Method for controlling micro-pollution residues of OLED organic evaporation cavity component

Abstract

The invention relates to the technical field of evaporation control, and discloses a control method for micro-pollution residues of an OLED organic evaporation cavity component. Comprising the following steps: acquiring a screening event, judging whether a micro-pollution residue control requirement is met or not, acquiring residue influence data of OLED evaporation equipment in a sub-interval, predicting a pollution state value of a next sub-interval, identifying target data from the residue influence data, and formulating a micro-pollution control instruction corresponding to the target data; compared with the prior art, the threshold of micro-pollution residue control can be improved, the acquisition amount of irrelevant data is greatly reduced, and meanwhile, a targeted control instruction can be timely and accurately formulated before the phenomenon of micro-pollution residue of the OLED evaporation equipment is about to occur, so that the control accuracy of the OLED evaporation equipment is improved. And it is ensured that the OLED evaporation equipment can be reasonably and timely optimized and controlled before the micro-pollution residual phenomenon occurs, and the micro-pollution residual phenomenon on the cavity component of the OLED evaporation equipment is prevented.

Description

Technical Field

[0001] The present invention relates to the field of evaporation control technology, and more specifically, to a method for controlling micro-contamination residues of OLED organic evaporation cavity components. Background Art

[0002] During the production and manufacturing process of OLED, as the evaporation parameters and environment in the cavity of the OLED evaporation equipment change, micro-pollution is likely to remain on the cavity components in the cavity of the OLED evaporation equipment, causing the cavity components to be contaminated and affecting the efficiency and quality of subsequent organic evaporation. In order to avoid the phenomenon of micro-pollution residue on the cavity components of the OLED evaporation equipment, it is necessary to control the micro-pollution residue in the OLED evaporation equipment.

[0003] The patent application with reference publication number CN116641036A discloses a control method for residual micro-pollution of OLED organic evaporation cavity components. The method simulates the vacuum working environment in the OLED organic evaporation process, compares and analyzes the physical and chemical properties of organic materials and cavity components, and factors such as the organic evaporation environment temperature and evaporation conditions. In the process, the vacuum baking temperature used is higher than the process temperature in the OLED display device manufacturing process. The reason is that the high temperature area in the evaporation cavity is concentrated at the evaporation source position, and the temperature at other positions in the cavity is relatively low and the film layer and pollutants are deposited more. Under this temperature condition, the removal effect of organic residues and micro-pollution on the surface of the cavity components can be ensured to be more ideal, thereby designing a control method suitable for large-scale popularization and minimizing the residual micro-pollution of organic evaporation cavity components.

[0004] When controlling the micro-pollution residue of the existing evaporation cavity parts, various evaporation parameters in the cavity of the OLED evaporation equipment are collected in real time, and the real-time pollution status of the OLED evaporation equipment is given after calculation and analysis, so as to perform corresponding control processing on the micro-pollution residue of the cavity parts of the OLED evaporation equipment. Since the real-time collection and control method can only be used to control the micro-pollution residue in the cavity of the OLED evaporation equipment in real time, there is a certain lag in the control of the micro-pollution residue. When a certain parameter in the cavity of the OLED evaporation equipment changes sharply in a short period of time, it is impossible to give early warning and control of the micro-pollution residue of the OLED evaporation equipment, which will lead to the phenomenon of micro-pollution residue in the cavity parts of the OLED evaporation equipment, thereby reducing the subsequent evaporation quality of the OLED evaporation equipment.

[0005] In view of this, the present invention proposes a method for controlling residual micro-contamination of OLED organic evaporation cavity components to solve the above-mentioned problem. Summary of the invention

[0006] In order to overcome the above-mentioned defects of the prior art and to achieve the above-mentioned purpose, the present invention provides the following technical solution: a method for controlling micro-contamination residues of OLED organic evaporation cavity components, comprising:

[0007] S1: Obtain basic identification parameters of the OLED evaporation equipment, and generate identification events based on the basic identification parameters, where the basic identification parameters include effective operation time value and safe temperature ratio value;

[0008] S2: Identify the event status of the screening event, and determine whether the micro-pollution residue control requirements are met based on the compliance criteria; if the micro-pollution residue control requirements are met, execute S3; if the micro-pollution residue control requirements are not met, repeat S1-S2;

[0009] S3: Divide the control period into sub-intervals, and obtain residual impact data of the OLED evaporation equipment in the sub-intervals, the residual impact data including vacuum attenuation value, humidity increase rate and gas purity value;

[0010] S4: Input the residual impact data into the pre-trained machine learning model, predict the pollution state value of the next sub-interval, and determine whether to issue a residual control prompt. The pollution state value includes a clean state and a polluted state. If a residual control prompt is issued, execute S5; if no residual control prompt is issued, repeat S3-S4.

[0011] S5: Identify target data from the residual impact data, and formulate micro-pollution control instructions corresponding to the target data. The micro-pollution control instructions include instructions for reducing vacuum attenuation value, instructions for reducing humidity increase rate, and instructions for increasing gas purity value.

[0012] Furthermore, the method for obtaining the effective running time value includes:

[0013] The operating power of the OLED evaporation equipment is detected by a power sensor to obtain the real-time power value;

[0014] The real-time power value greater than or equal to the preset power lower limit is recorded as the effective power value, and the time when the effective power value first appears is queried through the timestamp to obtain the starting time;

[0015] Between the start time and the current time, the time when the real-time power value is less than the preset power lower limit is recorded as the end time, and the duration between the start time and the end time is recorded as the sub-duration, obtaining 1 sub-duration;

[0016] After accumulating each sub-duration one by one, the effective running time value is obtained;

[0017] The expressions for effective runtime values ​​are:

[0018]

[0019] In the formula, SC yx is the effective running time value, SC za is the ath sub-period.

[0020] Furthermore, the method for obtaining the safety temperature ratio value includes:

[0021] Detect the operating temperature of the OLED evaporation equipment through the temperature sensor to obtain the real-time temperature value;

[0022] The real-time temperature value that is greater than or equal to the preset lower temperature limit is recorded as the effective temperature value, and the time when the effective temperature value first appears is queried through the timestamp and recorded as the temperature starting time;

[0023] After the temperature start time, when no valid temperature value appears within the preset time, the time when the last valid temperature value appears is recorded as the temperature end time, and the time between the temperature start time and the temperature end time is recorded as the safe temperature time, and w safe temperature time is obtained;

[0024] The safe temperature duration that is greater than the standard safe duration is recorded as the target duration, and p target durations are obtained;

[0025] The time when the real-time temperature value is first detected is recorded as the initial time, and the duration from the initial time to the current time is recorded as the total running time. The p target durations are accumulated one by one and compared with the total running time to obtain the safe temperature ratio value;

[0026] The expression of the safe temperature ratio is:

[0027]

[0028] Where WD zb is the safe temperature ratio, SC mbb is the bth target duration, SC zyx is the total running time.

[0029] Furthermore, the method for generating the screening event includes:

[0030] An event template having two event stations is selected from the event database, and the two event stations are recorded as a first event station and a second event station respectively according to the order of their positions;

[0031] Mark remark boxes in the first event station and the second event station respectively, and import the effective operation time value and the safety temperature ratio value into the remark boxes of the first event station and the second event station respectively, to obtain the first event station and the second event station after import;

[0032] After the import, number boxes are respectively established at the front end of the first event station and the second event station, and after inputting 01 and 02 in the two number boxes respectively, the identification event is obtained.

[0033] Furthermore, the event status includes the compliance status and the non-compliance status:

[0034] Methods for identifying compliance status and non-compliance status include:

[0035] The effective running time value SC in the first event station yx With the standard duration value SC bz Compare;

[0036] When SC yx Greater than or equal to SC bz When SC yx Smaller than SC bz When the event status of the first event station of the screening event is a non-standard state;

[0037] The safety temperature ratio WD in the second event station zb Compared with the standard ratio WD bz Compare;

[0038] When WD zb Greater than or equal to WD bz When WD zb Smaller than WD bz When , the event status of the second event station of the screening event is a non-standard state;

[0039] The criteria for determining compliance are: there is no non-compliance status in the screening event;

[0040] The methods for determining whether the requirements for micro-pollution residue control are met include:

[0041] Count the number of non-standard states at the first event station and the second event station in the identification event to obtain the non-standard value;

[0042] When the non-standard value is 0, it is determined that the micro-pollution residual control requirements are met;

[0043] When the non-standard value is greater than 0, it is determined that the micro-pollution residual control requirements are not met.

[0044] Furthermore, the method for obtaining the vacuum decay value includes:

[0045] The control period is equally divided into x subintervals, and the last moment of the previous subinterval and the first moment of the next subinterval are consecutive moments;

[0046] Mark the collection moments in the x subintervals one by one, and number all the collection moments in ascending order according to the chronological order, marking z collection moments with an interval of one number as the standard;

[0047] At z acquisition moments, the air pressure data in the cavity of the OLED evaporation equipment is detected by the air pressure sensor to obtain z air pressure values, and the z-1 air pressure difference value is obtained by subtracting the z-th air pressure value from the z-1-th air pressure value;

[0048] The expression for the pressure difference is:

[0049] QY czxz-1 =QY zxz -QY zxz-1 ;

[0050] In the formula, QY czxz-1 is the z-1th pressure difference value in the xth subinterval, QY zxz is the zth pressure value in the xth subinterval, QY zxz-1 is the z-1th air pressure value in the xth subinterval;

[0051] The vacuum decay value is obtained by accumulating z-1 pressure difference values ​​and averaging them;

[0052] The expression of vacuum decay value is:

[0053]

[0054] Where ZK sjx is the vacuum decay value of the x-th subinterval, QY czxc The cth pressure difference value in the xth subinterval.

[0055] Furthermore, the method for obtaining the humidity increase rate includes:

[0056] In the x subintervals, the collection time at the first position, the collection time at the middle position, and the collection time at the last position are marked respectively, and recorded as the first moment, the second moment, and the third moment;

[0057] Detecting humidity data in the cavity of the OLED evaporation device at a first moment, a second moment, and a third moment by a humidity sensor to obtain a first humidity value, a second humidity value, and a third humidity value;

[0058] After subtracting the second humidity value from the first humidity value, the resultant humidity value is compared with the duration between the first moment and the second moment to obtain x first increase rates;

[0059] The expression of the first rise rate is:

[0060]

[0061] In the formula, SG l1x is the first rising rate of the xth subinterval, SD z2x is the second humidity value of the xth subinterval, SD z1x The first humidity value of the xth subinterval, SC 12x is the duration between the first moment and the second moment of the x-th subinterval;

[0062] Subtract the third humidity value from the second humidity value and compare it with the duration from the second moment to the third moment to obtain x second increase rates;

[0063] The expression of the second rise rate is:

[0064]

[0065] Where, SG l2x is the second rising rate of the xth subinterval, SD z3x is the third humidity value of the xth subinterval, SC 23x is the duration between the second moment and the third moment of the x-th subinterval;

[0066] Add the x first increase rates and the x second increase rates and calculate the average to obtain x humidity increase rates;

[0067] The expression of humidity rise rate is:

[0068]

[0069] In the formula, SD sgx is the humidity increase rate of the xth subinterval.

[0070] Furthermore, the method for obtaining the gas purity value includes:

[0071] At the first moment, the second moment and the third moment, the purity of the inert gas at the gas inlet and the gas outlet of the gas pipeline of the OLED evaporation equipment is detected by the gas sensor respectively, and the first gas inlet purity value, the second gas inlet purity value, the third gas inlet purity value, the first gas outlet purity value, the second gas outlet purity value and the third gas outlet purity value are obtained respectively;

[0072] The first intake air purity value, the second intake air purity value and the third intake air purity value are added and averaged to obtain x intake air purity values;

[0073] The expression of the intake air purity value is:

[0074]

[0075] In the formula, JQ cdx is the intake air purity value of the xth sub-interval, JQ 1xis the first intake air purity value of the xth subinterval, JQ 2x is the second intake air purity value of the xth subinterval, JQ 3x is the third intake air purity value of the xth subinterval:

[0076] The first outlet gas purity value, the second outlet gas purity value and the third outlet gas purity value are added and averaged to obtain x outlet gas purity values;

[0077] The expression of outlet gas purity is:

[0078]

[0079] In the formula, CQ cdx is the outlet gas purity value of the xth sub-interval, CQ 1x is the first outlet gas purity value of the xth sub-interval, CQ 2x is the second outlet gas purity value of the xth sub-interval, CQ 3x is the third outlet gas purity value of the xth subinterval;

[0080] Add x inlet purity values ​​and x outlet purity values ​​and average them to obtain x gas purity values;

[0081] The expression of gas purity value is:

[0082]

[0083] Where QT cdx is the gas purity value of the xth subinterval.

[0084] Furthermore, the training method of the machine learning model includes:

[0085] Collect multiple sets of residual impact data of OLED evaporation equipment in clean and polluted states in advance;

[0086] The residual impact data is converted into multiple feature vectors using a sliding window method. The pollution state value is converted into a label corresponding to the residual impact data according to the sliding step size. The clean state is converted to 0, and the polluted state is converted to 1. One feature vector corresponds to one label and constitutes a set of training data. Multiple sets of training data constitute a training set. The residual impact data are arranged in the order of collection time, and the prediction time step Z, sliding step Q and sliding window length N are preset.

[0087] The feature vector is used as the input of the machine learning model, and the pollution state value of the next sub-interval after the time step Z is predicted as the output. The subsequent pollution state value of each training set is used as the prediction target. The sum of the minimized prediction errors is used as the training target. The machine learning model is trained to generate a machine learning model that predicts the pollution state value of the next sub-interval based on the residual impact data of the previous sub-interval.

[0088] The determination methods for whether to issue a residue control prompt include:

[0089] When the machine learning model output is 0, the pollution state value of the next subinterval is the clean state, and it is determined that no residual control prompt will be issued;

[0090] When the machine learning model output is 1, the contamination state value of the next subinterval is the contamination state, and it is determined that a residual control prompt is issued.

[0091] Furthermore, the target data identification method includes:

[0092] The vacuum decay value is compared with the vacuum safety value, and when the vacuum decay value is greater than the vacuum safety value, the vacuum decay value is identified as target data;

[0093] The humidity increase rate is compared with the humidity safety value. When the humidity increase rate is greater than the humidity safety value, the humidity increase rate is identified as target data.

[0094] Compare the gas purity value with the gas safety value, and when the gas purity value is less than the gas safety value, the gas purity value is identified as target data;

[0095] The method for formulating the vacuum decay value reduction instruction, the humidity increase rate reduction instruction and the gas purity value increase instruction includes:

[0096] When the target data is a vacuum decay value, an instruction to reduce the vacuum decay value is formulated;

[0097] When the target data is the humidity increase rate, an instruction to reduce the humidity increase rate is formulated;

[0098] When the target data is a gas purity value, an instruction to increase the gas purity value is formulated.

[0099] The technical effects and advantages of the method for controlling residual micro-contamination of OLED organic evaporation cavity components of the present invention are as follows:

[0100] The present invention obtains the basic identification parameters of the OLED evaporation equipment, generates identification events based on the basic identification parameters, identifies the event status of the identification event, and determines whether the micro-pollution residual control demand is met based on the standard-reaching judgment criterion, divides the control period into sub-intervals, and obtains the residual influence data of the OLED evaporation equipment in the sub-intervals, inputs the residual influence data into a pre-trained machine learning model, predicts the pollution state value of the next sub-interval, and determines whether to issue a residual control prompt, identifies the target data from the residual influence data, and formulates micro-pollution control instructions corresponding to the target data; compared with the prior art, the present invention can construct identification events based on the basic identification parameters, and effectively identify the basic operating status of the OLED evaporation equipment, effectively avoiding the OLED evaporation equipment in the initial operating period A large amount of irrelevant data is collected, thereby raising the threshold for micro-pollution residue control and greatly reducing the amount of irrelevant data collected. At the same time, the prediction method of residual impact data and machine learning model can be used to make early and accurate predictions of the pollution state value of OLED evaporation equipment at future moments. Therefore, before the OLED evaporation equipment is about to have micro-pollution residue, the target data causing the micro-pollution residue phenomenon can be discovered in time and accurately, and targeted control instructions can be formulated based on the target data to ensure that the OLED evaporation equipment can be reasonably and timely optimized and controlled before the micro-pollution residue phenomenon occurs, thereby preventing the occurrence of micro-pollution residue on the cavity components of the OLED evaporation equipment. Finally, the micro-pollution residue control effect of early discovery and early resolution is achieved, which greatly improves the evaporation work efficiency of the OLED evaporation equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0101] Figure 1 A schematic flow chart of a method for controlling residual micro-contamination of OLED organic evaporation cavity components provided in the first embodiment of the present invention;

[0102] Figure 2 A schematic diagram of a control system for residual micro-contamination of OLED organic evaporation cavity components provided in the second embodiment of the present invention. DETAILED DESCRIPTION

[0103] 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 only 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 ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0104] Example 1: Please refer to Figure 1 As shown, the method for controlling residual micro-contamination of an OLED organic evaporation cavity component in this embodiment includes:

[0105] S1: Obtain basic identification parameters of OLED evaporation equipment, and generate identification events according to the basic identification parameters;

[0106] Basic identification parameters refer to the initial operating data of the OLED evaporation equipment under operation, and serve as the data basis for whether to proceed to the next step of analysis and calculation, so as to accurately determine whether the OLED evaporation equipment needs to be analyzed and calculated for micro-pollution residual control;

[0107] Basic identification parameters include effective operating time value and safe temperature ratio value;

[0108] The effective operation time value refers to the numerical value of the operation time of the OLED evaporation device in the effective operation state, which can accurately represent the time of the OLED evaporation device in the effective operation state. When the effective operation time value is larger, it means that the numerical value of the operation time of the OLED evaporation device in the effective operation state is larger;

[0109] Methods for obtaining the effective runtime value include:

[0110] The operating power of the OLED evaporation equipment is detected by a power sensor to obtain the real-time power value;

[0111] The real-time power value greater than or equal to the preset power lower limit is recorded as the effective power value, and the time when the effective power value first appears is queried through the timestamp to obtain the starting time; the preset power lower limit refers to the minimum value of the real-time power value corresponding to the effective power value, which can limit the lower limit of the effective power value, thereby ensuring the recognition accuracy of the effective power value;

[0112] Between the start time and the current time, the time when the real-time power value is less than the preset power lower limit is recorded as the end time, and the duration between the start time and the end time is recorded as the sub-duration, obtaining 1 sub-duration;

[0113] After accumulating each sub-duration one by one, the effective running time value is obtained;

[0114] The expressions for effective runtime values ​​are:

[0115]

[0116] In the formula, SC yx is the effective running time value, SC za is the ath sub-period.

[0117] The safe temperature ratio refers to the ratio of the duration that the working temperature of the OLED evaporation equipment is within the safe temperature range to the total duration of the temperature value appearing, which can be used to numerically represent the duration that the OLED evaporation equipment reaches the safe temperature range. The larger the safe temperature ratio, the larger the ratio of the duration that the working temperature of the OLED evaporation equipment is within the safe temperature range to the total duration of the temperature value appearing.

[0118] Methods for obtaining the safe temperature ratio value include:

[0119] The operating temperature of the OLED evaporation equipment is detected by a temperature sensor to obtain the real-time temperature value;

[0120] The real-time temperature value that is greater than or equal to the preset lower temperature limit is recorded as the valid temperature value. The time when the valid temperature value first appears is queried through the timestamp and recorded as the temperature start time. The preset lower temperature limit refers to the minimum value of the real-time temperature value corresponding to the valid temperature value, which can limit the lower limit of the valid temperature value, thereby ensuring the recognition accuracy of the valid temperature value.

[0121] After the temperature start time, when no valid temperature value appears within the preset time, the time when the last valid temperature value appears is recorded as the temperature end time, and the time between the temperature start time and the temperature end time is recorded as the safe temperature time, and w safe temperature time is obtained; the preset time refers to the maximum span time between two adjacent valid temperature values, which can limit the collection time of two adjacent valid temperature values ​​to avoid the collection interval between two adjacent valid temperature values ​​being too long or too short;

[0122] The safe temperature duration longer than the standard safe duration is recorded as the target duration, and p target durations are obtained; the standard safe duration refers to the minimum value of the safe temperature duration identified as the target duration, which can limit the lower limit of the target duration;

[0123] The time when the real-time temperature value is first detected is recorded as the initial time, and the duration from the initial time to the current time is recorded as the total running time. The p target durations are accumulated one by one and compared with the total running time to obtain the safe temperature ratio value;

[0124] The expression of the safe temperature ratio is:

[0125]

[0126] Where WD zb is the safe temperature ratio, SC mbb is the bth target duration, SC zyx is the total running time.

[0127] After obtaining the effective operation time value and the safe temperature ratio value, the effective operation time value and the safe temperature ratio value can be processed so that the effective operation time value and the safe temperature ratio value can form a discrimination event. The discrimination event at this time can be used as a basis for subsequent judgment on whether the OLED evaporation equipment needs to perform micro-contamination residual analysis and control operations;

[0128] Methods for generating screening events include:

[0129] An event template with two event positions is selected from the event database, and the two event positions are recorded as the first event position and the second event position respectively according to the order of the positions; the event template refers to a blank event framework without specific data meaning, and serves as a basis for subsequent construction and identification of events; the first event position and the second event position refer to positions in the event template used to fill in and import specific data;

[0130] Mark remark boxes in the first event station and the second event station respectively, and import the effective operation time value and the safety temperature ratio value into the remark boxes of the first event station and the second event station respectively, to obtain the first event station and the second event station after import;

[0131] After importing, number boxes are set up at the front end of the first event station and the second event station, and after entering 01 and 02 in the two number boxes, the identification event is obtained. The number box is set up at the front end to facilitate the distinction between the first event station and the second event station, and also serves as the basis for the subsequent identification event determination.

[0132] It should be noted that the purpose of entering 01 and 02 in the two number boxes is to indicate the order of identification of the first event station and the second event station in the identification event, thereby ensuring that subsequent identification events can be identified and processed in an orderly and accurate manner.

[0133] S2: Identify the event status of the screening event and determine whether the micro-pollution residue control requirements are met based on the compliance criteria;

[0134] When a screening event is generated, the data in the screening event can be identified, so as to accurately obtain the true state of the screening event. The event state is used to represent the true state of the screening event and is used as a direct reflection of whether the micro-pollution residue control requirements are met. After identifying the event state of the screening event, it is necessary to make a judgment based on the compliance judgment criteria to ensure the accuracy of the judgment on whether the micro-pollution residue control requirements are met.

[0135] The event status includes the compliance status and non-compliance status; the compliance status refers to the status that the data on the event station in the screening event meets the preset standard value, and the non-compliance status refers to the status that the data on the event station in the screening event does not meet the preset standard value;

[0136] Methods for identifying compliance status and non-compliance status include:

[0137] Set the effective running time value SC in the first event station yx With the standard duration value SC bz Comparison; The standard duration value refers to the minimum value of the effective operation duration value when the micro-pollution residual control requirement is met, thereby providing a data basis for determining the time status of the first event station;

[0138] When SC yx Greater than or equal to SC bz When , it indicates that the effective operation duration value is greater than or equal to the standard duration value, and the event status of the first event station of the identified event is in the standard-compliant state;

[0139] When SC yx Smaller than SC bz When , it means that the effective operation time value is less than the standard time value. At this time, the event status of the first event station of the screening event is a non-standard state;

[0140] The safety temperature ratio WD in the second event station zb Compared with the standard ratio WD bz Comparison; The standard proportion value refers to the minimum value of the safe temperature proportion value when the micro-contamination residual control requirement is met, thereby providing a data basis for determining the time state of the second event station;

[0141] When WD zb Greater than or equal to WD bz When , it means that the safety temperature ratio is greater than or equal to the standard ratio. At this time, the event status of the second event station of the identification event is up to standard;

[0142] When WD zb Smaller than WD bz When , it means that the safety temperature ratio is less than the standard ratio. At this time, the event status of the second event station of the screening event is a non-standard status;

[0143] The criteria for determining compliance are: there is no non-compliance status in the screening event;

[0144] The methods for determining whether the requirements for micro-pollution residue control are met include:

[0145] Count the number of non-standard states at the first event station and the second event station in the identification event to obtain the non-standard value;

[0146] When the non-standard value is 0, it means that there is no non-standard state in the screening event, and it is determined that the micro-pollution residual control requirements are met;

[0147] When the non-standard value is greater than 0, it means that a non-standard state has occurred in the identification event, and it is determined that the micro-pollution residual control requirements are not met.

[0148] S3: Divide the control period into sub-intervals, and obtain residual impact data of the OLED evaporation equipment in the sub-intervals, the residual impact data including vacuum attenuation value, humidity increase rate and gas purity value;

[0149] When the requirements for residual micro-pollution control are met, relevant operation data of the OLED evaporation equipment during operation can be collected and analyzed to determine whether residual impact data affecting residual micro-pollution occurs during operation of the OLED evaporation equipment, and the collected residual impact data is used as a basis for subsequent determination of the operation status of the OLED evaporation equipment.

[0150] The control cycle refers to the minimum duration of a complete cycle of data iteration of the OLED evaporation equipment, so that each residual influence parameter within the control cycle can change a sufficient number of times and amplitudes, thereby facilitating the subsequent collection of residual influence data;

[0151] Residual impact data include vacuum decay value, humidity rise rate and gas purity value;

[0152] The vacuum attenuation value refers to the increase degree of the air pressure value in the OLED evaporation equipment within the sub-interval, which can be used to indicate the attenuation of the vacuum degree in the OLED evaporation equipment. When the vacuum attenuation value is larger, it means that the air pressure in the OLED evaporation equipment will be larger. At this time, the evaporation environment in the OLED evaporation equipment is worse, and the residual amount of micro-contamination is larger.

[0153] The methods for obtaining the vacuum decay value include:

[0154] The control cycle is divided into x subintervals, and the last moment of the previous subinterval and the first moment of the next subinterval are consecutive moments;

[0155] Mark the collection moments in the x subintervals one by one, and number all the collection moments in ascending order according to the chronological order, with one number interval as the standard, and mark z The use of interval numbering can not only reduce the number of collection moments and the subsequent calculation burden, but also avoid the adhesion of data at two adjacent collection moments, thereby improving the independence of data;

[0156] exist zAt a collection time, the air pressure data in the cavity of the OLED evaporation equipment is detected by the air pressure sensor to obtain z air pressure values, and the z-1 air pressure difference value is obtained by subtracting the z-th air pressure value from the z-1-th air pressure value;

[0157] The expression for the pressure difference is:

[0158] QY czxz-1 =QY zxz -QY zxz-1 ;

[0159] In the formula, QY czxz-1 is the z-1th pressure difference value in the xth subinterval, QY zxz is the zth pressure value in the xth subinterval, QY zxz-1 is the z-1th air pressure value in the xth subinterval;

[0160] The vacuum decay value is obtained by accumulating z-1 pressure difference values ​​and averaging them;

[0161] The expression of vacuum decay value is:

[0162]

[0163] Where ZK sjx is the vacuum decay value of the x-th subinterval, QY czxc The cth pressure difference value in the xth subinterval.

[0164] The humidity increase rate refers to the rate of increase of the humidity value in the OLED evaporation equipment within the sub-interval, which can be used to indicate the humidity level in the OLED evaporation equipment. The greater the humidity increase rate, the greater the rate of increase of the humidity value in the OLED evaporation equipment. At this time, the worse the evaporation environment in the OLED evaporation equipment is, the greater the residual amount of micro-contamination is.

[0165] Methods for obtaining the humidity rise rate include:

[0166] In the x subintervals, the collection time at the first position, the collection time at the middle position, and the collection time at the last position are marked respectively, and recorded as the first moment, the second moment, and the third moment;

[0167] Detecting humidity data in the cavity of the OLED evaporation device at a first moment, a second moment, and a third moment by a humidity sensor to obtain a first humidity value, a second humidity value, and a third humidity value;

[0168] After subtracting the second humidity value from the first humidity value, the resultant humidity value is compared with the duration between the first moment and the second moment to obtain x first increase rates;

[0169] The expression of the first rise rate is:

[0170]

[0171] Where, SG l1x is the first rising rate of the xth subinterval, SD z2x is the second humidity value of the xth subinterval, SD z1x The first humidity value of the xth subinterval, SC 12x is the duration between the first moment and the second moment of the x-th subinterval;

[0172] Subtract the third humidity value from the second humidity value and compare it with the duration from the second moment to the third moment to obtain x second increase rates;

[0173] The expression of the second rise rate is:

[0174]

[0175] Where, SG l2x is the second rising rate of the xth subinterval, SD z3x is the third humidity value of the xth subinterval, SC 23x is the duration between the second moment and the third moment of the x-th subinterval;

[0176] Add the x first increase rates and the x second increase rates and calculate the average to obtain x humidity increase rates;

[0177] The expression of humidity rise rate is:

[0178]

[0179] In the formula, SD sgx is the humidity increase rate of the xth subinterval.

[0180] The gas purity value refers to the numerical value of the gas cleanliness in the OLED evaporation equipment within the sub-interval, which can represent the gas purity in the OLED evaporation equipment. When the gas purity value is larger, it means that the gas purity in the OLED evaporation equipment is higher. At this time, the evaporation environment in the OLED evaporation equipment is better, and the residual amount of micro-contamination is less;

[0181] Methods for obtaining gas purity values ​​include:

[0182] At the first moment, the second moment and the third moment, the purity of the inert gas at the gas inlet of the gas pipeline of the OLED evaporation device is detected by the gas sensor respectively, and a first inlet gas purity value, a second inlet gas purity value and a third inlet gas purity value are obtained respectively;

[0183] At the first moment, the second moment and the third moment, the purity of the inert gas at the gas outlet of the gas pipeline of the OLED evaporation device is detected by the gas sensor respectively, and a first outlet gas purity value, a second outlet gas purity value and a third outlet gas purity value are obtained respectively;

[0184] The first intake air purity value, the second intake air purity value and the third intake air purity value are added and averaged to obtain x intake air purity values;

[0185] The expression of the intake air purity value is:

[0186]

[0187] In the formula, JQ cdx is the intake air purity value of the xth sub-interval, JQ 1x is the first intake air purity value of the xth subinterval, JQ 2x is the second intake air purity value of the xth subinterval, JQ 3x is the third intake air purity value of the xth subinterval:

[0188] The first outlet gas purity value, the second outlet gas purity value and the third outlet gas purity value are added and averaged to obtain x outlet gas purity values;

[0189] The expression of outlet gas purity is:

[0190]

[0191] In the formula, CQ cdx is the outlet gas purity value of the xth sub-interval, CQ 1x is the first outlet gas purity value of the xth sub-interval, CQ 2x is the second outlet gas purity value of the xth sub-interval, CQ 3x is the third outlet gas purity value of the xth subinterval;

[0192] Add x inlet purity values ​​and x outlet purity values ​​and average them to obtain x gas purity values;

[0193] The expression of gas purity value is:

[0194]

[0195] Where QT cdx is the gas purity value of the xth subinterval.

[0196] By collecting and calculating and analyzing the gas purity at the air inlet and outlet, the real gas purity environment in the gas pipeline and the OLED evaporation equipment cavity can be accurately simulated, thereby providing an accurate numerical basis for the calculation of the gas purity value.

[0197] S4: Input the residual impact data into the pre-trained machine learning model to predict the pollution state value of the next sub-interval and determine whether to issue a residual control prompt. The pollution state value includes clean state and polluted state.

[0198] After the vacuum decay value, humidity increase rate and gas purity value of the OLED evaporation equipment in the sub-interval are obtained, the vacuum decay value, humidity increase rate and gas purity value can be input into the pre-trained machine learning model, so that the machine learning model can predict the pollution state value of the next sub-interval according to the vacuum decay value, humidity increase rate and gas purity value of the previous sub-interval collected, which serves as a basis for judging whether the cavity components of the OLED evaporation equipment need to be controlled for micro-pollution residues;

[0199] The pollution state value includes clean state and pollution state; the pollution state refers to the residual micro-pollution phenomenon in the cavity parts of the OLED evaporation equipment, and the clean state refers to the absence of residual micro-pollution phenomenon in the cavity parts of the OLED evaporation equipment; the pollution state value is obtained by collecting a large number of vacuum decay values, humidity increase rates and gas purity values ​​corresponding to the clean state and pollution state in history;

[0200] Methods for training machine learning models include:

[0201] Collect multiple sets of residual impact data of OLED evaporation equipment in clean and polluted states in advance;

[0202] The residual impact data is converted into multiple feature vectors using a sliding window method. The pollution state value is converted into a label corresponding to the residual impact data according to the sliding step size. The clean state is converted to 0, and the polluted state is converted to 1. One feature vector corresponds to one label and constitutes a set of training data. Multiple sets of training data constitute a training set. The residual impact data are arranged in the order of collection time, and the prediction time step Z, sliding step Q and sliding window length N are preset.

[0203] The feature vector is used as the input of the machine learning model, and the pollution state value of the next sub-interval after the predicted time step Z is used as the output. The subsequent pollution state value of each training set is used as the prediction target. The machine learning model is trained with the minimized sum of prediction errors as the training target to generate a machine learning model that predicts the pollution state value of the next sub-interval based on the residual impact data of the previous sub-interval.

[0204] Exemplarily, the machine learning model is any one of CNN or AlexNet;

[0205] The calculation formula for the prediction error is:

[0206] zk=(ak-wk)2 ;

[0207] Where zk is the prediction error, k is the group number of the feature vector; ak is the predicted state value corresponding to the kth group of feature vectors, and Wk is the actual state value corresponding to the kth group of training data;

[0208] In the machine learning model, the feature vector is the residual impact data, and the state value is the contamination state value;

[0209] Other model parameters of the machine learning model, target loss value, optimization algorithm, training set test set validation set ratio, and loss function optimization are all achieved through actual engineering implementation and continuous experimental tuning.

[0210] The residual control prompt is used to issue an early warning prompt when micro-contamination residual occurs in the cavity parts of the OLED evaporation equipment, and serves as the basis for subsequent micro-contamination control of the cavity parts of the OLED evaporation equipment;

[0211] The determination methods for whether to issue a residue control prompt include:

[0212] When the machine learning model output is 0, it means that the pollution state value of the next sub-interval is a clean state. At this time, there is no micro-pollution residue in the cavity parts of the OLED evaporation equipment, and it is determined that no residual control prompt will be issued;

[0213] When the machine learning model output is 1, it means that the pollution state value of the next sub-interval is the pollution state. At this time, micro-pollution residues appear in the cavity parts of the OLED evaporation equipment, and it is determined to issue a residual control prompt.

[0214] S5: Identify target data from residual impact data and formulate micro-pollution control instructions corresponding to the target data;

[0215] The target data refers to the specific data in which anomalies appear in the residual impact data, so that the target data can be used as the object of subsequent micro-pollution residual control of the cavity components of the OLED evaporation equipment, so as to formulate appropriate micro-pollution control instructions;

[0216] Methods for identifying target data include:

[0217] Compare the vacuum decay value with the vacuum safety value; the vacuum safety value refers to the maximum value of the vacuum decay value when micro-contamination residues appear in the cavity parts of the OLED evaporation equipment, which can distinguish the actual size of the vacuum decay value, thereby ensuring the recognition accuracy of the target data;

[0218] When the vacuum decay value is greater than the vacuum safety value, it means that the vacuum decay value exceeds the maximum value of the vacuum decay value in the clean state, and the vacuum decay value is identified as the target data;

[0219] Compare the humidity increase rate with the humidity safety value; the humidity safety value refers to the maximum value of the humidity increase rate when micro-contamination residues appear in the cavity parts of the OLED evaporation equipment, which can distinguish the actual size of the humidity increase rate, thereby ensuring the recognition accuracy of the target data;

[0220] When the humidity increase rate is greater than the humidity safety value, it means that the humidity increase rate exceeds the maximum value of the humidity increase rate in the clean state, and the humidity increase rate is identified as the target data;

[0221] Compare the gas purity value with the gas safety value; the gas safety value refers to the minimum value of the gas purity value when micro-contamination residues appear in the cavity parts of the OLED evaporation equipment, which can distinguish the actual size of the gas purity value, thereby ensuring the recognition accuracy of the target data;

[0222] When the gas purity value is less than the gas safety value, it means that the gas purity value has not reached the minimum value of the gas purity value in a clean state. At this time, the gas purity value is identified as target data.

[0223] When the target data is identified, the corresponding micro-pollution control instructions can be formulated according to the identified target data, so that the micro-pollution control instructions can be used as specific measures for the OLED evaporation equipment to control the micro-pollution residue, thereby ensuring that the cavity parts of the OLED evaporation equipment will not have micro-pollution residues, so as to improve the evaporation efficiency of the OLED evaporation equipment;

[0224] Micro-pollution control instructions include instructions for reducing vacuum attenuation value, reducing humidity increase rate, and increasing gas purity value;

[0225] The method for formulating the vacuum decay value reduction instruction, the humidity increase rate reduction instruction and the gas purity value increase instruction includes:

[0226] When the target data is a vacuum attenuation value, it is necessary to control the vacuum attenuation value of the OLED evaporation equipment, and an instruction to reduce the vacuum attenuation value is formulated;

[0227] When the target data is the humidity increase rate, the humidity increase rate of the OLED evaporation equipment needs to be controlled, and an instruction to reduce the humidity increase rate is formulated;

[0228] When the target data is a gas purity value, it is necessary to control the gas purity value of the OLED evaporation equipment, and an instruction to increase the gas purity value is formulated.

[0229] It should be noted that when an instruction to reduce the vacuum attenuation value is formulated, the air pressure in the OLED evaporation equipment needs to be reduced, and the operating power of the vacuum pump needs to be increased. When an instruction to reduce the humidity increase rate is formulated, the humidity in the OLED evaporation equipment needs to be reduced, and the operating power of the dehumidifier needs to be increased. When an instruction to increase the gas purity value is formulated, the gas purity in the OLED evaporation equipment needs to be increased, and the filtration efficiency of the inert gas in the gas pipeline needs to be enhanced.

[0230] In the present embodiment, by obtaining the basic identification parameters of the OLED evaporation equipment, and generating identification events based on the basic identification parameters, identifying the event status of the identification event, and judging whether the micro-pollution residual control requirement is met based on the standard-reaching judgment criteria, the control period is divided into sub-intervals, and the residual impact data of the OLED evaporation equipment in the sub-intervals is obtained, the residual impact data is input into a pre-trained machine learning model, the pollution state value of the next sub-interval is predicted, and it is judged whether a residual control prompt is issued, the target data is identified from the residual impact data, and micro-pollution control instructions corresponding to the target data are formulated; compared with the prior art, the identification event can be constructed according to the basic identification parameters, and the basic operating status of the OLED evaporation equipment can be effectively identified, which effectively avoids the OLED evaporation equipment in the initial operation. A large amount of irrelevant data in a certain period of time has been collected, thereby raising the threshold for micro-pollution residue control and greatly reducing the amount of irrelevant data collected. At the same time, the prediction method of residual impact data and machine learning model can be used to make early and accurate predictions on the pollution state value of OLED evaporation equipment at future moments. Therefore, before the OLED evaporation equipment is about to have micro-pollution residue, the target data causing the micro-pollution residue phenomenon can be discovered in a timely and accurate manner, and targeted control instructions can be formulated based on the target data to ensure that the OLED evaporation equipment can be reasonably and timely optimized and controlled before the micro-pollution residue phenomenon occurs, thereby preventing the occurrence of micro-pollution residue on the cavity components of the OLED evaporation equipment. Finally, the micro-pollution residue control effect of early discovery and early resolution is achieved, which greatly improves the evaporation work efficiency of the OLED evaporation equipment.

[0231] Example 2: Please refer to Figure 2 As shown, the part not described in detail in this embodiment is described in the first embodiment, and a control system for residual micro-contamination of OLED organic evaporation cavity components is provided, which is used to implement a control method for residual micro-contamination of OLED organic evaporation cavity components, including an event identification module, a demand determination module, a data acquisition module, a model prediction module and an instruction formulation module, wherein each module is connected via a wired or wireless network;

[0232] The identification event module is used to obtain the basic identification parameters of the OLED evaporation equipment and generate identification events based on the basic identification parameters. The basic identification parameters include the effective operation time value and the safe temperature ratio value;

[0233] The demand determination module is used to identify the event status of the screening event and determine whether the micro-pollution residue control demand is met based on the compliance determination criteria;

[0234] A data acquisition module is used to divide the control period into sub-intervals and obtain the residual impact data of the OLED evaporation equipment in the sub-intervals, the residual impact data including vacuum decay value, humidity rise rate and gas purity value;

[0235] The model prediction module is used to input the residual impact data into the pre-trained machine learning model, predict the pollution state value of the next sub-interval, and determine whether to issue a residual control prompt. The pollution state value includes clean state and polluted state;

[0236] The instruction formulation module is used to identify target data from the residual impact data and formulate micro-pollution control instructions corresponding to the target data. The micro-pollution control instructions include instructions for reducing the vacuum attenuation value, reducing the humidity increase rate, and increasing the gas purity value.

[0237] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A method for controlling residual micro-contamination of OLED organic evaporation cavity components, characterized in that: include: S1: Obtain basic identification parameters of the OLED evaporation equipment, and generate identification events based on the basic identification parameters, where the basic identification parameters include effective operation time value and safe temperature ratio value; S2: Identify the event status of the screening event, and determine whether the micro-pollution residue control requirements are met based on the compliance criteria; if the micro-pollution residue control requirements are met, execute S3; if the micro-pollution residue control requirements are not met, repeat S1-S2; S3: Divide the control period into sub-intervals, and obtain residual impact data of the OLED evaporation equipment in the sub-intervals, the residual impact data including vacuum attenuation value, humidity increase rate and gas purity value; S4: Input the residual impact data into the pre-trained machine learning model, predict the pollution state value of the next sub-interval, and determine whether to issue a residual control prompt. The pollution state value includes a clean state and a polluted state. If a residual control prompt is issued, execute S5; if no residual control prompt is issued, repeat S3-S4. S5: Identify target data from the residual impact data, and formulate micro-pollution control instructions corresponding to the target data. The micro-pollution control instructions include instructions for reducing vacuum attenuation value, instructions for reducing humidity increase rate, and instructions for increasing gas purity value.

2. The method for controlling residual micro-contamination of OLED organic evaporation cavity components according to claim 1, characterized in that: The method for obtaining the effective running time value includes: The operating power of the OLED evaporation equipment is detected by a power sensor to obtain the real-time power value; The real-time power value greater than or equal to the preset power lower limit is recorded as the effective power value, and the time when the effective power value first appears is queried through the timestamp to obtain the starting time; Between the start time and the current time, the time when the real-time power value is less than the preset power lower limit is recorded as the end time, and the duration between the start time and the end time is recorded as the sub-duration, and i sub-durations are obtained; After accumulating the i sub-durations one by one, the effective running time value is obtained; The expressions for effective runtime values ​​are: In the formula, SC yx is the effective running time value, SC za is the ath sub-period.

3. The method for controlling residual micro-contamination of OLED organic evaporation cavity components according to claim 2, characterized in that: The method for obtaining the safety temperature ratio value includes: The operating temperature of the OLED evaporation equipment is detected by a temperature sensor to obtain the real-time temperature value; The real-time temperature value that is greater than or equal to the preset lower temperature limit is recorded as the effective temperature value, and the time when the effective temperature value first appears is queried through the timestamp and recorded as the temperature starting time; After the temperature start time, when no valid temperature value appears within the preset time, the time when the last valid temperature value appears is recorded as the temperature end time, and the time between the temperature start time and the temperature end time is recorded as the safe temperature time, and w safe temperature time is obtained; The safe temperature duration that is greater than the standard safe duration is recorded as the target duration, and p target durations are obtained; The time when the real-time temperature value is first detected is recorded as the initial time, and the duration from the initial time to the current time is recorded as the total running time. The p target durations are accumulated one by one and compared with the total running time to obtain the safe temperature ratio value; The expression of the safe temperature ratio is: Where, WD zb is the safe temperature ratio, SC mbb is the bth target duration, SC zyx is the total running time.

4. The method for controlling residual micro-contamination of OLED organic evaporation cavity components according to claim 3, characterized in that: The method for generating the screening event includes: An event template having two event stations is selected from the event database, and the two event stations are recorded as a first event station and a second event station respectively according to the order of their positions; Mark remark boxes in the first event station and the second event station respectively, and import the effective operation time value and the safety temperature ratio value into the remark boxes of the first event station and the second event station respectively, to obtain the first event station and the second event station after import; After the import, number boxes are respectively established at the front end of the first event station and the second event station, and after inputting 01 and 02 in the two number boxes respectively, the identification event is obtained.

5. The method for controlling residual micro-contamination of OLED organic evaporation cavity components according to claim 4, characterized in that: The event status includes the compliance status and non-compliance status: Methods for identifying compliance status and non-compliance status include: Set the effective running time value SC in the first event station yx With the standard duration value SC bz Compare; When SC yx Greater than or equal to SC bz When SC yx Smaller than SC bz When the event status of the first event station of the screening event is a non-standard state; The safety temperature ratio WD in the second event station zb Compared with the standard ratio WD bz Compare; When WD zb Greater than or equal to WD bz When WD zb Smaller than WD bz When , the event status of the second event station of the screening event is a non-standard state; The criteria for determining compliance are: there is no non-compliance status in the screening event; The methods for determining whether the requirements for micro-pollution residue control are met include: Count the number of non-standard states at the first event station and the second event station in the identification event to obtain the non-standard value; When the non-standard value is 0, it is determined that the micro-pollution residual control requirements are met; When the non-standard value is greater than 0, it is determined that the micro-pollution residual control requirements are not met.

6. The method for controlling residual micro-contamination of OLED organic vapor deposition cavity components according to claim 5, characterized in that: The method for obtaining the vacuum decay value comprises: The control period is equally divided into X subintervals, and the last moment of the previous subinterval and the first moment of the next subinterval are consecutive moments; Mark the collection moments in the X sub-intervals one by one, and number all the collection moments in ascending order according to the chronological order, marking z collection moments with an interval of one number as the standard; At Z acquisition moments, the air pressure data in the cavity of the OLED evaporation equipment is detected by the air pressure sensor to obtain z air pressure values, and the z-1 air pressure difference value is obtained by subtracting the z-th air pressure value from the z-1-th air pressure value; The expression for the pressure difference is: QY CZXZ-1 =QY ZXZ -QY ZXZ-1 ; In the formula, QY CZXZ-1 is the z-1th pressure difference value in the xth subinterval, QY ZXZ is the zth pressure value in the xth subinterval, QY ZXZ-1 is the z-1th air pressure value in the xth subinterval; The vacuum decay value is obtained by accumulating z-1 pressure difference values ​​and averaging them; The expression of vacuum decay value is: Where ZK sjx is the vacuum decay value of the x-th subinterval, QY czxc The cth pressure difference value in the xth subinterval.

7. The method for controlling residual micro-contamination of OLED organic evaporation cavity components according to claim 6, characterized in that: The method for obtaining the humidity increase rate includes: In the x subintervals, the collection time at the first position, the collection time at the middle position, and the collection time at the last position are marked respectively, and recorded as the first moment, the second moment, and the third moment; The humidity sensor detects humidity data in the cavity of the OLED evaporation device at the first moment, the second moment and the third moment respectively, to obtain a first humidity value, a second humidity value and a third humidity value; After subtracting the second humidity value from the first humidity value, the resultant humidity value is compared with the duration between the first moment and the second moment to obtain x first increase rates; The expression of the first rise rate is: In the formula, SG l1x is the first rising rate of the xth subinterval, SD z2x is the second humidity value of the xth subinterval, SD z1x The first humidity value of the xth subinterval, SC 12x is the duration between the first moment and the second moment of the x-th subinterval; Subtract the third humidity value from the second humidity value and compare with the duration from the second moment to the third moment to obtain X second increase rates; The expression of the second rise rate is: In the formula, SG l2x is the second rising rate of the xth subinterval, SD z3x is the third humidity value of the xth subinterval, SC 23x is the duration between the second moment and the third moment of the x-th subinterval; Add the X first increase rates and the X second increase rates and then calculate the average to obtain X humidity increase rates; The expression of humidity rise rate is: In the formula, SD sgx is the humidity increase rate of the xth subinterval.

8. The method for controlling residual micro-contamination of OLED organic vapor deposition cavity components according to claim 7, characterized in that: The method for obtaining the gas purity value comprises: At the first moment, the second moment, and the third moment, the purity of the inert gas at the gas inlet and the gas outlet of the gas pipeline of the OLED evaporation device is detected by the gas sensor respectively, and a first gas inlet purity value, a second gas inlet purity value, a third gas inlet purity value, a first gas outlet purity value, a second gas outlet purity value, and a third gas outlet purity value are obtained respectively; The first intake air purity value, the second intake air purity value and the third intake air purity value are added and averaged to obtain X intake air purity values; The expression of the intake air purity value is: In the formula, JQ cdx is the intake air purity value of the xth sub-interval, JQ 1x is the first intake air purity value of the xth subinterval, JQ 2x is the second intake air purity value of the xth subinterval, JQ 3x is the third intake air purity value of the x-th subinterval; The first outlet gas purity value, the second outlet gas purity value and the third outlet gas purity value are added and averaged to obtain X outlet gas purity values; The expression of outlet gas purity is: In the formula, CQ cdx is the outlet gas purity value of the xth sub-interval, CQ 1x is the first outlet gas purity value of the xth sub-interval, CQ 2x is the second outlet gas purity value of the xth sub-interval, CQ 3x is the third outlet gas purity value of the xth subinterval; Add x inlet purity values ​​and x outlet purity values ​​and average them to obtain x gas purity values; The expression of gas purity value is: Where QT cdx is the gas purity value of the xth subinterval.

9. The method for controlling residual micro-contamination of OLED organic vapor deposition cavity components according to claim 8, characterized in that: The training method of the machine learning model includes: Collect multiple sets of residual impact data of OLED evaporation equipment in clean and polluted states in advance; The residual impact data is converted into multiple feature vectors using a sliding window method. The pollution state value is converted into a label corresponding to the residual impact data according to the sliding step size. The clean state is converted to 0, and the polluted state is converted to 1. One feature vector corresponds to one label and constitutes a set of training data. Multiple sets of training data constitute a training set. The residual impact data are arranged in the order of collection time, and the prediction time step Z, sliding step Q and sliding window length N are preset. The feature vector is used as the input of the machine learning model, and the pollution state value of the next sub-interval after the time step Z is predicted as the output. The subsequent pollution state value of each training set is used as the prediction target. The sum of the minimized prediction errors is used as the training target. The machine learning model is trained to generate a machine learning model that predicts the pollution state value of the next sub-interval based on the residual impact data of the previous sub-interval. The determination methods for whether to issue a residue control prompt include: When the machine learning model output is 0, the pollution state value of the next subinterval is the clean state, and it is determined that no residual control prompt will be issued; When the machine learning model output is 1, the contamination state value of the next subinterval is the contamination state, and it is determined that a residual control prompt is issued.

10. The method for controlling residual micro-contamination of OLED organic vapor deposition cavity components according to claim 9, characterized in that: The target data identification method comprises: The vacuum decay value is compared with the vacuum safety value, and when the vacuum decay value is greater than the vacuum safety value, the vacuum decay value is identified as target data; The humidity increase rate is compared with the humidity safety value. When the humidity increase rate is greater than the humidity safety value, the humidity increase rate is identified as target data. Compare the gas purity value with the gas safety value, and when the gas purity value is less than the gas safety value, the gas purity value is identified as target data; The method for formulating the vacuum decay value reduction instruction, the humidity increase rate reduction instruction and the gas purity value increase instruction includes: When the target data is a vacuum decay value, an instruction to reduce the vacuum decay value is formulated; When the target data is the humidity increase rate, an instruction to reduce the humidity increase rate is formulated; When the target data is a gas purity value, an instruction to increase the gas purity value is formulated.

Images