A digital factory intelligent control system

Abstract

The present invention relates to the field of intelligent control, and specifically to a digital factory intelligent control system. The system includes a data acquisition module, a fan blade process test monitoring module, and a fan blade manufacturing process control module, which acquires fan blade data; analyzes the temperature data change characteristics of the fan blade data, and constructs a local cooling factor; constructs the overall icing change trend in combination with the temperature data distribution; constructs an ice-covering iteration influence coefficient; analyzes the correlation between temperature data and humidity data, and constructs a mutual influence coefficient; corrects the ice-covering thickness of each data point; and obtains process improvement control based on the corrected ice-covering thickness and fan blade data; thereby completing the control of the fan manufacturing process, effectively improving the manufacturing process accuracy of the fan blade, and ensuring the quality of the fan blade.

Description

Technical Field

[0001] The present invention relates to the field of intelligent control, and in particular to a digital factory intelligent control system. Background Art

[0002] In the context of the rapid development of science and technology, more and more factories are using digital means for intelligent control, which not only saves a lot of resources and manpower, but also improves production efficiency. Especially for wind farms located in remote areas, the implementation of intelligent control is of great significance.

[0003] Wind power generation has made a great contribution to the current energy structure, and its power supply is increasing year by year. With the increasing number of wind farms, its safety issues are also receiving more and more attention. In wind power scenarios, when the ambient temperature is low, ice will appear on the surface of the wind turbine blades, which will reduce the service life of the wind turbine, increase energy consumption, and even threaten the lives of surrounding residents. Therefore, it is urgent to improve the manufacturing process of wind turbine blades.

[0004] Generally, the traditional method of improving the manufacturing process of fan blades is to increase the thickness of the blades, use materials with higher material strength, optimize the surface morphology of the fan blades, etc. This improved process method is based on the study and analysis of the icing conditions, so as to understand the loads and influences that the fan blades bear under different icing conditions. However, the icing conditions on the surface of the fan blades will be affected by the actual icing conditions under conventional environmental monitoring. That is, although the ice layer is thicker during the temperature rise process, the ice layer on the surface of the fan blade is easy to remove; and when the temperature is relatively cold, although the ice layer is thinner, it is difficult to remove. The degree of icing under different external conditions is not the same for this ice layer thickness, and this error will cause the actual test value of the icing on the surface of the fan blade to be inaccurate, thereby causing certain errors in the process control during the fan blade manufacturing process. Summary of the invention

[0005] In order to solve the above technical problems, the purpose of the present invention is to provide a digital factory intelligent control system, and the technical solutions adopted are as follows:

[0006] The present invention proposes a digital factory intelligent control system, the system comprising:

[0007] Data acquisition module, which obtains fan blade data, including temperature data, humidity data, ice thickness data, fan speed data, and wind speed data;

[0008] The fan blade process test monitoring module groups the fan blade data; obtains the local cooling factor of each group of data according to the temperature data change characteristics of each group of fan blade data; obtains the overall icing change trend of each group of data according to the local cooling factor of each group of data combined with the temperature data distribution; obtains the icing iteration influence coefficient of each group of data according to the overall icing change trend of each group of data; analyzes the characteristics of the temperature data and humidity data of each group of data to obtain the temperature mark sequence and humidity mark sequence of each group of data; obtains the mutual influence coefficient of each group of data according to the temperature mark sequence and humidity mark sequence of each group of data; obtains the icing correction coefficient of each group of data according to the mutual influence coefficient of each group of data and the icing iteration influence coefficient; corrects the ice thickness of each data point according to the icing correction coefficient of each group of data;

[0009] The fan blade manufacturing process control module obtains PID controller parameters according to the corrected ice thickness and fan blade data; and completes the control of the fan manufacturing process according to the PID controller parameters.

[0010] Preferably, the local cooling factor of each group of data is obtained according to the temperature data variation characteristics of each group of fan blade data, specifically:

[0011] For each group of data, the temperature difference between each data point and the next data point is calculated; the minimum temperature difference is obtained; and the inverse of the sum of all the temperature differences in the group and the sum of the minimum temperature difference is used as the local cooling factor for each group of data.

[0012] Preferably, the overall icing change trend of each group of data is obtained according to the local cooling factor of each group of data in combination with the temperature data distribution, specifically:

[0013] Calculate the mean and variance of the temperature data of each group of data; calculate the inverse of the product of the mean, the variance, and the local cooling factor of each group of data; use the inverse as the exponent of an exponential function with a natural constant as the base; and use the calculation result of the exponential function as the overall freezing change trend of each group of data.

[0014] Preferably, the icing iterative influence coefficient of each group of data is obtained according to the overall icing change trend of each group of data, specifically:

[0015] The overall icing variation trend of the first group is taken as the icing iteration influence coefficient of the first group of data;

[0016] For each group of data except the first group; the Hurst index of each group of data and the previous group of data is obtained; the difference between the Hurst index of each group of data and the previous group of data is used as the exponent of the exponential function; the sum of the overall icing change trend of each group of data and the icing iteration influence coefficient of the previous group of data is calculated; and half of the product of the sum and the exponential function is used as the icing iteration influence coefficient of each group of data.

[0017] Preferably, the characteristics of the temperature data and humidity data of each group of data are analyzed to obtain the temperature label sequence and humidity label sequence of each group of data, specifically:

[0018] Set the step value;

[0019] The average value of each step value of temperature data is used as the temperature mark value of the temperature data; all the temperature mark values ​​in the group are combined into a temperature mark sequence of each group of data;

[0020] The same method is used to obtain the humidity mark sequence.

[0021] Preferably, the mutual influence coefficient of each group of data is obtained according to the temperature mark sequence and the humidity mark sequence of each group of data, and the acquisition method is:

[0022] Calculate the difference sequence between 1 and each element of the normalized humidity mark sequence; and use the Pearson correlation coefficient between the difference sequence and each group of normalized temperature mark sequences as the mutual influence coefficient of each group.

[0023] Preferably, the ice correction coefficient of each group of data is obtained according to the mutual influence coefficient of each group of data and the ice iterative influence coefficient, and the obtaining method is:

[0024] Calculate the normalized value of the product of the ice iterative influence coefficient and the mutual influence coefficient of each group of data; The sum of is taken as the ice correction coefficient for each set of data, where is the default value.

[0025] Preferably, the step of correcting the ice thickness of each data point according to the ice correction coefficient of each group of data includes:

[0026] The product of the ice thickness data of each data point and the ice correction coefficient of the group of data to which each data point belongs is taken as the corrected ice thickness.

[0027] Preferably, the PID controller parameters are obtained according to the corrected ice thickness and the fan blade data, and the acquisition method is:

[0028] The corrected ice cover data and temperature, humidity, wind speed, and fan speed data at the same time are used as the input of the BP neural network, and the output of the BP neural network is the PID controller parameters.

[0029] Preferably, the control of the fan manufacturing process according to the PID controller parameters includes:

[0030] The PID controller parameters are used as the input of the PID control system, and the output of the PID control system is the process improvement control; the process improvement control includes: increasing the thickness of the blade, using materials with higher material strength, and optimizing the surface morphology of the blade.

[0031] The present invention has the following beneficial effects:

[0032] The present invention obtains the local cooling factor by analyzing the temperature change, and obtains the overall icing change trend by combining the distribution of the data within the group. Through the iterative process, the current data can be affected by the previous data, so as to accurately adjust the data. The ice thickness data is corrected by analyzing the relationship between the temperature and humidity data to achieve the accuracy of the ice layer test value. By inputting the actual test value into the PID control, the fan blade manufacturing process control is achieved, so that the fan can reduce the blade icing effect during the actual operation of the fan from the manufacturing process, improve the accuracy of the process control, and realize intelligent control of the process of the fan blade manufacturing factory. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0034] Figure 1 A block diagram of a digital factory intelligent control system provided by an embodiment of the present invention;

[0035] Figure 2 Flowchart for obtaining the icing correction factor. DETAILED DESCRIPTION

[0036] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the specific implementation, structure, features and effects of a digital factory intelligent control system proposed according to the present invention in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0037] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0038] The specific scheme of a digital factory intelligent control system provided by the present invention is described in detail below with reference to the accompanying drawings.

[0039] See also Figure 1 , which shows a block diagram of a digital factory intelligent control system provided by an embodiment of the present invention. The system includes: a data acquisition module 101, a fan blade process test monitoring module 102, and a fan blade manufacturing process control module 103.

[0040] The data acquisition module 101 acquires relevant data during the operation of the wind turbine blades and pre-processes the data.

[0041] Factors that affect the icing of fan blades mainly include temperature, humidity, fan speed, etc. This embodiment focuses on monitoring the icing of fan blades based on the two factors that have the greatest impact on the icing of fan blades, temperature and humidity.

[0042] By setting temperature, humidity and pressure sensors on the surface of each blade of the fan, a sensor system is formed to obtain temperature data, humidity data and ice thickness data during the operation of the fan. At the same time, the fan speed data and wind speed data are obtained through the system background. The above-obtained data are used to monitor the intelligent control method of blade deicing of digital factory fans.

[0043] The data obtained from the sensor system is data information within the same time interval at the same moment. In the data information, there may be missing data in the data obtained by the sensor due to bad weather or other abnormal conditions. This embodiment uses the mean filling method to fill the missing data. The mean filling method is a well-known technology and will not be described in detail in this embodiment. Among them, the collection time interval is set to 5 minutes, and the number of collected data points is 1000. The implementer can adjust it according to the actual situation.

[0044] The fan blade process test monitoring module 102 analyzes the icing condition of the fan blade based on the temperature and humidity data, and constructs an icing correction coefficient to correct the measured icing thickness.

[0045] Most fans are installed in remote places with cold winters, harsh environments, and open areas, which can effectively utilize wind speed to operate the fans. However, it is precisely because of this characteristic that the fan surface is easily covered with ice, thus affecting the normal operation of the fan.

[0046] However, the ice thickness directly obtained by the pressure sensor is not accurate, that is, there are other objects on the surface of the fan blade, or the ice condition monitored by the pressure sensor on the surface of the fan blade is not accurate due to the push of the wind speed. At the same time, there are differences in the operation time of the de-icing system under the same ice thickness. That is, when the ice thickness is the same, the weather temperature is in a gradually rising state, and the ice layer may be gradually melting at this moment, resulting in inaccurate process control parameters of the fan blade, and the fan blade manufacturing process cannot be intelligently controlled according to the actual test values.

[0047] This embodiment collects information to obtain the types of ice on the fan blades, which are roughly divided into three categories. The various types of ice conditions are greatly affected by the ambient temperature, humidity, and wind speed, which will directly or indirectly reflect the ice conditions on the fan blade surface. Different types of ice conditions require deicing operations of different lengths of time, thereby achieving a digital intelligent control method.

[0048] This embodiment focuses on exploring the ice coverage on the surface of the fan blade from the two aspects of temperature and humidity. Because temperature and humidity data directly affect the ice coverage on the blade surface, and the impact on the fan ice coverage is more important, if the ice thickness data obtained from the pressure sensor is inaccurate, it can be assisted by external temperature and humidity data to obtain ice coverage data that is more in line with the actual situation, and more effectively realize the process control of fan blades.

[0049] In order to realize the intelligent control of the digitalization of fan blade manufacturing process and predict the icing phenomenon of fan blades, a suitable control strategy can be effectively made to reduce the possibility of icing of fan blades. Accordingly, this embodiment takes the change of temperature data based on icing as the basis, and explores the effect of the interaction between temperature data and humidity data on the blade icing phenomenon, so as to more accurately control the blade manufacturing system.

[0050] Icing on fan blades mostly occurs in cold weather, and is more likely to occur in areas with higher humidity than in dry deserts. Therefore, this embodiment analyzes the temperature and humidity in the air to explore the possibility of icing on fan blades from a fundamental level, so that the control of the fan blade process manufacturing system has a more accurate de-icing effect.

[0051] Since temperature and humidity have opposite effects on icing of wind turbine blades and the two have a great interaction, the lower the temperature and the higher the humidity, the more moisture in the air will condense into dew or ice, resulting in more severe icing on the surface of the wind turbine blades.

[0052] The data obtained above are grouped into M data points, where M is 50 and can be adjusted by the implementer. Regarding the effect of temperature on the icing of fan blades, when the temperature is lower than the dew point and the water vapor in the air is saturated, dew or freezing will form. At the same time, if the temperature drops suddenly from high temperature to low temperature, a certain amount of water or dew may accumulate on the surface of the blades under high temperature conditions. These water may quickly condense into ice, causing the blades to freeze. When the temperature rises from low temperature to high temperature, there is a small possibility that the icing phenomenon will become more serious. This situation will not be discussed in this embodiment. Based on this, regarding the effect of temperature on the icing of fan blades, each group of data is analyzed to construct a local cooling factor, which is expressed as:

[0053]

[0054]

[0055] In the formula, is the local cooling factor of the kth group of data, is the number of data in the group, is the ith temperature difference of the kth group of data, is the temperature difference sequence of the kth group of data, , They are the temperature values ​​of the i-th and i+1-th data points in the k-th group of data respectively.

[0056] When the temperature between adjacent data shows a decreasing trend, and there is a larger decreasing trend among them, this has a greater impact on the icing on the surface of the wind turbine blade, and the surface is more likely to freeze.

[0057] After analyzing from a local perspective, the overall change trend of the group also has great reference value. If the overall temperature of the group is low and the temperature change is large, combined with the temperature difference characteristics between the local adjacent data points of the group, the overall icing change trend of the group can be obtained, the data in the group can be evaluated for the icing situation on the surface of the wind turbine blades, and the overall icing change trend can be constructed. The expression is:

[0058]

[0059] In the formula, The overall icing trend of the kth group of data, , are the temperature mean and temperature variance of the kth group of data, is an exponential function with a natural constant as base, is the local cooling factor of the kth group of data.

[0060] Temperature data plays a decisive role in freezing. When the mean temperature is greater than or equal to 0°C, the exponent of the exponential function is negative, and the overall freezing trend is less than 1, indicating that the freezing trend is very small at this temperature and there is almost no freezing. When the mean temperature is less than 0°C, the exponent of the exponential function is positive, and the greater the local cooling factor, the greater the possibility of freezing.

[0061] When the average value in this group of data is lower and the variance is larger, that is, the temperature change is larger under low temperature conditions, due to the characteristics of the temperature change itself, whether it is in the direction of the same decrease or chaotic change, it has the effect of faster freezing. Therefore, the overall freezing change trend of the group can be judged by combining the local and overall perspectives.

[0062] And for the icing phenomenon of wind turbine blades, if the temperature is lower, the tendency of icing is greater, and the longer this tendency lasts, the more serious the icing phenomenon of the blades will be. Tracing back to the changes in each group, if the latter group has a longer-term trend than the former group, and this long-term trend is a tendency towards icing, then the degree of icing in the latter group of data will be more serious during the iteration of the former group of data. Based on this, the iterative influence is completed by traversing the influence of the former group of data on each group of data. For the first group of data, set , for any subsequent set of data, the following formula is used for calculation:

[0063]

[0064] In the formula, is the ice iterative influence coefficient of the k+1th group of data, is the overall icing trend of the k+1th group of data, is the ice-covering iteration influence coefficient of the kth group of data, is an exponential function with the natural constant e as base, , are the Hurst exponents of the first k+1 groups and the first k groups of data, respectively. The Hurst algorithm is a well-known technology and will not be described in detail in this embodiment.

[0065] The larger the icing iteration influence coefficient of the current group, and the larger the Hurst index of the group than that of the group not including the group, the longer the historical trend of the data including the group is. This long-term historical trend indicates that the group is more likely to freeze than the previous group on the previous group, which originally has a greater possibility of icing. The larger the icing iteration influence coefficient of the group is. Through iteration, the subsequent icing effect is affected layer by layer on the basis of the previous group, and the actual effect of wind turbine blade icing is analyzed more comprehensively from the perspective of historical influence, so that the blade deicing system does not only stay at the control of the process manufacturing system at the current moment, but more accurately controls the deicing effect of the process manufacturing system from a realistic level.

[0066] Since humidity changes less than temperature, and the humidity's effect on the ice on the surface of the fan blades needs to be combined with the temperature at the time to work, that is, no matter how high the humidity is, ice will not occur due to high temperature, but ice will occur when the temperature is low and the humidity is high.

[0067] The temperature data and humidity data are analyzed for correlation, and the relationship between the corresponding groups of data is used to obtain the effect of humidity data on the icing on the surface of the wind turbine blades. For each group of data, the step size of each temperature data and humidity data is set to 5, and the mean of every five data is used as the tag value of the five data. All the tag values ​​in each group form the tag sequence of the group. According to the element relationship between the corresponding tag sequences of each group, the mutual influence coefficient of each group is obtained:

[0068]

[0069] In the formula, is the mutual influence coefficient of the kth group of data, is the Pearson correlation coefficient, is the temperature label sequence of the kth group of data, is the humidity label sequence of the kth group of data.

[0070] Because the relationship between temperature data and humidity data for the icing on the fan blade surface is negatively correlated, the relationship between the corresponding groups is taken as positive, and the Pearson correlation coefficient between the corresponding groups is calculated to obtain the mutual influence relationship between the temperature data and the humidity data groups. If the mutual influence relationship between the groups is greater, it means that the humidity data at this time has a greater effect on the temperature data, otherwise, its effect on the temperature data is smaller.

[0071] Based on the mutual influence relationship between the above temperature data and humidity data, the ice thickness data is corrected by combining the ice coverage iterative influence coefficient of each group of temperature data to obtain the corrected ice thickness data:

[0072]

[0073] In the formula, is the corrected ice thickness of the i-th data point in the k+1-th data set, is the ice-covering iteration influence coefficient of the kth group of data, is the mutual influence coefficient of the kth group of data, is the ice thickness data of the i-th data point in the k-th data set. is the icing correction coefficient, and the icing correction coefficient acquisition process is as follows: Figure 2 shown.

[0074] The fan blade manufacturing process control module 103 obtains PID controller parameters through a BP neural network, and controls the fan blade manufacturing process in combination with the PID controller parameters.

[0075] By calculating the temperature changes, the degree of important influence on the icing of wind turbine blades is obtained. At the same time, the correlation between humidity data and temperature data is used to correct this important influence, so as to obtain a more accurate icing influence relationship of each group, and the manufacturing process of wind turbine blades is controlled based on the corrected icing thickness.

[0076] Therefore, this embodiment inputs data such as temperature, humidity, corrected ice thickness, wind speed, and fan speed at the same time into the BP neural network, uses the SGD optimization algorithm to optimize the network training process, outputs PID controller parameters, inputs the parameters into the PID control system, and optimizes and adjusts the fan blade process manufacturing control parameters in combination with the PID controller to obtain process improvement control, wherein the process improvement control includes: increasing the blade thickness, using materials with higher material strength, and optimizing the blade surface morphology, so that the fan can achieve the reduction of the blade ice effect during the actual operation of the fan from the manufacturing process. Among them, the BP neural network is a well-known technology and will not be repeated in this embodiment.

[0077] In summary, according to the embodiment of the present invention, the local cooling factor is obtained by analyzing the temperature change, and the overall icing change trend is obtained by combining the distribution of the data in the group. Through the iterative process, the current data can be affected by the previous data, so as to accurately adjust the data. The ice thickness data is corrected by analyzing the relationship between the temperature and humidity data to achieve the accuracy of the ice layer test value. By inputting the actual test value into the PID control, the fan blade manufacturing process control is achieved, so that the fan can reduce the blade icing effect during the actual operation of the fan from the manufacturing process, improve the accuracy of the process control, and realize intelligent control of the process of the fan blade manufacturing factory.

[0078] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0079] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

[0080] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A digital factory intelligent control system, characterized in that: The system comprises: Data acquisition module, which obtains fan blade data, including temperature data, humidity data, ice thickness data, fan speed data, and wind speed data; The fan blade process test monitoring module groups the fan blade data; obtains the local cooling factor of each group of data according to the temperature data change characteristics of each group of fan blade data; obtains the overall icing change trend of each group of data according to the local cooling factor of each group of data combined with the temperature data distribution; obtains the icing iteration influence coefficient of each group of data according to the overall icing change trend of each group of data; analyzes the characteristics of the temperature data and humidity data of each group of data to obtain the temperature mark sequence and humidity mark sequence of each group of data; obtains the mutual influence coefficient of each group of data according to the temperature mark sequence and humidity mark sequence of each group of data; obtains the icing correction coefficient of each group of data according to the mutual influence coefficient of each group of data and the icing iteration influence coefficient; corrects the ice thickness of each data point according to the icing correction coefficient of each group of data; The fan blade manufacturing process control module obtains PID controller parameters according to the corrected ice thickness and fan blade data; and controls the fan manufacturing process according to the PID controller parameters; The method of obtaining the local cooling factor of each group of data according to the temperature data variation characteristics of each group of fan blade data is specifically: For each group of data; calculating the temperature difference between each data point and the next data point; obtaining the minimum temperature difference; taking the inverse of the sum of all the temperature differences in the group and the sum of the minimum temperature difference as the local cooling factor for each group of data; The overall icing change trend of each group of data is obtained based on the local cooling factor of each group of data combined with the temperature data distribution, specifically: Calculate the mean and variance of the temperature data of each group of data; calculate the inverse of the product of the mean, the variance, and the local cooling factor of each group of data; use the inverse as the exponent of an exponential function with a natural constant as the base; use the calculation result of the exponential function as the overall icing change trend of each group of data; The icing iteration influence coefficient of each group of data is obtained according to the overall icing change trend of each group of data, specifically: The overall icing variation trend of the first group is taken as the icing iteration influence coefficient of the first group of data; For each group of data except the first group; the Hurst index of each group of data and the previous group of data is obtained; the difference between the Hurst index of each group of data and the previous group of data is used as the exponent of the exponential function; the sum of the overall icing change trend of each group of data and the icing iteration influence coefficient of the previous group of data is calculated; and half of the product of the sum and the exponential function is used as the icing iteration influence coefficient of each group of data; The characteristics of the temperature data and humidity data of each group of data are analyzed to obtain the temperature label sequence and humidity label sequence of each group of data, specifically: Set the step value; The average value of each step value of temperature data is used as the temperature mark value of the temperature data; all the temperature mark values ​​in the group are combined into a temperature mark sequence of each group of data; The same method is used to obtain the humidity mark sequence; The mutual influence coefficient of each group of data is obtained according to the temperature mark sequence and humidity mark sequence of each group of data, and the acquisition method is: Calculate the difference sequence between 1 and each element of the normalized humidity mark sequence; use the Pearson correlation coefficient between the difference sequence and each group of normalized temperature mark sequences as the mutual influence coefficient of each group; The ice correction coefficient of each group of data is obtained according to the mutual influence coefficient of each group of data and the ice iterative influence coefficient, and the obtaining method is: Calculate the normalized value of the product of the ice iterative influence coefficient and the mutual influence coefficient of each group of data; The sum of is taken as the ice correction coefficient for each set of data, where is the default value.

2. A digital factory intelligent control system according to claim 1, characterized in that: The step of correcting the ice thickness of each data point according to the ice correction coefficient of each group of data includes: The product of the ice thickness data of each data point and the ice correction coefficient of the group of data to which each data point belongs is taken as the corrected ice thickness.

3. A digital factory intelligent control system according to claim 1, characterized in that: The PID controller parameters are obtained according to the corrected ice thickness and fan blade data, and the acquisition method is: The corrected ice thickness and temperature, humidity, wind speed, and fan speed data at the same time are used as the input of the BP neural network, and the output of the BP neural network is the PID controller parameters.

4. A digital factory intelligent control system according to claim 1, characterized in that: The control of the fan manufacturing process is completed according to the PID controller parameters, including: The PID controller parameters are used as the input of the PID control system, and the output of the PID control system is the process improvement control; the process improvement control includes: increasing the thickness of the blade, using materials with higher material strength, and optimizing the surface morphology of the blade.

Images