Correlation integration method and apparatus for cascade system

Abstract

The present disclosure belongs to the technical field of control systems. Provided are a correlation integration method and apparatus for a cascade system. The correlation integration method for a cascade system is used for obtaining a gradient between an optimization variable and an objective function of a cascade system, wherein the cascade system comprises a plurality of cascade modules sequentially connected in series. The method comprises: concurrently acquiring a static gain matrix of each cascade module; and on the basis of the static gain matrices of all the cascade modules, acquiring a gradient between an optimization variable and an objective function of a cascade system. In the method, a static gain matrix of each cascade module is calculated concurrently, thereby shortening the calculation time of a gradient between an optimization variable and an objective function, and effectively improving the optimization efficiency of a cascade system.

Description

Cascade system correlation integration method and device Technical Field

[0001] The present disclosure relates to the technical field of control systems, and in particular to a method and device for integrating related cascade systems. Background Art

[0002] In process industries, materials are often processed using a cascade system consisting of multiple devices connected in series. This cascade system often takes more than ten hours or even several days to complete material processing.

[0003] To achieve optimal material processing results, it's necessary to optimize the processing parameters of the cascade system. These parameters can be considered optimization variables, and the optimized processing effect is called the optimization objective function. For cascade processing systems, the time constant from when the optimization variables are adjusted to when the objective function responds is very long. The data required for optimization typically takes 4 to 12 times this time constant, resulting in a significant time delay to obtain the necessary data. Therefore, if the cascade system to be optimized is too large and the time lag is too long, the optimization process will be time-consuming, inefficient, or even unacceptable.

[0004] Summary of the Invention

[0005] The present disclosure provides a cascade system correlation integration method and device.

[0006] According to a first aspect of the present disclosure, a cascade system correlation integration method is provided for obtaining a gradient between an optimization variable and an objective function of the cascade system, wherein the cascade system comprises a plurality of cascade modules connected in series, including:

[0007] Concurrently acquiring a static gain matrix of each of the cascaded modules;

[0008] Based on the static gain matrices of all the cascade modules, obtaining the gradient between the optimization variables and the objective function of the cascade system;

[0009] The static gain matrix of the cascade module is obtained by the following steps:

[0010] Obtaining input measurement values ​​and output measurement values ​​of the cascade module;

[0011] Obtaining an input autocorrelation integral matrix of the cascade module based on the input measurement value of the cascade module;

[0012] Based on the input measurement value and the output measurement value of the cascade module, obtaining a cross-correlation integral vector of the input and output of the cascade module;

[0013] The static gain matrix of the cascade module is obtained based on the input autocorrelation integral matrix of the cascade module and the cross-correlation integral vector of the input and output of the cascade module.

[0014] According to a second aspect of the present disclosure, a cascade system correlation integration device is provided for obtaining a gradient between an optimization variable and an objective function of the cascade system, wherein the cascade system comprises a plurality of cascade modules connected in series, including:

[0015] A first acquisition module, configured to concurrently acquire a static gain matrix of each of the cascaded modules;

[0016] A second acquisition module is used to obtain the gradient between the optimization variable and the objective function of the cascade system based on the static gain matrix of all the cascade modules;

[0017] The first acquisition module includes:

[0018] A first acquisition submodule, configured to acquire input measurement values ​​and output measurement values ​​of the cascade module;

[0019] A first obtaining submodule, configured to obtain an input autocorrelation integral matrix of the cascade module based on an input measurement value of the cascade module;

[0020] A second obtaining submodule, configured to obtain a cross-correlation integral vector of an input and an output of the cascade module based on an input measurement value and an output measurement value of the cascade module;

[0021] The second acquisition submodule is configured to acquire a static gain matrix of the cascade module based on an input autocorrelation integral matrix of the cascade module and a cross-correlation integral vector between an input and an output of the cascade module.

[0022] According to the cascade system correlation integration method disclosed in the present invention, the input measurement value and output measurement value of each cascade module in the cascade system are obtained respectively, the input autocorrelation integral matrix of the cascade module is obtained based on the input measurement value of the cascade module, and the cross-correlation integral vector of the input and output of the cascade module is obtained based on the input measurement value and the output measurement value of the cascade module. Based on the input autocorrelation integral matrix of the cascade module and the cross-correlation integral vector of the input and output of the cascade module, the static gain matrix of each cascade module can be obtained concurrently, that is, the acquisition of the static gain matrix of each cascade module can be carried out concurrently, which is much shorter than the time for obtaining the static gain matrix of the entire cascade system in a non-concurrent manner. Based on the static gain matrices of all cascade modules, the gradient between the optimization variable and the objective function of the cascade system can be quickly obtained, which reduces the time spent on obtaining the gradient between the optimization variable and the objective function of the cascade system, thereby effectively improving the optimization efficiency of the cascade system.

[0023] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The accompanying drawings are used to provide a further understanding of the present disclosure and constitute a part of the specification. Together with the embodiments of the present disclosure, they are used to explain the present disclosure and do not constitute a limitation of the present disclosure. By describing the detailed exemplary embodiments with reference to the accompanying drawings, the above and other features and advantages will become more apparent to those skilled in the art.

[0025] FIG1 is a flow chart of a cascade system correlation integration method provided by an embodiment of the present disclosure;

[0026] FIG2 is a flow chart of obtaining a static gain matrix of a cascade module according to an embodiment of the present disclosure;

[0027] FIG3 is a schematic diagram of a cascade system provided by an embodiment of the present disclosure;

[0028] FIG4 is a flow chart of obtaining an input autocorrelation integral matrix of a cascade module according to an embodiment of the present disclosure;

[0029] FIG5 is a flowchart of obtaining a cross-correlation integral vector of an input and output of a cascade module provided by an embodiment of the present disclosure;

[0030] FIG6 is a flow chart of obtaining a static gain matrix of a cascade module according to an embodiment of the present disclosure;

[0031] FIG7 is a block diagram of a cascade system-related integration device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0032] To enable those skilled in the art to better understand the technical solutions of the present disclosure, exemplary embodiments of the present disclosure are described below in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding. These details should be considered merely exemplary. Therefore, those skilled in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0033] In the absence of conflict, the various embodiments of the present disclosure and the various features therein may be combined with each other.

[0034] As used herein, the term "and / or" includes any and all combinations of one or more of the associated listed items.

[0035] The terms used herein are only used to describe specific embodiments and are not intended to limit the present disclosure. As used herein, the singular forms "a" and "the" are also intended to include the plural forms, unless the context clearly indicates otherwise. It will also be understood that when the terms "comprising" and / or "made of" are used in this specification, features, wholes, steps, operations, elements and / or components are specified to exist, but the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or groups thereof are not excluded. Similar words such as "connected" or "connected" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect.

[0036] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art. It will also be understood that terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant art and the present disclosure, and will not be interpreted as having an idealized or overly formal meaning unless expressly defined as such herein.

[0037] In a first aspect, an embodiment of the present disclosure provides a cascade system correlation integration method for obtaining the gradient between the optimization variable and the objective function of the cascade system, wherein the cascade system includes multiple cascade modules connected in series. FIG1 is a flow chart of a cascade system correlation integration method provided by an embodiment of the present disclosure. As shown in FIG1 , the cascade system correlation integration method provided by an embodiment of the present disclosure includes:

[0038] Step S11: concurrently obtain the static gain matrix of each cascade module.

[0039] In some embodiments, the cascade system includes a plurality of cascade modules, each cascade module having a corresponding static gain matrix. The static gain matrix of the cascade module represents the rate of change of the output measurement value of the cascade module relative to the input measurement value. Since the static gain matrix of each cascade module can be obtained concurrently, and the time consumed by obtaining the static gain matrix of each cascade module is shorter than the time consumed by obtaining the static gain matrix of the entire cascade system, the speed of concurrently obtaining the static gain matrix of each cascade module is faster than obtaining the static gain matrix of the entire cascade system. Since the gradient between the optimization variable and the objective function of the entire cascade system can be represented by the static gain matrix of the entire cascade system, after concurrently obtaining the static gain matrix of each cascade module, the gradient between the optimization variable and the objective function of the entire cascade system can be quickly obtained based on the static gain matrix of each cascade module.

[0040] FIG2 is a flow chart of a method for obtaining a static gain matrix of a cascaded module provided by an embodiment of the present disclosure. As shown in FIG2 , the method for obtaining a static gain matrix of a cascaded module includes steps S21-S24:

[0041] Step S21: Obtain input measurement values ​​and output measurement values ​​of the cascade module.

[0042] In some embodiments, a cascade system is composed of multiple cascade modules connected in series. Each cascade module processes input data and transmits the generated output data to the next cascade module. Except for the first cascade module, the input data of each cascade module is the output data of the previous cascade module. Each cascade module includes a static gain matrix, which represents the gain of the output data of the cascade module relative to the input data. The input data of the first cascade module is the input data of the cascade system, and the output data of the last cascade module is the output data of the cascade system.

[0043] FIG3 is a schematic diagram of a cascade system provided by an embodiment of the present disclosure. As shown in FIG3, the cascade system provided by an embodiment of the present disclosure includes q cascade modules connected in series. Except for the first cascade module, the input data of each cascade module is the output data of the previous cascade module. 1 S(t) is the input data of the first cascade module, 2 S(t) is the output data of the first cascade module, 2 S(t) is also the input data of the second cascade module. 3 S(t) is the output data of the second cascade module, 3 S(t) is also the input data of the third cascade module. The output data of each cascade module is the input data of the next cascade module. q S(t) is the output data of the q-1th cascade module, q S(t) is also the input data of the qth cascade module. q+1 S(t) is the output data of the qth cascade module. Each cascade module contains a static gain matrix. 1 G is the static gain matrix of the first cascade module, 2 G is the static gain matrix of the second cascade module, and so on. q G is the static gain matrix of the qth cascade module.

[0044] In some embodiments, the input and output of the cascade modules can be measured by common measuring equipment. The input measurement value and output measurement value of the cascade modules can have different dimensions. l S(t) and output measurement value l+1S(t) can be obtained by measuring the input and output measurements of different dimensions of the cascaded modules. For example, the input measurement value of the lth cascaded module is obtained l S(t) and output measurement value l+1 Methods for S(t) include:

[0045] Wherein, l = 1, 2, ..., q; l represents the lth cascade module; T represents transpose; S represents the measurement value of the cascade module; l S represents the measurement value of the lth cascade module; l+1 S(t) represents the measurement value output by the lth cascade module at time t; m represents the measurement value of the cascade module with m dimensions; m l+1 Indicates that the output measurement value of the lth cascade module is m l+1 dimensions; Indicates the mth module in the lth cascade module l+1 The output measurement of the dimension.

[0046] Step S22: Obtain an input autocorrelation integral matrix of the cascade module based on the input measurement values ​​of the cascade module.

[0047] In some embodiments, the input autocorrelation integral matrix of the cascade module can be obtained based on the input measurement values ​​of the cascade module using the correlation integral transformation, wherein the input autocorrelation integral matrix of the cascade module can represent the potential relationship between the input measurement values ​​of each dimension of the cascade module.

[0048] FIG4 is a flow chart of obtaining an input autocorrelation integral matrix of a cascade module according to an embodiment of the present disclosure. As shown in FIG4 , obtaining an input autocorrelation integral matrix of a cascade module based on input measurement values ​​of the cascade module includes:

[0049] Step S31: Obtain an input-related integral value of the cascade module based on the input measurement value of the cascade module.

[0050] In some embodiments, the input measurement value of the cascade module may include multiple groups of data, and the input correlation integral value of the cascade module represents different potential relationships between the multiple groups of data in the input measurement value of the cascade module. Wherein, the input correlation integral value includes an input autocorrelation integral value and an input cross-correlation integral value. The input autocorrelation integral value is the correlation integral value of the data in the same dimension input measurement value of the cascade module and the data, the input autocorrelation integral value represents the potential relationship between the data in the same dimension input measurement value of the cascade module and itself, and the input cross-correlation integral value is the correlation integral value of the data in the different dimensions input measurement values ​​of the cascade module, representing the potential relationship in the different dimensions input measurement values ​​of the cascade module.

[0051] Step S32: obtaining an input autocorrelation integral matrix of the cascade modules based on the input correlation integral values ​​of the cascade modules.

[0052] In some embodiments, by obtaining different potential relationships between multiple groups of data in the input measurement values ​​of the cascade module, the potential relationships between the input measurement values ​​of the cascade module can be obtained, that is, the input autocorrelation integral matrix of the cascade module. The input autocorrelation integral matrix of the cascade module can be obtained based on the input autocorrelation integral value and the input cross-correlation integral.

[0053] In some embodiments, obtaining an input autocorrelation integral matrix of the cascaded modules based on the input correlation integral values ​​of the cascaded modules includes:

[0054] Based on the input measurement value of the cascade module, the input autocorrelation integral matrix of the cascade module is obtained using formula (1) and formula (2);

[0055] Where i = 1, 2, ..., m l ; j=1,2,...,m l ; l=1,2,...,q; q means that the cascade system contains q cascade modules in total; l means the lth cascade module in the cascade system; m means that the measurement value of the cascade module has m dimensions in total; m l Indicates that the input measurement value of the lth cascade module has a total of m l dimensions;

[0056] l T i is the first integral constant of the i-th module; l M i is the second integral constant of the i-th module; λ is the first integral variable; τ is the second integral variable.

[0057] Represents the autocorrelation integral matrix; S represents the measurement value of the cascade module; S i Represents the measurement value of the i-th dimension in the cascade module; l S represents the input vector of the lth cascade module; l S i represents the input measurement value of the i-th dimension in the l-th cascade module; S j represents the measurement value of the jth dimension in the cascade module; l S j represents the input measurement value of the jth dimension in the lth cascade module; Represents the correlation integral value between the input of the i-th dimension and the input of the j-th dimension in the l-th cascade module; represents the input autocorrelation integral matrix of the lth cascade module; l s i (λ) represents the input measurement value of the l-th cascade module with respect to the i-th dimension of the variable λ; l s j (λ-τ) is the input measurement value of the l-th cascade module with respect to the j-th dimension of the variable λ-τ.

[0058] In some embodiments, the input measurement value of the cascade module may include data of multiple dimensions. Assume that the input measurement value of the cascade module includes m l The data of different dimensions can be obtained by using formula (1) and formula (2) respectively to obtain the m between the different dimension data in the input measurement value of the cascade module. l ×m l Then, based on the potential relationship between the different dimensional data in the input measurement values ​​of the cascade module, the potential relationship between the input measurement values ​​of the cascade module and the input measurement values ​​is generated, that is, the input autocorrelation integral matrix of the cascade module

[0059] Step S23: Based on the input measurement value and the output measurement value of the cascade module, obtain the cross-correlation integral vector of the input and output of the cascade module.

[0060] In some embodiments, a corresponding correlation integral calculation can be performed based on the input measurement values ​​and output measurement values ​​of the cascaded modules to obtain a cross-correlation integral vector of the input and output of the cascaded modules. The cross-correlation integral vector of the input and output of the cascaded modules represents the potential relationship between the input measurement values ​​and the output measurement values ​​of the cascaded modules.

[0061] FIG5 is a flow chart of obtaining a mutual correlation integral vector of the input and output of a cascade module provided by an embodiment of the present disclosure. As shown in FIG5 , obtaining a mutual correlation integral vector of the input and output of a cascade module includes steps S41-S42:

[0062] Step S41: Based on the input measurement value and the output measurement value of the cascade module, obtain the cross-correlation integral value of the input and output of the cascade module.

[0063] In some embodiments, the input measurement values ​​of the cascade module may include multiple sets of different data, and the output measurement values ​​may also include multiple sets of different data. The cross-correlation integral value of the input and output of the cascade module represents the potential relationship between a set of data in the input measurement values ​​of the cascade module and a set of data in the output measurement values ​​of the cascade module.

[0064] In some embodiments, based on multiple sets of different data in the input measurement values ​​of the cascade module and multiple sets of different data in the output measurement values, the potential relationship between the multiple sets of different data in the input measurement value group of the cascade module and the multiple sets of different data in the output measurement values, that is, the cross-correlation integral value of the input and output of the cascade module can be obtained.

[0065] Step S42: Obtain a cross-correlation integral vector between the input and output of the cascade module based on the cross-correlation integral value between the input and output of the cascade module.

[0066] In some embodiments, the cross-correlation integral vector of the input and output of the cascaded module represents the potential relationship between the input measurement value and the output measurement value of the cascaded module. The cross-correlation integral vector of the input and output of the cascaded module includes multiple sets of cross-correlation integral values ​​based on the input and output of the cascaded module. Therefore, based on the cross-correlation integral values ​​of the input and output of the cascaded module, the cross-correlation integral vector of the input and output of the cascaded module can be obtained.

[0067] In some embodiments, obtaining a cross-correlation integral vector of the input and output of the cascade module based on the cross-correlation integral value of the input and output of the cascade module includes:

[0068] Based on the input measurement value and output measurement value of the cascade module, the cross-correlation integral vector of the input and output of the cascade module is obtained using formula (3) and formula (4):

[0069] Where i = 1, 2, ..., m l ; j=1,2,...,m l+1 ; l=1,2,...,q; q means the cascade system contains q cascade modules; l means the lth cascade module in the cascade system; m means the measurement value of the cascade module has m dimensions; m l Indicates that the input vector of the lth cascade module has m l dimensions;

[0070] l T i is the first integral constant of the i-th module; l M i is the second integral constant of the i-th module; λ is the first integral variable; τ is the second integral variable;

[0071] represents the cross-correlation integral vector; S represents the measurement value of the cascade module; l S represents the measurement value of the lth cascade module; l S i Represents the input measurement value of the i-th dimension of the l-th cascade module;l+1 S j represents the output measurement value of the jth dimension in the lth cascade module; Indicates the mth l The input measurement of the dimension, Indicates the mth module in the lth cascade module l Input measurements of dimensions; represents the cross-correlation integral value between the input of the i-th dimension and the output of the j-th dimension in the l-th cascade module; Represents the cross-correlation integral vector between the input of the l-th cascade module and the output of the j-th dimension; l s i (λ) represents the input measurement value of the l-th cascade module with respect to the i-th dimension of the variable λ; l+1 s j (λ-τ) represents the output measurement value of the l-th cascade module with respect to the j-th dimension of the variable λ-τ.

[0072] In some embodiments, the input measurement value of the cascade module may include m l The output measurement value of the cascade module can contain m dimensions of data. l+1 dimensional data, based on the m input measurements of the cascade module l dimensional data and the output measurement value m of the cascade module l+1 Dimensional data, use formula (3) and formula (4) to obtain m l ×m l+1 The cross-correlation integral value of the input and output of the cascaded modules; based on m l ×m l+1 The mutual correlation integral values ​​of the input and output of the cascaded modules are obtained to obtain the mutual correlation integral vector of the input and output of the cascaded modules, wherein the mutual correlation integral vector of the input and output of the cascaded modules includes the mutual correlation integral values ​​of the input and output of multiple cascaded modules.

[0073] Step S24: Acquire a static gain matrix of the cascaded module based on the input autocorrelation integral matrix of the cascaded module and the cross-correlation integral vector of the input and output of the cascaded module.

[0074] In some embodiments, the static gain matrix of the cascade module represents the degree of change of the output measurement value of the cascade module relative to the input measurement value; the static gain matrix of the cascade module can be obtained by using the relevant matrix transformation based on the input autocorrelation integral matrix of the cascade module and the cross-correlation integral vector of the input and output of the cascade module.

[0075] FIG6 is a flow chart of obtaining a static gain matrix of a cascade module according to an embodiment of the present disclosure. As shown in FIG6 , obtaining a static gain matrix of a cascade module includes steps S51-S52:

[0076] Step S51 : obtaining a static gain column vector of the cascade module based on the input autocorrelation integral matrix of the cascade module and the cross-correlation integral vector of the input and output of the cascade module.

[0077] In some embodiments, the static gain column vector of the cascade module represents the static gain of the partial data of the output measurement value of the cascade module relative to the input measurement value; the static gain column vector of the cascade module can be obtained by using the relevant matrix transformation based on the input autocorrelation integral matrix of the cascade module and the cross-correlation integral vector of the input and output of the cascade module.

[0078] In some embodiments, obtaining a static gain column vector of the cascaded module based on an input autocorrelation integral matrix of the cascaded module and a cross-correlation integral vector of an input and an output of the cascaded module includes:

[0079] Cross-correlation integral vector of input and output based on cascaded modules and the input autocorrelation integral matrix of the cascade module Use formula (5) to obtain the static gain column vector of the cascade module

[0080] Where j = 1, 2, ..., m l+1 ;m l+1 Indicates that the output measurement value of the lth cascade module is m l+1 dimensions;

[0081] S represents the measurement value of the cascade module; l S represents the input value of the lth cascade module; S j represents the input measurement value of the jth dimension in the cascade module; l+1 S j represents the output measurement value of the jth dimension in the lth cascade module; Represents the cross-correlation integral vector between the input of the l-th cascade module and the output of the j-th dimension; represents the input autocorrelation integral matrix of the lth cascade module; Represents the static gain column vector of the lth cascade module.

[0082] In some embodiments, since the output measurement value of the cascade module may include m l+1 dimensional data, therefore, the cross-correlation integral vector based on the input and output of the cascade module is and the input autocorrelation integral matrix of the cascade module Formula (5) can be used to obtain the static gain column vectors of multiple dimensions in the cascade module, namely

[0083] Step S52: Obtain a static gain matrix of the cascaded modules based on the static gain column vector of the cascaded modules.

[0084] In some embodiments, the static gain matrix of the cascaded modules includes static gain column vectors of the plurality of cascaded modules.

[0085] In some embodiments, obtaining a static gain matrix of the cascaded modules based on the static gain column vector of the cascaded modules includes:

[0086] Static gain column vector based on cascaded modules Using formula (6), the static gain matrix of the cascaded module is obtained: l G

[0087] Wherein, l = 1, 2, ..., q; q means that the cascade system contains q cascade modules; l means the lth cascade module in the cascade system;

[0088] m l+1 Indicates that the output measurement value of the lth cascade module is m l+1 dimensions;

[0089] S represents the measurement value of the cascade module; Indicates the mth l+1 The measurement value of the dimension; Indicates the mth module in the lth cascade module l+1 Output measurement of dimension; Represents the static gain column vector of the lth cascade module; l G represents the static gain matrix of the lth cascade module.

[0090] In some embodiments, when obtaining the total m of the cascade module l+1 After the static gain column vector of dimensions is obtained, it can be based on m l+1 The static gain column vector of dimensions is used to obtain the complete static gain matrix of the cascaded module. l G.

[0091] In some embodiments, obtaining the gradient between the optimization variable and the objective function of the cascade system based on the static gain matrix of all cascade modules includes:

[0092] Static gain matrix based on all cascaded modules l G, using formula (7) to obtain the gradient between the optimization variable and the objective function of the cascade system;

[0093] Where K is the gradient between the optimization variable and the objective function of the cascade system; q means that the cascade system contains q cascade modules; l means the lth cascade module in the cascade system;l G represents the static gain matrix of the lth cascade module.

[0094] In some embodiments, since the static gain of the entire cascade system can be obtained by multiplying the static gain matrices of each cascade module, the static gain matrices of each cascade module can be obtained simultaneously through the concurrent calculation of the static gain matrices of each module. Afterwards, the static gain of the entire cascade system can be obtained using formula (7), that is, the gradient between the optimization variable of the cascade system and the objective function.

[0095] Step S12: Based on the static gain matrices of all cascaded modules, the gradient between the optimization variables and the objective function of the cascaded system is obtained.

[0096] In some embodiments, the optimization variables include at least one of controllable temperature, flow rate, pressure, and material composition. When the cascade system is a catalytic cracking unit, the controllable temperature includes at least one of the reaction temperature, the measured value of the reaction temperature controller, the feed temperature, and the measured value of the feed temperature controller; the flow rate includes at least one of the amount of fresh catalyst added, the pre-boost steam flow rate, and the measured value of the pre-boost steam flow rate; and the pressure includes the pressure in the catalytic cracking unit. The objective function is a function associated with the optimization variables.

[0097] The cascade system correlation integration method provided by the embodiment of the present disclosure obtains the input measurement value and output measurement value of each cascade module in the cascade system respectively, obtains the input autocorrelation integral matrix of the cascade module based on the input measurement value of the cascade module, obtains the cross-correlation integral vector of the input and output of the cascade module based on the input measurement value and the output measurement value of the cascade module, and based on the input autocorrelation integral matrix of the cascade module and the cross-correlation integral vector of the input and output of the cascade module, the static gain matrix of each cascade module can be obtained concurrently, that is, the static gain matrix of each cascade module can be obtained synchronously, and then based on the static gain matrices of all cascade modules, the gradient between the optimization variable and the objective function of the cascade system is obtained. Since the acquisition of the gain matrix of each cascade module can be carried out simultaneously, the time spent on obtaining the gradient between the optimization variable and the objective function of the cascade system is reduced, thereby effectively improving the optimization efficiency of the cascade system.

[0098] The steps of the various methods above are divided only for the purpose of clear description. During implementation, they can be combined into one step or some steps can be split and decomposed into multiple steps. As long as they include the same logical relationship, they are all within the scope of protection of this patent. Adding insignificant modifications or introducing insignificant designs to the algorithm or process without changing the core design of the algorithm and process are all within the scope of protection of this patent.

[0099] In a second aspect, embodiments of the present disclosure provide a cascade system-related integration device for obtaining the gradient between the optimization variable and the objective function of the cascade system. The cascade system includes multiple cascade modules connected in series. FIG7 is a block diagram of a cascade system-related integration device provided by an embodiment of the present disclosure. As shown in FIG7 , the cascade system-related integration device 500 includes:

[0100] The first acquisition module 501 is configured to concurrently acquire the static gain matrix of each cascaded module.

[0101] The second acquisition module 502 is configured to acquire the gradient between the optimization variable and the objective function of the cascade system based on the static gain matrices of all cascade modules.

[0102] The first acquisition module 501 includes:

[0103] The first acquisition submodule 51 is used to acquire input measurement values ​​and output measurement values ​​of the cascade module.

[0104] The first obtaining submodule 52 is configured to obtain an input autocorrelation integral matrix of the cascade module based on the input measurement values ​​of the cascade module.

[0105] In some embodiments, the first obtaining module is further configured to: obtain an input-related integral value of the cascade module based on an input measurement value of the cascade module;

[0106] An input autocorrelation integral matrix of the cascade modules is obtained based on the input correlation integral values ​​of the cascade modules.

[0107] In some embodiments, the first obtaining submodule 52 is further configured to obtain the input autocorrelation integral matrix of the cascade module using formula (1) and formula (2) based on the input measurement value of the cascade module.

[0108] Where i = 1, 2, ..., m l ; j=1,2,...,m l ; l=1,2,...,q; q means the cascade system contains q cascade modules; l means the lth cascade module in the cascade system; m means the measurement value of the cascade module has m dimensions; m l Indicates that the input measurement value of the lth cascade module has m l dimensions;

[0109] l T i is the first integral constant of the i-th module; l M i is the second integral constant of the i-th module; λ is the first integral variable; τ is the second integral variable;

[0110] Represents the autocorrelation integral matrix; S represents the measurement value of the cascade module; S i Represents the measurement value of the i-th dimension in the cascade module; l S represents the input vector of the lth cascade module; l S i represents the input measurement value of the i-th dimension in the l-th cascade module; S j represents the measurement value of the jth dimension in the cascade module; l S j represents the input measurement value of the jth dimension in the lth cascade module; Represents the correlation integral value of the input of the i-th dimension and the j-th dimension in the l-th cascade module; represents the input autocorrelation integral matrix of the lth cascade module; l s i (λ) represents the input measurement value of the l-th cascade module with respect to the i-th dimension of the variable λ; l s j (λ-τ) represents the input measurement value of the l-th cascade module with respect to the j-th dimension of the variable λ-τ.

[0111] The second obtaining submodule 53 is configured to obtain a cross-correlation integral vector between the input and output of the cascade module based on the input measurement value and the output measurement value of the cascade module.

[0112] In some embodiments, the second obtaining module is further configured to: obtain a cross-correlation integral value between the input and output of the cascade module based on the input measurement value and the output measurement value of the cascade module;

[0113] Based on the cross-correlation integral value of the input and output of the cascade module, a cross-correlation integral vector of the input and output of the cascade module is obtained.

[0114] In some embodiments, the second obtaining submodule 53 is further used to obtain the cross-correlation integral vector of the input and output of the cascade module using formula (3) and formula (4) based on the input measurement value and the output measurement value of the cascade module.

[0115] Where i = 1, 2, ..., m l ; j=1,2,...,m l+1 ; l=1,2,...,q; q means the cascade system contains q cascade modules; l means the lth cascade module in the cascade system; m means the measurement value of the cascade module has m dimensions; m l Indicates that the input vector of the lth cascade module has m l dimensions;

[0116] l T i is the first integral constant of the lth module; l M i is the second integral constant of the i-th module; λ is the first integral variable; τ is the second integral variable;

[0117] represents the cross-correlation integral vector; S represents the measurement value of the cascade module; l S represents the measurement value of the lth cascade module; l S i Represents the input measurement value of the i-th dimension of the l-th cascade module; l+1 S j represents the output measurement value of the jth dimension in the lth cascade module; Indicates the mth l The input measurement of the dimension, Indicates the mth module in the lth cascade module l Input measurements of dimensions; represents the cross-correlation integral value between the input of the i-th dimension and the output of the j-th dimension in the l-th cascade module; represents the cross-correlation integral vector of the input and output of the lth cascade module; l s i (λ) represents the input measurement value of the l-th cascade module with respect to the i-th dimension of the variable λ; l+1 s j (λ-τ) represents the output measurement value of the l-th cascade module with respect to the j-th dimension of the variable λ-τ.

[0118] The second acquisition submodule 54 is configured to acquire a static gain matrix of the cascade module based on the input autocorrelation integral matrix of the cascade module and the cross-correlation integral vector between the input and output of the cascade module.

[0119] In some embodiments, the second acquisition submodule 54 is further configured to: acquire a static gain column vector of the cascade module based on an input autocorrelation integral matrix of the cascade module and a cross-correlation integral vector of an input and output of the cascade module;

[0120] Obtain a static gain matrix of the cascaded modules based on the static gain column vector of the cascaded modules.

[0121] In some embodiments, the second acquisition submodule 54 is further configured to: calculate the cross-correlation integral vector of the input and output of the cascade module based on the cross-correlation integral vector of the input and output of the cascade module. and the input autocorrelation integral matrix of the cascade module Use formula (5) to obtain the static gain column vector of the cascade module

[0122] Where j = 1, 2, m l+1 ;m l+1 Indicates that the output measurement value of the lth cascade module is m l+1 dimensions;

[0123] S represents the measurement value of the cascade module; l S represents the input value of the lth cascade module; S j represents the input measurement value of the jth dimension in the cascade module; l+1 S j represents the output measurement value of the jth dimension in the lth cascade module; Represents the cross-correlation integral vector between the input of the l-th cascade module and the output of the j-th dimension; represents the input autocorrelation integral matrix of the lth cascade module; Represents the static gain column vector of the lth cascade module.

[0124] In some embodiments, the second acquisition submodule 54 is further configured to: obtain the static gain column vector of the cascade module based on the Using formula (6), the static gain matrix of the cascaded module is obtained: l G

[0125] Wherein, l = 1, 2, ..., q; q means that the cascade system contains q cascade modules; l means the lth cascade module in the cascade system;

[0126] m l+1 Indicates that the output measurement value of the lth cascade module is m l+1 dimensions;

[0127] S represents the measurement value of the cascade module; Indicates the mth l+1 The measurement value of the dimension; Indicates the mth module in the lth cascade module l+1 Output measurement of dimension; Represents the static gain column vector of the lth cascade module; l G represents the static gain matrix of the lth cascade module.

[0128] In some examples, the third acquisition module is further configured to: based on the static gain vector of all cascaded modules l G, using formula (7) to obtain the gradient between the optimization variable and the objective function of the cascade system;

[0129] Where K is the gradient between the optimization variable and the objective function of the cascade system; q means that the cascade system contains q cascade modules; l means the lth cascade module in the cascade system; l G represents the static gain matrix of the lth cascade module.

[0130] In some examples, the optimization variable includes at least one of controllable temperature, flow rate, pressure, and material composition. When the cascade system is a catalytic cracking unit, the controllable temperature includes at least one of the reaction temperature, a measured value of a reaction temperature controller, the feed temperature, and a measured value of a feed temperature controller; the flow rate includes at least one of the amount of fresh catalyst added, the pre-boost steam flow rate, and a measured value of the pre-boost steam flow rate; and the pressure includes the pressure in the catalytic cracking unit.

[0131] The objective function is a function associated with the optimization variables.

[0132] The cascade system correlation integration device provided by the embodiment of the present disclosure comprises the following steps: a first acquisition module respectively acquires the input measurement value and the output measurement value of each cascade module in the cascade system; the first acquisition module acquires the input autocorrelation integral matrix of the cascade module based on the input measurement value of the cascade module; the second acquisition module acquires the cross-correlation integral vector of the input and output of the cascade module based on the input measurement value and the output measurement value of the cascade module; the second acquisition module acquires the static gain matrix of each cascade module concurrently based on the input autocorrelation integral matrix of the cascade module and the cross-correlation integral vector of the input and output of the cascade module, and then acquires the gradient between the optimization variable and the objective function of the cascade system based on the static gain matrices of all cascade modules through the second acquisition module. By concurrently calculating the static gain matrices of each module, the time spent on acquiring the gradient between the optimization variable and the objective function of the cascade system is reduced, thereby effectively improving the optimization efficiency of the cascade system.

[0133] It should be noted that the cascade system-related integration device provided in the embodiment of the present disclosure can implement all steps of the cascade system-related integration method provided in the embodiment of the present disclosure. To save space, they will not be repeated here.

[0134] The functions or modules included in the device provided in the embodiments of the present disclosure can be used to execute the method described in the first aspect of the method embodiment above. Its specific implementation and technical effects can be referred to the description of the above method embodiment. For the sake of brevity, they will not be repeated here.

[0135] It should be noted that all modules involved in this embodiment are logical modules. In actual applications, a logical unit can be a physical unit, a part of a physical unit, or a combination of multiple physical units. In addition, to highlight the innovations of this disclosure, this embodiment does not include units that are not closely related to solving the technical problems proposed by this disclosure. However, this does not mean that other units do not exist in this embodiment.

[0136] It should be noted that, in this document, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or apparatus comprising the element.

[0137] Those skilled in the art will understand that although some embodiments described herein include certain features included in other embodiments but not other features, the combination of features from different embodiments is meant to be within the scope of the present embodiment and to form different embodiments.

[0138] It is understood that the above embodiments are merely exemplary embodiments for illustrating the principles of the present disclosure, and the present disclosure is not limited thereto. Those skilled in the art may make various modifications and improvements without departing from the spirit and substance of the present disclosure, and such modifications and improvements are also considered to be within the scope of protection of the present disclosure.

Claims

1. A cascade system correlation integration method, characterized in that: For obtaining the gradient between the optimization variable and the objective function of the cascade system, the cascade system comprises a plurality of cascade modules connected in series, including: Concurrently acquiring a static gain matrix of each of the cascaded modules; Based on the static gain matrices of all the cascade modules, obtaining the gradient between the optimization variables and the objective function of the cascade system; The static gain matrix of the cascade module is obtained by the following steps: Obtaining input measurement values ​​and output measurement values ​​of the cascade module; Obtaining an input autocorrelation integral matrix of the cascade module based on the input measurement value of the cascade module; Based on the input measurement value and the output measurement value of the cascade module, obtaining a cross-correlation integral vector of the input and output of the cascade module; The static gain matrix of the cascade module is obtained based on the input autocorrelation integral matrix of the cascade module and the cross-correlation integral vector of the input and output of the cascade module.

2. The method according to claim 1, characterized in that The step of obtaining an input autocorrelation integral matrix of the cascade module based on the input measurement value of the cascade module comprises: Obtaining an input-related integral value of the cascade module based on an input measurement value of the cascade module; An input autocorrelation integral matrix of the cascade modules is obtained based on the input correlation integral values ​​of the cascade modules.

3. The method according to claim 2, characterized in that The step of obtaining an input autocorrelation integral matrix of the cascaded modules based on the input correlation integral values ​​of the cascaded modules comprises: Based on the input measurement value of the cascade module, the input autocorrelation integral matrix of the cascade module is obtained using formula (1) and formula (2): Where i = 1, 2, ..., m l ; j=1,2,...,m l ; l=1,2,...,q; q means that the cascade system contains q cascade modules in total; l means the lth cascade module in the cascade system; m means that the measurement value of the cascade module has m dimensions in total; m l Indicates that the input measurement value of the lth cascade module has a total of m l dimensions; l T i is the first integral constant of the i-th module; l M i is the second integral constant of the i-th module; λ is the first integral variable; τ is the second integral variable; represents the autocorrelation integral matrix; S represents the measurement value of the cascade module; S i represents the measurement value of the i-th dimension in the cascade module; l S represents the input vector of the lth cascade module; l S i represents the input measurement value of the i-th dimension in the l-th cascade module; S j represents the measurement value of the j-th dimension in the cascade module; l S j represents the input measurement value of the j-th dimension in the l-th cascade module; represents the correlation integral value of the input of the i-th dimension and the j-th dimension in the l-th cascade module; represents the input autocorrelation integral matrix of the lth cascade module; l s i (λ) represents the input measurement value of the lth cascade module with respect to the i-th dimension of the variable λ; l s j (λ-τ) represents the input measurement value of the lth cascade module with respect to the jth dimension of the variable λ-τ.

4. The method according to claim 3, characterized in that The step of obtaining a cross-correlation integral vector of an input and an output of the cascade module based on the input measurement value and the output measurement value of the cascade module comprises: Obtaining a cross-correlation integral value between the input and output of the cascade module based on the input measurement value and the output measurement value of the cascade module; Based on the cross-correlation integral value of the input and output of the cascade module, a cross-correlation integral vector of the input and output of the cascade module is obtained.

5. The method according to claim 4, characterized in that The step of obtaining a cross-correlation integral vector of the input and output of the cascade module based on the cross-correlation integral value of the input and output of the cascade module comprises: Based on the input measurement value and output measurement value of the cascade module, the cross-correlation integral vector of the input and output of the cascade module is obtained using formula (3) and formula (4): in, j=1,2,...,m l+1 ; l=1,2,...,q; q means that the cascade system contains q cascade modules in total; l means the lth cascade module in the cascade system; m means that the measurement value of the cascade module has m dimensions in total; m l Indicates that the input vector of the lth cascade module has m l dimensions; l T i is the first integral constant of the i-th module; l M i is the second integral constant of the i-th module; λ is the first integral variable; τ is the second integral variable; represents the cross-correlation integral vector; S represents the measurement value of the cascade module; l S represents the measurement value of the lth cascade module; l S i represents the input measurement value of the i-th dimension of the l-th cascade module; l+1 S j represents the output measurement value of the j-th dimension in the l-th cascade module; Indicates the mth l The input measurement of the dimension, Indicates the mth in the lth cascade module l Input measurements of dimensions; represents the cross-correlation integral value between the input of the i-th dimension and the output of the j-th dimension in the l-th cascade module; represents the cross-correlation integral vector between the input of the lth cascade module and the output of the jth dimension; l s i (λ) represents the input measurement value of the lth cascade module with respect to the i-th dimension of the variable λ; l+1 s j (λ-τ) represents the output measurement value of the lth cascade module with respect to the jth dimension of the variable λ-τ.

6. The method according to claim 5, wherein the step of obtaining the static gain matrix of the cascaded modules based on the input autocorrelation integral matrix of the cascaded modules and the cross-correlation integral vector between the input and output of the cascaded modules comprises: Obtaining a static gain column vector of the cascade module based on an input autocorrelation integral matrix of the cascade module and a cross-correlation integral vector of an input and an output of the cascade module; Based on the static gain column vector of the cascade module, a static gain matrix of the cascade module is obtained.

7. The method according to claim 6, characterized in that The step of obtaining a static gain column vector of the cascade module based on an input autocorrelation integral matrix of the cascade module and a cross-correlation integral vector between an input and an output of the cascade module comprises: Based on the cross-correlation integral vector of the input and output of the cascade module and the input autocorrelation integral matrix of the cascade module The static gain column vector of the cascade module is obtained using formula (5): Where j = 1, 2, ..., m l+1 ;m l+1 Indicates that the output measurement value of the lth cascade module has m l+1 dimensions; S represents the measurement value of the cascade module; l S represents the input value of the lth cascade module; S j represents the input measurement value of the jth dimension in the cascade module; l+1 S j represents the output measurement value of the j-th dimension in the l-th cascade module; represents the cross-correlation integral vector between the input of the lth cascade module and the output of the jth dimension; represents the input autocorrelation integral matrix of the lth cascade module; Represents the static gain column vector of the lth cascade module.

8. The method according to claim 7, characterized in that The acquiring the static gain matrix of the cascaded modules based on the static gain column vector of the cascaded modules includes: Based on the static gain column vector of the cascade module Using formula (6), the static gain matrix of the cascade module is obtained: l G Wherein, l=1, 2, ..., q; q indicates that the cascade system includes q cascade modules in total; l indicates the lth cascade module in the cascade system; m l+1 Indicates that the output measurement value of the lth cascade module has m l+1 dimensions; S represents the measurement value of the cascade module; Indicates the mth l+1 The measurement value of the dimension; Indicates the mth module in the lth cascade module l+1 Output measurement of dimension; represents the static gain column vector of the lth cascade module; l G represents the static gain matrix of the lth cascade module.

9. The method according to claim 8, characterized in that The step of obtaining a gradient between an optimization variable and an objective function of the cascade system based on a static gain matrix of all the cascade modules includes: Based on the static gain matrix of all the cascaded modules l G, using formula (7) to obtain the gradient between the optimization variable and the objective function of the cascade system; Wherein, K is the gradient between the optimization variable and the objective function of the cascade system; q indicates that the cascade system includes a total of q cascade modules; l indicates the lth cascade module in the cascade system; l G represents the static gain matrix of the lth cascade module.

10. The method according to claim 1, characterized in that The optimization variables include at least one of controllable temperature, flow, pressure, and material components; The objective function is a function associated with the optimization variable.

11. A cascade system correlation integration device, characterized in that: For obtaining a gradient between an optimization variable and an objective function of a cascade system, the cascade system comprising a plurality of cascade modules connected in series, including: A first acquisition module, configured to concurrently acquire a static gain matrix of each of the cascaded modules; A second acquisition module is used to obtain the gradient between the optimization variable and the objective function of the cascade system based on the static gain matrix of all the cascade modules; The first acquisition module includes: A first acquisition submodule, configured to acquire input measurement values ​​and output measurement values ​​of the cascade module; A first obtaining submodule, configured to obtain an input autocorrelation integral matrix of the cascade module based on an input measurement value of the cascade module; A second obtaining submodule, configured to obtain a cross-correlation integral vector of an input and an output of the cascade module based on an input measurement value and an output measurement value of the cascade module; The second acquisition submodule is configured to acquire a static gain matrix of the cascade module based on an input autocorrelation integral matrix of the cascade module and a cross-correlation integral vector between an input and an output of the cascade module.

Images