CN113758503A - A process parameter estimation method, device, electronic device and storage medium - Google Patents

Abstract

Embodiments of the present application provide a process parameter estimation method, apparatus, electronic device, and storage medium. The method includes collecting a measurement value of a first variable according to a preset sampling interval; and obtaining a first parameter and a second parameter of the first variable according to the measurement value, wherein the first parameter is used to indicate that the noise signal has an impact on the first variable. The influence degree of the increment within the preset sampling interval, and the second parameter is used to indicate the degree of change of the measured value with time; according to the first parameter and the second parameter, the magnitude of the noise signal is determined; based on the magnitude of the noise signal, A posteriori estimated value of the second variable is determined; wherein, there is a functional relationship between the first variable and the second variable. Therefore, the solution of the present application can improve the estimation accuracy of the process parameters, thereby shortening the response time to reach the same accuracy.

Description

A process parameter estimation method, device, electronic device and storage medium

technical field

The present invention relates to the technical field of data processing, and in particular, to a process parameter estimation method, device, electronic device and storage medium.

Background technique

In the condition monitoring system, the process parameters reflect the working state of the entire system and are the basis and key to the normal operation of the system. Therefore, fast and accurate estimation of process parameters for condition monitoring is of great significance for safe production.

In the existing state monitoring process parameter estimation technology, the estimation accuracy and response time have high requirements on the accurate measurement of the initial stage signal. However, in some condition monitoring systems, the signal amplitude is small in the initial stage, which is greatly affected by noise, so that the process parameter estimation is greatly disturbed. That is to say, using a calculation method that relies on the signal in the initial stage to estimate the process parameters will cause the problem of poor stability of the calculation result when the signal amplitude is small in the initial stage, and the problem of long measurement response time and poor real-time performance when the signal amplitude is large in the initial stage.

It can be seen from the above that the process parameter estimation method in the prior art is greatly affected by system noise interference, so that the accuracy of the estimation result is poor, and the response time required to achieve higher accuracy is longer.

SUMMARY OF THE INVENTION

The embodiments of the present application provide a process parameter estimation method, apparatus, electronic device, and storage medium, so as to improve the estimation accuracy of the process parameters, thereby shortening the response time to reach the same accuracy.

In order to solve the above technical problems, this application is implemented as follows:

In a first aspect, an embodiment of the present application provides a method for estimating a process parameter, the method comprising:

collecting the measured value of the first variable according to the preset sampling interval;

Obtain a first parameter and a second parameter of the first variable according to the measured value, wherein the first parameter is used to represent the increase of the noise signal on the first variable within the preset sampling interval. The influence degree of the quantity, the second parameter is used to represent the degree of the change of the measured value with time;

determining the magnitude of the noise signal according to the first parameter and the second parameter;

determining a posteriori estimate of the second variable based on the magnitude of the noise signal;

Wherein, there is a functional relationship between the first variable and the second variable.

In a second aspect, an embodiment of the present application provides an apparatus for estimating a process parameter, and the apparatus includes:

a collection module, configured to collect the measured value of the first variable according to a preset sampling interval;

A parameter acquisition module, configured to acquire a first parameter and a second parameter of the first variable according to the measured value, wherein the first parameter is used to indicate that the noise signal has an effect on the first variable in the preset the influence degree of the increment within the sampling interval, the second parameter is used to represent the degree of the change of the measurement value with time;

a noise level determination module, configured to determine the magnitude of the noise signal according to the first parameter and the second parameter;

a first estimation module for determining a posteriori estimated value of the second variable based on the magnitude of the noise signal;

Wherein, there is a functional relationship between the first variable and the second variable.

In a third aspect, embodiments of the present application further provide an electronic device, including: a memory, a processor, and a computer program stored on the memory and executable on the processor, the computer program being processed by the processor The steps of the process parameter estimation method as described in the first aspect are implemented when the controller is executed.

In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the implementation is as described in the first aspect. The steps of the process parameter estimation method.

In the embodiment of the present application, the measured value of the first variable is collected according to the preset sampling interval, so as to obtain the first parameter and the second parameter of the first variable according to the measured value, wherein the first parameter is used to indicate that the noise signal affects the first variable. The degree of influence of the increment of a variable within the preset sampling interval, the second parameter is used to represent the degree of linear change of the measured value with time; and then the magnitude of the noise signal is determined according to the first parameter and the second parameter, And based on the magnitude of the noise signal, an a posteriori estimate of the second variable that is functionally related to the first variable is determined.

It can be seen that, in the embodiment of the present application, the magnitude of the noise signal is estimated according to the measured value of the first variable, so that the posterior estimation of the second variable having a functional relationship with the first variable is determined according to the magnitude of the noise signal. In this way, even if the system is unstable in the initial stage, the magnitude of the system noise can be accurately estimated, and then the process parameters can be more accurately estimated based on the estimated noise level (that is, after the second variable is more accurately determined. test estimate). Therefore, the embodiments of the present application can alleviate the influence of the system noise on the estimation of the process parameters, so as to achieve higher estimation accuracy, and further shorten the response time to achieve the same accuracy.

The above description is only an overview of the technical solution of the present application. In order to be able to understand the technical means of the present application more clearly, it can be implemented according to the content of the description, and in order to make the above-mentioned and other purposes, features and advantages of the present application more obvious and easy to understand , and the specific embodiments of the present application are listed below.

Description of drawings

In order to illustrate the technical solutions of the embodiments of the present application more clearly, the following briefly introduces the drawings that are used in the description of the embodiments of the present application. Obviously, the drawings in the following description are only some embodiments of the present application. , for those of ordinary skill in the art, other drawings can also be obtained based on these drawings without any creative effort.

1 is a flowchart of steps of a process parameter estimation method provided by an embodiment of the present application;

Fig. 2 is the contrast schematic diagram of the measured value of u _t and the measured value of lnu _t in the embodiment of the present application;

Fig. 3 is the schematic diagram of the deviation mean ratio Para and correlation coefficient ρ _{X, K} calculated in the embodiment of the present application;

4 is a schematic diagram of the comparison of the effects of using different algorithms to determine the posterior estimated value of T in the embodiment of the present application;

FIG. 5 is a structural block diagram of an apparatus for estimating a process parameter provided by an embodiment of the present application.

Detailed ways

The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the protection scope of the present application.

The load identification method for edge computing in this embodiment of the present application may run on a terminal device or a server. The terminal device may be a local terminal device. When the method runs on a server, it can be displayed for the cloud.

In an optional implementation manner, cloud display refers to an information display method based on cloud computing. In the operation mode of cloud display, the main body of the information processing program is separated from the main body of information screen presentation, the storage and operation of the display switching method are completed on the cloud display server, and the function of the cloud display client is to receive and send data. and the presentation of information screens. For example, the cloud display client can be a display device with a data transmission function close to the user side, such as a mobile terminal, a TV, a computer, a handheld computer, etc.; but the terminal device for information data processing is Cloud display server in the cloud. When browsing, the user operates the cloud display client to send operation instructions to the cloud display server, and the cloud display server displays information according to the operation instructions, encodes and compresses the data, returns to the cloud display client through the network, and finally performs the operation through the cloud display client. Decode and output the display content.

In another optional implementation manner, the terminal device may be a local terminal device. The local terminal device stores the application program and is used to present the application interface. The local terminal device is used for interacting with the user through a graphical user interface, that is, the conventional electronic device downloads, installs and runs the application program. The local terminal device may provide the graphical user interface to the user in various ways, for example, it may be rendered and displayed on the display screen of the terminal, or provided to the user through holographic projection. For example, the local terminal device may include a display screen for presenting a graphical user interface, the graphical user interface including an application screen, and a processor for running the application program, generating the graphical user interface, and controlling the graphical user interface The display of the interface on the display.

The process parameter estimation method provided by the embodiments of the present application will be described in detail below.

Referring to FIG. 1 , a flowchart of steps of a process parameter estimation method in an embodiment of the present application is shown, and the method includes the following steps 101 to 104 .

Step 101: Collect the measurement value of the first variable according to the preset sampling interval.

Wherein, the measured value of the first variable is a value containing system noise (ie, noise signal) of the system to be monitored. Therefore, the embodiment of the present application estimates the a posteriori estimated value of the second variable that has a functional relationship with the first variable according to the measured value of the first variable including the system noise.

Step 102: Acquire a first parameter and a second parameter of the first variable according to the measured value.

The first parameter is used to indicate the influence degree of the noise signal on the increment of the first variable within the preset sampling interval, and the second parameter is used to indicate the change of the measured value with time. degree.

Step 103: Determine the magnitude of the noise signal according to the first parameter and the second parameter.

In the embodiment of the present application, by calculating the first parameter and the second parameter of the first variable, the influence of the noise signal can be quantified, thereby providing a basis for the subsequent determination of the a posteriori estimated value of the second variable.

Step 104: Determine a posteriori estimated value of the second variable based on the magnitude of the noise signal.

Wherein, there is a functional relationship between the first variable and the second variable.

It can be seen from the above steps 101 to 104 that in the embodiment of the present application, the measurement value of the first variable can be collected according to the preset sampling interval, so as to obtain the first parameter and the second parameter of the first variable according to the measurement value, wherein the first parameter is the first variable. A parameter is used to indicate the degree of influence of the noise signal on the increment of the first variable within the preset sampling interval, and the second parameter is used to indicate the degree of linear variation of the measured value with time; and then according to the first parameter and the second parameter parameters, determine the magnitude of the noise signal, and based on the magnitude of the noise signal, determine an a posteriori estimate of the second variable that is functionally related to the first variable.

It can be seen that, in the embodiment of the present application, the magnitude of the noise signal is estimated according to the measured value of the first variable, so that the posterior estimation of the second variable having a functional relationship with the first variable is determined according to the magnitude of the noise signal. In this way, even if the system is unstable in the initial stage, the magnitude of the system noise can be accurately estimated, and then the process parameters can be more accurately estimated based on the estimated noise level (that is, after the second variable is more accurately determined. test estimate). Therefore, the embodiments of the present application can alleviate the influence of the system noise on the estimation of the process parameters, so as to achieve higher estimation accuracy, and further shorten the response time to achieve the same accuracy.

Optionally, before collecting the measured value of the first variable according to the preset sampling interval, the method further includes:

Obtain the monitoring and demand quantities of the system to be monitored;

Obtain a linear functional relationship expression between the monitored quantity and the to-be-determined quantity, wherein the dependent variable of the linear functional relationship expression includes the monitored quantity, and the coefficient of the independent variable of the linear functional relationship expression including said quantity to be requested;

The dependent variable of the linear functional relationship expression is determined as the first variable, and the coefficient of the independent variable of the linear functional relationship expression is determined as the second variable.

Among them, in a system to be monitored (such as a condition monitoring system), the accurate acquisition of the source signal (ie, the acquisition of the measured value of the monitored quantity) is the first step in the process parameter estimation, and is a crucial part. Usually, the monitoring quantity will show a linear trend with time, or a deformed form of linear change, such as exponential change. Also, the variable to be determined is included in the expression for the change of the monitoring amount over time. Among them, in order to simplify the process of determining the posterior estimated value of the quantity to be calculated according to the measured value of the monitored quantity (that is, to simplify the estimation process of the process parameters), in the embodiment of the present application, the functional relationship between the monitored quantity and the quantity to be calculated needs to be calculated. Converted to a linear form, and in the linear form of the functional relationship, the monitored quantity is part of the independent variable, and the quantity to be determined is part of the coefficient of the dependent variable.

It can be seen from this that, in the embodiments of the present application, there is a functional relationship between the monitoring quantity of the system to be monitored and the quantity to be determined, and the functional relationship may be a linear functional relationship or a nonlinear functional relationship. Among them, no matter whether the functional relationship between the monitored quantity of the system to be monitored and the quantity to be calculated is a linear functional relationship or a nonlinear functional relationship, the functional relationship between the monitored quantity and the quantity to be calculated can be converted into "monitoring quantity as a factor". part of the variable, and the quantity to be determined is part of the coefficient of the dependent variable" of the form.

In this way, in the transformed linear form of the functional relationship expression, the dependent variable including the monitored quantity is the first variable, and the coefficient including the quantity to be determined is the second variable. That is, in this case, the posterior estimated value of the second variable can be determined according to the measured value of the first variable, and then the posterior estimated value of the quantity to be calculated can be obtained according to the posterior estimated value of the second variable.

For example: V=2M*t+L, where V is the monitored quantity, M is the quantity to be determined, t is the time, and L is the known quantity, then V is the first variable and 2M is the second variable.

其中，u_t为监测量，T为待求量，t表示时间，u₀、N₀、N_t为已知量，则

可以转换为：

其中，

为等效噪声，则lnu_t为上述第一变量，

为上述第二变量。Or, for example:

Among them, u _t is the monitoring quantity, T is the quantity to be determined, t is the time, and u ₀ , N ₀ , and N _t are known quantities, then

can be converted to:

in,

is equivalent noise, then lnu _t is the first variable above,

is the second variable above.

Optionally, after the a posteriori estimated value of the second variable is determined based on the magnitude of the noise signal, the method further includes:

A posteriori estimated value of the to-be-determined quantity is determined according to the posteriori estimated value of the second variable.

待求量为T，则在确定出

的后验估计值后，可以得到T的后验估计值。Wherein, the second variable is the coefficient of the independent variable in the linear expression of the functional relationship conversion between the monitored quantity and the quantity to be calculated, and the coefficient of the independent variable includes the quantity to be calculated. Therefore, the posterior estimated value of the second variable is obtained. Afterwards, the posterior estimated value of the quantity to be calculated can be obtained further according to the posterior estimated value of the second variable. For example, in the preceding example, the second variable is

If the quantity to be sought is T, then after determining

After the posterior estimate of T, the posterior estimate of T can be obtained.

Optionally, obtaining the first parameter and the second parameter of the first variable according to the measured value includes:

Starting from the sampling start time of the first variable, the preset sliding window is moved according to the preset step size, and after each moving of the sliding window, according to the measurement value in the sliding window, calculate the the first parameter and the second parameter of the first variable;

Wherein, the preset step size includes a preset number of sampling points.

It can be seen from this that, in the embodiment of the present application, the preset sliding window is moved according to the preset step size, and the first parameter and the second parameter of the first variable are calculated once every time the sliding window is moved, so that according to this calculation The first parameter and the second parameter of , determine the magnitude of the noise signal of the current system to be monitored. That is, in the embodiment of the present application, each time the sliding window is moved, a noise estimation is performed, which realizes the dynamic estimation of the noise signal of the system to be monitored, so that the posterior estimation value of the second variable can be dynamically obtained according to the size of the noise. In this way, the accuracy of the posterior estimation value of the second variable is further improved.

Optionally, according to the measurement value in the sliding window, the process of calculating the first parameter of the first variable includes:

determining a differential signal of a discrete sequence, wherein the discrete sequence includes the measurements within the sliding window;

calculating the mean and standard deviation of the differential signal;

The ratio of the standard deviation to the mean is calculated to obtain the first parameter.

The signal difference can further reveal the influence of the noise signal in the measurement process on the measurement signal (ie, the first variable). For a linear signal, its mathematical basic form is a linear function. Therefore, the theoretical mathematical form of the differential signal is a constant, which reflects the growth rate of the measurement signal over time.

Assuming that the discrete sequence composed of the measurement values of the first variable in the sliding window is x(k), the expression of the differential signal is as follows:

Wherein, if the sampling interval is Δt, the differential signal is the increment of the linear signal (ie, the first variable) within the time of Δt. Generally, in the initial measurement stage, due to the great influence of noise, the fluctuation of the differential signal is also relatively large, which in turn has a great influence on the calculation of signal parameters.

In addition, after performing differential processing on the discrete sequence, the obtained differential signal reflects the influence degree of the noise signal on the first variable. Usually, the influence of the noise signal will gradually weaken with the passage of the measurement time. By calculating the first parameter and the second parameter of the first variable, the influence degree of the noise signal on the first variable can be adaptively quantified, which can be used for subsequent determination of the relationship with the first variable. The a posteriori estimate of the second variable that is functionally related to the first variable provides the basis.

In addition, the calculation process of the first parameter of the first variable is as follows:

标准差为

For example, the sliding window includes n sampling points, and the differential signal of a discrete sequence x(k) composed of measurement values in a sliding window is: y(k), where k is an integer from 0 to n-1. Then the average value of the differential signal is

The standard deviation is

反映了差分信号数值上的平均水平，而标准差σ_y则反映了差分信号在平均值附近的波动幅度，二者结合可以反映出噪声信号对所述第一变量在所述预设采样间隔时间内的增量的影响程度。因此，差分信号的平均值与标准差之比，可以作为上述第一参数。其中，差分信号的平均值与标准差之比也可以称为偏差均值比，记为Para。Among them, the average

It reflects the average level of the difference signal value, and the standard deviation σ _y reflects the fluctuation range of the difference signal near the average value. The combination of the two can reflect the effect of the noise signal on the first variable at the preset sampling interval. The degree of influence within the increment. Therefore, the ratio of the average value to the standard deviation of the differential signal can be used as the above-mentioned first parameter. Among them, the ratio of the mean value to the standard deviation of the differential signal can also be called the deviation mean ratio, denoted as Para.

In short, the above first parameter is recorded as the deviation-to-average ratio,

Optionally, according to the measurement value in the sliding window, the process of calculating the second parameter of the first variable includes:

A correlation coefficient between a discrete sequence and a sampling time of the measurement value is calculated, and the correlation coefficient is determined as the second parameter, wherein the discrete sequence includes the measurement value within the sliding window.

其中，Var(·)表示方差计算函数，Cov(·)表示协方差计算函数，其定义式为：Cov(X,K)＝E(XK)-E(X)E(K)。E(·)表示期望(即均值)计算函数。Among them, the correlation coefficient is a measure of the degree of linear correlation between the variables under study. For the discrete sequence x(k) and discrete measurement time point k, the correlation coefficient between the two reflects the degree of linear variation of discrete with time. The larger the absolute value of the correlation coefficient (that is, close to 1), the more linear the relationship between discrete sequence and time will be. . If the random variable X represents a discrete sequence and the random variable K represents a discrete time series, the expression of the correlation coefficient between the two is:

Among them, Var(·) represents a variance calculation function, and Cov(·) represents a covariance calculation function, and its definition formula is: Cov(X,K)=E(XK)-E(X)E(K). E(·) represents the expectation (ie mean) calculation function.

Therefore, the calculation process of the correlation coefficient of random variables X and K is organized as follows:

It can be seen from the above that by calculating the deviation mean ratio Para and the correlation coefficient ρ _X,K , an enhanced adaptive preprocessing can be performed for the discrete sequence to reduce the influence of the noise signal on the process parameter estimation in the initial stage.

Optionally, the determining the magnitude of the noise signal according to the first parameter and the second parameter includes:

The magnitude Q of the noise signal is calculated according to a first preset formula |Para*(1-ρ)|=Q, where Para represents the first parameter and ρ represents the second parameter.

Optionally, the determining a posteriori estimated value of the second variable based on the magnitude of the noise signal includes:

The magnitude of the noise parameter and the predetermined initial parameters of the Kalman equation are substituted into the predetermined Kalman filter equation to obtain a posteriori estimated value of the second variable at different sampling times.

Wherein, starting from the sampling start time of the first variable, the preset sliding window is moved according to the preset step size, and after each moving of the sliding window, according to the measurement value in the sliding window , when calculating the first parameter and the second parameter of the first variable, the magnitude of the noise signal at different sampling times (ie, Q(k) can be obtained according to the first parameter and the second parameter obtained each time , k is an integer from 0 to n-1), so that based on the magnitude of the noise signal at each moment and the initial parameters of the predetermined Kalman filter equation, substituted into the predetermined Kalman filter equation to obtain the second variable in Posterior estimates at different sampling instants.

In addition, the Kalman filter equation includes the following five equations:

B＝0，T_c为上述预设采用间隔时间，其中，该状态预测方程用于采用上一时刻的状态估计当前状态；State prediction equation:

B=0, T _c is the above-mentioned preset adoption interval, wherein, the state prediction equation is used to estimate the current state by using the state at the previous moment;

Q(k)表示第k个采样时刻的噪声信号的量级；其中，该均方误差预测方程用于利用上一时刻的误差估计当前状态误差；Mean squared error prediction equation:

Q(k) represents the magnitude of the noise signal at the kth sampling moment; wherein, the mean square error prediction equation is used to estimate the current state error by using the error at the previous moment;

H＝[1,0]，R(k)表示第k个采样时刻的传感器测量噪声的协方差矩阵，且R(k)为已知量；其中，该滤波增益方程用来衡量测量值的准确性；Filter gain equation:

H=[1,0], R(k) represents the covariance matrix of the sensor measurement noise at the kth sampling time, and R(k) is a known quantity; wherein, the filter gain equation is used to measure the accuracy of the measurement value sex;

y(k)＝Hx(k)+v(k),H＝[1,0],v(k)表示第k个采样时刻的传感器测量噪声的量级，且v(k)为已知量；其中，该滤波估计方程用以获取当前时刻状态变量；Filter estimation equation:

y(k)=Hx(k)+v(k), H=[1,0], v(k) represents the magnitude of the sensor measurement noise at the kth sampling time, and v(k) is a known quantity ; Wherein, the filter estimation equation is used to obtain the state variable at the current moment;

其中，该状态迭代方程用于根据当前状态变量更新误差项。State iteration equation:

Among them, the state iteration equation is used to update the error term according to the current state variable.

V(k)表示上述所述的第一变量，M_k表示上述所述的第二变量，上述

表示第k个采样时刻状态变量的先验估计，

表示第k个采样时刻状态变量的后验估计，

表示第k个采样时刻的后验估计的协方差矩阵，

表示第k个采样时刻的先验估计的协方差矩阵。Among them, the state variables in the above Kalman filter equation

V(k) represents the above-mentioned first variable, M _k represents the above-mentioned second variable, and the above-mentioned

represents the prior estimate of the state variable at the kth sampling time,

represents the posterior estimate of the state variable at the kth sampling time,

represents the covariance matrix of the posterior estimate at the kth sampling instant,

Represents the covariance matrix of the prior estimate at the kth sampling instant.

其中，

即为第一变量在第0个采样时刻的测量值，

取预先确定的值；并且，令

分别取预先确定的值，即预先确定卡尔曼滤波方程的初始参数：

For example, when k=0, that is, the 0th sampling time, let

in,

is the measured value of the first variable at the 0th sampling time,

take a predetermined value; and, let

Take the pre-determined values respectively, that is, pre-determine the initial parameters of the Kalman filter equation:

根据均方误差预测方程得到

根据滤波增益方程得到

根据滤波估计方程得到

其中，

y(1)＝Hx(1)+v(1)；根据状态迭代方程得到

Based on the above content, when k=1, that is, the first sampling time, it can be obtained according to the state prediction equation

According to the mean square error prediction equation, we get

According to the filter gain equation, we get

According to the filter estimation equation, we get

in,

y(1)=Hx(1)+v(1); obtained according to the state iteration equation

根据均方误差预测方程得到

根据滤波增益方程得到

根据滤波估计方程得到

其中，

y(2)＝Hx(2)+v(2)；根据状态迭代方程得到

When k=2, that is, the second sampling time, it can be obtained according to the state prediction equation

According to the mean square error prediction equation, we get

According to the filter gain equation, we get

According to the filter estimation equation, we get

in,

y(2)=Hx(2)+v(2); obtained according to the state iteration equation

其中，

表示上述第一变量的后验估计值，

表示上述第二变量的后验估计值，因此，最终可以得到第二变量在不同采样时刻的后验估计值。Similarly, at each subsequent sampling time, according to the above method, the Kalman filter equation is used to iterate, so as to obtain

in,

represents the posterior estimate of the first variable above,

represents the a posteriori estimated value of the second variable above, therefore, the posterior estimated value of the second variable at different sampling times can be finally obtained.

Optionally, the determining process of the Kalman filter equation includes:

determining a stochastic state space equation according to the functional relationship between the first variable and the second variable;

Based on the stochastic state space equation, the Kalman filter equation is determined.

In the first aspect, according to the functional relationship between the first variable and the second variable, the process of determining the stochastic state space equation is as follows:

The functional relationship between the first variable and the second variable can be converted into a linear form. For a linear signal, its continuous expression is: the first expression V=M·t+L, the variable to be determined (ie the second variable) is usually included in the parameter M reflecting the growth rate and the parameter L reflecting the bias condition.

令M_k+1＝M_k+o(M_k)，o(M_k)为远小于M_k的增量。其中，对于线性信号，前一采样时刻和下一采样时刻的斜率变化在很短时内很小，因此o(M_k)是一个很小的量，即o(M_k)表在物理意义上表示测量信号随时间变化的增长(或下降)速度基本是恒定的。Wherein, by discretizing the above-mentioned first expression, the second expression can be obtained: V(k+1)=V(k)+T _c ·M _k . Hypothetical state variable

Let _Mk+1 = _Mk +o( _Mk ), o( _Mk ) be an increment much smaller than _Mk . Among them, for a linear signal, the slope change between the previous sampling moment and the next sampling moment is very small in a very short time, so o(M _k ) is a very small quantity, that is, o(M _k ) is represented in the physical sense. Indicates that the rate of increase (or decrease) of the measured signal over time is substantially constant.

Let w ₁ (k)=o(M _k ), then the third expression x ₁ (k+1)=x ₁ (k)+T _c ·M _k and the fourth expression x ₂ (k+1) are obtained =x ₂ (k)+w ₁ (k). Among them, measurement noise is usually introduced in the measurement process, and x ₁ (k) is the real value without noise, then if v(k) represents the measurement noise, y(k) represents the real value superimposed noise, that is, the observation vector , then the fifth expression y(k)=x ₁ (k)+v(k) can be obtained.

其中，

where x ₂ (k)=M _k , therefore, the above third expression can be converted into the sixth expression: x ₁ (k+1)=x ₁ (k)+T _c ·x ₂ (k). Thus, from the sixth expression x ₁ (k+1)=x ₁ (k)+T _c ·x ₂ (k) and the fourth expression x ₂ (k+1)=x ₂ (k)+w ₁ (k) can get the seventh expression:

in,

In addition, according to the fifth expression y(k)=x ₁ (k)+v(k), the eighth expression: y(k)=[1 0]×x(k)+v(k) can be obtained.

其中，

So far, the stochastic state space equation is obtained

in,

In the second aspect, according to the stochastic state space equation, the process of determining the Kalman filter equation is as follows:

表示根据上一次迭代计算结果产生的估计值，称为先验估计，先验估计误差为

定义

表示根据当前计算结果产生的估计值，称为后验估计，后验估计误差为

After the stochastic state-space equations are established, linear recursive fitting requires the definition of prior and posterior estimates and their errors. Among them, the definition

Represents the estimated value generated according to the calculation result of the previous iteration, which is called a priori estimation, and the prior estimation error is

definition

Represents the estimated value generated according to the current calculation result, which is called a posteriori estimation, and the posterior estimation error is

则

定义后验估计误差的协方差矩阵为

则

The covariance matrix that defines the prior estimation error is

but

The covariance matrix that defines the posterior estimation error is

but

使得

最小。The objective function of the state space linear recursive fitting model is to find the optimal estimate of the system state vector at time k given the state observation vector y(k) at time k when the system structure is known.

make

minimum.

计算出k时刻(即第k个采样时刻)的先验估计

其中，在计算k时刻的先验估计(即

)时，应该使用经过修正的后验估计数值，即

又因为此处是进行线性递归拟合，所以可以得到

其中，前面得到的随机状态空间方程为：

因此，可以得到

进而可以确定

B＝0(表示没有外部控制因素干预)。First, it is necessary to

Calculate the prior estimate of the k time (ie the kth sampling time)

Among them, when calculating the prior estimate at time k (ie

), a revised posterior estimate should be used, i.e.

And because the linear recursive fitting is performed here, we can get

Among them, the stochastic state space equation obtained earlier is:

Therefore, it can be obtained

to determine

B=0 (indicates no external control factors intervene).

的递推公式中需要加入噪声干扰Q(k)，且Q(k)量级与abs(Para·(1-ρ_X,K))相同，其中abs(·)为绝对值函数。即有：

But due to the existence of w(k), the covariance matrix of the prior estimation error

Noise interference Q(k) needs to be added to the recursive formula of , and the magnitude of Q(k) is the same as abs(Para·(1-ρ _X,K )), where abs(·) is an absolute value function. That is:

计算出k时刻的观测向量的估计

即定义

其中，H＝[1 0]。Then, it can be estimated from the prior

Calculate the estimate of the observation vector at time k

i.e. definition

where H=[1 0].

与估计值

差值，以修正先验估计

得到后验估计

即

其中，y(k)＝Hx(k)+v(k)，

则可以得到

进而可以得到

进而得到

其中，v(k)为传感器测量噪声，量级一般由传感器制造商给出，I为单位矩阵。Again, the measured value can be calculated

with estimated value

difference, to correct the prior estimate

get a posterior estimate

which is

Among them, y(k)=Hx(k)+v(k),

then you can get

which can be obtained

to get

Among them, v(k) is the sensor measurement noise, the magnitude is generally given by the sensor manufacturer, and I is the identity matrix.

因此，可以得到

其中，R(k)为v(k)的协方差。while the aforementioned has been

Therefore, it can be obtained

where R(k) is the covariance of v(k).

最小，则令

得到：

从而将

代入

可以得到最优估计状态的

为：

Again, the model estimation principle is to make the optimal state estimation

minimum, then let

get:

thus will

substitute

The optimal estimated state can be obtained

for:

So far, the above five equations included in the Kalman filter equation have been derived.

In addition, it should be noted that the five equations included in the above-mentioned Kalman filter equation (which may also be referred to as the five core equations included in the enhanced adaptive recursive fitting algorithm) are divided into time information update equations and measurement parameter update equations. equation.

)和均方误差预测方程(即

)。测量信息参数更新方程包括上述所述的滤波增益方程(即

)和滤波估计方程(即

)，以及状态迭代方程(即

)。Specifically, the time information parameter update equation includes the above-mentioned state prediction equation (ie

) and the mean squared error prediction equation (ie

). The measurement information parameter update equation includes the filter gain equation described above (ie

) and the filtered estimation equation (i.e.

), and the state iteration equation (ie

).

To sum up, the specific implementation of the process parameter estimation method in the embodiment of the present application may be as follows:

其中，T是测量信号(即启动源信号u)的e倍增周期，其数值大小反映了该系统的运行状态。而在实际信号监测过程中，通常会引入测量噪声等干扰，而噪声分布一般服从高斯分布或者泊松分布，假设噪声为N，则测量信号的表达式为：

其中v₀，v_t为真值，u₀，u_t为测量值，N₀，N_t为噪声。化简测量信号的表达式则可得：

进而可以得到：

其中，

为等效噪声。For example, in a monitoring system, the variation law of the starting source signal u with time t presents an exponential form (ideally), that is,

Among them, T is the e multiplication period of the measurement signal (ie, the starting source signal u), and its numerical value reflects the operating state of the system. In the actual signal monitoring process, interference such as measurement noise is usually introduced, and the noise distribution generally obeys the Gaussian distribution or the Poisson distribution. Assuming that the noise is N, the expression of the measurement signal is:

Where v ₀ , v _t are true values, u ₀ , u _t are measured values, and N ₀ , N _t are noise. Simplifying the expression of the measurement signal can get:

And then you can get:

in,

is the equivalent noise.

可以得到lnu_t为本申请实施例中的第一变量，

为本申请实施例中的第二变量。It can be seen that, in the above monitoring system, the a posteriori estimated value of the quantity _T to be obtained can be obtained according to the monitoring quantity ut. And by the functional relationship between the monitored quantity and the quantity to be sought

It can be obtained that lnu _t is the first variable in the embodiment of the present application,

is the second variable in this embodiment of the present application.

Thus, the process of determining the a posteriori estimate of the quantity T to be sought may include steps L1 to L5 as described below.

Step L1: collect the measured value of _ut according to the preset sampling interval time, and then calculate the measured value of _lnut , wherein the collected measured value of _ut can be as shown in the first graph 201 and the third graph 203 in FIG. 2 . shown, the measured value of _lnut can be shown in the second graph 202 and the fourth graph 204 in FIG. 2;

Step L2: Calculate the differential signal of the discrete sequence composed of the measured values of lnu _t ;

Step L3: From the sampling start time, move the preset sliding window according to the preset step size, and after each moving of the sliding window, calculate according to the differential signal obtained from the measured value group of lnu _t in the sliding window. The first parameter (that is, the deviation mean ratio Para) and the second parameter (that is, the correlation coefficient ρ _X,K ), that is, the values of the deviation mean ratio Para and the correlation coefficient ρ _{X, K} at different sampling times are obtained, as shown in Figure 3. Then, the magnitude of the noise signal at different sampling moments can be obtained;

在不同采样时刻的后验估计值；Step L4: Substitute the magnitude of the noise parameter and the initial parameters of the predetermined Kalman equation into the predetermined Kalman filter equation to obtain

Posterior estimates at different sampling times;

在不同采样时刻的后验估计值，得到T在不同采样时刻的后验估计值，如图4中的第五图401所示。Step L5: According to

Posterior estimated values of T at different sampling times are obtained, as shown in the fifth graph 401 in FIG. 4 .

In addition, in order to verify the effectiveness of the process parameter estimation method provided by the embodiments of the present application, other signal filtering or curve fitting methods are used to determine the posterior estimation value of T of the aforementioned reactor monitoring system. These methods mainly include direct estimation methods, random Sampling consensus algorithm, least square adaptive filtering algorithm, maximum likelihood algorithm and Kalman filtering algorithm, etc. Specifically, the sixth graph 402 as shown in FIG. 4 represents the posterior estimated value of T determined by the direct estimation method, the seventh graph 403 represents the posterior estimated value of T determined by the random sampling consensus algorithm, and the eighth graph 404 Represents the posterior estimated value of T determined using the least squares adaptive filtering algorithm, the ninth graph 405 represents the posterior estimated value of T determined using the maximum likelihood algorithm, and the tenth graph 406 represents the T determined using the Kalman filtering algorithm The posterior estimate of .

Among them, in the simulation experiment, the calculation accuracy (error), the response time and the algorithm running time are used as the evaluation indicators to evaluate the performance of the calculation method. Among them, the calculation accuracy is reflected in the form of percent error, which is the basic requirement for the algorithm; and the response time is the time required when the output of the quantity to be calculated becomes stable. The specified calculation accuracy is 1.00% (in the form of percentile error), and the response time and running time results of the above six comparison algorithms are shown in Table 1. Among them, the calculation accuracy of the least square adaptive filtering algorithm cannot meet the specified requirements.

Table 1 Comparison of calculation accuracy and response time of various signal processing algorithms

It can be seen from the above Table 1 and FIG. 4 that the process parameter estimation method provided by the embodiment of the present application has the shortest response time to achieve the same calculation accuracy, and has the best effect.

5, which is a structural block diagram of a process parameter estimation apparatus provided by an embodiment of the present application, the process parameter estimation apparatus 500 may include the following modules:

a collection module 501, configured to collect the measured value of the first variable according to a preset sampling interval;

A parameter obtaining module 502, configured to obtain a first parameter and a second parameter of the first variable according to the measured value, wherein the first parameter is used to indicate that the noise signal has a Suppose the influence degree of the increment within the sampling interval, and the second parameter is used to represent the degree of the change of the measurement value with time;

a noise level determination module 503, configured to determine the magnitude of the noise signal according to the first parameter and the second parameter;

a first estimation module 504, configured to determine a posteriori estimated value of the second variable based on the magnitude of the noise signal;

Wherein, there is a functional relationship between the first variable and the second variable.

Optionally, the device further includes:

The variable acquisition module is used to acquire the monitoring quantity and the quantity to be demanded of the system to be monitored;

A relationship acquisition module for acquiring a linear functional relationship expression between the monitored quantity and the to-be-determined quantity, wherein the dependent variable of the linear functional relationship expression includes the monitored quantity, and the linear functional relationship expression The coefficient of the independent variable of the formula includes the quantity to be calculated;

A variable determination module, configured to determine the dependent variable of the linear functional relationship expression as the first variable, and determine the coefficient of the independent variable of the linear functional relationship expression as the second variable.

Optionally, the device further includes:

The second estimation module is configured to determine the a posteriori estimated value of the quantity to be calculated according to the posterior estimated value of the second variable.

Optionally, the parameter acquisition module is specifically used for:

Starting from the sampling start time of the first variable, the preset sliding window is moved according to the preset step size, and after each moving of the sliding window, according to the measurement value in the sliding window, calculate the the first parameter and the second parameter of the first variable;

Wherein, the preset step size includes a preset number of sampling points.

Optionally, when the parameter acquisition module calculates the first parameter of the first variable according to the measurement value in the sliding window, it is specifically used for:

determining a differential signal of a discrete sequence, wherein the discrete sequence includes the measurements within the sliding window;

calculating the mean and standard deviation of the differential signal;

The ratio of the standard deviation to the mean is calculated to obtain the first parameter.

Optionally, when the parameter acquisition module calculates the second parameter of the first variable according to the measurement value in the sliding window, it is specifically used for:

A correlation coefficient between a discrete sequence and a sampling time of the measurement value is calculated, and the correlation coefficient is determined as the second parameter, wherein the discrete sequence includes the measurement value within the sliding window.

Optionally, the noise level determination module is specifically used for:

The magnitude Q of the noise signal is calculated according to a first preset formula |Para*(1-ρ)|=Q, where Para represents the first parameter and ρ represents the second parameter.

Optionally, the first estimation module is specifically used for:

The magnitude of the noise parameter and the predetermined initial parameters of the Kalman equation are substituted into the predetermined Kalman filter equation to obtain a posteriori estimated value of the second variable at different sampling times.

Optionally, the device further includes:

a state space equation determination module, configured to determine a random state space equation according to the functional relationship between the first variable and the second variable;

A Kalman filter equation determination module, configured to determine the Kalman filter equation based on the stochastic state space equation.

It can be seen from the above that in the embodiment of the present application, the magnitude of the noise signal is estimated according to the measured value of the first variable, so that the posterior estimation of the second variable having a functional relationship with the first variable is determined according to the magnitude of the noise signal. In this way, even if the system is unstable in the initial stage, the magnitude of the system noise can be accurately estimated, and then the process parameters can be more accurately estimated based on the estimated noise level (that is, after the second variable is more accurately determined. test estimate). Therefore, the embodiments of the present application can alleviate the influence of the system noise on the estimation of the process parameters, so as to achieve higher estimation accuracy, and further shorten the response time to achieve the same accuracy.

As for the apparatus embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and reference may be made to the partial description of the method embodiment for related parts.

The embodiment of the present application also provides an electronic device, including:

One or more processors; and one or more machine-readable media having instructions stored thereon, which, when executed by the one or more processors, cause the electronic device to perform the methods described in the embodiments of the present application .

The embodiments of the present application further provide one or more machine-readable media on which instructions are stored, and when executed by one or more processors, cause the processors to execute the methods described in the embodiments of the present application.

The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments, and the same and similar parts between the various embodiments may be referred to each other.

Those skilled in the art should understand that the embodiments of the embodiments of the present application may be provided as methods, apparatuses, or computer program products. Accordingly, the embodiments of the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, embodiments of the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.

The embodiments of the present application 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 application. It will be understood that each process and/or block in the flowchart illustrations and/or block diagrams, and combinations of processes and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to the processor of a general purpose computer, special purpose computer, embedded processor or other programmable data processing terminal equipment to produce a machine that causes the instructions to be executed by the processor of the computer or other programmable data processing terminal equipment Means are created for implementing the functions specified in a flow or flows of the flowcharts and/or a block or blocks of the block diagrams.

These computer program instructions may also be stored in a computer readable memory capable of directing a computer or other programmable data processing terminal equipment to operate in a particular manner, such that the instructions stored in the computer readable memory result in an article of manufacture comprising instruction means, the The instruction means implement the functions specified in the flow or flow of the flowcharts and/or the block or blocks of the block diagrams.

These computer program instructions can also be loaded on a computer or other programmable data processing terminal equipment, so that a series of operational steps are performed on the computer or other programmable terminal equipment to produce a computer-implemented process, thereby executing on the computer or other programmable terminal equipment The instructions executed on the above provide steps for implementing the functions specified in the flowchart or blocks and/or the block or blocks of the block diagrams.

Although the preferred embodiments of the embodiments of the present application have been described, those skilled in the art may make additional changes and modifications to these embodiments once the basic inventive concepts are known. Therefore, the appended claims are intended to be construed to include the preferred embodiments as well as all changes and modifications that fall within the scope of the embodiments of the present application.

Finally, it should also be noted that in this document, relational terms such as first and second are used only to distinguish one entity or operation from another, and do not necessarily require or imply these entities or that there is any such actual relationship or sequence between operations. Moreover, the terms "comprising", "comprising" or any other variation thereof are intended to encompass non-exclusive inclusion, such that a process, method, article or terminal device comprising a list of elements includes not only those elements, but also a non-exclusive list of elements. other elements, or also include elements inherent to such a process, method, article or terminal equipment. Without further limitation, an element defined by the phrase "comprises a..." does not preclude the presence of additional identical elements in the process, method, article or terminal device comprising said element.

A method and device for displaying a cover image provided by the present application have been described above in detail. The principles and implementations of the present application are described with specific examples in this paper. The method of the application and its core idea; at the same time, for those skilled in the art, according to the idea of the application, there will be changes in the specific implementation and application scope. In summary, the content of this description should not be understood to limit this application.

Claims (12)

1. A method of process parameter estimation, the method comprising:

collecting a measured value of a first variable according to a preset sampling interval time;

acquiring a first parameter and a second parameter of the first variable according to the measured value, wherein the first parameter is used for representing the influence degree of a noise signal on the increment of the first variable in the preset sampling interval time, and the second parameter is used for representing the change degree of the measured value along with the time;

determining the magnitude of the noise signal according to the first parameter and the second parameter;

determining a posteriori estimate of a second variable based on the magnitude of the noise signal;

wherein the first variable has a functional relationship with the second variable.

2. The method of process parameter estimation according to claim 1, wherein prior to collecting the measured value of the first variable at the predetermined sampling interval, the method further comprises:

acquiring the monitoring quantity and the quantity to be calculated of a system to be monitored;

acquiring a linear function relation expression between the monitored quantity and the quantity to be solved, wherein a dependent variable of the linear function relation expression comprises the monitored quantity, and a coefficient of an independent variable of the linear function relation expression comprises the quantity to be solved;

and determining the dependent variable of the linear function relational expression as the first variable, and determining the coefficient of the independent variable of the linear function relational expression as the second variable.

3. The method of claim 2, wherein after determining the a posteriori estimate of the second variable based on the magnitude of the noise signal, the method further comprises:

and determining the posterior estimated value of the quantity to be solved according to the posterior estimated value of the second variable.

4. The method of claim 1, wherein said obtaining a first parameter and a second parameter of the first variable from the measured value comprises:

moving a preset sliding window according to a preset step length from the sampling start time of the first variable, and calculating the first parameter and the second parameter of the first variable according to the measured value in the sliding window after moving the sliding window once;

the preset step length comprises a preset number of sampling points.

5. The process parameter estimation method of claim 4, wherein the process of calculating the first parameter of the first variable from the measurements within the sliding window comprises:

determining a discrete sequence of differential signals, wherein the discrete sequence comprises the measurements within the sliding window;

calculating the mean and standard deviation of the differential signal;

and calculating the ratio of the standard deviation to the average value to obtain the first parameter.

6. The process parameter estimation method of claim 4, wherein the process of calculating the second parameter of the first variable from the measured values within the sliding window comprises:

calculating a correlation coefficient of a discrete sequence comprising the measurement values within the sliding window with a sampling time of the measurement values and determining the correlation coefficient as the second parameter.

7. The method of claim 1, wherein determining the magnitude of the noise signal based on the first parameter and the second parameter comprises:

the magnitude Q of the noise signal is calculated according to a first preset formula | Para (1- ρ) | ═ Q, where Para represents the first parameter and ρ represents the second parameter.

8. The process parameter estimation method of claim 1, wherein determining a posteriori estimates of a second variable based on the magnitude of the noise signal comprises:

and substituting the magnitude of the noise parameter and the predetermined initial parameter of the Kalman equation into the predetermined Kalman filtering equation to obtain the posterior estimation values of the second variable at different sampling moments.

9. The method of process parameter estimation according to claim 8, wherein the determining of the kalman filter equation comprises:

determining a random state space equation according to the functional relation between the first variable and the second variable;

and determining the Kalman filtering equation based on the random state space equation.

10. A process parameter estimation apparatus, the apparatus comprising:

the acquisition module is used for acquiring the measured value of the first variable according to the preset sampling interval time;

a parameter obtaining module, configured to obtain a first parameter and a second parameter of the first variable according to the measured value, where the first parameter is used to indicate a degree of influence of a noise signal on an increment of the first variable in the preset sampling interval time, and the second parameter is used to indicate a degree of change of the measured value with time;

a noise magnitude determination module, configured to determine a magnitude of the noise signal according to the first parameter and the second parameter;

a first estimation module for determining a posteriori estimate of a second variable based on the magnitude of the noise signal;

wherein the first variable has a functional relationship with the second variable.

11. An electronic device, comprising: memory, a processor and a computer program stored on the memory and executable on the processor, the computer program, when executed by the processor, implementing the steps of the process parameter estimation method according to any one of claims 1 to 9.

12. A computer-readable storage medium, characterized in that the computer-readable storage medium has stored thereon a computer program which, when being executed by a processor, carries out the steps of the process parameter estimation method according to one of claims 1 to 9.

Images