CN117891160B - Intelligent control system and method for switch cabinet - Google Patents

Abstract

The present invention relates to the field of data processing technology. Specifically, it relates to a switch cabinet intelligent control system and method. The method includes: real-time collection of the operation data of the intelligent switch cabinet, and setting a sliding window, each time the operation data is collected, the sliding window slides to the right once; the quality of the operation data at the center of the sliding window is calculated based on all the operation data in the sliding window, and the learning rate of the BP neural network model is adjusted according to the quality of the operation data at the center of the sliding window to optimize the BP neural network model; the PID algorithm is optimized using the optimized BP neural network model, and the operation data of the intelligent switch cabinet is controlled using the optimized PID algorithm. The method of the present invention can more accurately control the operation data of the switch cabinet, and can improve the adaptability and robustness of the switch cabinet intelligent control system.

Description

A switch cabinet intelligent control system and method

Technical Field

The present invention generally relates to the field of data processing technology. More specifically, the present invention relates to a switch cabinet intelligent control system and method.

Background technique

Switchgear is an electrical equipment and control system, which is mainly used in the power generation, transmission, distribution and power conversion process of the power system. As an important part of the power system, the accuracy and reliability of its current distribution and control directly affect the safe and stable operation of the power system. PID (proportional-integral-differential) control algorithm is the most widely used feedback control method. PID algorithm is simple and effective. It can adjust the 、/> and/> Three parameters are used to control the system output to achieve the desired control effect. However, when facing complex systems or nonlinear systems, traditional PID control algorithms often need to rely on experience and repeated trials to adjust parameters, which may be inefficient and imprecise in some cases. With the development of artificial intelligence technology, especially the application of neural networks, new possibilities are provided to solve the limitations of traditional PID control in complex systems. BP (back propagation) neural network, as a classic deep learning model, can optimize parameters by learning the complex relationship between input and output. Applying BP neural network to the optimization of PID control parameters, that is, learning and adjusting the PID controller through neural network/> 、/> and/> Three parameters can achieve more intelligent and precise control.

The problem with the existing BP neural network is that in the switch cabinet touch screen intelligent control scenario, there is a lot of electromagnetic interference in the surrounding environment, which will cause a lot of noise in the operating data collected by the sensor. These noises are also random and uncertain, which will affect the selection of the learning rate of the BP neural network in the back propagation. Because the quality of the input operating data is uncertain, the fixed learning rate may not be applicable to all current data, resulting in the obtained PID controller. 、/> and/> The inaccuracy of the three parameters is insufficient, which will reduce the accuracy of the operating data control, leading to unstable and unreliable control of the switch cabinet, and reducing the control accuracy and efficiency of the switch cabinet touch screen intelligent control system.

Summary of the invention

To solve one or more of the above technical problems, the present invention provides solutions in the following aspects.

In a first aspect, the present invention provides a switch cabinet intelligent control method, comprising:

Collect the operation data of the intelligent switch cabinet in real time. A sliding window is preset before the collection. Each time the operation data is collected, the sliding window slides to the right once.

The quality of the operating data at the center of the sliding window is calculated based on all the operating data in the sliding window, and the learning rate of the BP neural network model is adjusted based on the quality of the operating data at the center of the sliding window to optimize the BP neural network model, and the learning rate of the BP neural network model is inversely proportional to the quality of the operating data; the BP neural network model is used to optimize the parameters of the PID algorithm; the PID algorithm is used to control the operating data of the intelligent switch cabinet; the quality of the operating data is used to characterize the closeness between the collected operating data and the true value of the operating data;

Using the optimized BP neural network model to adjust the proportional coefficient of the PID algorithm , integral adjustment coefficient/> and differential adjustment coefficient/> The three parameters are optimized, and the optimized PID algorithm is used to control the operation data of the intelligent switch cabinet.

In one embodiment, the quality of the running data of the sliding window center is calculated including:

Calculate the noise representation of the running data at the center of the sliding window;

The noise performance degree of the operating data at the center of the sliding window is corrected according to the variance of the absolute operating data within the time range corresponding to the sliding window, and the larger the variance, the smaller the corrected noise performance degree; the absolute operating data refers to the operating data setting value on the switch cabinet touch screen;

The quality of the running data at the center of the sliding window is calculated based on the corrected noise representation.

In one embodiment, calculating the noise performance of the running data at the center of the sliding window includes:

Calculate the noise performance degree of the operating data at the center of the sliding window in the sliding window according to all the operating data in the sliding window, and record it as the first noise performance degree of the operating data at the center of the sliding window;

Calculate the mean of all running data within the sliding window;

Remove the data point with the largest deviation from the mean in the sliding window, calculate the noise performance degree of the operating data at the center of the sliding window in the sliding window according to the remaining operating data in the sliding window, and record it as the second noise performance degree of the operating data at the center of the sliding window;

The noise performance degree of the operating data at the center of the sliding window is calculated according to the first noise performance degree and the second noise performance degree of the operating data at the center of the sliding window; the calculation expression is:

In the formula, Indicates the degree of noise in the running data at the center of the sliding window, /> represents the first noise performance level of the operating data at the center of the sliding window, /> Indicates the second noise level of the operating data at the center of the sliding window.

In one embodiment, calculating the first noise performance level of the operating data at the center of the sliding window includes:

Fit all data points in the sliding window to obtain a fitting curve;

The first noise performance degree of the operating data at the center of the sliding window is calculated based on the fitting curve and the values of all data points in the sliding window; the calculation expression is:

In the formula, represents the first noise performance level of the operating data at the center of the sliding window, /> Indicates the number of data points in the sliding window, /> represents the integral of the jth group of adjacent data points on the fitting curve, /> represents the jth running data point in the sliding window,/> It represents the value of the jth running data point, /> Indicates the j+1th running data point in the sliding window,/> Indicates the value of the j+1th running data point in the sliding window, /> It represents the trapezoidal area of the original data corresponding to the jth group of adjacent data points/> Represents the loss of two adjacent data points during curve fitting within the sliding window.

In one embodiment, the modified noise performance degree calculation expression is as follows:

In the formula, With/> Respectively represent the i-th running data point/> The degree of noise performance before and after correction; /> Represents the variance of the absolute running data points within the time range corresponding to the sliding window.

In one embodiment, the quality calculation expression of the running data at the center of the sliding window is:

,

in Indicates the i-th running data point/> The quality of For the i-th running data point/> The corrected noise performance level.

In one embodiment, adjusting the learning rate of the BP neural network model includes:

The learning rate of the BP neural network model is calculated based on the quality of the running data at the center of the sliding window. The calculation expression is:

In the formula, Represents the learning rate of the BP neural network model,/> Indicates the quality of the running data at the center of the sliding window.

In one embodiment, the number of hidden layers of the BP neural network model is 3, the activation function adopts the ReLU function, the number of neurons in the output layer is 3, the back propagation process adopts the gradient descent algorithm, and the initial value of the learning rate is 0.5.

In a second aspect, the present invention provides a switch cabinet intelligent control system, comprising a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the switch cabinet intelligent control method of the present invention is implemented.

The technical effect of the present invention is as follows: the intelligent control method of the switch cabinet of the present invention adaptively adjusts the learning rate by judging the quality of the operating data points. A smaller learning rate should be set for the operating data points with large quality, aiming to ensure accurate model training results; a larger learning rate is set for the operating data points with small quality, and the learning rate is adaptively adjusted by analyzing the quality of the operating data used for BP neural network training. Compared with the fixed-size learning rate of the traditional BP neural network, it can ensure the search for the optimal solution while accelerating the convergence speed during the training process, thereby making the control of the operating data of the switch cabinet more accurate; furthermore, the BP neural network can learn nonlinear relationships, while the traditional PID controller is linear. Through the neural network, the complex nonlinear relationship between the current data and the PID parameters can be better captured, thereby improving the adaptability, robustness, stability and reliability of the intelligent control system of the switch cabinet; in addition, since there is no noise in the operating data setting value of the switch cabinet, the operating data setting value is the absolute operating data, and the collected operating data is the relative operating data. If the absolute operating data point does not change within the corresponding time range, it means that the noise performance of the relative operating data point at this time is absolutely reliable, because in this case, the cause of the change in the relative operating data is definitely caused by noise; on the contrary, if the absolute operating data point changes within the corresponding time range, it means that the change in the relative operating data may be caused by the change in the operating data setting value, and the reliability of its noise performance is lower. Therefore, according to the variance of the absolute operating data within the corresponding time range of the sliding window, the noise performance of the operating data at the center of the sliding window is corrected, which can make the calculated noise performance more accurate, thereby making the quality of the calculated operating data point more accurate, thereby further improving the accuracy, stability and reliability of the switch cabinet control.

BRIEF DESCRIPTION OF THE DRAWINGS

By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

FIG1 is a flow chart schematically showing a switch cabinet intelligent control method according to an embodiment of the present invention;

FIG2 is a flow chart schematically illustrating a method for calculating the quality of operating data of a sliding window center according to an embodiment of the present invention;

3 is a flow chart schematically illustrating a method for calculating the noise performance degree of operating data at the center of a sliding window according to an embodiment of the present invention;

4 is a flow chart schematically illustrating a method for calculating a first noise performance degree of operating data at the center of a sliding window according to an embodiment of the present invention;

FIG5 is a schematic diagram schematically showing a fitting error of an embodiment of the present invention;

FIG. 6 is a schematic diagram showing the structure of a switch cabinet intelligent control system according to an embodiment of the present invention.

Detailed ways

The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

Switch cabinet intelligent control method embodiment:

As shown in FIG1 , the switch cabinet intelligent control method of the present invention includes:

S101, collecting the operation data of the intelligent switch cabinet in real time, and a sliding window is preset before the collection, and each time the operation data is collected, the sliding window slides to the right once;

The collected operation data of the intelligent switch cabinet can be the operation parameters of the intelligent switch cabinet such as current, voltage or power. If the collected operation data is the current of the intelligent switch cabinet, a current sensor can be used; if the collected operation data is the voltage of the intelligent switch cabinet, a voltage sensor can be used. The collection frequency of the operation data of the intelligent switch cabinet is 10HZ.

The length of the sliding window is an odd number of data points. Preferably, in this embodiment, the length of the sliding window is set to 11 data points.

S102, optimizing the BP neural network model, specifically: calculating the quality of the operating data at the center of the sliding window according to all the operating data in the sliding window, and adjusting the learning rate of the BP neural network model according to the quality of the operating data at the center of the sliding window to optimize the BP neural network model, wherein the learning rate of the BP neural network model is inversely proportional to the quality of the operating data; the BP neural network model is used to optimize the parameters of the PID algorithm; the PID algorithm is used to control the operating data of the intelligent switch cabinet;

The input of the BP neural network model is the operation data collection value of the intelligent switch cabinet, the operation data setting value of the intelligent switch cabinet, and the proportional adjustment coefficient of the PID algorithm. Initial value, integral adjustment coefficient/> The initial value and differential adjustment coefficient of The number of hidden layers of the BP neural network model is 3, the activation function uses the ReLU function, the number of neurons in the output layer is 3, and the back propagation process uses the gradient descent algorithm. The initial value of the learning rate of the BP neural network model can be set to 0.5.

The quality of the collected operating data refers to the authenticity of the operating data, that is, the degree of closeness between the collected operating data and the actual operating data of the switch cabinet. Generally, the authenticity of the collected operating data is mainly affected by noise. The greater the noise, the worse the authenticity. Therefore, the quality of the collected operating data can be calculated based on the noise performance of the operating data.

Because the data acquisition device in the intelligent switch cabinet scenario will be subject to electromagnetic interference, resulting in noise in the collected operating data. At the same time, when training the BP neural network model, a slight change in the learning rate will cause the stability and accuracy of the model. Due to the different quality of the operating data, the uniform learning rate has different training accuracy for operating data of different quality, which affects the accuracy of the subsequently trained BP neural network model, resulting in the deterioration of the accuracy of the parameters of the output PID algorithm, and then the deterioration of the accuracy of the control of the switch cabinet operating parameters. This step analyzes the quality of the operating data to adaptively adjust the learning rate of the BP neural network model, thereby improving the accuracy of the BP neural network model training while ensuring the stability of the model.

S103, using the optimized BP neural network model to optimize the PID algorithm, and using the optimized PID algorithm to control the operation data of the intelligent switch cabinet, specifically: using the optimized BP neural network model to adjust the proportional coefficient of the PID algorithm , integral adjustment coefficient/> and differential adjustment coefficient/> The three parameters are optimized, and the optimized PID algorithm is used to control the operation data of the intelligent switch cabinet.

It should be noted that when the switch cabinet intelligent control method of the present invention is used to control the switch cabinet, before the PID algorithm is optimized for the first time, the PID algorithm before the first optimization is used to control the operating data of the intelligent switch cabinet, and before the BP neural network model is optimized for the first time, the BP neural network model before the first optimization is used to optimize the parameters of the PID algorithm.

Before the first optimization, the PID algorithm is set with a proportional adjustment coefficient , integral adjustment coefficient/> and differential adjustment coefficient/> The initial value of the proportional adjustment coefficient /> The initial value can be set to 0.2, the integral adjustment coefficient/> The initial value can be set to 0.01, the differential adjustment coefficient/> The initial value of can be set to 0.2.

The switch cabinet intelligent control method of the present invention adaptively adjusts the learning rate by judging the quality of the operating data points. A smaller learning rate should be set for operating data points with large quality, aiming to ensure accurate model training results; a larger learning rate is set for operating data points with small quality, and the learning rate is adaptively adjusted by analyzing the quality of the operating data used for BP neural network training. Compared with the fixed-size learning rate of the traditional BP neural network, it can ensure the search for the optimal solution while accelerating the convergence speed during the training process, thereby making the control of the switch cabinet operating data more precise; in addition, the BP neural network can learn nonlinear relationships, while the traditional PID controller is linear. Through the neural network, the complex nonlinear relationship between current data and PID parameters can be better captured, thereby improving the adaptability and robustness of the switch cabinet intelligent control system.

In one embodiment, as shown in FIG2 , the quality of the running data of the sliding window center is calculated including:

S201, calculating the noise performance level of the operating data at the center of the sliding window;

The degree of noise representation can be calculated based on the fitting loss generated by curve fitting all data points in the sliding window.

If the running data in the center of the sliding window is noise, then removing a data point that deviates most from the mean of all running data in the sliding window will inevitably greatly affect the value of the noise performance degree. On the contrary, if it is a normal data point, removing a data point will not cause a large change in the value of the above noise performance degree. Therefore, a more accurate noise performance degree can be calculated by the difference in the noise performance degree before and after removing the data. The larger the difference, the higher the noise performance degree.

S202. Correcting the noise performance level of the operating data at the center of the sliding window, specifically: correcting the noise performance level of the operating data at the center of the sliding window according to the variance of the absolute operating data within the time range corresponding to the sliding window, and the larger the variance, the smaller the corrected noise performance level; the absolute operating data refers to the operating data setting value on the switch cabinet touch screen.

Since the operating data of the switch cabinet can be set manually on the switch cabinet touch screen, and there is no noise in the operating data setting value, the operating data setting value can be recorded as absolute operating data, and the collected operating data can be recorded as relative operating data. Since the change of the operating data setting value itself will cause the change of the collected operating data, the credibility of the noise performance degree can be calculated based on the change of the absolute operating data of the relative operating data point within the corresponding time range of its sliding window, so as to make corrections to the noise performance degree of the relative operating data point. If the absolute operating data point does not change within the corresponding time range, it means that the noise performance degree of the relative operating data point at this time is absolutely credible, because in this case, the cause of the change in the relative operating data must be caused by noise; on the contrary, if the absolute operating data point changes within the corresponding time range, it means that the change in the relative operating data may be caused by the change in the operating data setting value, and the credibility of its noise performance degree is lower.

Therefore, the variance of the absolute operating data within the corresponding time range can characterize the credibility of the noise performance of the operating data in the center of the sliding window. The larger the variance, the more drastic the data change of the absolute operating data point in the local range, and the lower the credibility of the noise performance of the relative operating data point; conversely, the smaller the variance, the slower the data change of the absolute operating data point in the local range, and the higher the credibility of the noise performance of the relative operating data point.

S203: Calculate the quality of the operating data at the center of the sliding window according to the corrected noise performance level.

The higher the noise performance of a data point, the greater the possibility that the data point has noise. Therefore, the quality of the running data at the center of the sliding window is negatively correlated with the corrected noise performance.

It can be seen from the above embodiments that a relatively accurate noise performance degree can be calculated by removing the difference between the noise performance degrees before and after the data. As shown in FIG3 , in one embodiment, calculating the noise performance degree of the running data at the center of the sliding window includes:

S301, calculating a first noise performance degree of the operating data at the center of the sliding window, specifically: calculating the noise performance degree of the operating data at the center of the sliding window in the sliding window according to all the operating data in the sliding window, and recording it as the first noise performance degree of the operating data at the center of the sliding window;

Because noise is random, that is, the manifestation of noise in the data is diverse, which may be mutations of individual data points, irregular changes in frequency and amplitude in a local range, etc., but the changes in these data will make all data points fit on the fitting curve in the process of data fitting, that is, the loss in the fitting process is large. Therefore, the loss when fitting the curve for the data points in the sliding window can be used to characterize the degree of noise manifestation of the running data at the center of the sliding window in the sliding window. The greater the loss, the higher the degree of noise manifestation.

S302, calculating the mean of all running data in the sliding window;

S303, calculating the second noise performance degree of the operating data at the center of the sliding window, specifically: removing the data point with the largest deviation from the mean in the sliding window, calculating the noise performance degree of the operating data at the center of the sliding window in the sliding window based on the remaining operating data in the sliding window, and recording it as the second noise performance degree of the operating data at the center of the sliding window;

In order to improve the credibility of the calculation of the noise performance degree, the present invention removes the point in the sliding window that deviates most from the average value of the data points in the sliding window, and then calculates the noise performance degree again in the sliding window after removing the data point, and subsequently calculates the noise performance degree based on the difference in the noise performance degree before and after removing the data point.

S304, calculating the noise performance degree of the operating data at the center of the sliding window according to the first noise performance degree and the second noise performance degree of the operating data at the center of the sliding window; the calculation expression is:

(1)

In the formula, Indicates the degree of noise in the running data at the center of the sliding window, /> represents the first noise performance level of the operating data at the center of the sliding window, /> Indicates the second noise level of the operating data at the center of the sliding window.

It can be seen from the above embodiments that the larger the variance is, the smaller the noise performance degree after correction is. In one embodiment, the calculation expression of the corrected noise performance degree is as follows:

(2)

In the formula, With/> Respectively represent the i-th running data point/> The degree of noise performance before and after correction; exp() is an exponential function; /> Represents the variance of the absolute running data points within the time range corresponding to the sliding window.

It can be seen from the above embodiments that the quality of the operating data at the center of the sliding window is negatively correlated with the corrected noise performance level. In one embodiment, the quality calculation expression of the operating data at the center of the sliding window is:

(3)

in Indicates the i-th running data point/> The quality of For the i-th running data point/> The corrected noise performance level.

The first noise performance level and the second noise performance level of the operating data at the center of the sliding window are calculated in the same manner. As shown in FIG4 , in one embodiment, calculating the first noise performance level of the operating data at the center of the sliding window includes:

S401, fitting all data points in the sliding window to obtain a fitting curve;

The least square method can be used to fit all data points in the sliding window to obtain a fitting curve.

S402, calculating the first noise performance degree of the operating data at the center of the sliding window according to the fitting curve and the values of all data points in the sliding window; the calculation expression is:

(4)

In the formula, represents the first noise performance level of the operating data at the center of the sliding window, /> Indicates the number of data points in the sliding window, /> It represents the integral of the jth group of adjacent data points on the fitting curve. Its physical meaning represents the trapezoidal area of the curve and the abscissa axis corresponding to x=j, x=j+1. represents the jth running data point in the sliding window,/> It represents the value of the jth running data point, /> represents the j+1th running data point in the sliding window, Indicates the value of the j+1th running data point in the sliding window, /> It represents the trapezoidal area of the original data corresponding to the jth group of adjacent data points/> Represents the loss of two adjacent data points during curve fitting within the sliding window.

As shown in Figure 5, the loss in the curve fitting process can be represented by the difference between the integral between two adjacent points in the fitting curve and the trapezoidal area corresponding to the two adjacent points. In the figure, C and D are two adjacent points, and the trapezoidal area is the area of the figure ABCD, and the integral is the area of the surface trapezoid ABDE. The difference between the two can characterize the loss of the fitting process. Indicates the loss of two adjacent data points during curve fitting within the sliding window,/> It represents the average value of the fitting loss corresponding to all adjacent data points. This value can be used to characterize the degree of noise performance of the running data at the center of the sliding window within the sliding window.

In one embodiment, adjusting the learning rate of the BP neural network model includes:

The learning rate of the BP neural network model is calculated based on the quality of the running data at the center of the sliding window. The calculation expression is:

(5)

In the formula, Represents the learning rate of the BP neural network model,/> Indicates the quality of the running data at the center of the sliding window.

Using the above expression to calculate the learning of the BP neural network model can ensure that the learning rate is less than 0.5, thereby avoiding the divergence problem caused by excessive learning rate, reducing the substantial update of weights, and making the model more stable.

Switch cabinet intelligent control system implementation example:

The present invention also provides a switch cabinet intelligent control system. As shown in FIG6 , the switch cabinet intelligent control system includes a processor and a memory, the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the switch cabinet intelligent control method according to the first aspect of the present invention is implemented; the switch cabinet intelligent control system also includes a touch screen arranged on the outer surface of the switch cabinet, and the switch cabinet controlled by the switch cabinet intelligent control system of the present invention includes a tray arranged inside the cabinet, and a sliding mechanism is also arranged at the bottom of the tray, and the switch cabinet intelligent control system controls the connection to the sliding mechanism, and various electronic components are installed on the tray. After clicking a specific button on the touch screen, the switch cabinet intelligent control system controls the sliding mechanism to move, thereby automatically pulling the tray out of the cabinet.

The switch cabinet intelligent control system also includes other components familiar to those skilled in the art, such as a communication bus and a communication interface. The configuration and functions of these components are known in the art and will not be described in detail here.

In the present invention, the aforementioned memory may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, apparatus or device. For example, a computer-readable storage medium may be any appropriate magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory RRAM (Resistive Random Access Memory), a dynamic random access memory DRAM (Dynamic Random Access Memory), a static random access memory SRAM (Static Random-Access Memory), an enhanced dynamic random access memory EDRAM (Enhanced Dynamic Random Access Memory), a high-bandwidth memory HBM (High-Bandwidth Memory), a hybrid memory cube HMC (Hybrid Memory Cube), etc., or any other medium that can be used to store the required information and can be accessed by an application, a module, or both. Any such computer storage medium may be part of a device or accessible or connectable to a device. Any application or module described in the present invention may be implemented using computer-readable/executable instructions that may be stored or otherwise maintained by such a computer-readable medium.

In the description of this specification, "plurality" or "several" means at least two, such as two, three or more, etc., unless otherwise clearly and specifically defined.

Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

Claims (5)

1. A switch cabinet intelligent control method, characterized by comprising: Collect the operation data of the intelligent switch cabinet in real time. A sliding window is preset before the collection. Each time the operation data is collected, the sliding window slides to the right once. The quality of the operating data at the center of the sliding window is calculated based on all the operating data in the sliding window, and the learning rate of the BP neural network model is adjusted based on the quality of the operating data at the center of the sliding window to optimize the BP neural network model, and the learning rate of the BP neural network model is inversely proportional to the quality of the operating data; the BP neural network model is used to optimize the parameters of the PID algorithm; the PID algorithm is used to control the operating data of the intelligent switch cabinet; the quality of the operating data is used to characterize the degree of proximity between the collected operating data and the true value of the operating data; the calculation of the quality of the operating data at the center of the sliding window includes: calculating the noise performance degree of the operating data at the center of the sliding window; correcting the noise performance degree of the operating data at the center of the sliding window based on the variance of the absolute operating data within the corresponding time range of the sliding window, and the larger the variance, the smaller the corrected noise performance degree; the absolute operating data refers to the operating data setting value on the switch cabinet touch screen; calculating the quality of the operating data at the center of the sliding window based on the corrected noise performance degree; adjusting the learning rate of the BP neural network model includes: calculating the learning rate of the BP neural network model based on the quality of the operating data at the center of the sliding window, and its calculation expression is: ; In the formula, Represents the learning rate of the BP neural network model,/> Indicates the quality of the running data at the center of the sliding window; Using the optimized BP neural network model to adjust the proportional coefficient of the PID algorithm , integral adjustment coefficient/> and differential adjustment coefficient/> The three parameters are optimized, and the optimized PID algorithm is used to control the operation data of the intelligent switch cabinet; The noise performance degree of the operating data of the center of the sliding window is calculated as follows: Calculate the noise performance degree of the operating data at the center of the sliding window in the sliding window according to all the operating data in the sliding window, and record it as the first noise performance degree of the operating data at the center of the sliding window; Calculate the mean of all running data within the sliding window; Remove the data point with the largest deviation from the mean in the sliding window, calculate the noise performance degree of the operating data at the center of the sliding window in the sliding window according to the remaining operating data in the sliding window, and record it as the second noise performance degree of the operating data at the center of the sliding window; The noise performance degree of the operating data at the center of the sliding window is calculated according to the first noise performance degree and the second noise performance degree of the operating data at the center of the sliding window; the calculation expression is: ; In the formula, Indicates the degree of noise in the running data at the center of the sliding window, /> represents the first noise performance level of the operating data at the center of the sliding window, /> A second noise performance level of the operating data representing the center of the sliding window; Calculating a first noise representation level of the running data at the center of the sliding window includes: Fit all data points in the sliding window to obtain a fitting curve; The first noise performance degree of the operating data at the center of the sliding window is calculated based on the fitting curve and the values of all data points in the sliding window; the calculation expression is: ; In the formula, Indicates the number of data points in the sliding window, /> represents the integral of the jth group of adjacent data points on the fitting curve, /> represents the jth running data point in the sliding window,/> It represents the value of the jth running data point, /> Indicates the j+1th running data point in the sliding window, /> Indicates the value of the j+1th running data point in the sliding window,/> It represents the trapezoidal area of the original data corresponding to the jth group of adjacent data points. Represents the loss of two adjacent data points during curve fitting within the sliding window. 2. The switch cabinet intelligent control method according to claim 1, characterized in that the corrected noise performance degree calculation expression is as follows: ; In the formula, With/> Respectively represent the i-th running data point/> The degree of noise performance before and after correction; /> Represents the variance of the absolute running data points within the time range corresponding to the sliding window. 3. The switch cabinet intelligent control method according to claim 1, characterized in that the quality calculation expression of the operating data at the center of the sliding window is: ; in Indicates the i-th running data point/> The quality of For the i-th running data point/> The corrected noise performance level. 4. The switch cabinet intelligent control method according to any one of claims 1 to 3 is characterized in that the number of hidden layers of the BP neural network model is 3, the activation function adopts the ReLU function, the number of neurons in the output layer is 3, the back propagation process adopts the gradient descent algorithm, and the initial value of the learning rate is 0.5. 5. A switch cabinet intelligent control system, comprising a processor and a memory, wherein the memory stores computer program instructions, and wherein when the computer program instructions are executed by the processor, the switch cabinet intelligent control method according to any one of claims 1 to 4 is implemented.