A vacuum pump operation control method and system

Abstract

The invention relates to the technical field of vacuum pumps, and discloses a vacuum pump operation control method and system, which obtain actual values ​​of the vacuum degree of a target space at multiple moments, and obtain error rates and error change rates of the vacuum degree of the target space at multiple moments according to expected values ​​of the vacuum degree of the target space; calibrate the error rate and the error change rate according to the outer air pressure of the target space to obtain a standard error rate and a standard error change rate; set a PID controller of the vacuum pump, initialize parameters of the PID controller, and determine adjustable ranges for the parameters of the PID controller; construct a first membership degree and a second membership degree for the standard error rate and the standard error change rate, respectively; match the first membership degree and the second membership degree to a plurality of preset membership degree matrices to obtain a plurality of membership degree sets, and obtain adjustment values ​​of the parameters of the PID controller for the membership degree sets.

Description

Technical Field

[0001] The present invention relates to the field of vacuum pumps, and more particularly to a vacuum pump operation control method and system. Background Art

[0002] A vacuum pump is a device used to achieve a near-vacuum environment and is widely used in multiple industrial and scientific fields. However, in actual use, the control accuracy of the vacuum pump is often affected by external air pressure fluctuations. Traditional vacuum pump control systems usually rely on manual adjustment of controller parameters to cope with changes in external conditions. However, when the external air pressure fluctuates frequently, manual adjustment is difficult to compensate for these changes in a timely and effective manner, resulting in difficulty in maintaining stable control accuracy of the vacuum pump.

[0003] There are some methods in the prior art that attempt to achieve dynamic adjustment of vacuum pump controller parameters. For example, some methods use optimization algorithms to iteratively optimize controller parameters, treating controller parameters as encoded individuals and continuously adjusting them to approach the optimal control effect. However, in the initial stage, due to the characteristics of random optimization, this method may introduce a large control deviation before the parameters converge, resulting in the system's control accuracy being far below expectations for a period of time. Summary of the invention

[0004] The present invention provides a vacuum pump operation control method and system to solve the technical problems raised in the background technology.

[0005] In a first aspect, the present invention provides a vacuum pump operation control method, comprising:

[0006] Step 1, obtaining the actual value of the vacuum degree of the target space at multiple times, and obtaining the error rate and error change rate of the vacuum degree of the target space at multiple times according to the expected value of the vacuum degree of the target space;

[0007] Step 2, calibrating the error rate and the error change rate according to the outer air pressure of the target space to obtain the standard error rate and the standard error change rate;

[0008] Step 3, setting the PID controller of the vacuum pump, initializing the parameters of the PID controller, and determining the adjustable ranges of the parameters of the PID controller respectively;

[0009] Step 4, constructing the first membership degree and the second membership degree for the standard error rate and the standard error change rate respectively;

[0010] Step 5, matching the first membership degree and the second membership degree to a plurality of preset membership degree matrices to obtain a plurality of membership degree sets, and obtaining adjustment values ​​of parameters of the PID controller for the membership degree sets;

[0011] Step 6: According to the adjustment results of the adjustment values ​​of the parameters of the PID controller, effective adjustment values ​​are screened, and the corresponding parameters are used as training samples, and the effective adjustment values ​​are used as sample labels to build an adjustment model.

[0012] Furthermore, the error rate and error change rate of the target space vacuum degree at multiple times are obtained, including:

[0013] According to the expected value of the vacuum degree of the target space , calculate the The error rate of the target space vacuum degree at a certain moment is calculated as follows:

[0014] ;in, Indicates The error rate of the target space vacuum degree at a moment, Indicates the vacuum degree of the target space The actual value at the moment;

[0015] According to the actual values ​​of the target space vacuum degree at multiple moments, the error change rate of the target space vacuum degree at the tth moment is calculated. The calculation formula is as follows:

[0016] ;in, represents the error change rate of the target space vacuum degree at the tth moment, Indicates the vacuum degree of the target space The actual value at a moment.

[0017] Furthermore, the outer air pressure of the target space is used to calibrate the error rate and error change rate, including:

[0018] Step 7, constructing a simulation experiment, the simulation experiment includes a first test space and a second test space;

[0019] Step 8, the second test space is located inside the first test space, and a vacuum pump is provided in the first test space to evacuate the second test space according to a fixed expected value;

[0020] Step 9, randomly adjusting the air pressure in the first test space within a preset range;

[0021] Step 10, calculating the error rate and error change rate of the second test space at fixed time intervals, and matching the error rate and error change rate under different air pressures;

[0022] Step 11, setting the standard atmospheric pressure, and taking the error rate and error change rate under the standard atmospheric pressure as the reference error rate and reference error change rate;

[0023] Step 12, fitting the error rate and error change rate under different air pressures to the reference error rate and reference error change rate to obtain the pressure deviation fitting factor of the reference error rate and reference error change rate;

[0024] Step 13, calibrating the error rate and the error change rate according to the pressure deviation fitting factor and the outer air pressure of the target space.

[0025] Further, the PID controller of the vacuum pump is set, including:

[0026] The formula for the PID controller is as follows:

[0027] ;in, Represents the output of the PID controller, which is used to control the speed of the motor; , and Respectively represent the proportional term coefficient, integral term coefficient and differential term coefficient; Represents the time from the 0th moment to the tth moment To perform the integration operation, represents the integral variable, , It represents the time integral unit in the integral, which is used to divide the time period from the 0th moment to the tth moment into several time intervals;

[0028] in, , and The initialization value represents the parameters of the PID controller working under standard atmospheric pressure. The adjustable range is set by experts through experimental tests and experience according to the control requirements of the PID controller.

[0029] Furthermore, the first membership degree and the second membership degree are constructed for the standard error rate and the standard error change rate, respectively, including:

[0030] Step 14, determining the value range of the standard error rate and the value range of the standard error change rate based on historical data;

[0031] Step 15, defining the fuzzy linguistic variables of the standard error rate, including negative N, zero Z and positive P; the fuzzy linguistic variables of the standard error change rate, including decreasing D, stable S and increasing U;

[0032] Step 16, assigning a membership function to each fuzzy linguistic variable, and defining the low end, center, and high end of each membership function, associating the center of the membership function with zero Z or stable S, associating the low end of the membership function with negative N or decreasing D; associating the high end of the membership function with positive P or increasing U;

[0033] Step 17, mapping the standard error rate to the corresponding fuzzy linguistic variable through a membership function to obtain a first membership of the corresponding fuzzy linguistic variable;

[0034] The standard error change rate is mapped to the corresponding fuzzy linguistic variable through the membership function to obtain the second membership of the corresponding fuzzy linguistic variable.

[0035] Furthermore, the first membership degree and the second membership degree are matched to a plurality of preset membership degree matrices to obtain a plurality of membership degree sets, and adjustment values ​​of parameters of the PID controller are obtained for the membership degree sets, including:

[0036] Step 18, construct multiple membership matrices, the rows of the membership matrix represent decreasing D, stable S and increasing U respectively, and the columns of the membership matrix represent negative N, zero Z and positive P respectively, as follows:

[0037] The first membership matrix is ​​used to control the proportional term coefficient of the PID controller ;

[0038] The second membership matrix is ​​used to control the integral term coefficient of the PID controller ;

[0039] The third membership matrix is ​​used to control the differential term coefficients of the PID controller ;

[0040] Step 19, based on historical experience, define the rules of the membership matrix through experts; wherein each cell in the membership matrix represents a comparative term coefficient , integral term coefficient Or the differential coefficient The fuzzy control rules for adjustment;

[0041] Step 20, querying each membership matrix according to the first membership degree and the second membership degree to obtain a query output of each membership matrix;

[0042] Step 21, perform fuzzy clarification on the query output to obtain the proportional term coefficient of the PID controller , integral term coefficient and the differential term coefficient adjustment value.

[0043] Furthermore, according to the adjustment results of the adjustment values ​​of the parameters of the PID controller, effective adjustment values ​​are screened, and the corresponding parameters are used as training samples, and the effective adjustment values ​​are used as sample labels to construct an adjustment model, including:

[0044] Step 22: Obtain the proportional term coefficient of the vacuum pump PID controller of each area based on the cloud , integral term coefficient and the differential term coefficient Adjustment value, and the corresponding expected vacuum value, actual vacuum value and deviation from standard atmospheric pressure;

[0045] Step 23, set a screening threshold, screen the difference between the expected vacuum value and the actual vacuum value, and if the difference between the expected vacuum value and the actual vacuum value is less than the screening threshold, the corresponding proportional term coefficient , integral term coefficient and the differential term coefficient The adjustment value, as well as the corresponding expected vacuum value, actual vacuum value and deviation value from the standard atmospheric pressure are used as training data;

[0046] Step 24: Set the proportional term coefficient , integral term coefficient and the differential term coefficient The adjustment value is used as a sample label of the adjustment model training data, and the expected value of the vacuum degree, the actual value of the vacuum degree and the deviation value from the standard atmospheric pressure are used as the adjustment model training samples;

[0047] Step 25, reversely update the weight parameters and bias parameters of the hidden layer of the adjustment model through the mean error variance loss function.

[0048] In a second aspect, a vacuum pump operation control system is applied to the method described, comprising:

[0049] The data acquisition module is used to obtain the actual value of the vacuum degree of the target space at multiple times, and obtain the error rate and error change rate of the vacuum degree of the target space at multiple times according to the expected value of the vacuum degree of the target space;

[0050] A data calibration module is used to calibrate the error rate and the error change rate according to the outer air pressure of the target space to obtain a standard error rate and a standard error change rate;

[0051] A controller module is used to set a PID controller of a vacuum pump, initialize parameters of the PID controller, and determine adjustable ranges of the parameters of the PID controller;

[0052] A membership mapping module, used for constructing a first membership degree and a second membership degree for a standard error rate and a standard error change rate, respectively;

[0053] A parameter adjustment module, used for matching the first membership degree and the second membership degree to a plurality of preset membership degree matrices to obtain a plurality of membership degree sets, and obtaining adjustment values ​​of parameters of the PID controller for the membership degree sets;

[0054] The adjustment model module is used to screen effective adjustment values ​​according to the adjustment results of the adjustment values ​​of the parameters of the PID controller, and use the corresponding parameters as training samples and the effective adjustment values ​​as sample labels to build an adjustment model.

[0055] In a third aspect, a vacuum pump operation control device is provided, the vacuum pump control device comprising:

[0056] A memory, a processor, and a vacuum pump control program stored in the memory and executable on the processor, wherein the vacuum pump control program is configured to implement the method as described above.

[0057] In a fourth aspect, a storage medium is provided, wherein a vacuum pump control program is stored on the storage medium, and the vacuum pump control program implements the method described above when executed by a processor.

[0058] The beneficial effect of the present invention is that the PID controller of the vacuum pump is fuzzily updated in the early stage of the system by means of fuzzy control, and then the vacuum pump is preliminarily optimized, avoiding most of the influence of atmospheric pressure changes on the control accuracy of the vacuum pump. Then, the adjustment model is constructed by the effective parameters obtained by the preliminary optimization, and the fuzzy update is converted into a precise update, which further improves the control accuracy of the vacuum pump. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 is a flow chart of a vacuum pump operation control method of the present invention;

[0060] Figure 2 It is a module diagram of a vacuum pump operation control system of the present invention. DETAILED DESCRIPTION

[0061] The subject matter described herein will now be discussed with reference to example implementations. It should be understood that the discussion of these implementations is only to enable those skilled in the art to better understand and implement the subject matter described herein, and the functions and arrangements of the elements discussed may be changed without departing from the scope of protection of the contents of this specification. Various examples may omit, replace, or add various processes or components as needed. In addition, the features described in some examples may also be combined in other examples.

[0062] like Figure 1 As shown, a vacuum pump operation control method includes:

[0063] Step 1, obtaining the actual value of the vacuum degree of the target space at multiple times, and obtaining the error rate and error change rate of the vacuum degree of the target space at multiple times according to the expected value of the vacuum degree of the target space;

[0064] Step 2, calibrating the error rate and the error change rate according to the outer air pressure of the target space to obtain the standard error rate and the standard error change rate;

[0065] Step 3, setting the PID controller of the vacuum pump, initializing the parameters of the PID controller, and determining the adjustable ranges of the parameters of the PID controller respectively;

[0066] Step 4, constructing the first membership degree and the second membership degree for the standard error rate and the standard error change rate respectively;

[0067] Step 5, matching the first membership degree and the second membership degree to a plurality of preset membership degree matrices to obtain a plurality of membership degree sets, and obtaining adjustment values ​​of parameters of the PID controller for the membership degree sets;

[0068] Step 6: According to the adjustment results of the adjustment values ​​of the parameters of the PID controller, effective adjustment values ​​are screened, and the corresponding parameters are used as training samples, and the effective adjustment values ​​are used as sample labels to build an adjustment model.

[0069] In one embodiment of the present invention, the error rate and error change rate of the target space vacuum degree at multiple times are obtained, including:

[0070] According to the expected value of the vacuum degree of the target space , calculate the The error rate of the target space vacuum degree at a certain moment is calculated as follows:

[0071] ;in, Indicates The error rate of the target space vacuum degree at a moment, Indicates the vacuum degree of the target space The actual value at the moment;

[0072] According to the actual values ​​of the target space vacuum degree at multiple moments, the error change rate of the target space vacuum degree at the tth moment is calculated. The calculation formula is as follows:

[0073] ;in, represents the error change rate of the target space vacuum degree at the tth moment, Indicates the vacuum degree of the target space The actual value at a moment.

[0074] In one embodiment of the present invention, the outer air pressure of the target space is used to calibrate the error rate and the error change rate, including:

[0075] Step 7, constructing a simulation experiment, the simulation experiment includes a first test space and a second test space;

[0076] Step 8, the second test space is located inside the first test space, and a vacuum pump is provided in the first test space to evacuate the second test space according to a fixed expected value;

[0077] Step 9, randomly adjusting the air pressure in the first test space within a preset range;

[0078] Step 10, calculating the error rate and error change rate of the second test space at fixed time intervals, and matching the error rate and error change rate under different air pressures;

[0079] Step 11, setting the standard atmospheric pressure, and taking the error rate and error change rate under the standard atmospheric pressure as the reference error rate and reference error change rate;

[0080] Step 12, fitting the error rate and error change rate under different air pressures to the reference error rate and reference error change rate to obtain the pressure deviation fitting factor of the reference error rate and reference error change rate;

[0081] Step 13, calibrating the error rate and the error change rate according to the pressure deviation fitting factor and the outer air pressure of the target space.

[0082] For example, several data were obtained, some of which are as follows:

[0083] At multiple moments in the first preset time period, the air pressure of the first test space is 90 kPa, the average error rate of the second test space at multiple moments is -10%, and the average error change rate of the second test space at multiple moments is 5%.

[0084] At multiple moments in the second preset time period, the air pressure in the first test space was 100 kPa, the average error rate of the second test space at multiple moments was -5%, and the average error change rate of the second test space at multiple moments was 3%.

[0085] At multiple moments in the third preset time period, the air pressure in the first test space was 110 kPa, the average error rate of the second test space at multiple moments was 2%, and the average error change rate of the second test space at multiple moments was -3%.

[0086] Step 31, setting 100 kPa as the standard atmospheric pressure, the reference error rate is -5%, and the reference error change rate is 3%.

[0087] Step 32, initialize the error rate fitting function: .

[0088] Step 33, for the first preset time period, the pressure difference is 10 kPa, and the error rate difference is 0.05; then .

[0089] Step 34, for the second preset time period, the pressure difference is -10 kPa, and the error rate difference is -0.07, then .

[0090] Step 35, combine to get ; Represents the formula obtained based on the remaining data.

[0091] Step 36, solve the above series equation and get , and The value of ; then the pressure deviation fitting factor of the error rate is obtained.

[0092] Step 37, when the external air pressure of the target space is Y, the error rate is , then according to the difference between air pressure Y and standard atmospheric pressure, the air pressure deviation fitting factor S1 is obtained, and the error rate after calibration is .

[0093] Step 38, according to steps 31 to 37, the pressure deviation fitting factor of the error deviation rate is obtained in the same way, and the pressure deviation fitting factor is used to Perform calibration.

[0094] In one embodiment of the present invention, setting a PID controller of a vacuum pump comprises:

[0095] The formula for the PID controller is as follows:

[0096] ;in, Represents the output of the PID controller, which is used to control the speed of the motor; , and Respectively represent the proportional term coefficient, integral term coefficient and differential term coefficient; Represents the time from the 0th moment to the tth moment To perform the integration operation, represents the integral variable, , It represents the time integral unit in the integral, which is used to divide the time period from the 0th moment to the tth moment into several time intervals;

[0097] in, , and The initialization value represents the parameters of the PID controller working under standard atmospheric pressure. The adjustable range is set by experts through experimental tests and experience according to the control requirements of the PID controller.

[0098] Specifically, the PID controller coordinates the three parts of proportion, integration and differentiation to enable the vacuum pump system to quickly and accurately reach the required vacuum degree under standard atmospheric pressure, thereby maintaining a stable vacuum environment.

[0099] It should be noted that the PID controller is a common technology used in the field of vacuum pump control and will not be described in detail here.

[0100] In one embodiment of the present invention, a first degree of membership and a second degree of membership are constructed for the standard error rate and the standard error change rate, respectively, including:

[0101] Step 14, determining the value range of the standard error rate and the value range of the standard error change rate based on historical data;

[0102] Step 15, defining the fuzzy linguistic variables of the standard error rate, including negative N, zero Z and positive P; the fuzzy linguistic variables of the standard error change rate, including decreasing D, stable S and increasing U;

[0103] Step 16, assigning a membership function to each fuzzy linguistic variable, and defining the low end, center, and high end of each membership function, associating the center of the membership function with zero Z or stable S, associating the low end of the membership function with negative N or decreasing D; associating the high end of the membership function with positive P or increasing U;

[0104] Step 17, mapping the standard error rate to the corresponding fuzzy linguistic variable through a membership function to obtain a first membership of the corresponding fuzzy linguistic variable;

[0105] The standard error change rate is mapped to the corresponding fuzzy linguistic variable through the membership function to obtain the second membership of the corresponding fuzzy linguistic variable.

[0106] Step 51, assuming that in the historical data, the standard error rate range is ; The range of standard error change rate is .

[0107] Step 52, for the standard error rate, three fuzzy linguistic variables are defined, including negative N, zero Z and positive P;

[0108] For the standard error change rate, three fuzzy linguistic variables are defined, namely, decreasing D, stable S, and increasing U;

[0109] Step 53, defining membership functions for negative N, zero Z, positive P, falling D, stable S and rising U respectively;

[0110] Assume that the triangular membership function is defined for negative N as follows:

[0111] ;in, represents the low end of the triangular membership function, represents the center of the triangular membership function, Represents the high end of the triangular membership function;

[0112] The trapezoidal membership function is defined for zero Z as follows:

[0113] ;in, represents the low end of the trapezoidal membership function, represents the center of the trapezoidal membership function, Represents the high end of the trapezoidal membership function;

[0114] The Gaussian membership function is defined for positive P as follows:

[0115] ;in, represents the mean of the Gaussian function, represents the standard deviation of the Gaussian function; represents the center of the Gaussian membership function, As the lower end of the Gaussian membership function, As the high end of the Gaussian membership function;

[0116] Step 54, through three membership functions, Mapped to negative N, zero Z and positive P fuzzy linguistic variables respectively, assuming that the first membership degree is ;

[0117] Step 55, according to step 51 to step 54, the second membership degree is obtained in the same way. Assume that the second membership degree is .

[0118] In one embodiment of the present invention, the first membership and the second membership are matched to a plurality of preset membership matrices to obtain a plurality of membership sets, and the adjustment values ​​of the parameters of the PID controller are obtained for the membership sets, including:

[0119] Step 18, construct multiple membership matrices, the rows of the membership matrix represent decreasing D, stable S and increasing U respectively, and the columns of the membership matrix represent negative N, zero Z and positive P respectively, as follows:

[0120] The first membership matrix is ​​used to control the proportional term coefficient of the PID controller ;

[0121] The second membership matrix is ​​used to control the integral term coefficient of the PID controller ;

[0122] The third membership matrix is ​​used to control the differential term coefficients of the PID controller ;

[0123] Step 19, based on historical experience, define the rules of the membership matrix through experts; wherein each cell in the membership matrix represents a comparative term coefficient , integral term coefficient Or the differential coefficient The fuzzy control rules for adjustment;

[0124] Step 20, querying each membership matrix according to the first membership degree and the second membership degree to obtain a query output of each membership matrix;

[0125] Step 21, perform fuzzy clarification on the query output to obtain the proportional term coefficient of the PID controller , integral term coefficient and the differential term coefficient adjustment value.

[0126] Specifically, they include:

[0127] Step 61, constructing a first membership matrix, as follows:

[0128]

[0129] Step 62, based on the first degree of membership and the second degree of membership Query the first membership matrix, and the query output formula is as follows:

[0130] ;in, Represents the output of the query output formula, represents the union operation of the set, 9 represents the 9 cells in the first membership matrix, Represents the index of 9 units, represents the first degree of membership, represents the transpose operation of a vector, represents the mth unit in the first membership matrix, Indicates a user-defined fuzzy operation, including maximum and minimum operations or maximum product operations, etc.

[0131] Step 63, based on the calculation result of the query output formula, the query output of the first membership matrix is ​​obtained. The query output is usually a polygon edge set of a two-dimensional coordinate graph. An algorithm is randomly selected to perform fuzzy sharpening on the polygon edge set to obtain the proportional term coefficient The adjustment value of the fuzzy sharpening algorithm includes the maximum membership method, the minimum membership method, the average membership method and the area center method, etc.

[0132] Step 64, based on the proportional term coefficient The adjustment value of the proportional term coefficient Adjust the integral coefficient according to steps 61 to 64. and the differential term coefficient Make adjustments and get an updated PID controller.

[0133] It should be noted that the random selection of an algorithm for blur sharpening in step 63 is to obtain diversified training data for constructing an adjustment model in step 6.

[0134] In one embodiment of the present invention, according to the adjustment result of the adjustment value of the parameter of the PID controller, the effective adjustment value is screened, and the corresponding parameter is used as a training sample, and the effective adjustment value is used as a sample label to construct an adjustment model, including:

[0135] Step 22: Obtain the proportional term coefficient of the vacuum pump PID controller of each area based on the cloud , integral term coefficient and the differential term coefficient Adjustment value, and the corresponding expected vacuum value, actual vacuum value and deviation from standard atmospheric pressure;

[0136] Step 23, set a screening threshold, screen the difference between the expected vacuum value and the actual vacuum value, and if the difference between the expected vacuum value and the actual vacuum value is less than the screening threshold, the corresponding proportional term coefficient , integral term coefficient and the differential term coefficient The adjustment value, as well as the corresponding expected vacuum value, actual vacuum value and deviation value from the standard atmospheric pressure are used as training data;

[0137] Step 24: Set the proportional term coefficient , integral term coefficient and the differential term coefficient The adjustment value is used as a sample label of the adjustment model training data, and the expected value of the vacuum degree, the actual value of the vacuum degree and the deviation value from the standard atmospheric pressure are used as the adjustment model training samples;

[0138] Step 25, reversely update the weight parameters and bias parameters of the hidden layer of the adjustment model through the mean error variance loss function.

[0139] Assume that the vacuum pump using the method described in the present invention is sold to different regions around the world. After the vacuum pump arrives at the destination, the system can automatically obtain the local air pressure value, and according to the difference in air pressure, perform fuzzy update on the PID controller through the operations of step 1 to step 22, thereby realizing adaptive adjustment of control parameters to improve control accuracy. During the fuzzy update process, the vacuum pump adjusts parameters through fuzzy logic and rule base to obtain new PID controller parameters suitable for the current air pressure environment. However, this PID controller obtained by fuzzy update can only improve the control accuracy to a certain extent, and its improvement effect has certain limitations and unevenness. Specifically: in some cases, the accuracy improvement of the PID controller after fuzzy update is about 5%, that is, the deviation between the actual vacuum degree and the target vacuum degree is reduced by 5%; while in other cases, the accuracy improvement of the PID controller after fuzzy update may reach 8%, that is, the deviation is reduced by 8%. Due to the difference in this accuracy improvement effect, it is necessary to screen and verify the update effect. To this end, the present invention proposes a screening mechanism to set a screening threshold: first, set an allowable accuracy deviation range as the screening threshold. The threshold is used to determine whether the updated PID controller meets the accuracy requirements. Calculate deviation: For each vacuum pump that has been updated, measure the actual value of its vacuum degree under the current air pressure environment, compare it with the expected value of the vacuum degree, and calculate the deviation between the two. Screen effective updates: When the deviation between the actual vacuum degree and the expected vacuum degree of a vacuum pump after the update is less than the screening threshold, it is determined that the updated data can be used as training data for further optimization and adjustment of the model.

[0140] The expected value of vacuum degree, the actual value of vacuum degree and the deviation value from standard atmospheric pressure in the training data are combined to construct an input vector of the adjustment model.

[0141] The proportionality coefficient , integral term coefficient and the differential term coefficient The adjustment value combinations construct the sample labels for the adjustment model.

[0142] The loss is calculated by the mean error variance of the output of the adjustment model, and the weight parameters and bias parameters of the hidden layer of the adjustment model are reversely updated by the gradient descent method according to the loss. In this way, the trained adjustment model is obtained, and the parameters of the PID controller of vacuum pumps in various places are dynamically scheduled by the adjustment model step by step instead of the fuzzy update.

[0143] The beneficial effect of the present invention is that the PID controller of the vacuum pump is fuzzily updated in the early stage of the system by fuzzy control, and then the vacuum pump is preliminarily optimized, avoiding most of the influence of atmospheric pressure changes on the control accuracy of the vacuum pump. Then, the adjustment model is constructed by the effective parameters obtained by the preliminary optimization, and the fuzzy update is converted into a precise update, which further improves the control accuracy of the vacuum pump.

[0144] In one embodiment of the present invention, the hidden layer of the adjustment model is constructed based on a support vector machine.

[0145] In one embodiment of the present invention, the weight parameters and bias parameters of the hidden layer of the adjustment model are optimized and updated by the mean error variance loss function, that is, the variance of the output of the adjustment model for the training samples and the sample labels is calculated as the loss of the adjustment model.

[0146] like Figure 2 As shown, a vacuum pump operation control system is applied to the method described above, comprising:

[0147] The data acquisition module is used to obtain the actual value of the vacuum degree of the target space at multiple times, and obtain the error rate and error change rate of the vacuum degree of the target space at multiple times according to the expected value of the vacuum degree of the target space;

[0148] A data calibration module is used to calibrate the error rate and the error change rate according to the outer air pressure of the target space to obtain a standard error rate and a standard error change rate;

[0149] A controller module is used to set a PID controller of a vacuum pump, initialize parameters of the PID controller, and determine adjustable ranges of the parameters of the PID controller;

[0150] A membership mapping module, used for constructing a first membership degree and a second membership degree for a standard error rate and a standard error change rate, respectively;

[0151] A parameter adjustment module, used for matching the first membership degree and the second membership degree to a plurality of preset membership degree matrices to obtain a plurality of membership degree sets, and obtaining adjustment values ​​of parameters of the PID controller for the membership degree sets;

[0152] The adjustment model module is used to screen effective adjustment values ​​according to the adjustment results of the adjustment values ​​of the parameters of the PID controller, and use the corresponding parameters as training samples and the effective adjustment values ​​as sample labels to build an adjustment model;

[0153] A vacuum pump operation control device, the vacuum pump control device comprising:

[0154] A memory, a processor, and a vacuum pump control program stored in the memory and executable on the processor, wherein the vacuum pump control program is configured to implement the method as described above.

[0155] A storage medium stores a vacuum pump control program, and the vacuum pump control program implements the method when executed by a processor.

[0156] The above describes an embodiment of the present embodiment, but the present embodiment is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present embodiment, ordinary technicians in this field can also make many forms, all of which are within the protection of the present embodiment.

Claims

1. A vacuum pump operation control method, characterized in that: include: Step 1: obtain the actual value of the vacuum degree of the target space at multiple times, and obtain the error rate and error change rate of the vacuum degree of the target space at multiple times according to the expected value of the vacuum degree of the target space, as follows: According to the expected value of the vacuum degree of the target space , calculate the The error rate of the target space vacuum degree at a certain moment is calculated as follows: ;in, Indicates The error rate of the target space vacuum degree at a moment, Indicates the vacuum degree of the target space The actual value at the moment; According to the actual values ​​of the target space vacuum degree at multiple moments, the error change rate of the target space vacuum degree at the tth moment is calculated. The calculation formula is as follows: ;in, represents the error change rate of the target space vacuum degree at the tth moment, Indicates the vacuum degree of the target space The actual value at the moment; Step 2, calibrating the error rate and the error change rate according to the outer air pressure of the target space to obtain the standard error rate and the standard error change rate; Step 3, setting the PID controller of the vacuum pump, initializing the parameters of the PID controller, and determining the adjustable ranges of the parameters of the PID controller respectively; Step 4, constructing the first membership degree and the second membership degree for the standard error rate and the standard error change rate respectively; Step 5, matching the first membership degree and the second membership degree to a plurality of preset membership degree matrices to obtain a plurality of membership degree sets, and obtaining adjustment values ​​of parameters of the PID controller for the membership degree sets; Step 6, according to the adjustment results of the adjustment values ​​of the parameters of the PID controller, the effective adjustment values ​​are screened, and the corresponding parameters are used as training samples, and the effective adjustment values ​​are used as sample labels to build an adjustment model; The outside air pressure of the target space is used to calibrate the error rate and error change rate, including: Step 7, constructing a simulation experiment, the simulation experiment includes a first test space and a second test space; Step 8, the second test space is located inside the first test space, and a vacuum pump is provided in the first test space to evacuate the second test space according to a fixed expected value; Step 9, randomly adjusting the air pressure in the first test space within a preset range; Step 10, calculating the error rate and error change rate of the second test space at fixed time intervals, and matching the error rate and error change rate under different air pressures; Step 11, setting the standard atmospheric pressure, and taking the error rate and error change rate under the standard atmospheric pressure as the reference error rate and reference error change rate; Step 12, fitting the error rate and error change rate under different air pressures to the reference error rate and reference error change rate to obtain the pressure deviation fitting factor of the reference error rate and reference error change rate; Step 13, calibrating the error rate and the error change rate according to the pressure deviation fitting factor and the outer air pressure of the target space.

2. A vacuum pump operation control method according to claim 1, characterized in that: Set up the PID controller for the vacuum pump, including: The formula for the PID controller is as follows: ;in, Represents the output of the PID controller, which is used to control the speed of the motor; , and Respectively represent the proportional term coefficient, integral term coefficient and differential term coefficient; Represents the time from the 0th moment to the tth moment To perform the integration operation, represents the integral variable, , It represents the time integral unit in the integral, which is used to divide the time period from the 0th moment to the tth moment into several time intervals; in, , and The initialization value represents the parameters of the PID controller working under standard atmospheric pressure. The adjustable range is set by experts through experimental tests and experience according to the control requirements of the PID controller.

3. A vacuum pump operation control method according to claim 1, characterized in that: The first membership and the second membership are constructed for the standard error rate and the standard error change rate, respectively, including: Step 14, determining the value range of the standard error rate and the value range of the standard error change rate based on historical data; Step 15, defining the fuzzy linguistic variables of the standard error rate, including negative N, zero Z and positive P; the fuzzy linguistic variables of the standard error change rate, including decreasing D, stable S and increasing U; Step 16, assigning a membership function to each fuzzy linguistic variable, and defining the low end, center, and high end of each membership function, associating the center of the membership function with zero Z or stable S, associating the low end of the membership function with negative N or decreasing D; associating the high end of the membership function with positive P or increasing U; Step 17, mapping the standard error rate to the corresponding fuzzy linguistic variable through a membership function to obtain a first membership of the corresponding fuzzy linguistic variable; The standard error change rate is mapped to the corresponding fuzzy linguistic variable through the membership function to obtain the second membership of the corresponding fuzzy linguistic variable.

4. A vacuum pump operation control method according to claim 1, characterized in that: Matching the first membership degree and the second membership degree to a plurality of preset membership degree matrices to obtain a plurality of membership degree sets, and obtaining adjustment values ​​of parameters of the PID controller for the membership degree sets, including: Step 18, construct multiple membership matrices, the rows of the membership matrix represent decreasing D, stable S and increasing U respectively, and the columns of the membership matrix represent negative N, zero Z and positive P respectively, as follows: The first membership matrix is ​​used to control the proportional term coefficient of the PID controller ; The second membership matrix is ​​used to control the integral term coefficient of the PID controller ; The third membership matrix is ​​used to control the differential term coefficients of the PID controller ; Step 19, based on historical experience, define the rules of the membership matrix through experts; wherein each cell in the membership matrix represents a comparative term coefficient , integral term coefficient Or the differential coefficient The fuzzy control rules for adjustment; Step 20, querying each membership matrix according to the first membership degree and the second membership degree to obtain a query output of each membership matrix; Step 21, perform fuzzy clarification on the query output to obtain the proportional term coefficient of the PID controller , integral term coefficient and the differential term coefficient adjustment value.

5. A vacuum pump operation control method according to claim 1, characterized in that: According to the adjustment results of the adjustment values ​​of the parameters of the PID controller, the effective adjustment values ​​are screened, and the corresponding parameters are used as training samples, and the effective adjustment values ​​are used as sample labels to build an adjustment model, including: Step 22: Obtain the proportional term coefficient of the vacuum pump PID controller of each area based on the cloud , integral term coefficient and the differential term coefficient Adjustment value, and the corresponding expected vacuum value, actual vacuum value and deviation from standard atmospheric pressure; Step 23, set a screening threshold, screen the difference between the expected vacuum value and the actual vacuum value, and if the difference between the expected vacuum value and the actual vacuum value is less than the screening threshold, the corresponding proportional term coefficient , integral term coefficient and the differential term coefficient The adjustment value, as well as the corresponding expected vacuum value, actual vacuum value and deviation value from the standard atmospheric pressure are used as training data; Step 24: Set the proportional term coefficient , integral term coefficient and the differential term coefficient The adjustment value is used as a sample label of the adjustment model training data, and the expected value of the vacuum degree, the actual value of the vacuum degree and the deviation value from the standard atmospheric pressure are used as the adjustment model training samples; Step 25, reversely update the weight parameters and bias parameters of the hidden layer of the adjustment model through the mean error variance loss function.

6. A vacuum pump operation control system, characterized in that: The method according to any one of claims 1 to 5, comprising: The data acquisition module is used to obtain the actual value of the vacuum degree of the target space at multiple times, and obtain the error rate and error change rate of the vacuum degree of the target space at multiple times according to the expected value of the vacuum degree of the target space; A data calibration module is used to calibrate the error rate and the error change rate according to the outer air pressure of the target space to obtain the standard error rate and the standard error change rate; A controller module is used to set a PID controller of a vacuum pump, initialize parameters of the PID controller, and determine adjustable ranges of the parameters of the PID controller; A membership mapping module, used for constructing a first membership degree and a second membership degree for a standard error rate and a standard error change rate, respectively; A parameter adjustment module, used for matching the first membership degree and the second membership degree to a plurality of preset membership degree matrices to obtain a plurality of membership degree sets, and obtaining adjustment values ​​of parameters of the PID controller for the membership degree sets; The adjustment model module is used to screen effective adjustment values ​​according to the adjustment results of the adjustment values ​​of the parameters of the PID controller, and use the corresponding parameters as training samples and the effective adjustment values ​​as sample labels to build an adjustment model.

7. A vacuum pump operation control device, characterized in that: The vacuum pump operation control device comprises: A memory, a processor, and a vacuum pump control program stored in the memory and executable on the processor, wherein the vacuum pump control program is configured to implement the method according to any one of claims 1 to 5.

8. A storage medium, characterized in that: The storage medium stores a vacuum pump control program, and when the vacuum pump control program is executed by a processor, the method according to any one of claims 1 to 5 is implemented.

Images