Valve viscosity detection method and device

Abstract

The invention relates to a valve viscosity detection method and device. The method comprises the steps that target historical data of a preprocessed control loop in a past preset time period is obtained; adopting a valve viscosity model to process MV historical data; establishing a process feature model based on the PV historical data and the real opening value of the valve, and obtaining an expression of a first parameter and an expression of a second parameter based on the linear process feature model; performing simulation by using the expression of the first parameter and the expression of the second parameter to obtain a simulated PV estimated value, establishing an expression of a mean square error of the PV historical data and the PV estimated value, and establishing a target function with the minimum mean square error as an optimization target; solving the target function to obtain an optimal viscosity parameter; according to the optimal viscosity parameter, the target valve viscosity level corresponding to the optimal viscosity parameter is determined, and compared with the prior art, the calculation amount is reduced, and the accuracy of valve viscosity detection can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of valve sticking detection, and in particular to a valve sticking detection method and device. Background Art

[0002] Valve sticking is a common fault in process control, which may cause problems such as oscillation of the controlled variable in the process, thereby causing the control performance of the control loop to deteriorate, affecting production stability and product quality. Therefore, an appropriate non-invasive method, namely the valve sticking detection algorithm, can detect the valve sticking condition through the daily operation data of the control loop, and give a timely evaluation of the valve sticking condition, helping operators to make timely decisions and reduce its impact on control performance.

[0003] At present, the existing technology performs valve sticking detection by means of graphic fitting.

[0004] However, in actual industrial processes, the computational complexity of the graphic fitting method is relatively large and the accuracy is relatively low. Summary of the invention

[0005] 1. Technical issues to be resolved In view of the above-mentioned shortcomings and deficiencies of the prior art, the present invention provides a valve viscosity detection method and device, which solves the technical problem of relatively low accuracy in the prior art.

[0006] (II) Technical solution In order to achieve the above object, the main technical solutions adopted by the present invention include: In a first aspect, an embodiment of the present invention provides a valve stickiness detection method, comprising: obtaining target historical data of a control loop after preprocessing in the past preset time period; wherein the target historical data includes PV historical data and MV historical data; processing the MV historical data using a valve stickiness model to obtain a valve real opening value; wherein the valve real opening value is obtained by a stickiness parameter, and the stickiness parameter is used to determine a target valve stickiness level; establishing a linear process characteristic model including a first parameter and a second parameter based on the PV historical data and the valve real opening value, and obtaining an expression of the first parameter and an expression of the second parameter based on the linear process characteristic model; wherein the input data of the linear process characteristic model is the valve real opening value, and the output data of the linear process characteristic model is the PV historical data; performing simulation using the expression of the first parameter and the expression of the second parameter to obtain a simulated PV estimation value, and establishing an expression of the mean square error of the PV historical data and the PV estimation value, and establishing an objective function with the minimum mean square error as the optimization target; solving the objective function to obtain an optimal stickiness parameter; and determining the corresponding target valve stickiness level according to the optimal stickiness parameter.

[0007] In a possible embodiment, the viscosity parameter includes a static friction parameter and a dynamic friction parameter; the calculation expression of the actual valve opening value is as follows: ; In the formula, Indicates the current time t The actual valve opening value; Represents the MV historical data at time t; cum_u Indicates the external force currently acting on the valve that is balanced with the friction force; represents the static friction parameter; represents the dynamic friction parameter; abs Indicates absolute value; Indicates the last moment t- 1 is the actual valve opening value.

[0008] In one possible embodiment, the linear process characteristic model is expressed as follows: ; In the formula, to Respectively represent the PV history data from the 1st to the N PV historical data; to Respectively represent the first moment to the N- The actual valve opening value at time 1; a Represents the first parameter; b Represents the second parameter; to They represent the errors from the second to the first due to noise. N An error due to noise.

[0009] In one possible embodiment, the expression of mean square error is as follows: ; ; In the formula, MSE represents mean square error; Representation and The corresponding estimated value; Representation and The corresponding estimated value; Representation and The corresponding estimated value.

[0010] In a possible embodiment, solving the objective function to obtain the optimal viscosity parameter includes: using a grid search algorithm to explore a rough area in the search space of the viscosity parameter; and using a Bayesian optimization algorithm to refine the search in the rough area to obtain the optimal viscosity parameter.

[0011] In a possible embodiment, a target valve viscosity level corresponding to an optimal viscosity parameter is determined, including: calculating a valve viscosity quantification index according to the optimal viscosity parameter; determining a target index interval in which the valve viscosity quantification index is located, and using the valve viscosity level corresponding to the target index interval as the target valve viscosity level.

[0012] In a second aspect, an embodiment of the present invention provides a valve viscosity detection device, comprising: an acquisition module, used to acquire target historical data of a preprocessed control loop in the past preset time period; wherein the target historical data includes PV historical data and MV historical data; a processing module, used to process the MV historical data using a valve viscosity model to obtain a valve actual opening value; wherein the valve actual opening value is obtained by a viscosity parameter, and the viscosity parameter is used to determine a target valve viscosity level; an establishment module, used to establish a linear process characteristic model including a first parameter and a second parameter based on the PV historical data and the valve actual opening value, and based on the linear process characteristic A linear process characteristic model is used to obtain an expression of the first parameter and an expression of the second parameter; wherein the input data of the linear process characteristic model is the actual valve opening value, and the output data of the linear process characteristic model is the PV historical data; a simulation module is used to perform simulation using the expression of the first parameter and the expression of the second parameter to obtain a simulated PV estimation value, and to establish an expression of the mean square error between the PV historical data and the PV estimation value, and to establish an objective function with the minimum mean square error as the optimization target; a solution module is used to solve the objective function to obtain an optimal viscosity parameter; a determination module is used to determine the corresponding target valve viscosity level according to the optimal viscosity parameter.

[0013] In a possible embodiment, the viscosity parameter includes a static friction parameter and a dynamic friction parameter; the calculation expression of the actual valve opening value is as follows: ; In the formula, Indicates the current time t The actual valve opening value; Represents the MV historical data at time t; cum_u Indicates the external force currently acting on the valve that is balanced with the friction force; represents the static friction parameter; represents the dynamic friction parameter; abs Indicates absolute value; Indicates the last moment t- 1 is the actual valve opening value.

[0014] In one possible embodiment, the linear process characteristic model is expressed as follows: ; In the formula, to Respectively represent the PV history data from the 1st to the N PV historical data; to Respectively represent the first moment to the N- The actual valve opening value at time 1; a Represents the first parameter; b Represents the second parameter; to They represent the errors from the second to the first due to noise. N An error due to noise.

[0015] In one possible embodiment, the expression of mean square error is as follows: ; ; In the formula, MSE represents mean square error; Representation and The corresponding estimated value; Representation and The corresponding estimated value; Representation and The corresponding estimated value.

[0016] (III) Beneficial effects The beneficial effects of the present invention are: The embodiment of the present application provides a valve viscosity detection method and device, which obtains target historical data of a preprocessed control loop in a preset time period in the past, wherein the target historical data includes PV historical data and MV historical data, and processes the MV historical data using a valve viscosity model to obtain a valve true opening value, wherein the valve true opening value is obtained by a viscosity parameter, and the viscosity parameter is used to determine a target valve viscosity level, and a linear process characteristic model including a first parameter and a second parameter is established based on the PV historical data and the valve true opening value, and an expression of the first parameter and the second parameter is obtained based on the linear process characteristic model. An expression; wherein, the input data of the linear process characteristic model is the actual valve opening value, the output data of the linear process characteristic model is the PV historical data, and the expression of the first parameter and the expression of the second parameter are used for simulation to obtain the simulated PV estimated value, and an expression of the mean square error of the PV historical data and the PV estimated value is established, and an objective function with the minimum mean square error as the optimization target is established, and the objective function is solved to obtain the optimal viscosity parameter, and according to the optimal viscosity parameter, the corresponding target valve viscosity level is determined. Compared with the prior art, it not only reduces the amount of calculation, but also can improve the accuracy of valve viscosity detection.

[0017] In order to make the above-mentioned objectives, features and advantages to be achieved by the embodiments of the present application more obvious and understandable, the following specifically cites preferred embodiments and describes them in detail with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments of the present application will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0019] Figure 1 A flow chart of a valve sticking detection method provided by an embodiment of the present application is shown; Figure 2 A schematic diagram of a valve viscosity model provided in an embodiment of the present application is shown; Figure 3 A schematic diagram of a search space provided in an embodiment of the present application is shown; Figure 4 A structural block diagram of a valve viscosity detection device provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0020] In order to better explain the present invention and facilitate understanding, the present invention is described in detail below through specific implementation modes in conjunction with the accompanying drawings.

[0021] In addition to the graph fitting method, the existing technology can also use neural network algorithms for valve sticking detection. In addition, if a deep neural network algorithm is to be trained, a large amount of labeled data is required for model training. However, in actual industrial processes, obtaining high-quality labeled data is extremely costly and difficult to obtain.

[0022] Based on this, an embodiment of the present application provides a valve viscosity detection method and device, which obtains target historical data of a preprocessed control loop in a preset time period in the past, wherein the target historical data includes PV historical data and MV historical data, and processes the MV historical data using a valve viscosity model to obtain a true valve opening value, wherein the true valve opening value is obtained by a viscosity parameter, and the viscosity parameter is used to determine a target valve viscosity level, and a linear process characteristic model including a first parameter and a second parameter is established based on the PV historical data and the true valve opening value, and an expression of the first parameter and the second parameter is obtained based on the linear process characteristic model. Parameter expression; wherein, the input data of the linear process characteristic model is the actual valve opening value, the output data of the linear process characteristic model is the PV historical data, and simulation is performed using the expression of the first parameter and the expression of the second parameter to obtain the simulated PV estimated value, and an expression of the mean square error of the PV historical data and the PV estimated value is established, and an objective function with the minimum mean square error as the optimization target is established, and the objective function is solved to obtain the optimal viscosity parameter, and according to the optimal viscosity parameter, the corresponding target valve viscosity level is determined. Compared with the existing technology, it not only reduces the amount of calculation, but also can improve the accuracy of valve viscosity detection.

[0023] In order to better understand the above technical solution, exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided to enable a clearer and more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

[0024] In order to facilitate understanding of the embodiments of the present application, some terms involved in the embodiments of the present application are explained as follows: "PV": It is the English abbreviation of Process Variable, and its Chinese name is measurement value; "MV": It is the abbreviation of Manipulated Variable. Its Chinese name is the output value after the controller is calculated in the automatic state. "Quality Code": Its English name is Tag Quality, and it indicates whether the bit number is of good value.

[0025] See also Figure 1 , Figure 1 FIG. 2 is a flow chart of a valve sticking detection method provided by an embodiment of the present application. Figure 1As shown, the valve sticking detection method can be performed by a valve sticking detection device, and the specific device of the valve sticking detection device can be set according to actual needs. For example, the valve sticking detection device can be a controller, etc. Specifically, the valve sticking detection method includes: Step S110, obtaining the target historical data of the preprocessed control loop in the past preset time period, wherein the target historical data includes PV historical data and MV historical data.

[0026] It should be understood that the specific process of preprocessing can be set according to actual needs, and the embodiments of the present application are not limited to this.

[0027] Optionally, historical data of the control loop in the past preset time period is obtained, and then the PV historical data and MV historical data of the loop are quality checked. At the same time, each value in the historical data has a quality code, and when the quality code is a good value, it is considered to be valid data, and the longest section of data with both the PV historical data and the MV historical data having a good quality code is extracted for processing. Among them, the specific time period of the past preset time period can be set according to actual needs, and the embodiment of the present application is not limited to this. For example, the past preset time period can be 24 hours, etc.

[0028] Also, after the extraction is completed, the data segment must meet the following conditions: ensure that there is a certain fluctuation in the data segment (non-linear), and the PV value and MV value are not always constant; the amount of data in this segment is not less than 200 data points.

[0029] It should be noted here that if the above conditions are not met, the valve sticking detection will not be performed on the circuit, that is, the following steps will not be performed.

[0030] In addition, when the amount of data is too large (for example, the amount of data is 200 to 10,000, etc.), appropriate resampling can be performed to reduce the amount of calculation and further improve the calculation efficiency while meeting the data volume.

[0031] In addition, the PV historical data and MV historical data can be normalized separately to reduce the problems caused by the difference in numerical values. At the same time, the normalization formula is as follows: ; In the formula, i Indicates i Values; x i Indicates i Normalized data (for example, normalized PV historical data or normalized MV historical data); X i Indicates i Historical data (for example, iPV historical data or i MV historical data); represents the mean of historical data. For example, x i Indicates i PV historical data, Represents the mean of PV historical data; for example, x i Indicates i MV historical data, Represents the mean of MV historical data; Represents the standard deviation of historical data. For example, x i Indicates i PV historical data, Indicates the standard deviation of PV historical data; for example, x i Indicates i MV historical data, Indicates the standard deviation of MV historical data.

[0032] Step S120, using the valve viscosity model to process the MV historical data to obtain the valve actual opening value, wherein the valve actual opening value is obtained from the viscosity parameter, and the viscosity parameter is used to determine the target valve viscosity level.

[0033] Specifically, valve sticking means that the output of the valve cannot change smoothly with the change of the signal from the controller. The valve output may remain unchanged or suddenly jump when the controller signal changes. Due to the existence of sticking, there is a difference in the output of the valve between the round trips of the valve. When the direction of valve movement changes, due to the existence of dead zone and static friction, the valve output will remain unchanged. The range of this unchanged is S, that is, S represents the amount of change of the controller signal, and within this range of change, the output of the valve will remain unchanged; when the controller signal is large enough and the valve output suddenly jumps, the size of the jump is J, that is, J represents the valve jump range when the valve output suddenly jumps. Also, considering the normalized static friction and the normalized dynamic friction The parameters have the following relationship: S = + ; J= - ; In the formula, S、 , and JAll can be expressed as a percentage of the valve range. And, this application uses process simulation to calculate the static friction of the valve and dynamic friction size, thereby confirming whether there is viscosity and quantifying the viscosity index.

[0034] And, if Figure 2 As shown in Figure 2, the processing process of the valve viscosity model is as follows: Get Among them, the Represents the MV historical data at time t; according to ,calculate , and the The calculation expression is as follows: ; In the formula, cum_u Indicates the external force currently acting on the valve that is balanced with the friction force, which comes from the signal controlling it; Indicates the residual force acting on the valve that does not cause the valve to move. When the simulation starts, it is assumed that there is no residual force on the valve, and no external force is added, that is, the initial is 0; Indicates the last moment t- 1 is the actual valve opening value; judge Is it greater than ;in, abs Indicates absolute value; If confirmed Greater than , then calculated by the following formula and , specifically: ; In the formula, Indicates the current time t The actual valve opening value; Indicates the static friction parameter of the valve; Indicates the dynamic friction parameter of the valve; If confirmed Less than or equal to , then calculated by the following formula and , specifically: .

[0035] That is, according to the valve sticking model and the MV history data of the control loop (i.e. Figure 2 In ), the actual valve opening value can be obtained (i.e. the output of the actual valve position), specifically: ; In the formula, Indicates the current time t The actual valve opening value; Represents the MV historical data at time t; cum_ u Indicates the external force currently acting on the valve that is balanced with the friction force; represents the static friction parameter; represents the dynamic friction parameter; abs Indicates absolute value; Indicates the last moment t- 1 is the actual valve opening value.

[0036] Step S130, based on the PV historical data and the actual valve opening value, a linear process characteristic model including a first parameter and a second parameter is established, and based on the linear process characteristic model, an expression of the first parameter and an expression of the second parameter are obtained. The input data of the linear process characteristic model is the actual valve opening value, and the output data of the linear process characteristic model is the PV historical data.

[0037] Specifically, the PV historical data (i.e., the actual measured value PV of the process) can be used as the output y of the linear process characteristic model, and the actual valve opening value of the valve after the viscous model with parameters specified by the valve viscous algorithm can be obtained. u As the input signal of the linear system. And, for the convenience of calculation, the linear process characteristic model can adopt the first-order ARX model, assuming that the output of the process y ,enter u exist t The moments have the following relationship (ignoring the influence of time lag), specifically: ; Similarly, at other times: ; ; ... ; And, the system of equations can be written in matrix form: ; In the formula, to Respectively represent the PV history data from the 1st to the N PV historical data; to Respectively represent the first moment to the N- The actual valve opening value at time 1; a Represents the first parameter;b Represents the second parameter; to They represent the errors from the second to the first due to noise. N The error caused by noise is to is the error caused by noise.

[0038] make , , and , the equation is simplified as: .

[0039] In addition, since the actual opening value of the valve yes and The first parameter can be obtained by the least squares method. a The expression and the second parameter b The expression of the first parameter a The expression and the second parameter b The expressions are all the same as and Related functions.

[0040] Step S140, using the expression of the first parameter and the expression of the second parameter to perform simulation, obtain the simulated PV estimation value, establish an expression for the mean square error between the PV historical data and the PV estimation value, and establish an objective function with the minimum mean square error as the optimization target.

[0041] Specifically, after obtaining the expression of the first parameter and the expression of the second parameter, the obtained parameters are used for simulation. At a certain time t, the estimated output of the simulation is: ; In the formula, Representation and The corresponding estimated value.

[0042] And the calculation error of the simulation is: ; And the overall error equation is: ; And, the mean square error of the calculation MSE for: ; In the formula, MSE Represents the mean square error between PV historical data and PV estimated value; n yes Length; express The corresponding estimated value; express The corresponding estimated value; express The corresponding estimated value.

[0043] In addition, the mean square error MSE The minimum setting is the optimization target, and When known, it can be calculated by least squares , then the objective function can be written as: .

[0044] Step S150, solving the objective function to obtain the optimal viscosity parameter.

[0045] Specifically, a grid search algorithm is used to explore a rough region in the search space of the viscosity parameter; and a Bayesian optimization algorithm is used to refine the search in the rough region to obtain the optimal viscosity parameter.

[0046] For example, you can set the search space as follows: First of all, the valve cannot be completely fixed, otherwise there will be dynamic friction of the valve. With static friction The sum is less than the range between the maximum and minimum values ​​of MV S max = MV max - MV min , + ≤ S max .in, and Both are greater than or equal to 0.

[0047] Secondly, static friction is not less than dynamic friction, that is, ≤ .

[0048] And, according to the above conditions, we can construct Figure 3 The search space of the viscosity parameter is shown in FIG. , and the rough area of ​​the search space can be as follows: Figure 3 The plane triangle range shown. And, in and In the search space, according to the given accuracy, generally 1%, the search space is divided into a number of grid points.

[0049] Then, the selection step is performed: after setting the search space, 4 sets of data in the parameter space can be randomly selected for parameter calculation and saved. A Gaussian process model is trained based on these data as initial training data, and the Gaussian process model can use a square exponential kernel, and the hyperparameters of the square exponential kernel are estimated by maximizing the marginal likelihood function.

[0050] Then, an evaluation step is performed: a Gaussian process model and an acquisition function can be used to select a new parameter combination for evaluation, and the next sampling point is added to the training data. Among them, the acquisition function can select the expected improvement EI function.

[0051] Finally, the iterative process is performed: the selection step and the evaluation step are repeated until the maximum number of iterations is reached, the optimization is stopped, and the optimal viscosity parameters are finally obtained. Among them, the viscosity parameters include the static friction parameters and dynamic friction parameters .

[0052] Step S160: determining the corresponding target valve viscosity level according to the optimal viscosity parameter.

[0053] Optionally, according to the optimal viscosity parameter, the valve viscosity quantitative index is calculated; the target index interval of the valve viscosity quantitative index is determined, and the valve viscosity level corresponding to the target index interval is used as the target valve viscosity level. The specific intervals of each index interval can be set according to actual needs, and the embodiments of the present application are not limited thereto.

[0054] For example, when the valve is determined to have a serious sticking problem, the sticking index should be larger, otherwise the sticking index should be smaller. And, the formula for quantifying the sticking index of the valve is as follows: ; .

[0055] In the formula, Indicates the quantitative index of valve viscosity; represents the viscosity sensitivity index; f Indicates the total valve friction.

[0056] In addition, if the valve viscosity quantification index is greater than or equal to 0 and less than or equal to 0.3, the valve viscosity level can be judged as none; if the valve viscosity quantification index is greater than 0.3 and less than or equal to 0.5, the valve viscosity level can be judged as weak; if the valve viscosity quantification index is greater than 0.5 and less than or equal to 0.7, the valve viscosity level can be judged as medium; if the valve viscosity quantification index is greater than 0.7 and less than or equal to 1, the valve viscosity level can be judged as strong.

[0057] In summary, with the help of the technical solution of the present application, the present application obtains the target historical data of the preprocessed control loop in the past preset time period, wherein the target historical data includes PV historical data and MV historical data, and uses the valve viscosity model to process the MV historical data to obtain the valve actual opening value, wherein the valve actual opening value is obtained by the viscosity parameter, and the viscosity parameter is used to determine the target valve viscosity level, and a linear process characteristic model including a first parameter and a second parameter is established based on the PV historical data and the valve actual opening value, and based on the linear process characteristic model, an expression of the first parameter and an expression of the second parameter are obtained. Expression; wherein, the input data of the linear process characteristic model is the actual valve opening value, the output data of the linear process characteristic model is the PV historical data, and the expression of the first parameter and the expression of the second parameter are used for simulation to obtain the simulated PV estimation value, and an expression of the mean square error of the PV historical data and the PV estimation value is established, and an objective function with the minimum mean square error as the optimization target is established, and the objective function is solved to obtain the optimal viscosity parameter, and according to the optimal viscosity parameter, the corresponding target valve viscosity level is determined. Compared with the prior art, it not only reduces the amount of calculation, but also can improve the accuracy of valve viscosity detection.

[0058] It should be understood that the above-mentioned valve viscosity detection method is only exemplary, and those skilled in the art can make various modifications based on the above-mentioned method, and the modified scheme also belongs to the protection scope of this application.

[0059] See also Figure 4 , Figure 4 The structure block diagram of a valve sticking detection device 400 provided in an embodiment of the present application is shown. It should be understood that the valve sticking detection device 400 can execute each step in the above method embodiment. The specific functions of the valve sticking detection device 400 can be found in the description above. To avoid repetition, the detailed description is appropriately omitted here. The valve sticking detection device 400 includes at least one software function module that can be stored in a memory in the form of software or firmware or solidified in the operating system (OS) of the valve sticking detection device 400. Specifically, the valve sticking detection device 400 includes: The acquisition module 410 is used to acquire the target historical data of the preprocessed control loop in the past preset time period; wherein the target historical data includes PV historical data and MV historical data; The processing module 420 is used to process the MV historical data using a valve viscosity model to obtain a valve real opening value; wherein the valve real opening value is obtained by a viscosity parameter, and the viscosity parameter is used to determine a target valve viscosity level; Establishing module 430, for establishing a linear process characteristic model including a first parameter and a second parameter based on the PV historical data and the actual valve opening value, and obtaining an expression of the first parameter and an expression of the second parameter based on the linear process characteristic model; wherein the input data of the linear process characteristic model is the actual valve opening value, and the output data of the linear process characteristic model is the PV historical data; A simulation module 440 is used to perform simulation using the expression of the first parameter and the expression of the second parameter to obtain a simulated PV estimate, establish an expression of a mean square error between the PV historical data and the PV estimate, and establish an objective function with the minimum mean square error as an optimization target; A solution module 450 is used to solve the objective function to obtain the optimal viscosity parameter; The determination module 460 is used to determine the corresponding target valve viscosity level according to the optimal viscosity parameter.

[0060] In a possible embodiment, the viscosity parameter includes a static friction parameter and a dynamic friction parameter; the calculation expression of the actual valve opening value is as follows: ; In the formula, Indicates the current time t The actual valve opening value; Represents the MV historical data at time t; cum_u Indicates the external force currently acting on the valve that is balanced with the friction force; represents the static friction parameter; represents the dynamic friction parameter; abs Indicates absolute value; Indicates the last moment t- 1 is the actual valve opening value.

[0061] In one possible embodiment, the linear process characteristic model is expressed as follows: ; In the formula, to Respectively represent the PV history data from the 1st to the N PV historical data; to Respectively represent the first moment to the N- The actual valve opening value at time 1; a Represents the first parameter; b Represents the second parameter; to They represent the errors from the second to the first due to noise. N An error due to noise.

[0062] In one possible embodiment, the expression of mean square error is as follows: ; ; In the formula, MSE represents mean square error; Representation and The corresponding estimated value; Representation and The corresponding estimated value; Representation and The corresponding estimated value.

[0063] Since the device described in the above embodiment of the present invention is a device used to implement the method of the above embodiment of the present invention, based on the method described in the above embodiment of the present invention, a person skilled in the art can understand the specific structure and deformation of the device, so it is not described here. All devices used in the method of the above embodiment of the present invention belong to the scope of protection of the present invention.

[0064] It should be understood by those skilled in the art that embodiments of the present invention may be provided as methods, systems or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0065] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present invention. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions.

[0066] It should be noted that the word "a" or "an" preceding a component does not exclude the presence of a plurality of such components. The invention may be implemented by means of hardware comprising a number of different components and by means of a suitably programmed computer. Among the several means listed, several of these means may be embodied by the same hardware. The use of the words first, second, third, etc., is for convenience of expression only and does not indicate any order. These words may be understood as part of the name of the component.

[0067] In addition, it should be noted that, in the description of this specification, the description of the terms "one embodiment", "some embodiments", "embodiment", "example", "specific example" or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, unless they are contradictory.

[0068] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments after knowing the basic creative concept. Therefore, the technical solution should be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present invention.

[0069] Obviously, those skilled in the art can make various modifications and variations to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention belong to the scope of the technical solution of the present invention and its equivalent technology, the present invention should also include these modifications and variations.

Claims

1. A valve sticking detection method, characterized in that: include: Obtaining target historical data of the preprocessed control loop in the past preset time period; wherein the target historical data includes PV historical data and MV historical data; The MV historical data is processed using a valve viscosity model to obtain a true valve opening value; wherein the true valve opening value is obtained by a viscosity parameter, and the viscosity parameter is used to determine a target valve viscosity level; A linear process characteristic model including a first parameter and a second parameter is established based on the PV historical data and the actual valve opening value, and an expression of the first parameter and an expression of the second parameter are obtained based on the linear process characteristic model; wherein the input data of the linear process characteristic model is the actual valve opening value, and the output data of the linear process characteristic model is the PV historical data; Perform simulation using the expression of the first parameter and the expression of the second parameter to obtain a simulated PV estimated value, establish an expression of a mean square error between the PV historical data and the PV estimated value, and establish an objective function with the minimum mean square error as an optimization target; Solving the objective function to obtain an optimal viscosity parameter; According to the optimal viscosity parameter, the corresponding target valve viscosity level is determined.

2. The valve sticking detection method according to claim 1, characterized in that: The viscosity parameters include static friction parameters and dynamic friction parameters; the calculation expression of the actual valve opening value is as follows: ； In the formula, Indicates the current time t The actual valve opening value; Represents the MV historical data at time t; cum_u Indicates the external force currently acting on the valve that is balanced with the friction force; represents the static friction parameter; represents the dynamic friction parameter; abs Indicates absolute value; Indicates the last moment t- 1 is the actual valve opening value.

3. The valve sticking detection method according to claim 2, characterized in that: The expression of the linear process characteristic model is as follows: ； In the formula, to Respectively represent the PV history data from the 1st to the N PV historical data; to Respectively represent the first moment to the N- The actual valve opening value at time 1; a represents the first parameter; b represents the second parameter; to They represent the errors from the second to the first due to noise. N An error due to noise.

4. The valve sticking detection method according to claim 3, characterized in that: The expression of the mean square error is as follows: ； ； In the formula, MSE represents the mean square error; Representation and The corresponding estimated value; Indicates and The corresponding estimated value; Indicates and The corresponding estimated value.

5. The valve sticking detection method according to claim 1, characterized in that: Solving the objective function to obtain the optimal viscosity parameter includes: exploring a rough region within the search space of the viscosity parameter using a grid search algorithm; A Bayesian optimization algorithm is used to refine the search in the rough area to obtain the optimal viscosity parameter.

6. The valve sticking detection method according to claim 1, characterized in that: Determining the corresponding target valve viscosity level according to the optimal viscosity parameter includes: Calculating a valve viscosity quantitative index according to the optimal viscosity parameter; A target index interval in which the valve viscosity quantitative index is located is determined, and a valve viscosity level corresponding to the target index interval is used as the target valve viscosity level.

7. A valve sticking detection device, characterized in that: include: An acquisition module, used for acquiring target historical data of the preprocessed control loop in the past preset time period; wherein the target historical data includes PV historical data and MV historical data; A processing module, used for processing the MV historical data by using a valve viscosity model to obtain a valve real opening value; wherein the valve real opening value is obtained by a viscosity parameter, and the viscosity parameter is used to determine a target valve viscosity level; An establishing module is used to establish a linear process characteristic model including a first parameter and a second parameter based on the PV historical data and the actual valve opening value, and obtain an expression of the first parameter and an expression of the second parameter based on the linear process characteristic model; wherein the input data of the linear process characteristic model is the actual valve opening value, and the output data of the linear process characteristic model is the PV historical data; A simulation module, used for performing simulation using the expression of the first parameter and the expression of the second parameter to obtain a simulated PV estimated value, and establishing an expression of a mean square error between the PV historical data and the PV estimated value, and establishing an objective function with the minimum mean square error as an optimization target; A solution module, used for solving the objective function to obtain an optimal viscosity parameter; The determination module is used to determine the corresponding target valve viscosity level according to the optimal viscosity parameter.

8. The valve sticking detection device according to claim 7, characterized in that: The viscosity parameters include static friction parameters and dynamic friction parameters; the calculation expression of the actual valve opening value is as follows: ； In the formula, Indicates the current time t The actual valve opening value; Represents the MV historical data at time t; cum_u Indicates the external force currently acting on the valve that is balanced with the friction force; represents the static friction parameter; represents the dynamic friction parameter; abs Indicates absolute value; Indicates the last moment t- 1 is the actual valve opening value.

9. The valve sticking detection device according to claim 8, characterized in that: The expression of the linear process characteristic model is as follows: ； In the formula, to Respectively represent the PV history data from the 1st to the N PV historical data; to Respectively represent the first moment to the N- The actual valve opening value at time 1; a represents the first parameter; b represents the second parameter; to They represent the errors from the second to the first due to noise. N An error due to noise.

10. The valve sticking detection device according to claim 9, characterized in that: The expression of the mean square error is as follows: ； ； In the formula, MSE represents the mean square error; Representation and The corresponding estimated value; Indicates and The corresponding estimated value; Indicates and The corresponding estimated value.

Images