenance costs, and promoting intelligent transformation.

[0003] In the 9th issue of Computer Age journal in 2023, there is an article entitled "Research on Time Series Prediction Method Based on Historical Trends and Prediction Errors". The core idea of ​​this prediction method is: on the one hand, determine the trend type of the recent time series (growth, decay or irregular), on the other hand, obtain the range of the current prediction value based on the prediction error and the current true value, and finally determine the final prediction value by combining the sequence trend and the prediction range. After that, by repeatedly repeating "prediction-correction-prediction", continuous prediction of continuous states can be achieved, which can be applied to many fields such as financial stocks, weather forecasts, sensor warnings, etc.

[0004] However, the above-mentioned existing prediction methods use a fixed prediction error factor in the prediction process. The parameter data of metal powder production equipment has different data change patterns. The time series adaptive prediction method based on historical trends and prediction errors (Historical Trends Forecast Errors, HTFE) uses a fixed prediction error factor and cannot be accurately changed based on the data change pattern, resulting in low accuracy of the obtained prediction results.

[0005] Based on this, how to obtain fault warning results of metal powder production equipment based on accurate prediction results is a technical problem that needs to be solved urgently by technical personnel in this field. Summary of the invention

[0006] In order to solve the above-mentioned technical problem of how to obtain a fault warning result of a metal powder production equipment based on an accurate prediction value, the present invention provides a fault warning method and system for a metal powder production equipment.

[0007] In a first aspect, the present invention provides a fault warning method for metal powder production equipment, which adopts the following technical solution:

[0008] A fault early warning method for metal powder production equipment comprises the following steps:

[0009] The length of the historical pressure data segment is preset, and the actual value and predicted value of each data point in the historical pressure data segment of the data point to be monitored in the production equipment pressure time series sequence are obtained, and the predicted value is calculated using the prediction error factor through the HTFE algorithm; the slope of each data point is obtained, and the change rate of the historical pressure data segment is determined, and the change rate is positively correlated with the absolute value of the slope of each data point in the historical pressure data segment; the prediction error index of the historical pressure data segment of the data point to be monitored is determined, and the prediction error index is positively correlated with the absolute value of the difference between the actual value and the predicted value of each data point in the historical pressure data segment; the prediction error factor of the data point to be monitored is calculated:

[0010] ;

[0011] In the formula, represents the prediction error factor of the data point to be monitored, represents the prediction error factor for the historical pressure data segment, Represents the forecast error indicator for the historical pressure data segment, Indicates the rate of change of the historical pressure data segment, represents the rate threshold, It represents an exponential function with e as the base; the fault monitoring result of the production equipment is determined based on the prediction error factor of the data point to be monitored, and an early warning is issued for abnormal results.

[0012] The present invention takes into account that a fixed prediction error factor cannot adapt to the diverse data change patterns in the pressure time series. Therefore, the change rate of the historical pressure data segment is obtained by acquiring the data change speed in the historical pressure data segment of the data point to be monitored, and the direction of adjustment of the prediction error factor is determined according to the change rate of the historical pressure data segment. Combined with the prediction error of the historical pressure data segment, the prediction error factor of the data point to be monitored can be accurately obtained, and then the predicted value of the data point can be accurately obtained, which can effectively improve the accuracy of the fault warning of the metal powder production equipment.

[0013] According to a fault warning method for metal powder production equipment provided by the present invention, the actual value and predicted value of each data point in the historical pressure data segment of the data point to be monitored in the pressure time series sequence of the production equipment are obtained, and before that, it also includes: obtaining pressure data during the operation of the production equipment, and preprocessing the pressure data to obtain the pressure time series sequence of the production equipment.

[0014] The present invention takes into account that the overall quality of the collected original data is low, and therefore improves the overall quality of the data through various preprocessing methods to prepare for subsequent data processing.

[0015] According to a fault warning method for metal powder production equipment provided by the present invention, the determination of the change rate of the historical pressure data segment includes: taking any data point in the pressure time series sequence of the production equipment as the target data point, constructing a preset local range of the target data point, and determining the difference between the target data point and the mean value of the data points other than the target data point in the local range to obtain the noise performance degree of the target data point, wherein the noise performance degree is positively correlated with the absolute value of the difference; determining the change rate of the historical pressure data segment, wherein the change rate of the historical pressure data segment is negatively correlated with the mean value of the noise performance degree of each data point in the historical pressure data segment.

[0016] The present invention takes into account the possible presence of noise data in the pressure data, which affects the calculation result of the change rate of the historical pressure data segment. Therefore, by obtaining the noise performance degree of each data point, it is determined whether the source of the data change speed is noise data or normal data change, thereby effectively improving the accuracy of the change rate of the historical pressure data segment obtained.

[0017] According to a fault warning method for metal powder production equipment provided by the present invention, the noise performance degree of the target data point satisfies the relationship:

[0018] ;

[0019] In the formula, Indicates The noise level of each data point is Indicates The number of data points within the local range of a data point, Indicates In the local range of the data points, except Data points outside of The actual value of the data point, Indicates The actual value of the data point, Represents the absolute value symbol.

[0020] According to a fault warning method for metal powder production equipment provided by the present invention, the change rate of the historical pressure data segment satisfies the relationship:

[0021] ;

[0022] In the formula, Indicates the rate of change of the historical pressure data segment, Indicates the number of data points in the historical pressure data segment, represents the slope of the ith data point in the historical pressure data segment, Indicates the mean value of the noise performance of each data point in the historical pressure data segment, represents the exponential function with base e, Represents the absolute value symbol.

[0023] The present invention provides a method for calculating the change rate of a historical pressure data segment accurately. By obtaining the slope of each data point to represent the speed of data change, and using the noise performance degree of the data point as the credibility of the speed of data change, the change rate of the historical pressure data segment can be accurately obtained and the adjustment direction of the prediction error factor can be determined.

[0024] According to a fault warning method for metal powder production equipment provided by the present invention, the prediction error index of the historical pressure data segment of the data point to be monitored satisfies the relationship:

[0025] ;

[0026] In the formula, Represents the forecast error indicator for the historical pressure data segment, Indicates the number of data points in the historical pressure data segment, represents the actual value of the i-th data point in the historical pressure data segment, represents the predicted value of the ith data point in the historical pressure data segment, Indicates the degree of noise performance of the ith data point in the historical pressure data segment, represents the linear normalization function, represents the exponential function with base e, Represents the absolute value symbol.

[0027] The present invention provides a method for calculating an accurate prediction error index for a historical pressure data segment. The prediction error in the historical pressure data segment is obtained while taking into account the influence of noise data that may exist in the data on the prediction error index result. The noise performance degree of the data point is used as the prediction error of the data point to determine the weight of the prediction error index for the historical pressure data segment. The prediction error index for the historical pressure data segment can be accurately obtained.

[0028] According to a fault warning method for metal powder production equipment provided by the present invention, the fault monitoring result of the production equipment is determined based on the prediction error factor of the data point to be monitored, including: taking the data point to be monitored as the starting point, obtaining a future data segment to be monitored, and the length of the future data segment to be monitored is equal to the length of the historical pressure data segment; using the prediction error factor of the data point to be monitored as the prediction error factor of each data point in the future data segment to be monitored, and using the prediction error factor of each data point in the future data segment to be monitored through the HTFE algorithm to obtain the predicted value of each data point in the future data segment to be monitored; and obtaining the fault monitoring result based on the comparison result of the predicted value of each data point in the future data segment to be monitored with the pressure threshold.

[0029] According to a fault warning method for metal powder production equipment provided by the present invention, the fault monitoring result is obtained based on the comparison result of the predicted value of each data point in the future data segment to be monitored and the pressure threshold, including: a preset pressure threshold; if the predicted value of the data point is greater than or equal to the pressure threshold, the fault monitoring result of the data point is abnormal; if the predicted value of the data point is less than the pressure threshold, the fault monitoring result of the data point is normal.

[0030] In a second aspect, the present invention provides a fault warning system for metal powder production equipment, which adopts the following technical solution:

[0031] A fault warning system for metal powder production equipment 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 fault warning method for metal powder production equipment is implemented.

[0032] By adopting the above technical solution, the fault warning method of the above-mentioned metal powder production equipment is generated into a computer program and stored in a memory to be loaded and executed by a processor, so that a terminal device is manufactured according to the memory and the processor for easy use.

[0033] The present invention has the following technical effects:

[0034] Based on the above technical scheme, the present invention obtains the change rate of the historical pressure data segment by acquiring the speed of data change in the historical pressure data segment of the data point to be monitored, determines the direction of adjustment of the prediction error factor according to the change rate of the historical pressure data segment, and combines the prediction error of the historical pressure data segment to avoid the influence of noise data to accurately obtain the prediction error factor of the data point to be monitored, and then accurately obtains the predicted value of the data point, which can effectively improve the accuracy of the fault warning of the metal powder production equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] 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.

[0036] Figure 1 A schematic flow chart of a fault warning method for metal powder production equipment provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0037] 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.

[0038] 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.

[0039] It should be noted that the existing adaptive prediction algorithm for time series based on historical trends and prediction errors (Historical Trends Forecast Errors, HTFE) obtains prediction results by using a fixed prediction error factor. However, in the process of processing the parameter data of metal powder production equipment to obtain prediction results, the HTFE algorithm has different data change patterns in the data sequence due to factors such as noise data and instantaneous fluctuations in the metal powder production equipment. The fixed prediction error factor cannot be adaptively changed based on the data change pattern, resulting in low accuracy of the prediction results. Ultimately, it is impossible to issue a safety warning based on the low-accuracy prediction results, which poses a major safety hazard.

[0040] Based on this, an embodiment of the present invention discloses a fault warning method for metal powder production equipment. By obtaining the prediction error of the historical pressure data segment of the data point to be monitored and the speed of data change, the prediction error factor of the data point to be monitored is adjusted, so that the predicted value and fault monitoring result of the data point to be monitored can be accurately obtained.

[0041] See also Figure 1 As shown, Figure 1 A schematic flow chart of a fault warning method for metal powder production equipment provided by an embodiment of the present invention, the method comprising the following steps:

[0042] S1: Preset the length of the historical pressure data segment, and obtain the actual value and predicted value of each data point in the historical pressure data segment of the data point to be monitored in the pressure time series of the production equipment.

[0043] It should be noted that the prediction error factor plays an important role in the HTFE algorithm. If the prediction error factor is too large, the algorithm may be too flexible and overly sensitive to noise or instantaneous fluctuations in the data, resulting in unstable prediction results. If the prediction error factor is too small, the algorithm may be too conservative and insufficiently responsive to rapid changes in the data. Based on this, the embodiment of the present invention adjusts the prediction error factor of the data point to be monitored by analyzing the prediction error in the historical data before the data point to be monitored and the speed of data change.

[0044] It is understandable that metal powder production equipment will generate a variety of parameter data during operation, such as temperature, pressure, gas flow, electrical parameters, etc. For ease of analysis, the present invention embodiment takes the processing of any one of these parameters as an example for illustration. During the operation of production equipment, pressure is a key indicator that can reflect the operating status of the equipment and the stability of the production process. The following specifically describes the steps for processing pressure data to obtain fault warning results:

[0045] For example, in an embodiment of the present invention, the actual value and predicted value of each data point in the historical pressure data segment of the data point to be monitored in the production equipment pressure time series sequence are obtained, and the process also includes: obtaining the pressure data during the operation of the production equipment, and preprocessing the pressure data to obtain the production equipment pressure time series sequence. The preprocessing method can be data cleaning, data transformation, etc., which can be set according to actual needs.

[0046] Specifically, a pressure sensor is used to collect pressure data generated during the operation of the metal powder production equipment, and a series of data points collected and arranged in time order are combined into a pressure time series sequence. The obtained pressure time series sequence contains multiple data points, and the pressure value actually collected at each data point is used as the actual value of the data point. In the pressure time series sequence, the data point to be monitored is obtained and the length of the historical pressure data segment is preset. The length of the historical pressure data segment is the number of data points included in the historical pressure data segment. Continuous data points are obtained from the historical pressure data before the data point to be monitored to form a historical pressure data segment to determine the data point to be monitored, and the predicted value of each data point in the historical pressure data segment of the data point to be monitored is determined.

[0047] The data point to be monitored may be the latest data point to be monitored, or may be any data point in the pressure time series, and may be specifically set according to actual needs.

[0048] For example, the predicted value of each data point in the historical pressure data segment of the data point to be monitored can be calculated by the HTFE algorithm using the prediction error factor. The actual value and the predicted value of each data point in the historical pressure data segment finally obtained correspond to each other.

[0049] Preferably, the length of the historical pressure data segment may be 20; the length of the historical pressure data segment may be set according to actual needs, and the embodiment of the present invention does not impose too many restrictions on this.

[0050] For example, when the predicted value of each data point in the historical pressure data segment is obtained by the HTFE algorithm, the prediction interval error factor can be set to 0.5, the number of historical trend observation steps can be set to 3, and the prediction error factor of the historical pressure data segment can be set to Set it to 0.1, and you can set it according to actual needs.

[0051] After obtaining the actual value and predicted value of each data point in the historical pressure data segment of the data point to be monitored based on the above steps, continue to perform the following steps:

[0052] S2: Obtain the slope of each data point and determine the rate of change of the historical pressure data segment.

[0053] It should be noted that in order to determine the prediction error factor of each data point, it is necessary to obtain the data change rate in the historical pressure data segment of the data point to be monitored, and the absolute value of the slope of each data point can represent the speed of data change at the data point. The larger the absolute value of the slope, the faster the change. Therefore, the average of the absolute value of the slope can be used to represent the speed of data change in a segment of data. Based on this, the embodiment of the present invention obtains the change rate of the historical pressure data segment by obtaining the slope of each data point in the historical pressure data segment, and the change rate is positively correlated with the absolute value of the slope of each data point in the historical pressure data segment.

[0054] It is understandable that due to the influence of noise data that may exist in the data points, the obtained prediction results are inaccurate. Based on this, the noise performance degree of each data point can be obtained first to characterize the possibility that the data point is noise data.

[0055] For example, in an embodiment of the present invention, when determining the rate of change of a historical pressure data segment, any data point in the pressure time series sequence of the production equipment can be used as a target data point, and a preset local range of the target data point is constructed. The difference between the target data point and the mean value of the data points other than the target data point in the local range is determined to obtain the noise performance degree of the target data point, and the noise performance degree is positively correlated with the absolute value of the difference; the rate of change of the historical pressure data segment is determined, and the rate of change of the historical pressure data segment is negatively correlated with the mean value of the noise performance degree of each data point in the historical pressure data segment.

[0056] The local range of the target data point is the number of data points contained in the local range. In order to evenly obtain data points from both sides of the target data point, the number of data points contained in the local range can be set to an odd number. The local range can be set specifically according to actual needs.

[0057] It is understandable that noise data generally presents a larger data deviation than the surrounding data. In order to accurately obtain the noise performance degree of the data point, the data points before and after the target data point can be obtained in the pressure time series sequence, and the local range of the target data point can be jointly constructed. The deviation of the data change situation between the target data point and its local range can be analyzed, so as to accurately obtain the noise performance degree of the data point.

[0058] Specifically, when constructing the local range of the target data point, the target data point can be taken as the center, and a number of data points can be equally obtained on both sides of the target data point to form the local range of the target data point. The obtained local range includes both the target data point and the data points around the target data point. The specific degree of noise performance of each data point can be found in the following relationship:

[0059] ;

[0060] In the formula, Indicates The noise level of each data point is Indicates The number of data points within the local range of a data point, Indicates In the local range of the data points, except data points outside The actual value of the data point, Indicates The actual value of the data point, Represents the absolute value symbol.

[0061] In the above formula, Indicates that the local range is The mean of all data points except the data point itself. This value can represent the overall data change degree of the data points in the local range. The mean is The larger the difference between the data points, the The greater the deviation of a data point from the overall data, the The greater the probability that the data point is noise data.

[0062] It should be noted that when determining the noise performance of data points based on the local range, there may be some edge data points, such as the starting data point and the end data point, which do not have enough data points around them to construct their local range. In this case, for the edge data points that cannot construct the local range due to insufficient surrounding data points, it is considered that their reference value is low and they are not processed in the technical solution provided in the embodiment of the present invention.

[0063] After the noise performance level of each data point is obtained based on the above steps, the change rate of the historical pressure data segment of the data point to be monitored can be calculated.

[0064] It should be noted that, considering the possible existence of noise data in the data points, the noise performance degree of each data point obtained above can be used as the credibility of the speed of change. If the mean value of the noise performance degree of the data points in the historical pressure data segment is larger, it means that the data changes generated by the data points in the historical pressure data segment may be caused by noise data.

[0065] Specifically, when determining the rate of change of the historical pressure data segment, the least square method can be used to first use the time point corresponding to each data point in the pressure time series as the horizontal coordinate and the actual value of the data point as the vertical coordinate to perform curve fitting, thereby obtaining the slope of each data point. Then, the mean value of the noise performance degree of each data point in the historical pressure data segment is used as the credibility of the speed of change of each data point in the historical pressure data segment, and finally the rate of change of the historical pressure data segment is obtained. For details, please refer to the following relationship:

[0066] ;

[0067] In the formula, Indicates the rate of change of the historical pressure data segment, Indicates the number of data points in the historical pressure data segment, represents the slope of the ith data point in the historical pressure data segment, Indicates the mean value of the noise performance of each data point in the historical pressure data segment, represents the exponential function with base e, Represents the absolute value symbol.

[0068] In the above formula, Indicates how fast the data in the historical pressure data segment changes. The larger the value, the faster the data in the historical pressure data segment changes; conversely, the smaller the value, the slower the data in the historical pressure data segment changes.

[0069] By obtaining the data change rate in the historical pressure data segment, the direction in which the prediction error factor of the data point to be monitored needs to be adjusted can be obtained. The adjustment direction includes increasing or decreasing. In order to accurately obtain the specific adjusted data value, the prediction error index of the historical pressure data segment can be determined based on the following steps, thereby obtaining the prediction error factor of the data point to be monitored.

[0070] S3: Determine the prediction error index of the historical pressure data segment of the data point to be monitored.

[0071] It should be noted that the difference between the actual value of the data point and the predicted value obtained based on the prediction error factor is the prediction error. By analyzing the difference between the actual value and the predicted value of each data point in the historical pressure data segment, the prediction error index of the historical pressure data segment can be determined. The prediction error index is positively correlated with the absolute value of the difference between the actual value and the predicted value of each data point in the historical pressure data segment. Considering that there may be the influence of noise data in the data point, resulting in a large error between the actual value and the predicted value of the data point, the noise performance degree of each data point can be obtained as the weight of the prediction error index of the historical pressure data segment for each data point in the historical pressure data segment.

[0072] Specifically, when determining the prediction error index of the historical pressure data segment of the data point to be monitored, the actual value, predicted value and noise performance degree of each data point in the historical pressure data segment can be obtained first, and the prediction error at each data point in the historical pressure data segment can be determined. Then, based on the prediction error of each data point, the prediction error index of the historical pressure data segment of the data point to be monitored can be determined. For details, please refer to the following relationship:

[0073] ;

[0074] In the formula, Represents the forecast error indicator for the historical pressure data segment, Indicates the number of data points in the historical pressure data segment, represents the actual value of the i-th data point in the historical pressure data segment, represents the predicted value of the ith data point in the historical pressure data segment, Indicates the degree of noise performance of the ith data point in the historical pressure data segment, represents the linear normalization function, represents the exponential function with base e, Represents the absolute value symbol.

[0075] In the above formula, Indicates the prediction error of the i-th data point in the historical pressure data segment. The larger the value, the larger the prediction error at the i-th data point. In order to further exclude the possibility that the prediction error of this data point is caused by noise, the noise performance degree of the i-th data point can be obtained. If the noise performance degree of the i-th data point is higher, the possibility that the data point is noise data is higher, and the possibility that the prediction error of the i-th data point is caused by noise is also greater. Therefore, when determining the prediction error index of the historical pressure data segment, the noise performance degree of the data point can be used as the weight of the data point to reduce the influence of the data point on the prediction error index of the historical pressure data segment.

[0076] After respectively obtaining the change rate and prediction error index of the historical pressure data segment of the data point to be monitored based on the above steps, the prediction error factor of the data point to be monitored can be obtained based on the change rate and prediction error index of the historical pressure data segment:

[0077] S4: Calculate the prediction error factor of the data point to be monitored.

[0078] It should be noted that if the change rate of the historical pressure data segment of the data point to be monitored is high, the corresponding prediction error index is higher, which means that the prediction error factor of the historical pressure data segment is smaller. In order to improve the accuracy of the prediction results of the data point to be monitored, the prediction error factor can be increased to improve the flexibility of the algorithm; conversely, if the change rate of the historical pressure data segment of the data point to be monitored is low, the corresponding prediction error index is higher, which means that the prediction error factor of the historical pressure data segment is larger. In order to improve the accuracy of the prediction results of the data point to be monitored, the prediction error factor can be reduced to improve the prediction accuracy of the algorithm.

[0079] Preferably, the rate threshold may be preset to 0.35, and may be specifically set according to actual needs.

[0080] Specifically, a change rate greater than or equal to the rate threshold is defined as faster, and a change rate less than the rate threshold is defined as slower. The prediction error factor and prediction error index of the historical pressure data segment are combined to determine the prediction error factor of the data point to be monitored. For details, see the following relationship:

[0081] ;

[0082] In the formula, represents the prediction error factor of the data point to be monitored, represents the prediction error factor for the historical pressure data segment, Represents the forecast error indicator for the historical pressure data segment, Indicates the rate of change of the historical pressure data segment, represents the rate threshold, represents an exponential function with base e;

[0083] In the above formula, when When , it means that the data of the historical pressure data segment changes faster, and the prediction error factor of the subsequent data points to be monitored needs to be increased. The prediction error factor of the data point to be monitored is positively correlated with the prediction error factor of the historical pressure data segment.

[0084] when When , it means that the data change speed of the historical pressure data segment is slow, and the prediction error factor of the subsequent data points to be monitored needs to be reduced. The prediction error factor of the data points to be monitored is negatively correlated with the prediction error factor of the historical pressure data segment.

[0085] After obtaining the prediction error factor of the data point to be monitored based on the above method, the prediction result of the future data segment to be monitored with the data point to be monitored as the starting point can be obtained, that is, the following steps are performed:

[0086] S5: Determine the fault monitoring results of the production equipment based on the prediction error factor of the data point to be monitored, and issue an early warning for abnormal results.

[0087] It should be noted that, when obtaining the prediction error factor of the data point to be monitored, the historical pressure data segment before the data point to be monitored is used as the target for analysis. Therefore, the prediction error factor of the data point to be monitored can be used as the prediction error factor of the data point in the future data segment to be monitored, so as to obtain the fault monitoring results of each data point in the future data segment to be monitored.

[0088] For example, when determining the fault monitoring result of the production equipment based on the prediction error factor of the data point to be monitored, the data point to be monitored can be used as the starting point to obtain the future data segment to be monitored, and the length of the future data segment to be monitored is equal to the length of the historical pressure data segment; the prediction error factor of the data point to be monitored is used as the prediction error factor of each data point in the future data segment to be monitored, and the prediction error factor of each data point in the future data segment to be monitored is used through the HTFE algorithm to obtain the prediction value of each data point in the future data segment to be monitored; the fault monitoring result is obtained based on the comparison result of the prediction value of each data point in the future data segment to be monitored and the pressure threshold.

[0089] It is understandable that the first data point in the future data segment to be monitored obtained based on the above method is the data point to be monitored. After the predicted value of each data point in the future data segment to be monitored is obtained by the HTFE algorithm, the fault monitoring result of the metal powder production equipment can be obtained.

[0090] For example, when obtaining the fault monitoring result based on the comparison result between the predicted value of each data point in the future monitored data segment and the pressure threshold, the pressure threshold can be preset; if the predicted value of the data point is greater than or equal to the pressure threshold, the fault monitoring result of the data point is abnormal; if the predicted value of the data point is less than the pressure threshold, the fault monitoring result of the data point is normal.

[0091] The pressure threshold may be set according to actual needs, and the embodiment of the present invention does not impose too many restrictions on this.

[0092] For example, when warnings are issued for abnormal results, different levels of warnings can be set according to the number of data points with abnormal monitoring results in the future data segment to be monitored.

[0093] The warning method can be sound warning or signal light warning, etc., which can be set according to actual needs.

[0094] It can be seen that in an embodiment of the present invention, when obtaining the fault result of the metal powder production equipment, the length of the historical pressure data segment can be preset, and the actual value and predicted value of each data point in the historical pressure data segment of the data point to be monitored in the pressure time series sequence of the production equipment can be obtained, and the predicted value is calculated using the prediction error factor through the HTFE algorithm; the slope of each data point is obtained, and the change rate of the historical pressure data segment is determined, and the change rate is positively correlated with the absolute value of the slope of each data point in the historical pressure data segment; the prediction error index of the historical pressure data segment of the data point to be monitored is determined, and the prediction error index is positively correlated with the absolute value of the difference between the actual value and the predicted value of each data point in the historical pressure data segment; the prediction error factor of the data point to be monitored is calculated; based on the prediction error factor of the data point to be monitored, the fault monitoring result of the production equipment is determined, and an early warning is issued for abnormal results.

[0095] In this way, the present invention obtains the change rate of the historical pressure data segment by acquiring the speed of data change in the historical pressure data segment of the data point to be monitored, determines the direction of adjustment of the prediction error factor according to the change rate of the historical pressure data segment, and combines the prediction error of the historical pressure data segment to accurately obtain the prediction error factor of the data point to be monitored, and then accurately obtain the predicted value of the data point, which can effectively improve the accuracy of the fault warning of the metal powder production equipment.

[0096] An embodiment of the present invention further discloses a fault warning system for metal powder production equipment, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a fault warning method for metal powder production equipment according to the present invention is implemented.

[0097] 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.

[0098] In the present invention, the aforementioned memory may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium may be any suitable magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory RRAM, a dynamic random access memory DRAM, a static random access memory SRAM, an enhanced dynamic random access memory EDRAM, a high bandwidth memory HBM, a hybrid memory cube HMC, etc., or any other medium that can be used to store the required information and can be accessed by an application, a module, or both. Any such computer storage medium may be part of a device or accessible or connectable to a device.

[0099] 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.

[0100] 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. A fault warning method for metal powder production equipment, characterized in that: include: The length of the historical pressure data segment is preset, and the actual value and predicted value of each data point in the historical pressure data segment of the data point to be monitored are obtained in the pressure time series of the production equipment, and the predicted value is calculated by using the prediction error factor through the HTFE algorithm; Obtaining the slope of each data point and determining the rate of change of the historical pressure data segment includes: taking any data point in the production equipment pressure time series as a target data point, constructing a preset local range of the target data point, and determining the difference between the target data point and the mean value of the data points other than the target data point in the local range to obtain the noise performance degree of the target data point, wherein the noise performance degree is positively correlated with the absolute value of the difference; Determining a rate of change of the historical pressure data segment, wherein the rate of change of the historical pressure data segment is negatively correlated with a mean value of a noise performance degree of each data point in the historical pressure data segment, and the rate of change is positively correlated with an absolute value of a slope of each data point in the historical pressure data segment; The rate of change of the historical pressure data segment satisfies the relationship: ; In the formula, Indicates the rate of change of the historical pressure data segment, Indicates the number of data points in the historical pressure data segment, represents the slope of the ith data point in the historical pressure data segment, Indicates the mean value of the noise performance of each data point in the historical pressure data segment, represents the exponential function with base e, Indicates absolute value symbol Determine the prediction error index of the historical pressure data segment of the data point to be monitored, wherein the prediction error index is positively correlated with the absolute value of the difference between the actual value and the predicted value of each data point in the historical pressure data segment; calculate the prediction error factor of the data point to be monitored: ； In the formula, represents the prediction error factor of the data point to be monitored, represents the prediction error factor for the historical pressure data segment, Represents the forecast error indicator for the historical pressure data segment, Indicates the rate of change of the historical pressure data segment, represents the rate threshold, represents an exponential function with base e; The fault monitoring results of the production equipment are determined based on the prediction error factors of the data points to be monitored, and early warnings are issued for abnormal results.

2. A fault warning method for metal powder production equipment according to claim 1, characterized in that: The step of obtaining the actual value and predicted value of each data point in the historical pressure data segment of the data point to be monitored in the production equipment pressure time series sequence also includes: The pressure data during the operation of the production equipment is acquired, and the pressure data is preprocessed to obtain a pressure time series sequence of the production equipment.

3. A fault warning method for metal powder production equipment according to claim 1, characterized in that: The noise performance degree of the target data point satisfies the relationship: ； In the formula, Indicates The noise level of each data point is Indicates The number of data points within the local range of a data point, Indicates In the local range of the data points, except data points outside The actual value of the data point, Indicates The actual value of the data point, Represents the absolute value symbol.

4. The fault warning method for metal powder production equipment according to claim 1, characterized in that: The prediction error index of the historical pressure data segment of the data point to be monitored satisfies the relationship: ； In the formula, Represents the forecast error indicator for the historical pressure data segment, Indicates the number of data points in the historical pressure data segment, represents the actual value of the i-th data point in the historical pressure data segment, represents the predicted value of the ith data point in the historical pressure data segment, Indicates the degree of noise performance of the ith data point in the historical pressure data segment, represents the linear normalization function, represents the exponential function with base e, Represents the absolute value symbol.

5. The fault warning method for metal powder production equipment according to claim 1, characterized in that: The method of determining the fault monitoring result of the production equipment based on the prediction error factor of the data point to be monitored includes: Taking the data point to be monitored as a starting point, obtaining a future data segment to be monitored, wherein the future data segment to be monitored is equal to the length of the historical pressure data segment; Using the prediction error factor of the data point to be monitored as the prediction error factor of each data point in the future data segment to be monitored, and using the prediction error factor of each data point in the future data segment to be monitored through the HTFE algorithm to obtain the prediction value of each data point in the future data segment to be monitored; The fault monitoring result is obtained based on the comparison result between the predicted value of each data point in the future data segment to be monitored and the pressure threshold.

6. A fault warning method for metal powder production equipment according to claim 5, characterized in that: The fault monitoring result is obtained based on the comparison result of the predicted value of each data point in the future data segment to be monitored and the pressure threshold, including: Preset pressure threshold; If the predicted value of the data point is greater than or equal to the pressure threshold, the fault monitoring result of the data point is abnormal; If the predicted value of the data point is less than the pressure threshold, the fault monitoring result of the data point is normal.

7. A fault warning system for metal powder production 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, a fault warning method for metal powder production equipment according to any one of claims 1 to 6 is implemented.

Images