A closed-loop control system performance degradation detection and recovery method

Abstract

The present invention discloses a closed-loop control system performance degradation detection and recovery method, which is used to realize the closed-loop control system performance degradation detection and recovery of process industrial processes, and belongs to the field of industrial process monitoring technology. The method includes: collecting process data of the closed-loop control system to be detected; performing performance degradation detection on the closed-loop control system based on the gap measurement technology; constructing a feedforward controller based on the closed-loop subspace prediction function and a state feedback controller based on the observer; when the closed-loop control system is detected to have performance degradation, the feedforward controller based on the closed-loop subspace prediction function and the state feedback controller based on the observer are used to optimize the performance of the closed-loop control system and realize the performance recovery of the closed-loop control system. The technical solution of the present invention can ensure the system operation reliability and product quality stability of the process industrial production process.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial process monitoring, and in particular to a method for detecting and restoring performance degradation of a closed-loop control system of an industrial process. Background Art

[0002] Faced with increasingly fierce international market competition and increasing product quality requirements, modern process industry production processes are constantly developing in the direction of large-scale, integrated and complex, and higher requirements are put forward for safe, reliable, high-quality and efficient production. At present, feedback controllers have been widely used in various process industry production processes to meet the requirements of safe and reliable operation of process industry production processes. However, in actual process industry applications, due to various situations such as changes in process operating conditions, mismatch between control system and controller parameters, unknown interference or process failures, the performance of closed-loop control systems is easily degraded; if they cannot be handled in time, it will lead to a chain reaction, which will cause fluctuations and abnormalities in the operating conditions of the entire process industry production process, affect product quality and production efficiency, and cause huge property losses and personal injuries. Performance monitoring of closed-loop control systems is an effective means to ensure the safe and reliable operation of process industry production processes. By real-time monitoring of closed-loop control systems, abnormal conditions in the production process can be discovered in time, potential failure risks can be predicted, and corresponding measures can be taken to adjust and optimize, ensure the stable operation of the production process, and improve production efficiency and product quality. Therefore, ensuring the safe and reliable operation of process industry production processes through performance monitoring of closed-loop control systems has become an important part of the sustainable and intelligent development of process industries. It will have important strategic significance for maximizing the operating potential of process industries and enhancing their international competitiveness.

[0003] Research in recent years has shown that in the field of performance monitoring of closed-loop control systems, existing performance monitoring technologies mainly focus on unknown interference or certain types of process component failures in the production process, such as sensor failures or actuator failures, and there are few reports on system-level performance degradation detection and recovery technologies for closed-loop control systems. In view of the major demands of modern process industrial production processes for safe and reliable operation and stable product quality, it is necessary to comprehensively utilize the process mechanism, prior knowledge and process data of the production process to study the performance degradation detection and recovery strategy of closed-loop control systems, so that the global performance of the process industrial production process can be efficiently improved, and provide a theoretical basis and technical guarantee for ensuring the safe and stable operation of the process industrial production process.

[0004] As we all know, gap measurement plays an important role in the stability analysis of closed-loop control systems, especially in the field of robust control. Simply put, gap measurement is used to measure the distance between two linear time-invariant systems under closed-loop conditions, which helps us understand the changes in the performance of closed-loop control systems when the system changes. However, in practical applications, there are some problems with existing gap measurement techniques. On the one hand, due to the uncertainty and complexity of process industry production processes, it is very difficult to establish an accurate closed-loop control system model, which limits the model-based gap measurement method. On the other hand, existing gap measurement techniques are mainly used for offline analysis and monitoring of system performance, and lack the ability to detect real-time performance degradation of closed-loop control systems. Therefore, it is necessary to use real-time collected process data to study the gap measurement data-driven implementation method of closed-loop control systems, and then realize real-time detection of performance degradation of closed-loop control systems.

[0005] At the same time, in the actual process of industrial production, in order to ensure the economic benefits of the entire production process, the system is usually not shut down for maintenance immediately when performance degradation is detected. Instead, it is hoped that a fast and accurate performance recovery strategy will be implemented when performance degradation of the closed-loop control system is detected. While shortening system maintenance and downtime, it reduces the operating cost of the production process and ensures the safe and stable operation of the production process.

[0006] However, as far as we know, most of the existing technologies are devoted to solving the active / passive fault-tolerant control implementation of specific types of faults in closed-loop control systems, and pay relatively little attention to the performance recovery technology of closed-loop control systems. In addition, the existing technologies still have shortcomings in model dependence, real-time performance, fault identification, complex system processing and robustness, and need to be further improved to meet the needs of practical applications. Summary of the invention

[0007] The present invention provides a closed-loop control system performance degradation detection and recovery method to solve the deficiencies of the prior art in terms of model dependence, real-time performance and robustness.

[0008] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0009] In one aspect, the present invention provides a closed-loop control system performance degradation detection and recovery method for realizing closed-loop control system performance degradation detection and recovery in a process industry process, the method comprising:

[0010] Collecting process data of the closed-loop control system to be tested; wherein the process data includes external input data, process input data and process output data of the closed-loop control system;

[0011] Using the process data of the closed-loop control system and based on the gap measurement technology, the closed-loop control system is tested for performance degradation to determine whether the closed-loop control system has performance degradation;

[0012] A feedforward controller based on a closed-loop subspace prediction function and a state feedback controller based on an observer are constructed; when performance degradation of the closed-loop control system is detected, based on a preset recovery strategy, the performance of the closed-loop control system is optimized by using the feedforward controller based on a closed-loop subspace prediction function and the state feedback controller based on an observer, thereby achieving performance recovery of the closed-loop control system.

[0013] Furthermore, the closed-loop control system is subjected to performance degradation detection based on the gap measurement technology by using the process data of the closed-loop control system to determine whether the closed-loop control system has performance degradation, including:

[0014] Based on the process data, construct the Hankel data matrix U f ,Y f ,as follows:

[0015]

[0016]

[0017] in, is time k 1 External input data, where k 1 =k-2s-N,...,ks-1; is time k 2 External input data, u(k 2 ) is the time k 2 The process input data, y(k 2 ) is the time k 2 The process output data, where k 2 =ks-N+1,...,ks,...,k; k is the current moment, s and N are both preset positive integers, and s is greater than the control system order, and N is greater than s;

[0018] The Hankel data matrix is ​​decomposed using the LQ decomposition technique, and the decomposition results are as follows:

[0019]

[0020]

[0021] Among them, L u and L y is the lower triangular matrix obtained by LQ decomposition; Q u and Q yis the standard orthogonal vector group matrix obtained by LQ decomposition; L u,11 Because L u The i-th 1 Row j 1 The submatrix consisting of column elements, i 1 =1,2,...,s+1,j 1 =1,2,...,s+1;L u,21 Because L u The i-th 2 Row j 2 The submatrix consisting of column elements, i 2 =s+2,...,2s+2,j 2 =1,2,...,s+1;L u,22 Because L u The i-th 3 Row j 3 The submatrix consisting of column elements, i 3 =s+2,...,2s+2,j 3 =s+2,...,2s+2; L u,31 Because L u The i-th 4 Row j 4 The submatrix consisting of column elements, i 4 =2s+3,...,3s+3,j 4 =1,...,s+1;L u,32 Because L u The i-th 5 Row j 5 The submatrix consisting of column elements, i 5 =2s+3,...,3s+3,j 5 =s+2,...,2s+2; L u,33 Because L u The i-th 6 Row j 6 The submatrix consisting of column elements, i 6 =2s+3,...,3s+3,j 6 =2s+3,...,3s+3; Q u,1 For Q u The i-th 7 Row j 7 The submatrix consisting of column elements, i 7 =1,...,s+1,j 7 =1,...,N;Q u,2 For Q u The i-th 8 Row j 8 The submatrix consisting of column elements, i 8=s+2,...,2s+2,j 8 =1,...,N;Q u,3 For Q u The i-th 9 Row j 9 The submatrix consisting of column elements, i 9 =2s+3,...,3s+3,j 9 =1,...,N;L y,11 Because L y The i-th 1 Row j 1 The submatrix composed of column elements; L y,21 Because L y The i-th 2 Row j 2 The submatrix composed of column elements; L y,22 Because L y The i-th 3 Row j 3 The submatrix composed of column elements; L y,31 Because L y The i-th 4 Row j 4 The submatrix composed of column elements; L y,32 Because L y The i-th 5 Row j 5 The submatrix composed of column elements; L y,33 Because L y The i-th 6 Row j 6 The submatrix consisting of column elements; Q y,1 For Q y The i-th 7 Row j 7 The submatrix consisting of column elements; Q y,2 For Q y The i-th 8 Row j 8 The submatrix consisting of column elements; Q y,3 For Q y The i-th 9 Row j 9 The submatrix consisting of column elements;

[0022] Using the stable image expression of closed-loop control system Get its data-driven implementation:

[0023]

[0024] in, A data-driven form of image expression for closed-loop control systems; and Respectively represent the relationship between process input and output signals and external input signals;

[0025] Using equations (1), (2) and (3) to realize the data-driven identification of the stable image expression of the closed-loop control system, the data-driven realization form of the stable image expression of the closed-loop control system is obtained as follows:

[0026]

[0027] Decomposing the data-driven implementation form of the closed-loop control system stability image expression, we get:

[0028]

[0029] in, Reason The i 10 Row j 10 The submatrix consisting of column elements, i 10 =1,...,2s+2,j 10 =1,...,s+1; Reason The i 11 Row j 11 The submatrix consisting of column elements, i 11 =1,...,2s+2,j 10 =s+1,...,2s+2;

[0030] Pair Matrix Perform singular value decomposition and get Among them, [U 1 U 2 ] is a left singular vector, ∑ 1 is a diagonal matrix, is a right singular vector; let The standardized data-driven implementation of the stable image expression is as follows:

[0031]

[0032] When new process data is collected, the iterative form of LQ decomposition in equations (1) and (2) is obtained as follows:

[0033]

[0034]

[0035] Among them, L u,new , Q u,new , L y,new and Q y,new L u , Qu , L y and Q y The updated value of It is a column vector composed of the data of the external input signal at the past moment; u is a column vector composed of the past data of the external input signal; f (k) is the column vector composed of the future data of the process input signal; y f (k) is a column vector composed of the future data of the process output signal; Reason and u f (k) is a column vector composed of; Reason and f (k) is a column vector composed of;

[0036] Using the Givens rotation technique, we get L u,new and L y,new The iterative update form is as follows:

[0037]

[0038]

[0039] Among them, Q u,givens and Q y,givens is the Givens matrix;

[0040] The standardized data-driven implementation form of online updating of stable image expression using equations (4), (5) and (6)

[0041] Using process data collected offline, calculate the nominal and Using the process data collected online, calculate the current and Then calculate the closed-loop control system G under nominal conditions and the closed-loop control system G at the current moment k Gap distance:

[0042]

[0043] Among them, σ max (·) are the singular values ​​of the matrix;

[0044] Set the following online evaluation function:

[0045]

[0046] Based on the online evaluation function, the performance degradation detection logic of the closed-loop control system is designed:

[0047]

[0048] Among them, J th,0 is the first preset threshold.

[0049] Furthermore, the feedforward controller based on the closed-loop subspace prediction function is constructed in the following manner: a data-driven form of image expression of a closed-loop control system is utilized to realize the design of a feedforward controller based on the closed-loop subspace prediction function.

[0050] Furthermore, the data-driven form of the stable image expression using the closed-loop control system is used to implement the design of a feedforward controller based on a closed-loop subspace prediction function, including:

[0051] Based on the relationship between external input data and process output data, the prediction matrix of future process output data is defined as:

[0052]

[0053]

[0054] in, is the subspace matrix related to the past moment data of the external input data; is the subspace matrix related to the future time data of the external input data; and are the Hankel data matrices of the external input data in the past and future respectively;

[0055] Using formula (14), we can get the prediction sequence of future process output signals:

[0056]

[0057] The external input data is expressed as the following integral form:

[0058]

[0059] in, is the external input data at time k-1; is the external input signal at time k The incremental value of

[0060] The external input data sequence of the future s prediction interval and its increment sequence And the process output data sequence for the future s prediction interval and its increment sequence Defined as:

[0061]

[0062]

[0063] in, k 3 External input data at time, k 3 The incremental value of the external input data at the moment, k 3 The process output prediction data at time k 3 =k,1,...,k+s; T represents matrix transpose;

[0064] Using equations (16), (17) and (18), we can obtain the incremental prediction output sequence related to the future process output data prediction sequence:

[0065]

[0066] in, k 4 The incremental value of the external input data at time k 4 =ks-1,...,k-1;

[0067] The forecast sequence of future process output data is rewritten as:

[0068]

[0069]

[0070] y k =[y T (k) y T (k) … y T (k)] T

[0071] Among them, y(k) is the process output data at time k; For k y The identity matrix of order, k y The dimension of the process output signal;

[0072] The feedforward control based on the closed-loop subspace prediction function is described as the following optimization problem:

[0073]

[0074] Among them, v f,s For future reference sequences, satisfy the following form:

[0075] v f,s =[v T (k+1) … vT (k+s+1)] T (twenty four)

[0076] Wherein, v(k+1) is the reference signal data at time k+1; v(k+s+1) is the reference signal data at time k+s+1; and, T y and is a weight matrix that satisfies the following form:

[0077]

[0078] Among them, λ y,1 ,λ y,2 ,λ y,s+1 and These are element values ​​determined according to system performance requirements;

[0079] Substituting equation (21) into equation (23), the optimization problem can be further written as:

[0080]

[0081] set up get:

[0082]

[0083] At each sampling moment, using The first element Able to obtain the optimal external input signal value

[0084] Furthermore, the observer-based state feedback controller is constructed in the following manner: a data-driven form of a closed-loop control system stabilizing the kernel expression is used to implement the observer-based state feedback controller design.

[0085] Furthermore, the data-driven form of the closed-loop control system stabilizing the kernel expression is used to implement the observer-based state feedback controller design, including:

[0086] Considering that the stable kernel expression of the closed-loop control system satisfies:

[0087]

[0088] Among them, u s (k)=[u T (ks) … u T (k)] T is a column vector of process input data, y s (k) = [y T (ks) … y T (k)] TA column vector consisting of process output data;

[0089] Substituting formula (3) into formula (28), we get:

[0090]

[0091] in, A column vector consisting of external input data;

[0092] Obviously, there is make If and only if the stable kernel expression and stable image expression of the closed-loop control system satisfy the following relationship:

[0093]

[0094] in, for The orthogonal complement of

[0095] When the stable image expression of the closed-loop control system is calculated by formula (4), the stable kernel expression of the closed-loop control system satisfies the following relationship:

[0096]

[0097] and, Able to pass the matrix The singular value decomposition of is solved;

[0098]

[0099] Among them, ∑ is a diagonal matrix;

[0100] Select the following equivalent vectors:

[0101]

[0102] Among them, [β s,0 β s,1 … β s,s ] and [α s,0 α s,1 … α s,s ] is a matrix Elements in a row vector;

[0103] The following state observer is obtained:

[0104]

[0105] Among them, x z (k) is the state estimation signal at time k, x z (k+1) is the state estimation signal at time k+1, r(k) is the residual signal at time k, Az ,B z ,C z ,G z ,D z ,L z is the parameter matrix of the designed state observer and satisfies: C z =[0 … 0 1],G z =α s,s ,L z =-[α s,0 α s,1 … α s,s-1 ] T ;

[0106] Rewrite formula (34) into the following form:

[0107]

[0108] in,

[0109]

[0110] Based on formula (35), the following state feedback controller is designed:

[0111]

[0112] Where u(z) is the z-transform of the process input signal; x z (z) is the z-transform of the state estimation signal; is the z-transform of the external input signal; F z is the state feedback gain to be designed, which is determined by the following quadratic cost function:

[0113]

[0114] in, and R u is a weight matrix that satisfies

[0115] The state feedback gain F of the state feedback controller z Determined as:

[0116]

[0117] Among them, the P matrix is ​​a positive solution of the Riccati equation, satisfying the following form

[0118]

[0119] Furthermore, the preset recovery strategy uses a feedforward controller based on a closed-loop subspace prediction function and a state feedback controller based on an observer to optimize the performance of the closed-loop control system and achieve performance recovery of the closed-loop control system, including:

[0120] If J th,0 ≤J(k)≤J th,1 , the feedforward controller based on the closed-loop subspace prediction function is applied to the closed-loop control system to optimize the performance of the closed-loop control system and realize the performance recovery of the closed-loop control system;

[0121] If J(k)>J th,1 , applying the observer-based state feedback controller to the closed-loop control system, optimizing the performance of the closed-loop control system, and realizing the performance recovery of the closed-loop control system, J th,1 is the second preset threshold, and J th,1 >J th,0 .

[0122] On the other hand, the present invention further provides an electronic device, comprising a processor and a memory; wherein the memory stores at least one instruction, and the instruction is loaded and executed by the processor to implement the above method.

[0123] In yet another aspect, the present invention further provides a computer-readable storage medium, wherein at least one instruction is stored in the storage medium, and the instruction is loaded and executed by a processor to implement the above method.

[0124] The beneficial effects brought about by the technical solution provided by the present invention include at least:

[0125] The performance degradation detection and recovery method of the closed-loop control system proposed in the present invention uses the input and output data of the closed-loop control system of the process industry process to detect the performance degradation of the closed-loop control system of the process industry process. At the same time, the performance recovery of the closed-loop control system is achieved by using a feedforward controller based on the closed-loop subspace prediction function and a state feedback controller based on the observer. The performance degradation detection and recovery method of the closed-loop control system of the process industry process designed by the present invention has the following advantages compared with the existing performance monitoring methods:

[0126] 1) It does not need to rely on the precise mechanism model of the closed-loop control system. It can realize performance degradation detection and recovery by only using the input and output data of the closed-loop control system of the process industry, which has stronger engineering applicability;

[0127] 2) The designed closed-loop control system performance degradation detection method based on gap measurement can utilize the process data collected in real time to realize real-time detection of performance degradation of the closed-loop control system, understand the operating status of the process industry process, and improve the operating efficiency of the equipment.

[0128] 3) The designed feedforward controller based on closed-loop subspace prediction function and observer-based state feedback controller can apply a fast and accurate performance recovery strategy when the performance degradation of the closed-loop control system is detected, shortening the system maintenance and downtime, while reducing the operating cost of the process industry process and ensuring the safe and stable operation of the process industry production process.

[0129] In summary, the closed-loop control system performance degradation detection and recovery method proposed in the present invention has improvements in model dependence, real-time performance, complex system processing and robustness, which better meets the needs of practical applications in process industries and can ensure the system operation reliability and product quality stability of process industry production processes. BRIEF DESCRIPTION OF THE DRAWINGS

[0130] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0131] Figure 1 is a flow chart of a closed-loop control system performance degradation detection and recovery method provided by an embodiment of the present invention;

[0132] Figure 2 It is a schematic diagram of a performance degradation detection and recovery strategy for a closed-loop control system for a process industry process provided by an embodiment of the present invention;

[0133] Figure 3 It is a system block diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0134] In order to make the objectives, technical solutions and advantages of the present invention more clear, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

[0135] First of all, it should be noted that in the embodiments of the present invention, words such as "exemplarily" and "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "exemplary" in the present invention should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of the word "exemplarily" is intended to present the concept in a concrete way. In addition, in the embodiments of the present invention, the meaning expressed by "and / or" can be both, or it can be either of the two.

[0136] First embodiment

[0137] In view of the deficiencies of the prior art in terms of model dependence, real-time and robustness, this embodiment provides a closed-loop control system performance degradation detection and recovery method for realizing closed-loop control system performance degradation detection and recovery in process industrial processes. This method can utilize process data collected in real time to realize online data-driven calculation of gap metrics and real-time detection of performance degradation of closed-loop control systems without relying on precise models of closed-loop control systems, and on this basis, apply a fast and accurate performance recovery strategy to achieve performance optimization of closed-loop control systems. This method can be implemented by an electronic device, which can be a terminal or a server. Specifically, the execution flow of this method is as follows: Figure 1 As shown, the following steps are included:

[0138] S1, collecting process data of the closed-loop control system to be detected; wherein the process data includes external input data, process input data and process output data of the closed-loop control system;

[0139] It should be noted that the method of this embodiment can be applied to the performance degradation detection and recovery of closed-loop control systems in various process industries; for example: the thickness control system of a plate and strip mill in the production process of hot strip rolling; when applied to this system, the external input data collected are: the rack outlet thickness setting value h ref ; The process input data is: the output signal u of the thickness controller; the process output data is: the measured value h of the thickness at the rack outlet.

[0140] S2, using the process data of the closed-loop control system and based on the gap measurement technology, to perform performance degradation detection on the closed-loop control system to determine whether the closed-loop control system has performance degradation;

[0141] Specifically, in this embodiment, the implementation process of the performance degradation detection based on the gap metric is as follows:

[0142] Step 1: Collect and store external input data of the closed-loop control system through sensors, communication networks and computing units The process input and output data u(k) and y(k) construct the Hankel data matrix:

[0143]

[0144]

[0145] in, For external input data The Hankel matrix at the past moment, For external input data The Hankel matrix at the future time, U fis the Hankel matrix of the process input data u(k) at the future time, Y f The Hankel matrix of the process output data y(k) at the future time. is the external input data at time k-2s-N; is the external input data at time k-2s-1; It is the external input data at time ksN; is the external input data at time ks-1; It is the external input data at time ks-N+1; is the external input data at time ks; is the external input data at time k-N+1; u(ks-N+1) is the process input data at time ks-N+1; u(ks) is the process input data at time ks; u(k-N+1) is the process input data at time k-N+1; y(ks-N+1) is the process output data at time ks-N+1; y(ks) is the process output data at time ks; y(k-N+1) is the process output data at time k-N+1; s and N are parameters that need to be determined according to the actual situation of the closed-loop control system, s and N are both positive integers, and s is greater than the control system order, and N is greater than s;

[0146] Step 2: Decompose the above Hankel data matrix using LQ decomposition technique:

[0147]

[0148]

[0149] Among them, L u and L y is the lower triangular matrix obtained by LQ decomposition, Q u and Q y is the standard orthogonal vector group matrix obtained by LQ decomposition; L u,11 For L u The submatrix consisting of the elements from row 1 to row s+1 and column 1 to column s+1 of the matrix; L u,21 For L u The submatrix of the matrix consists of the elements from the s+2th row to the 2s+2th row and the 1st column to the s+1th column; L u,22 For L u The submatrix consisting of the elements from the s+2th row to the 2s+2th row and the s+2th column to the 2s+2th column of the matrix; L u,31 For L u The submatrix consisting of the elements from the 2s+3th row to the 3s+3th row and the 1st column to the s+1th column of the matrix; L u,32 For L uThe submatrix consisting of the elements from the 2s+3th row to the 3s+3th row and the s+2th column to the 2s+2th column of the matrix; L u,33 For L u The submatrix composed of the elements from the 2s+3rd row to the 3s+3rd row and the 2s+3rd column to the 3s+3rd column of the matrix; Q u,1 Q u The submatrix composed of the elements from the 1st row to the s+1th row and the 1st column to the Nth column of the matrix; Q u,2 Q u The submatrix of the matrix consists of the elements from the s+2th row to the 2s+2th row and the 1st column to the Nth column; Q u,3 Q u The submatrix consisting of the elements from the 2s+3th row to the 3s+3th row and the 1st column to the Nth column of the matrix; L y,11 For L y The submatrix consisting of the elements from row 1 to row s+1 and column 1 to column s+1 of the matrix; L y,21 For L y The submatrix of the matrix consists of the elements from the s+2th row to the 2s+2th row and the 1st column to the s+1th column; L y,22 For L y The submatrix consisting of the elements from the s+2th row to the 2s+2th row and the s+2th column to the 2s+2th column of the matrix; L y,31 For L y The submatrix consisting of the elements from the 2s+3th row to the 3s+3th row and the 1st column to the s+1th column of the matrix; L y,32 For L y The submatrix consisting of the elements from the 2s+3th row to the 3s+3th row and the s+2th column to the 2s+2th column of the matrix; L y,33 For L y The submatrix composed of the elements from the 2s+3rd row to the 3s+3rd row and the 2s+3rd column to the 3s+3rd column of the matrix; Q y,1 Q y The submatrix composed of the elements from the 1st row to the s+1th row and the 1st column to the Nth column of the matrix; Q y,2 Q y The submatrix of the matrix consists of the elements from the s+2th row to the 2s+2th row and the 1st column to the Nth column; Q y,3 Q y The submatrix composed of the elements of the 2s+3rd to 3s+3rd rows and the 1st to Nth columns of the matrix; Step 3: Use the stable image expression of the closed-loop control system Get its data-driven implementation:

[0150]

[0151] in, A data-driven form of closed-loop control system stability expression, and Describes the process input and output signals u and y and the external input signal The relationship between them.

[0152] Step 4: Using equations (1), (2) and (3), realize data-driven identification of the stable image expression of the closed-loop control system:

[0153]

[0154] Step 5: Decompose the data-driven implementation form of the closed-loop control system stability image expression:

[0155]

[0156] in, for The submatrix consisting of the elements from row 1 to row 2s+2 and column 1 to column s+1 of the matrix; for The submatrix consisting of the elements from the 1st to the 2s+2th rows and the s+1th to the 2s+2th columns of the matrix.

[0157] Step 6: Matrix Perform singular value decomposition (SVD): (Among them, [U 1 U 2 ] is a left singular vector, ∑ 1 is a diagonal matrix, is the right singular vector, where the dimensions of each matrix are related to the order n of the system, which will not be elaborated), let ( for ), we can get a standardized data-driven implementation of stable image expression:

[0158]

[0159] Step 7: When new measurement data is collected When u(k) and y(k), the iterative form of LQ decomposition in equation (1) and equation (2) can be obtained:

[0160]

[0161]

[0162] Among them, L u,new , Q u,new , L y,new and Q y,new is L in formula (1) and formula (2)u , Q u , L y and Q y Updated values ​​of the matrix; For external input signal A column vector consisting of past time data, For external input signal Column vector consisting of past time data, u f (k) is the column vector of the future data of the process input signal u, y f (k) is a column vector composed of the future data of the process output signal y; for and u f (k) is a column vector composed of; for and f (k) is a column vector composed of

[0163]

[0164]

[0165] Step 8: Use Givens rotation technique to get L u,new and L y,new The iterative update form is:

[0166]

[0167]

[0168] Among them, Q u,givens and Q y,givens is the Givens matrix. On this basis, the standardized data-driven implementation form of the stable image expression can be updated online using equations (4), (5) and (6):

[0169] Step 9: Calculate the nominal value using the process data collected offline and Using the process data collected online, calculate the current and Based on this, calculate the closed-loop control system G under nominal conditions and the closed-loop control system G at the current moment k Gap distance

[0170]

[0171] Among them, max(·) is the maximum value of the matrix; is the closed-loop control system G under nominal conditions and the closed-loop control system G at the current moment k Direct clearance distance between is the closed-loop control system G at the current moment k The direct gap distance between the closed-loop control system G under nominal conditions; σ max (·) are the singular values ​​of the matrix.

[0172] Step 10: Since the closed-loop control system G under the nominal condition and the closed-loop control system G at the current moment k The gap distance between can provide evaluation information for the performance degradation degree of the closed-loop control system, and the following online evaluation function J(k) can be determined:

[0173]

[0174] Step 11: Based on the online performance evaluation function (12) of the closed-loop control system, design the performance degradation detection logic of the closed-loop control system:

[0175]

[0176] Among them, J th,0 It is a pre-designed threshold that can be used to trigger subsequent performance recovery strategies.

[0177] S3, construct a feedforward controller based on closed-loop subspace prediction function and a state feedback controller based on observer;

[0178] Specifically, in this embodiment, a feedforward controller based on a closed-loop subspace prediction function is constructed, and the implementation steps of implementing the feedforward control based on the closed-loop subspace prediction function are as follows:

[0179] Step 1: Online data-driven implementation based on stable image expression of closed-loop control system External input signal The relationship between the process output signal y and the prediction matrix of the future process output signal It can be defined as:

[0180]

[0181]

[0182] in, is the subspace matrix related to the past time data of the external input signal, L vf is the subspace matrix related to the future data of the external input signal, and Respectively, external input signals Hankel data matrices for past and future moments.

[0183] Step 2: Use formula (14) to obtain the predicted sequence of future process output signals:

[0184]

[0185] Step 3: In order to reduce the tracking error of the feedback control system, it is necessary to add an integral link to the designed feedforward controller. Therefore, the external input signal It can be expressed in the following integral form:

[0186]

[0187] in, is the external input data at time k-1; is the external input signal at time k The incremental value of .

[0188] Step 4: At the same time, the external input signal sequence of the future s prediction interval Its incremental sequence The process output signal sequence for the future s prediction interval and its increment sequence It can be defined as:

[0189]

[0190]

[0191] in, is the external input data at time k+s; is the incremental value of the external input signal at time k+s; Output forecast data for the process at time k; Output forecast data for the process at time k+s; The incremental value of the prediction signal output by the process at time k; The incremental value of the prediction signal output by the process at time k+s.

[0192] Step 5: Using equations (16)-(18), we can obtain the incremental prediction output sequence related to the future process output signal prediction sequence:

[0193]

[0194] in, is the external input signal sequence of the past s interval, which can be expressed as:

[0195]

[0196] and, It is the incremental value of the external input signal at time ks-1; is the incremental value of the external input signal at time k-1.

[0197] Step 6: The prediction sequence of future process output signals can be rewritten as:

[0198]

[0199]

[0200] y k =[y T (k) y T (k) … y T (k)] T

[0201] in, For k y The identity matrix of order; k y The dimension of the process output signal; is the process output signal sequence of the prediction interval s at time k+1; y k is a sequence of process output signals at time k; ∏ is a data matrix composed of unit matrices; and is the key parameter matrix, which can be calculated by formula (22); Step 7: Feedforward control based on closed-loop subspace prediction function can be described as the following optimization problem:

[0202]

[0203] in, is the designed quadratic performance index function; v f,s is the reference signal sequence of the future s prediction interval, satisfying the following form:

[0204]

[0205] v(k+1) is the reference signal data at time k+1; v(k+s+1) is the reference signal data at time k+s+1.

[0206] And, T y and is a weight matrix that satisfies the following form:

[0207] Among them, λ y,1 ,λ y,2 ,λ y,s+1 and are all elements on the diagonal of the weight matrix and are determined according to the system performance requirements. It is worth pointing out that the weight matrix Ty Mainly affects tracking error, weight matrix Mainly affects the change amplitude of the external input signal.

[0208] Step 8: Substitute equation (21) into equation (23). Equation (23) can be further written as:

[0209]

[0210] Step 9: Set You can get:

[0211]

[0212] Step 10: At each sampling moment, use The first element The optimal external input signal value can be obtained:

[0213]

[0214] So far, through the above steps 1 to 10, the data-driven form of the closed-loop control system stable image expression is utilized to realize the design of the closed-loop control system feedforward controller based on the closed-loop subspace prediction function, thereby realizing the system performance recovery when the performance of the closed-loop control system is degraded, and ensuring the tracking performance of the closed-loop control system.

[0215] Furthermore, in this embodiment, an observer-based state feedback controller is constructed, and the implementation steps of the observer-based state feedback control are as follows:

[0216] Step 1: Consider the stable kernel expression of the closed-loop control system satisfy:

[0217]

[0218] Among them, u s (k)=[u T (ks) … u T (k)] T is a column vector of process input data, y s (k) = [y T (ks) … y T (k)] T A column vector containing the output data of the procedure.

[0219] Substituting the stability image expression (3) of the closed-loop control system into equation (28), we can obtain:

[0220]

[0221] in, A column vector consisting of external input data.

[0222] Obviously, there is make If and only if the stable kernel expression and stable image expression of the closed-loop control system satisfy the following relationship:

[0223]

[0224] in, for The orthogonal complement of

[0225] Step 2: Further, when the stable image expression of the closed-loop control system is calculated by formula (4), the stable kernel expression of the closed-loop control system satisfies the following relationship:

[0226]

[0227] and, Through the matrix The SVD decomposition of is solved to get:

[0228]

[0229] Step 3: Considering the closed-loop control system stability kernel expression, the data-driven form of expression (32) can be used to design a data-driven state observer by selecting the following equivalent vector β s and α s :

[0230]

[0231] Among them, [β s,0 β s,1 … β s,s ] and [α s,0 α s,1 … α s,s ] is a matrix The elements in a row vector have no substantive definition.

[0232] The following state observer can be obtained:

[0233]

[0234] Among them, x z (k) is the state estimation signal at time k, x z (k+1) is the state estimation signal at time k+1, r(k) is the residual signal at time k, A z ,B z ,C z ,Gz ,D z ,L z is the parameter matrix of the designed state observer and satisfies the following form C z =[0 … 0 1],G z =α s,s ,L z =-[α s,0 α s,1 … α s,s-1 ] T

[0235] Step 4. Further, considering the one-to-one correspondence between the state observer based on the equivalent space and the residual generator based on the diagnostic observer, equation (34) can be further rewritten as follows:

[0236]

[0237] Among them, A c ,B c ,C c ,D c ,L c is the parameter matrix of the designed diagnostic state observer, which satisfies the following form

[0238]

[0239] Step 5: Based on the state observer expression (35), the following state feedback controller can be designed:

[0240]

[0241] Where u(z) is the z-transform of the process input signal; x z (z) is the z-transform of the state estimation signal; is the z-transform of the external input signal; F z The state feedback gain that needs to be designed can be determined by the following quadratic cost function:

[0242]

[0243] in, and R u is a weight matrix that satisfies R u >0.

[0244] Step 6: State feedback gain F of the state feedback controller z It can be determined by minimizing the cost function (38):

[0245]

[0246] Among them, the P matrix is ​​the positive solution of the Riccati equation and satisfies the following form:

[0247]

[0248] So far, through the above steps one to six, the data-driven form of the closed-loop control system stable kernel expression is used to realize the design of the closed-loop control system state feedback controller based on the observer, realize the system performance recovery when the closed-loop control system performance is degraded, and ensure the stable performance of the closed-loop control system.

[0249] S4, when the performance degradation of the closed-loop control system is detected, based on a preset recovery strategy, a feedforward controller based on a closed-loop subspace prediction function and a state feedback controller based on an observer are used to optimize the performance of the closed-loop control system and achieve performance recovery of the closed-loop control system;

[0250] Among them, closed-loop control system performance degradation detection and recovery strategies are as follows: Figure 2 As shown, the details are as follows:

[0251] 1) If J th,0 ≤J(k)≤J th,1 , the designed feedforward controller based on closed-loop subspace prediction function is applied to the closed-loop control system. Without changing the original controller structure and parameters of the closed-loop control system, it is used to optimize the performance of the closed-loop control system and improve the tracking performance of the closed-loop control system; Among them, J th,0 and J th,1 are all preset thresholds, and J th,1 >J th,0 ;

[0252] 2) If J(k)>J th,1 , the designed observer-based state feedback controller is applied to the closed-loop control system to restore the performance of the closed-loop control system and improve the stability of the closed-loop control system; if necessary, the designed feedforward controller can be further optimized to optimize the tracking performance of the closed-loop control system.

[0253] In summary, this embodiment provides a closed-loop control system performance degradation detection and recovery method, using process input and output data, based on the offline and online data-driven implementation of the stable image expression of the closed-loop control system, a closed-loop control system online performance degradation detection method based on gap measurement is proposed; a feedforward controller based on the closed-loop subspace prediction function and an observer-based state feedback controller are constructed for the performance recovery of the closed-loop control system; based on the data-driven implementation of the stable image expression and the stable kernel expression of the closed-loop control system, a closed-loop control system performance degradation detection and recovery strategy for process industry processes is developed. Compared with the prior art, the closed-loop control system performance degradation detection and recovery method of this embodiment has been improved in terms of model dependence, real-time, complex system processing and robustness, and can better meet the needs of actual application of process industry, and can ensure the system operation reliability and product quality stability of process industry production processes.

[0254] Second embodiment

[0255] This embodiment provides an electronic device, such as Figure 3 As shown, the electronic device includes: a processor and a memory; wherein the processor and the memory can be connected via a communication bus; the memory stores at least one instruction, and the instruction is loaded and executed by the processor to implement the method of the first embodiment. In addition, the electronic device may also include a transceiver, the processor and the transceiver can be connected via a communication bus, and the transceiver is used to communicate with other devices.

[0256] Next, combine Figure 3 The following is a detailed introduction to the various components of the electronic device:

[0257] Among them, the processor is the control center of the electronic device, and the electronic device may include multiple processors, each of which may be a single-core processor (single-CPU) or a multi-core processor (multi-CPU). The processor here may be a processor or a general term for multiple processing elements. For example, the processor is one or more central processing units (CPUs), or other general-purpose processors, application specific integrated circuits (ASICs), or one or more integrated circuits configured to implement an embodiment of the present invention, such as one or more microprocessors (digital signal processors, DSPs), or one or more field programmable gate arrays (field programmable gate arrays, FPGAs), or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc. The processor may execute various functions of the electronic device by running or executing software programs stored in the memory and calling data stored in the memory.

[0258] In a specific implementation, as an embodiment, the processor may include one or more CPUs, such as Figure 3 The CPU0 and CPU1 shown in the figure are, of course, only exemplary.

[0259] The memory is used to store the software program for executing the solution of the present invention, and the execution is controlled by the processor. The specific implementation method can refer to the above method embodiment and will not be repeated here.

[0260] Optionally, the memory may be a read-only memory (ROM) or other types of static storage devices that can store static information and instructions, a random access memory (RAM) or other types of dynamic storage devices that can store information and instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital versatile discs, Blu-ray discs, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory may be integrated with the processor or exist independently and accessed through the interface circuit ( Figure 3 (not shown) is coupled to the processor, which is not specifically limited in this embodiment of the present invention.

[0261] The transceiver may include a receiver and a transmitter ( Figure 3 The receiver is used to implement the receiving function, and the transmitter is used to implement the sending function. The transceiver can be integrated with the processor or exist independently and communicate with the electronic device through the interface circuit ( Figure 3 (not shown) is coupled to the processor, which is not specifically limited in this embodiment of the present invention.

[0262] In addition, it should be noted that Figure 3 The structure of the electronic device shown in the figure does not constitute a limitation on the device, and the actual device may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently. In addition, the technical effects achieved by the electronic device when executing the method of the first embodiment above can refer to the technical effects described in the first embodiment above, so they are not repeated here.

[0263] Third embodiment

[0264] This embodiment provides a computer-readable storage medium, which stores at least one instruction, and the instruction is loaded and executed by a processor to implement the method of the first embodiment. The computer-readable storage medium may be a ROM, a random access memory, a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc. The instructions stored therein may be loaded by a processor in a terminal to execute the method.

[0265] In addition, it should be noted that the present invention can be provided as a method, an apparatus or a computer program product. Therefore, the embodiment of the present invention can be in the form of a full or partial hardware embodiment, a full or partial software embodiment or an embodiment combining software and hardware. Moreover, when implemented using software, the embodiment of the present invention can be in the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program codes. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center containing one or more available media sets. The available medium may be a magnetic medium (eg, a floppy disk, a hard disk, a magnetic tape), an optical medium (eg, a DVD), or a semiconductor medium. The semiconductor medium may be a solid state hard disk.

[0266] The embodiments of the present invention are described with reference to the flowcharts and / or block diagrams of the methods, terminal devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, an embedded processor, or other programmable data processing terminal device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0267] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing terminal device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1These computer program instructions can also be loaded onto a computer or other programmable data processing terminal device, so that a series of operation steps are executed on the computer or other programmable terminal device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable terminal device provide for implementing the process in the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0268] It should also be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. The terms "include", "comprises" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal device. In the absence of further restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, article or terminal device including the elements.

[0269] In several embodiments provided by the present invention, it should be understood that the disclosed equipment, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of functional modules / units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point, the coupling or direct coupling or communication connection between each other shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms. The unit described as a separate component may or may not be physically separated, and the component displayed as a unit may or may not be a physical unit, that is, it may be located in one place, or it may be distributed on multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the scheme of this embodiment. In addition, each functional unit in each embodiment of the present invention can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit.

[0270] If the method is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0271] Finally, it should be noted that the above is only a preferred embodiment of the present invention. It should be pointed out that although the preferred embodiment of the present invention has been described, for ordinary technicians in this technical field, once the basic creative concept of the present invention is known, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the protection scope of the present invention. Therefore, the attached claims are intended to be interpreted as including the preferred embodiment and all changes and modifications that fall within the scope of the embodiments of the present invention.

Claims

1. A closed-loop control system performance degradation detection and recovery method, used to implement closed-loop control system performance degradation detection and recovery in process industry, characterized in that: include: Collecting process data of the closed-loop control system to be tested; wherein the process data includes external input data, process input data and process output data of the closed-loop control system; Using the process data of the closed-loop control system and based on the gap measurement technology, the closed-loop control system is tested for performance degradation to determine whether the closed-loop control system has performance degradation; Construct a feedforward controller based on a closed-loop subspace prediction function and an observer-based state feedback controller; When performance degradation of the closed-loop control system is detected, based on a preset recovery strategy, a feedforward controller based on a closed-loop subspace prediction function and a state feedback controller based on an observer are used to optimize the performance of the closed-loop control system and achieve performance recovery of the closed-loop control system; Using the process data of the closed-loop control system and based on the gap measurement technology, the closed-loop control system is tested for performance degradation to determine whether the closed-loop control system has performance degradation, including: Based on the process data, construct the Hankel data matrix U f ,Y f ,as follows: in, is the external input data at time k1, where k1=k-2s-N,...,ks-1; is the external input data at time k2, u(k2) is the process input data at time k2, y(k2) is the process output data at time k2, where k2=ks-N+1,...,ks,...,k; k is the current time, s and N are both preset positive integers, and s is greater than the control system order, and N is greater than s; The Hankel data matrix is ​​decomposed using the LQ decomposition technique, and the decomposition results are as follows: Among them, L u and L y is the lower triangular matrix obtained by LQ decomposition; Q u and Q y is the standard orthogonal vector group matrix obtained by LQ decomposition; L u,11 Because L u The submatrix composed of the elements in the i1th row and j1th column in L, i1=1,2,...,s+1, j1=1,2,...,s+1; u,21 Because L u The submatrix composed of the elements in the i2th row and j2th column in L, i2=s+2,...,2s+2, j2=1,2,...,s+1; u,22 Because L u The submatrix composed of the elements in the i3th row and j3th column in L, i3=s+2,...,2s+2, j3=s+2,...,2s+2; u,31 Because L u The submatrix composed of the elements in the i4th row and j4th column in L, i4=2s+3,...,3s+3, j4=1,...,s+1; u,32 Because L u The submatrix composed of the elements in the i5th row and j5th column in L, i5=2s+3,...,3s+3, j5=s+2,...,2s+2; u,33 Because L u The submatrix composed of the elements in the i6th row and j6th column in Q, i6 = 2s+3, ..., 3s+3, j6 = 2s+3, ..., 3s+3; u,1 For Q u The submatrix composed of the elements in the i7th row and j7th column, i7=1,...,s+1, j7=1,...,N; Q u,2 For Q u The submatrix composed of the elements in the i8th row and j8th column in Q, i8=s+2,...,2s+2, j8=1,...,N; u,3 For Q u The submatrix composed of the elements in the i9th row and j9th column in L, i9=2s+3,...,3s+3, j9=1,...,N; y,11 Because L y The submatrix composed of the elements in the i1th row and j1th column in L y,21 Because L y The submatrix composed of the elements in the i2th row and j2th column in L y,22 Because L y The submatrix composed of the elements in the i3th row and j3th column in L y,31 Because L y The submatrix consisting of the elements in the i4th row and j4th column in L y,32 Because L y The submatrix composed of the elements in the i5th row and j5th column in L y,33 Because L y The submatrix consisting of the elements in the i6th row and j6th column of Q y,1 For Q y The submatrix consisting of the elements in the i7th row and j7th column of Q y,2 For Q y The submatrix consisting of the elements in the i8th row and j8th column of Q y,3 For Q y The submatrix consisting of the elements in the i9th row and j9th column of ; Using the stable image expression of the closed-loop control system, its data-driven implementation form is obtained as follows: in, A data-driven form of image expression for closed-loop control systems; and Respectively represent the relationship between process input and output signals and external input signals; Using equations (1), (2) and (3) to realize the data-driven identification of the stable image expression of the closed-loop control system, the data-driven realization form of the stable image expression of the closed-loop control system is obtained as follows: Decomposing the data-driven implementation form of the closed-loop control system stability image expression, we get: in, Reason The i 10 Row j 10 The submatrix consisting of column elements, i 10 =1,...,2s+2,j 10 =1,...,s+1;I d,f Reason The i 11 Row j 11 The submatrix consisting of column elements, i 11 =1,...,2s+2,j 10 =s+1,...,2s+2; Pair Matrix Perform singular value decomposition and get Among them, [U1 U2] is the left singular vector, ∑1 is the diagonal matrix, is a right singular vector; let The standardized data-driven implementation of the stable image expression is as follows: When new process data is collected, the iterative form of LQ decomposition in equations (1) and (2) is obtained as follows: Among them, L u,new , Q u,new , L y,new and Q y,new L u , Q u , L y and Q y The updated value of It is a column vector composed of the data of the external input signal at the past moment; u is a column vector composed of the past data of the external input signal; f (k) is the column vector composed of the future data of the process input signal; y f (k) is a column vector composed of the future data of the process output signal; Reason and u f (k) is a column vector composed of; Reason and f (k) is a column vector composed of; Using the Givens rotation technique, we get L u,new and L y,new The iterative update form is as follows: Among them, Q u,givens and Q y,givens is the Givens matrix; The standardized data-driven implementation form of online updating of stable image expression using equations (4), (5) and (6) Using process data collected offline, calculate the nominal and Using the process data collected online, calculate the current and Then calculate the closed-loop control system G under nominal conditions and the closed-loop control system G at the current moment k Gap distance: Among them, σ max (·) are the singular values ​​of the matrix; Set the following online evaluation function: Based on the online evaluation function, the performance degradation detection logic of the closed-loop control system is designed: Among them, J th,0 is the first preset threshold.

2. The closed-loop control system performance degradation detection and recovery method according to claim 1, characterized in that: The construction method of the feedforward controller based on the closed-loop subspace prediction function is: using the data-driven form of the closed-loop control system stable image expression to realize the design of the feedforward controller based on the closed-loop subspace prediction function.

3. The closed-loop control system performance degradation detection and recovery method according to claim 2, characterized in that: The data-driven form of the stable image expression using the closed-loop control system to realize the design of a feedforward controller based on a closed-loop subspace prediction function includes: Based on the relationship between external input data and process output data, the prediction matrix of future process output data is defined as: in, is the subspace matrix related to the past moment data of the external input data; is the subspace matrix related to the future time data of the external input data; Using formula (14), we can get the prediction sequence of future process output signals: The external input data is expressed as the following integral form: in, is the external input data at time k-1; is the external input signal at time k The incremental value of The external input data sequence of the future s prediction interval and its increment sequence And the process output data sequence for the future s prediction interval and its increment sequence Defined as: in, is the external input data at time k3, is the incremental value of the external input data at time k3, Output prediction data for the process at time k3, k3 = k, ..., k + s; T represents matrix transposition; Using equations (16), (17) and (18), we can obtain the incremental prediction output sequence related to the future process output data prediction sequence: in, is the incremental value of the external input data at time k4, k4 = ks-1, ..., k-1; The forecast sequence of future process output data is rewritten as: and k =[and T (k) and T (k) … and T (k)] T Among them, y(k) is the process output data at time k; For k y The identity matrix of order, k y The dimension of the process output signal; The feedforward control based on the closed-loop subspace prediction function is described as the following optimization problem: Among them, v f,s For future reference sequences, satisfy the following form: v f,s =[v T (k+1) … v T (k+s+1)] T (24) Wherein, v(k+1) is the reference signal data at time k+1; v(k+s+1) is the reference signal data at time k+s+1; and, T y and T v is a weight matrix that satisfies the following form: Among them, λ y,1 ,λ y,2 ,λ y,s+1 and These are element values ​​determined according to system performance requirements; Substituting equation (21) into equation (23), the optimization problem can be further written as: set up get: At each sampling moment, using The first element of Able to obtain the optimal external input signal value 4. The closed-loop control system performance degradation detection and recovery method according to claim 3, characterized in that: The observer-based state feedback controller is constructed in the following manner: a data-driven form of a closed-loop control system stabilizing the kernel expression is utilized to implement the observer-based state feedback controller design.

5. The closed-loop control system performance degradation detection and recovery method according to claim 4, characterized in that: The data-driven form of the closed-loop control system stabilizing the kernel expression is used to implement the observer-based state feedback controller design, including: Considering the stable nuclear expression of the closed-loop control system satisfy: Among them, u s (k)=[u T (ks) … u T (k)] T is a column vector of process input data, y s (k) = [y T (ks)…y T (k)] T A column vector consisting of process output data; Substituting formula (3) into formula (28), we get: in, A column vector consisting of external input data; Obviously, there is make If and only if the stable kernel expression and stable image expression of the closed-loop control system satisfy the following relationship: in, for The orthogonal complement of When the stable image expression of the closed-loop control system is calculated by formula (4), the stable kernel expression of the closed-loop control system satisfies the following relationship: and, Able to pass the matrix The singular value decomposition of is solved; Among them, Σ is a diagonal matrix; Choose the following equivalent vectors: Among them, [β s,0 β s,1 … β s,s ] and [α s,0 α s,1 … α s,s ] is a matrix Elements in a row vector; The following state observer is obtained: Among them, x z (k) is the state estimation signal at time k, x z (k+1) is the state estimation signal at time k+1, r(k) is the residual signal at time k, A z ,B z ,C z ,G z ,D z ,L z is the parameter matrix of the designed state observer and satisfies: C z =[0 … 0 1],G z =α s,s ,L z =-[α s,0 α s,1 … α s,s-1 ] T ; Rewrite formula (34) into the following form: in, Based on formula (35), the following state feedback controller is designed: Where u(z) is the z-transform of the process input signal; x z (z) is the z-transform of the state estimation signal; is the z-transform of the external input signal; F z is the state feedback gain to be designed, which is determined by the following quadratic cost function: in, and R u is a weight matrix that satisfies R u >0; The state feedback gain F of the state feedback controller z Determined as: Among them, the P matrix is ​​a positive solution of the Riccati equation, satisfying the following form 6. The closed-loop control system performance degradation detection and recovery method according to claim 1, characterized in that: The preset recovery strategy uses a feedforward controller based on a closed-loop subspace prediction function and a state feedback controller based on an observer to optimize the performance of the closed-loop control system and achieve performance recovery of the closed-loop control system, including: If J th,0 ≤J(k)≤J th,1 , the feedforward controller based on the closed-loop subspace prediction function is applied to the closed-loop control system to optimize the performance of the closed-loop control system and realize the performance recovery of the closed-loop control system; If J(k)>J th,1 , applying the observer-based state feedback controller to the closed-loop control system, optimizing the performance of the closed-loop control system, and realizing the performance recovery of the closed-loop control system; Among them, J th,1 is the second preset threshold, and J th,1 >J th,0 .

Images