The present invention belongs to the field of product quality management and control technologies, and specifically to real-time prediction and regulation methods and systems of product quality based on a process dynamic pattern.
To vigorously develop industry, Germany has proposed “Industry 4.0” and China has proposed the strategy of “Made in China 2025”, clearly proposing to adhere to the unity of intelligent manufacturing and improving product quality. In the context of rapid global development and high-speed competition, product quality determines the life of products and is directly related to the comprehensive economic benefits of factories. High-level product quality has become the foundation of enterprises in foreign markets and the basis of development.
In actual production, real-time access to product quality helps production personnel to adjust process variables in time and better control production processes to ensure product quality. However, since product quality cannot be detected online, it is often necessary to predict product quality by modeling. Currently, modeling methods mainly include mechanistic modeling and data-driven modeling. A mechanistic model can reflect the dynamic characteristic relationship between product quality and process variables in industrial processes well and has strong interpretability. However, mechanistic modeling is costly and difficult, it is difficult to guarantee accuracy and reliability, and there are often problems such as low model accuracy and easy mismatch. Data-driven models do not need to focus on the inner mechanism of processes, have strong applicability, and can better reflect changes in product quality in actual processes. However, existing data-driven models are inadequate to portray the dynamic change characteristics of product quality and the stochastic characteristics of processes, are only applicable to open-loop systems, and cannot consider the influence of control inputs on changes in product quality.
Embodiments of the present invention provide real-time prediction and regulation methods and systems of product quality based on a process dynamic pattern, to resolve the problem in the prior art that existing data-driven models are inadequate to portray the dynamic change characteristics of product quality and the stochastic characteristics of processes, are only applicable to open-loop systems, and cannot consider the influence of control inputs on changes in product quality.
Embodiments of the present invention provide a real-time prediction method of product quality based on a process dynamic pattern. The method includes:
Preferably, the state space probability model of quality pattern dynamic motion equation constructed in step S1 is a linear time-invariant system or a nonlinear time-variant system.
Preferably, a method for constructing a state space probability model of quality mode dynamic motion equation in step S1 includes:
t
k
=At
k−1
+Bu
k−1
+w
k,
x
k
=Qt
k
+e
k, and
y
k
=Ct
k+εk,
where tk is the process pattern, xk is the process variable, yk is the product quality, uk is the control input, A, B, Q, and C are the model parameters requiring optimized learning, wk, ek, and εk denote noise, and k=0, 1, 2, . . . , N, denoting different sampling moments.
Preferably, a type of the noise wk, ek, and εk in the state space probability model of quality pattern dynamic motion equation is selected from independent and identically distributed Gaussian noise, Poisson noise, multiplicative noise, salt-and-pepper noise, and gamma distribution noise.
Preferably, the calculating a probability density function distribution of a quality pattern in step S2 specifically includes the following steps:
Preferably, a method of the optimized learning in step S3 is selected from an expectation maximum likelihood algorithm, a genetic algorithm, a particle swarm optimization algorithm, a simulated annealing algorithm, a greedy algorithm, and a neighborhood search algorithm.
Preferably, a method for performing online prediction on product quality indicators in step S4 is selected from a Kalman filter method, an iterative least squares method, a maximum a posteriori estimation method, a polynomial interpolation method, and a finite impulse response method.
Preferably, the performing online prediction on product quality indicators based on the optimized state space probability model of quality pattern dynamic motion equation in step S4 specifically includes the following steps:
p(tknew|x0:knew,u0:knew)=N(ĝk,Ĝk),
ĝ
k
=A
new
g
{circumflex over ( )}
+B
new
u
k−1
new
+{circumflex over (K)}[x
k
new
−Q
new(Anewgk−1{circumflex over ( )}+Bnewuk−1new)],
Ĝ
k=(I−{circumflex over (K)}Qnew)(AnewĜk−1AnewT+Φwnew),
{circumflex over (K)}=(AnewĜk−1AnewT+Φw)QnewT[Qnew(AnewGk−1AnewT+Φwnew)QnewT+Φenew]−1,
ĝ
0
=h
0
new
+H
0
new
Q
newT(QnewH0newQnewT+Φenew)−1(x0new−Qnewh0new), and
Ĝ
0
=H
0
new
−H
0
new
Q
newT(QnewH0newQnewT+Φenew)−1QnewH0new; and
y
k
new
=C
new
t
k
new,
where h0new is an optimized mean value of the Gaussian distribution, H0new is an optimized variance of the Gaussian distribution, tknew is an optimized process pattern, xknew is an optimized process variable, uknew is an optimized control input, Anew, Bnew, Qnew, and Cnew are the optimized model parameters, Φwnew, Φenew, and Φεnew, are optimized variances of Gaussian noises wk, ek, and εk, gk is a mean value of tknew, Gk is a variance of tknew, K is a filter gain, I is an identity matrix, and k=0, 1, 2, . . . , N, denoting different sampling moments.
The embodiments of the present invention further provide an online regulation method of product quality based on a process dynamic pattern. The method includes the real-time prediction method of product quality based on a process dynamic pattern, and performing online regulation on product quality indicators based on an optimized state space probability model of quality pattern dynamic motion equation.
Preferably, any one of pattern feedback control, PID control, fuzzy control, adaptive control, prediction control, and robust control is used in the online regulation method.
Preferably, the performing online regulation on product quality indicators based on an optimized state space probability model of quality pattern dynamic motion equation specifically includes the following steps:
u
k
new
=
t
k
new
+r(ykopt);
t
k
new=(Anew+Bnew)tk−1new+Bnewr(yk−1opt)+wknew; and
The embodiments of the present invention provide a real-time prediction system of product quality based on a process dynamic pattern. The system includes:
Compared with the prior art, the present invention has the following beneficial effects:
Embodiments of the present invention provide real-time prediction and regulation methods and systems of product quality based on a process dynamic pattern. In the present invention, by means of a Bayesian network, a quality pattern dynamic online prediction method under supervised learning is proposed from the perspective of probability, and a probability dynamic timing motion equation between a quality pattern and process variables and control inputs is established, so that online prediction of product quality can be implemented, and online regulation can be performed on the product quality by using an analytical relationship between the control input and the quality pattern, to ensure that the product quality remains at an optimal level.
To describe the technical solutions in the embodiments of the present invention or in the prior art more clearly, the accompanying drawings that need to be used in the embodiments are briefly described below. The features and advantages of the present invention will be more clearly understood by referring to the accompanying drawings, which are schematic and should not be construed as limiting the present invention in any way. A person of ordinary skill in the art can obtain other accompanying drawings without creative efforts based on these accompanying drawings. Where:
To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following clearly and completely describes the technical solutions in embodiments of the present invention with reference to the accompanying drawings in embodiments of the present invention. Apparently, the described embodiments are some rather than all of the embodiments of the present invention.
All other embodiments obtained by persons of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.
As shown in
The performing optimized learning on parameters of the state space probability model of quality pattern dynamic motion equation, to obtain optimal model parameters and an optimized state space probability model of quality pattern dynamic motion equation specifically includes the following steps.
The present invention constructs a probability dynamic timing motion equation between a quality pattern and process variables and control inputs, can implement online prediction of product quality, and can perform online regulation on the product quality by using an analytical relationship between the control input and the quality pattern, to ensure that the product quality remains at an optimal level.
Further, the state space probability model of quality pattern dynamic motion equation constructed in step S101 is a linear time-invariant system or a nonlinear time-variant system.
A method for constructing a state space probability model of quality pattern dynamic motion equation in step S101 is:
t
k
=At
k−1
+Bu
k−1
+w
k,
x
k
=Qt
k
+e
k, and
y
k
=Ct
k+εk,
A type of the noise wk, ek, and εk in the state space probability model of quality pattern dynamic motion equation is selected from independent and identically distributed Gaussian noise, Poisson noise, multiplicative noise, salt-and-pepper noise, and gamma distribution noise.
Further, the calculating a probability density function distribution of a quality pattern in step S102 specifically includes the following steps.
Further, a method of the optimized learning in step S103 is selected from an expectation maximum likelihood algorithm, a genetic algorithm, a particle swarm optimization algorithm, a simulated annealing algorithm, a greedy algorithm, and a neighborhood search algorithm.
The performing optimized learning on parameters of the state space probability model of quality pattern dynamic motion equation, to obtain optimal model parameters and an optimized state space probability model of quality pattern dynamic motion equation in step S103 specifically includes the following steps.
Further, it is designed that a method for performing real-time prediction on product quality indicators in step S104 is any one of a Kalman filter method, an iterative least squares method, a maximum a posteriori estimation method, a polynomial interpolation method, and a finite impulse response method.
The performing online prediction on product quality indicators based on the optimized state space probability model of quality pattern dynamic motion equation in step S104 specifically includes the following steps.
p(tknew|x0:knew,u0:knew)=N(ĝk,Ĝk),
ĝ
k
=A
new
g
{circumflex over ( )}
+B
new
u
k−1
new
+{circumflex over (K)}[x
k
new
−Q
new(Anewgk−1{circumflex over ( )}+Bnewuk−1new)],
Ĝ
k=(I−{circumflex over (K)}Qnew)(AnewĜk−1AnewT+Φwnew),
{circumflex over (K)}=(AnewĜk−1AnewT+Φw)QnewT[Qnew(AnewGk−1AnewT+Φwnew)QnewT+Φenew]−1,
ĝ
0
=h
0
new
+H
0
new
Q
newT(QnewH0newQnewT+Φenew)−1(x0new−Qnewh0new), and
Ĝ
0
=H
0
new
−H
0
new
Q
newT(QnewH0newQnewT+Φenew)−1QnewH0new; and
y
k
new
=C
new
t
k
new,
A diagram of prediction errors of product quality may be obtained through computer simulation, as shown in
The embodiments of the present invention provide an online regulation method of product quality based on a process dynamic pattern. The method includes the foregoing real-time prediction method of product quality based on a process dynamic pattern, and performing online regulation on product quality indicators based on an optimized state space probability model of quality pattern dynamic motion equation.
Further, any one of pattern feedback control, PID control, fuzzy control, adaptive control, prediction control, and robust control is used in the online regulation method.
Further, the performing online regulation on product quality indicators based on an optimized state space probability model of quality pattern dynamic motion equation specifically includes the following steps:
u
k
new
=
t
k
new
+r(ykopt);
t
k
new=(Anew+Bnew)tk−1new+Bnewr(yk−1opt)+wknew
A diagram of control errors of product quality may be obtained through computer simulation, as shown in
The embodiments of the present invention provide a real-time prediction system of product quality based on a process dynamic pattern. As shown in
The system is configured to implement the real-time prediction method of product quality based on a process dynamic pattern in Embodiment 1 above. To avoid redundancy, details are not described again herein.
It should be noted that, the above description only provides preferred embodiments of the present invention and the employed technical principles. It should be appreciated by those skilled in the art that the present invention is not limited to the particular embodiments described herein. Those skilled in the art may make various obvious changes, readjustments, and replacements without departing from the scope of protection of the present invention. Therefore, while the present invention is illustrated in detail in combination with the above embodiments, the present invention is not only limited to the above embodiments, and can further include more other equivalent embodiments without departing from the concept of the present invention.
The scope of the present invention is defined by the scope of the appended claims.
A person skilled in the art should understand that the embodiments of the present application may be provided as a method, a system or a computer program product.
Therefore, the present application may use a form of hardware only embodiments, software only embodiments, or embodiments with a combination of software and hardware. Moreover, the present application may use a form of a computer program product that is implemented on one or more computer-usable storage media (including but not limited to a disk memory, a CD-ROM, an optical memory, and the like) that include computer usable program code.
The present application is described with reference to the flowcharts and/or block diagrams of the method, the device (system), and the computer program product according to the embodiments of the present application. It should be understood that computer program instructions may be used to implement each process and/or each block in the flowcharts and/or the block diagrams and a combination of a process and/or a block in the flowcharts and/or the block diagrams. These computer program instructions may be provided for a general-purpose computer, a dedicated computer, an embedded processor, or a processor of any other programmable data processing device to generate a machine, so that the instructions executed by a computer or a processor of any other programmable data processing device generate an apparatus for implementing a specific function in one or more processes in the flowcharts and/or in one or more blocks in the block diagrams.
These computer program instructions may be stored in a computer readable memory that can instruct the computer or any other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory generate an artifact that includes an instruction apparatus. The instruction apparatus implements a specific function in one or more processes in the flowcharts and/or in one or more blocks in the block diagrams.
These computer program instructions may be loaded onto a computer or another programmable data processing device, so that a series of operations and steps are performed on the computer or the another programmable device, thereby generating computer-implemented processing. Therefore, the instructions executed on the computer or the another programmable device provide steps for implementing a specific function in one or more processes in the flowcharts and/or in one or more blocks in the block diagrams.
Obviously, the foregoing embodiments are merely examples for clear description, rather than a limitation to implementations. For a person of ordinary skill in the art, other changes or variations in different forms may also be made based on the foregoing description. All implementations cannot and do not need to be exhaustively listed herein. Obvious changes or variations that are derived there from still fall within the protection scope of the invention of the present invention.
Number | Date | Country | Kind |
---|---|---|---|
202210826869.6 | Jul 2022 | CN | national |