A method for predicting vacuum dry pump failures based on data distribution

Abstract

The present invention discloses a method for predicting a vacuum dry pump fault based on data distribution, comprising: step S1, constructing a vacuum dry pump operation state vector set; step S2, constructing a vacuum dry pump operation data distribution vector set according to the vacuum dry pump operation state vector set; step S3, constructing a data distribution vector selection model according to the vacuum dry pump operation data distribution vector set; step S4, constructing a historical data index list according to the vacuum dry pump operation data distribution vector set and the data distribution vector selection model; step S5, constructing a vacuum dry pump fault prediction training set based on data distribution according to the vacuum dry pump operation data distribution vector set and the historical data index list; step S6, using a classifier to establish a vacuum dry pump fault classification prediction model for the vacuum dry pump fault prediction training set based on data distribution, and performing vacuum dry pump fault prediction on other downtime pump data through the vacuum dry pump fault classification prediction model. The present invention achieves the goal of vacuum dry pump fault prediction.

Description

Technical Field

[0001] The present invention relates to the technical field of health monitoring of vacuum dry pumps, and in particular to a vacuum dry pump fault prediction method based on data distribution. Background Art

[0002] In order to ensure the quality and performance of semiconductor devices, the operational maintenance of vacuum dry pumps is particularly important. Establishing a reliable health detection system is the key to ensuring the smooth operation of electromechanical equipment. With the increase in the number of equipment detection points and sampling frequency, mechanical health monitoring has entered the "big data" era.

[0003] The traditional method based on manual inspection not only requires high expert experience, but is also time-consuming and labor-intensive and cannot guarantee universality. It can no longer meet the requirements of mechanical big data.

[0004] Therefore, a vacuum dry pump fault prediction method based on data distribution is provided. Summary of the invention

[0005] The purpose of the present invention is to overcome the existing defects and provide a vacuum dry pump fault prediction method based on data distribution to achieve the goal of vacuum dry pump fault prediction.

[0006] The technical solution to achieve the above purpose is:

[0007] A vacuum dry pump fault prediction method based on data distribution, comprising:

[0008] Step S1, constructing a vacuum dry pump operation state vector set VDPRS;

[0009] Step S2, constructing a vacuum dry pump operation data distribution vector set DPDDS according to the vacuum dry pump operation state vector set VDPRS;

[0010] Step S3, constructing a data distribution vector selection model DDSM according to the vacuum dry pump operation data distribution vector set DPDDS;

[0011] Step S4, constructing a historical data index list HDIS according to the vacuum dry pump operation data distribution vector set DPDDS and the data distribution vector selection model DDSM;

[0012] Step S5, constructing a vacuum dry pump fault prediction training set BDDPD based on data distribution according to the vacuum dry pump operation data distribution vector set DPDDS and the historical data index list HDIS;

[0013] Step S6, using the Logistic classifier to establish a vacuum dry pump fault classification prediction model DVDPS for the vacuum dry pump fault prediction training set BDDPD based on data distribution, and using the vacuum dry pump fault classification prediction model DVDPS to predict vacuum dry pump faults for other downtime pump data.

[0014] Preferably, the step S1 comprises:

[0015] Step S11, obtaining a set of operating data of vacuum dry pumps of the same type based on actual applications of vacuum dry pumps in the market;

[0016] Step S12, constructing a vacuum dry pump operation data subset with different numbers under the same vacuum dry pump type;

[0017] Step S13, constructing a second-level vector of the dry pump operation status {Time, DryPumpPower, BoosterPower, DPInverterSpeed, MBInverterSpeed, DPTemperature, MBTemperature, N2 Flow, ExhaustPressure} based on the dry pump motor power, motor speed, sensor temperature, nitrogen flow and exhaust pressure parameters;

[0018] Step S14, the running status data set of the vacuum dry pump that has been shut down is processed by the step S13 to obtain a shut down running status vector set ERSVS = [rsv1, rsv2, …, rsv i ], where i=1,2,3,…,n;

[0019] Step S15, the operating status data set corresponding to all moments in the vacuum dry pump to be tested is processed by the step S13 to obtain the operating status vector set TRSVS=[rsv1, rsv2, …, rsv j ], where j=1,2,3,…,m;

[0020] Step S16, constructing the downtime operation status tag set ETAG=[etag1,etag2, …, etag x ], where x=1,2,3,…,n; the label set corresponds to each vacuum dry pump in the downtime operation state vector set ERSVS in step S14;

[0021] Step S17, constructing a vacuum dry pump operation state vector set VDPRS={ERSVS, ETAG,TRSVS} through the obtained downtime operation state vector set ERSVS, the vacuum dry pump operation state vector set TRSVS to be tested and the downtime operation state tag set ETAG.

[0022] Preferably, in step S13, Time represents the current time when the equipment is running, DryPumpPower represents the working power of the lower dry pump, BoosterPower represents the working power of the upper dry pump, DPInverterSpeed ​​represents the working frequency of the lower dry pump, MBInverterSpeed ​​represents the working frequency of the upper dry pump, DPTemperature represents the working temperature of the lower dry pump, MBTemperature represents the working temperature of the upper dry pump, N2 Flow represents the nitrogen flow rate introduced into the dry pump, and ExhaustPressure represents the exhaust pressure of the dry pump.

[0023] Preferably, the step S2 comprises:

[0024] Step S21, setting the basic unit for data distribution vector processing, the basic unit includes minutes, hours and days;

[0025] Step S22, reading the operating status data prsd of a single vacuum dry pump;

[0026] Step S23, the read operation status data prsd is processed by data distribution vector to obtain a unit vector set prsdb of a single dry pump, and the time corresponding to the first row of data in the unit vector set prsdb is used as the time of the prsdb, recorded as prsdb_t;

[0027] Step S24, reading a running state feature pf, calculating the minimum value, maximum value, average value, standard deviation and median of the running state feature pf, and taking these five values ​​as the data distribution vector of the feature, recorded as pfdv;

[0028] Step S25, traverse each operating state feature pf in the unit vector set prsdb of a single dry pump, and obtain the vacuum dry pump operating data distribution vector through the processing in step S24, which is in the form of [pfdv1, pfdv2,…, pfdv k ], where k=1,2,3,…,8;

[0029] Step S26, processing the downtime operation state vector set ERSVS in the vacuum dry pump operation state vector set VDPRS in the step S17 through the steps S21 to S25 to obtain a downtime operation state data distribution vector set DPDDS-E;

[0030] Step S27, processing the vacuum dry pump operation state vector set TRSVS to be tested in the vacuum dry pump operation state vector set VDPRS in the step S17 through the steps S21 to S25 to obtain a dry pump operation state data distribution vector set DPDDS-T to be predicted;

[0031] Step S28, marking the time corresponding to the first row of data of each processing unit in the downtime operation status data distribution vector set DPDDS-E as the time of the unit distribution vector, and obtaining the downtime operation status label set ELABEL;

[0032] Step S29, constructing a dry pump operation data distribution vector set DPDDS={DPDDS-E, ELABEL, DPDDS-T} through the obtained downtime operation status data distribution vector set DPDDS-E, the dry pump operation status data distribution vector set to be predicted DPDDS-T and the downtime operation status label set ELABEL.

[0033] Preferably, step S3 comprises:

[0034] Step S31, constructing a selection index empty list SIL of the dry pump to be predicted;

[0035] Step S32, selecting a data distribution vector trsv of the operating state of a dry pump to be predicted at a certain moment in the vacuum dry pump operating data distribution vector set DPDDS;

[0036] Step S33, calculating the Euclidean distance between each dry pump operation data distribution vector in the subset of the downtime operation status data distribution vector set DPDDS-E in the vacuum dry pump operation data distribution vector set DPDDS and the operation status data distribution vector trsv of the dry pump to be predicted at a certain moment, and returning the index eindex corresponding to the minimum Euclidean distance;

[0037] Step S34, deleting the index eindex in the subset of the downtime operation status data distribution vector set DPDDS-E;

[0038] Step S35, adding the deleted index eindex to the empty list SIL of selection indexes of the dry pump to be predicted;

[0039] Step S36, set the length of the empty list SIL of selection indexes of the dry pump to be predicted to L, and execute the steps S31-S35 in a loop until the length of the empty list SIL of selection indexes of the dry pump to be predicted meets L, and return the list SIL of selection indexes of the dry pump to be predicted, thereby obtaining the data distribution vector selection model DDSM.

[0040] Preferably, step S4 comprises:

[0041] Step S41, obtaining a downtime subset subs in the downtime operation status data distribution vector set DPDDS-E, and selecting an operation status data distribution vector trsv of a dry pump to be predicted at a certain moment in the vacuum dry pump operation data distribution vector set DPDDS;

[0042] Step S42, the downtime subset subs is processed by obtaining a data distribution vector selection model DDSM to obtain a historical data index list hds of a certain moment operation status data distribution vector trsv of the dry pump to be predicted under the downtime subset subs;

[0043] Step S43, traverse each downtime subset subs in the downtime operation status data distribution vector set DPDDS-E, and process according to the step S42 to obtain a historical data index list HDIS={hds1, hds2, …, hds y}, where y=1,2,3,…,n.

[0044] Preferably, step S5 comprises:

[0045] Step S51, using the historical data index list HDIS to select the downtime operation status data distribution vector set DPDDS-E to obtain the vacuum dry pump fault prediction feature set DPF-D based on data distribution;

[0046] Step S52, using the historical data index list HDIS to select the downtime operation status label set ELABEL to obtain the vacuum dry pump fault prediction label set DPF-L based on data distribution;

[0047] Step S53, constructing a vacuum dry pump fault prediction training set BDDPD={DPF-D, DPF-L} based on data distribution by using the obtained vacuum dry pump fault prediction feature set DPF-D and vacuum dry pump fault prediction label set DPF-L.

[0048] The beneficial effects of the present invention are as follows: the present invention uses multiple equipment fault data sets to construct a training data set to make up for the insufficiency of a single equipment fault data set, takes the electrical parameters of a vacuum dry pump as the research object, and proposes a vacuum dry pump fault prediction method based on data distribution. In the process of establishing a training sample library, different historical database samples are selected by dynamically adjusting the number of sample libraries and calculating the Euclidean distance between samples. This method can establish a vacuum dry pump fault prediction model, thereby achieving the goal of vacuum dry pump fault prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is a flow chart of a vacuum dry pump fault prediction method based on data distribution of the present invention;

[0050] Figure 2 It is a specific flow chart of constructing the vacuum dry pump operation state vector set VDPRS in the present invention;

[0051] Figure 3 It is a specific flow chart of constructing a vacuum dry pump operation data distribution vector set DPDDS in the present invention;

[0052] Figure 4 It is a specific flow chart of constructing the data distribution vector selection model DDSM in the present invention;

[0053] Figure 5 It is a specific flow chart of constructing the historical data index list HDIS in the present invention;

[0054] Figure 6 It is a specific flow chart of constructing a vacuum dry pump fault prediction training set BDDPD based on data distribution in the present invention;

[0055] Figure 7 It is a schematic diagram of the number of historical distribution vectors selected under different k in the present invention;

[0056] Figure 8 It is a schematic diagram of the early warning situation obtained by using the Logistic classifier in the present invention. DETAILED DESCRIPTION

[0057] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. In the description of the present invention, it should be noted that the terms "center", "up", "down", "left", "right", "vertical", "horizontal", "inside", "outside" and the like indicate directions or positional relationships based on the directions or positional relationships shown in the accompanying drawings, which are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore cannot be understood as a limitation on the present invention. In addition, the terms "first", "second", and "third" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance.

[0058] The present invention will be further described below in conjunction with the accompanying drawings.

[0059] like Figure 1 As shown, a vacuum dry pump fault prediction method based on data distribution includes:

[0060] Step S1, constructing a vacuum dry pump operation state vector set VDPRS.

[0061] like Figure 2 As shown, step S1 specifically includes:

[0062] Step S11, obtaining a set of operating data of vacuum dry pumps of the same type based on actual applications of vacuum dry pumps in the market.

[0063] Step S12, constructing a vacuum dry pump operation data subset with different numbers under the same vacuum dry pump type.

[0064] Step S13, constructing the second-level vector of the dry pump operation status {Time, DryPumpPower, BoosterPower, DPInverterSpeed, MBInverterSpeed, DPTemperature, MBTemperature, N2 Flow, ExhaustPressure} based on the dry pump motor power, motor speed, sensor temperature, nitrogen flow and exhaust pressure parameters.

[0065] In the embodiment, Time represents the current time when the device is running, DryPumpPower represents the working power of the lower dry pump, BoosterPower represents the working power of the upper dry pump, DPInverterSpeed ​​represents the working frequency of the lower dry pump, MBInverterSpeed ​​represents the working frequency of the upper dry pump, DPTemperature represents the working temperature of the lower dry pump, MBTemperature represents the working temperature of the upper dry pump, N2 Flow represents the nitrogen flow rate introduced into the dry pump, and ExhaustPressure represents the exhaust pressure of the dry pump.

[0066] Step S14, the running status data set of the vacuum dry pump that has been shut down is processed by the step S13 to obtain a shut down running status vector set ERSVS = [rsv1, rsv2, …, rsv i ], where i=1,2,3,…,n.

[0067] Step S15, the operating status data set corresponding to all moments in the vacuum dry pump to be tested is processed by the step S13 to obtain the operating status vector set TRSVS=[rsv1, rsv2, …, rsv j ], where j=1,2,3,…,m.

[0068] Step S16, constructing the downtime operation status tag set ETAG=[etag1,etag2, …, etag x ], where x=1,2,3,…,n; the label set corresponds to each vacuum dry pump in the downtime operation state vector set ERSVS in step S14.

[0069] Step S17, constructing a vacuum dry pump operation state vector set VDPRS={ERSVS, ETAG,TRSVS} through the obtained downtime operation state vector set ERSVS, the vacuum dry pump operation state vector set TRSVS to be tested and the downtime operation state tag set ETAG.

[0070] Step S2, constructing a vacuum dry pump operation data distribution vector set DPDDS according to the vacuum dry pump operation state vector set VDPRS.

[0071] like Figure 3 As shown, step S2 specifically includes:

[0072] Step S21, setting the basic unit for data distribution vector processing, the basic unit includes minutes, hours and days.

[0073] Step S22, reading the operating status data prsd of a single vacuum dry pump.

[0074] Step S23, the read operation status data prsd is processed through the data distribution vector to obtain the unit vector set prsdb of the single dry pump, and the time corresponding to the first row of data in the unit vector set prsdb is used as the time of the prsdb, recorded as prsdb_t.

[0075] Step S24, read a running status feature pf, calculate the minimum value, maximum value, average value, standard deviation and median of the running status feature pf, and use these five values ​​as the data distribution vector of the feature, recorded as pfdv.

[0076] Step S25, traverse each operating state feature pf in the unit vector set prsdb of a single dry pump, and obtain the vacuum dry pump operating data distribution vector through the processing in step S24, which is in the form of [pfdv1, pfdv2,…, pfdv k ], where k=1,2,3,…,8.

[0077] Step S26, processing the downtime operation state vector set ERSVS in the vacuum dry pump operation state vector set VDPRS in step S17 through steps S21 to S25 to obtain a downtime operation state data distribution vector set DPDDS-E.

[0078] Step S27, processing the vacuum dry pump operation state vector set TRSVS to be tested in the vacuum dry pump operation state vector set VDPRS in step S17 through steps S21 to S25 to obtain a dry pump operation state data distribution vector set DPDDS-T to be predicted.

[0079] Step S28, marking the time corresponding to the first row of data of each processing unit in the downtime operation status data distribution vector set DPDDS-E as the time of the unit distribution vector, and obtaining the downtime operation status label set ELABEL.

[0080] Step S29, constructing a dry pump operation data distribution vector set DPDDS={DPDDS-E, ELABEL, DPDDS-T} through the obtained downtime operation status data distribution vector set DPDDS-E, the dry pump operation status data distribution vector set to be predicted DPDDS-T and the downtime operation status label set ELABEL.

[0081] Step S3, constructing a data distribution vector selection model DDSM according to the vacuum dry pump operation data distribution vector set DPDDS.

[0082] like Figure 4 As shown, step S3 specifically includes:

[0083] Step S31, constructing a selection index empty list SIL of the dry pump to be predicted.

[0084] Step S32: selecting a data distribution vector trsv of the operating state of a dry pump to be predicted at a certain moment in the vacuum dry pump operating data distribution vector set DPDDS.

[0085] Step S33, calculate the Euclidean distance between each dry pump operation data distribution vector in the subset of the downtime operation status data distribution vector set DPDDS-E in the vacuum dry pump operation data distribution vector set DPDDS and the operation status data distribution vector trsv of the dry pump to be predicted at a certain moment, and return the index eindex corresponding to the minimum Euclidean distance.

[0086] Step S34: deleting the index eindex in the subset of the downtime operation status data distribution vector set DPDDS-E.

[0087] Step S35, adding the deleted index eindex to the empty list SIL of selection indexes of the dry pump to be predicted.

[0088] Step S36, set the length of the empty list SIL of selection indexes of the dry pump to be predicted to L, and execute the steps S31-S35 in a loop until the length of the empty list SIL of selection indexes of the dry pump to be predicted meets L, and return the list SIL of selection indexes of the dry pump to be predicted, thereby obtaining the data distribution vector selection model DDSM.

[0089] Step S4, constructing a historical data index list HDIS according to the vacuum dry pump operation data distribution vector set DPDDS and the data distribution vector selection model DDSM.

[0090] like Figure 5 As shown, step S4 specifically includes:

[0091] Step S41, obtaining a downtime subset subs in the downtime operation status data distribution vector set DPDDS-E, and selecting an operation status data distribution vector trsv of a dry pump to be predicted at a certain moment in the vacuum dry pump operation data distribution vector set DPDDS.

[0092] Step S42, the downtime subset subs is processed by obtaining a data distribution vector selection model DDSM to obtain a historical data index list hds of a certain moment operation status data distribution vector trsv of the dry pump to be predicted under the downtime subset subs.

[0093] Step S43, traverse each downtime subset subs in the downtime operation status data distribution vector set DPDDS-E, and process according to the step S42 to obtain a historical data index list HDIS={hds1, hds2, …, hds y}, where y=1,2,3,…,n.

[0094] Step S5, constructing a vacuum dry pump fault prediction training set BDDPD based on data distribution according to the vacuum dry pump operation data distribution vector set DPDDS and the historical data index list HDIS.

[0095] like Figure 6 As shown, step S5 specifically includes:

[0096] Step S51, using the historical data index list HDIS to select the downtime operation status data distribution vector set DPDDS-E, to obtain the vacuum dry pump fault prediction feature set DPF-D based on data distribution.

[0097] Step S52: Use the historical data index list HDIS to select the downtime operation status label set ELABEL to obtain the vacuum dry pump fault prediction label set DPF-L based on data distribution.

[0098] Step S53, constructing a vacuum dry pump fault prediction training set BDDPD={DPF-D, DPF-L} based on data distribution by using the obtained vacuum dry pump fault prediction feature set DPF-D and vacuum dry pump fault prediction label set DPF-L.

[0099] Step S6, using the Logistic classifier to establish a vacuum dry pump fault classification prediction model DVDPS for the vacuum dry pump fault prediction training set BDDPD based on data distribution, and using the vacuum dry pump fault classification prediction model DVDPS to predict vacuum dry pump faults for other downtime pump data.

[0100] The number of data distribution vectors k selected for each downtime data is set to 24*14. Under this selected number, the vacuum dry pump fault prediction training set BDDPD={DPF-D, DPF-L} based on data distribution is obtained; the Logistic classifier is used to predict the vacuum dry pump fault classification prediction model DVDPS based on the dry pump fault prediction training set based on data distribution. The experimental results show that the model can use other downtime pump data to predict vacuum dry pump faults; thus demonstrating the effectiveness of the vacuum dry pump fault prediction method based on data distribution.

[0101] The data distribution vector of the vacuum dry pump at 17:00 on January 14, 2024 is used as the prediction vector. The number of historical distribution vectors selected under different k is as follows: Figure 7 As shown, where k=1,…,30.

[0102] The 62-day operation data of the vacuum dry pump is used as the model input, and the early warning situation obtained by using the Logistic classifier when k is set to 24*14 is as follows Figure 8 shown.

[0103] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions described in the above embodiments may still be modified, or some or all of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A vacuum dry pump fault prediction method based on data distribution, characterized in that: include: Step S1, constructing a vacuum dry pump operation state vector set VDPRS; Step S2, constructing a vacuum dry pump operation data distribution vector set DPDDS according to the vacuum dry pump operation state vector set VDPRS; Step S3, constructing a data distribution vector selection model DDSM according to the vacuum dry pump operation data distribution vector set DPDDS; Step S4, constructing a historical data index list HDIS according to the vacuum dry pump operation data distribution vector set DPDDS and the data distribution vector selection model DDSM; Step S5, constructing a vacuum dry pump fault prediction training set BDDPD based on data distribution according to the vacuum dry pump operation data distribution vector set DPDDS and the historical data index list HDIS; Step S6, using a Logistic classifier to establish a vacuum dry pump fault classification prediction model DVDPS for the vacuum dry pump fault prediction training set BDDPD based on data distribution, and using the vacuum dry pump fault classification prediction model DVDPS to perform vacuum dry pump fault prediction on other downtime pump data; The step S1 comprises: Step S11, obtaining a set of operating data of vacuum dry pumps of the same type based on actual applications of vacuum dry pumps in the market; Step S12, constructing a vacuum dry pump operation data subset with different numbers under the same vacuum dry pump type; Step S13, constructing a second-level vector of the dry pump operation status {Time, DryPumpPower, BoosterPower, DPInverterSpeed, MBInverterSpeed, DPTemperature, MBTemperature, N2 Flow, ExhaustPressure} based on the dry pump motor power, motor speed, sensor temperature, nitrogen flow and exhaust pressure parameters; Step S14, processing the running status data set of the vacuum dry pump that has been shut down through the step S13 to obtain a shut down running status vector set ERSVS = [rsv1, rsv2, …, rsvi, …, rsvn], where i = 1, 2, 3, …, n; Step S15, processing the operating status data set corresponding to all moments of the vacuum dry pump to be tested through the step S13 to obtain an operating status vector set TRSVS=[rsv1, rsv2, …, rsvj, …, rsvm] of the vacuum dry pump to be tested, where j=1,2,3,…,m; Step S16, constructing a downtime operation status tag set ETAG=[etag1,etag2, …, etagx, …, etagn] according to the historical data of the downtime operation of the dry pump, where x=1,2,3,…,n; the tag set corresponds to each vacuum dry pump in the downtime operation status vector set ERSVS in step S14; Step S17, constructing a vacuum dry pump operation state vector set VDPRS={ERSVS, ETAG,TRSVS} through the obtained downtime operation state vector set ERSVS, the vacuum dry pump operation state vector set TRSVS to be tested and the downtime operation state tag set ETAG; In the step S13, Time represents the current time of the device operation, DryPumpPower represents the working power of the lower dry pump, BoosterPower represents the working power of the upper dry pump, DPInverterSpeed ​​represents the working frequency of the lower dry pump, MBInverterSpeed ​​represents the working frequency of the upper dry pump, DPTemperature represents the working temperature of the lower dry pump, MBTemperature represents the working temperature of the upper dry pump, N2 Flow represents the nitrogen flow rate introduced into the dry pump, and ExhaustPressure represents the exhaust pressure of the dry pump; The step S2 comprises: Step S21, setting the basic unit for data distribution vector processing, the basic unit includes minutes, hours and days; Step S22, reading the operating status data prsd of a single vacuum dry pump; Step S23, the read operation status data prsd is processed by data distribution vector to obtain a unit vector set prsdb of a single dry pump, and the time corresponding to the first row of data in the unit vector set prsdb is used as the time of the prsdb, recorded as prsdb_t; Step S24, reading a running state feature pf, calculating the minimum value, maximum value, average value, standard deviation and median of the running state feature pf, and taking these five values ​​as the data distribution vector of the feature, recorded as pfdv; Step S25, traverse each operating state feature pf in the unit vector set prsdb of a single dry pump, and obtain a vacuum dry pump operating data distribution vector through the processing in step S24, which is in the form of [pfdv1, pfdv2, …, pfdvk, …, pfdv8], where k=1,2,3,…,8; Step S26, processing the downtime operation state vector set ERSVS in the vacuum dry pump operation state vector set VDPRS in the step S17 through the steps S21 to S25 to obtain a downtime operation state data distribution vector set DPDDS-E; Step S27, processing the vacuum dry pump operation state vector set TRSVS to be tested in the vacuum dry pump operation state vector set VDPRS in the step S17 through the steps S21 to S25 to obtain a dry pump operation state data distribution vector set DPDDS-T to be predicted; Step S28, marking the time corresponding to the first row of data of each processing unit in the downtime operation status data distribution vector set DPDDS-E as the time of the unit distribution vector, and obtaining the downtime operation status label set ELABEL; Step S29, constructing a dry pump operation data distribution vector set DPDDS={DPDDS-E, ELABEL, DPDDS-T} through the obtained downtime operation status data distribution vector set DPDDS-E, the to-be-predicted dry pump operation status data distribution vector set DPDDS-T and the downtime operation status label set ELABEL; The step S3 comprises: Step S31, constructing a selection index empty list SIL of the dry pump to be predicted; Step S32, selecting a data distribution vector trsv of the operating state of a dry pump to be predicted at a certain moment in the vacuum dry pump operating data distribution vector set DPDDS; Step S33, calculating the Euclidean distance between each dry pump operation data distribution vector in the subset of the downtime operation status data distribution vector set DPDDS-E in the vacuum dry pump operation data distribution vector set DPDDS and the operation status data distribution vector trsv of the dry pump to be predicted at a certain moment, and returning the index eindex corresponding to the minimum Euclidean distance; Step S34, deleting the index eindex in the subset of the downtime operation status data distribution vector set DPDDS-E; Step S35, adding the deleted index eindex to the empty list SIL of selection indexes of the dry pump to be predicted; Step S36, setting the length of the empty list SIL of the selection index of the dry pump to be predicted to L, looping through the steps S31 to S35 until the length of the empty list SIL of the selection index of the dry pump to be predicted meets L, returning the list SIL of the selection index of the dry pump to be predicted, and then obtaining the data distribution vector selection model DDSM; The step S4 comprises: Step S41, obtaining a downtime subset subs in the downtime operation status data distribution vector set DPDDS-E, and selecting an operation status data distribution vector trsv of a dry pump to be predicted at a certain moment in the vacuum dry pump operation data distribution vector set DPDDS; Step S42, the downtime subset subs is processed by obtaining a data distribution vector selection model DDSM to obtain a historical data index list hds of a certain moment operation status data distribution vector trsv of the dry pump to be predicted under the downtime subset subs; Step S43, traverse each downtime subset subs in the downtime operation status data distribution vector set DPDDS-E, and process according to the step S42 to obtain a historical data index list HDIS={hds1, hds2, …, hdsy, …, hdsn}, where y=1,2,3,…,n; The step S5 comprises: Step S51, using the historical data index list HDIS to select the downtime operation status data distribution vector set DPDDS-E to obtain the vacuum dry pump fault prediction feature set DPF-D based on data distribution; Step S52, using the historical data index list HDIS to select the downtime operation status label set ELABEL to obtain the vacuum dry pump fault prediction label set DPF-L based on data distribution; Step S53, constructing a vacuum dry pump fault prediction training set BDDPD={DPF-D, DPF-L} based on data distribution by using the obtained vacuum dry pump fault prediction feature set DPF-D and vacuum dry pump fault prediction label set DPF-L.

Images