Transformer drying process control system and method based on artificial intelligence technology

Abstract

The application discloses a transformer drying process control system and method based on artificial intelligence technology, and relates to the technical field of power equipment manufacturing. By acquiring adjustable parameters and real-time data of the drying process, index parameters for evaluating drying efficiency and quality are extracted for weighted processing to determine a comprehensive evaluation factor of the drying process. The adjustable parameters and the comprehensive evaluation factor are subjected to correlation analysis, the correlation degree between the adjustable parameters and the comprehensive evaluation factor is calculated, and adjustable parameters within a set correlation degree threshold range are selected as strong correlation parameters of the comprehensive evaluation factor. By using historical data of the acquired comprehensive evaluation factor and strong correlation parameters, an adjustable parameter optimization model is constructed, the optimization model is used to acquire optimal values of each strong correlation parameter under different initial conditions, the acquired optimal values are used to set adjustable parameters in each stage of the drying process, and the drying process is subjected to adaptive control according to the set adjustable parameters.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power equipment manufacturing, and particularly relates to a transformer drying process control system and method based on artificial intelligence technology. BACKGROUND

[0002] Transformer drying processing is a crucial link in the process of power equipment manufacturing, and has a direct impact on the insulation performance, service life and operation safety of the transformer. The quality of the internal insulation system of the transformer is crucial, and is a key factor to ensure the safe and reliable operation of the transformer under high voltage and high charge. Drying processing is one of the basic links to ensure the quality of the insulation system in the manufacturing and maintenance process.

[0003] The traditional transformer drying process includes hot air circulation, vacuum drying and other methods, which can effectively remove the moisture content in the insulation material, but has the significant shortcomings of low efficiency, long drying time and high energy consumption. The VD-109 type closed cycle hot air vacuum drying equipment currently applied in actual production combines closed cycle hot air and vacuum drying technology, so that the drying efficiency of transformers with voltage grade below 35kV is improved, and high-temperature humid gas does not need to be discharged to the workshop during the drying process, improving the production environment. The transformer drying process of the equipment includes six stages of closed cycle hot air, preheating, step-by-step vacuum, intermittent vacuum, high vacuum and end point judgment.

[0004] However, the existing process for judging the drying end point relies on the fixed parameter settings of the equipment manufacturer, and lacks scientific and intelligent models to support real-time data analysis and dynamic judgment of the transformer. This leads to the following technical problems in actual drying operation: 1) low drying efficiency: the drying end point judgment process is tedious, and the complete drying process of a transformer usually requires repeated vacuum operation and temperature cycling, and long-term stage monitoring. The static operation mode judgment and repeated operation of the existing equipment make the drying process time-consuming, resulting in low production efficiency; 2) energy waste: during the drying process, long-time high-temperature heating and vacuum pumping operation consumes a large amount of energy, resulting in high power consumption. The static preset mode of the existing equipment cannot accurately control the parameter settings and end point time according to the specific transformer and environmental conditions, which may cause unnecessary extension of the drying time, resulting in energy waste and increasing the production cost of enterprises; 3) unstable drying quality: since the parameters of each stage are set based on past experience, there is a lack of real-time feedback control mechanism, so that the drying effect is easily affected by environmental conditions and the initial state of the transformer core, resulting in fluctuations in drying quality.

[0005] In summary, in order to realize dynamic modeling and optimization control of key parameters in the drying process, thereby improving the efficiency and precision of the transformer in the closed cycle hot air-transformer vacuum drying process, and reducing unnecessary energy consumption in the production process, we propose a transformer drying process control system and method based on artificial intelligence technology. SUMMARY

[0006] The main purpose of the present application is to provide a transformer drying process control system and method based on artificial intelligence technology, which can effectively solve the problems in the background art.

[0007] To achieve the above purpose, the technical solution adopted by the present application is,

[0008] The transformer drying process control method based on artificial intelligence technology comprises:

[0009] The drying process of the transformer equipment to be controlled is divided into n stages, and the jth adjustable parameter x ij of the ith stage is obtained. h Real-time data in the drying process are collected, index parameters for evaluating drying efficiency and quality are extracted for weighted processing, and a comprehensive evaluation factor Ce ij of the hth drying process is determined.

[0010] Correlation analysis is performed on the adjustable parameter x h and the comprehensive evaluation factor Ce ij , the correlation degree r min between the adjustable parameter and the comprehensive evaluation factor is calculated, the correlation degree threshold r ij is set, and the adjustable parameter r min ≥ r μ is selected as the strong correlation parameter of the comprehensive evaluation factor.

[0011] The historical data of the obtained comprehensive evaluation factor and strong correlation parameter are used to construct an adjustable parameter optimization model, wherein the expression of the optimization model is:

[0012]

[0013]

[0014] In the formula, A represents the adjustable space of the strong correlation parameter; represents the mapping relationship between the qth strong correlation parameter and the comprehensive evaluation factor when the initial state is s μ ; x q represents the qth strong correlation parameter; x qmin represents the adjustable lower limit of the qth strong correlation parameter; x qmax represents the adjustable upper limit of the qth strong correlation parameter; and S represents the space set of the initial state.

[0015] The optimal values of each of the strong correlation parameters under different initial conditions are obtained by using the optimization model, and the adjustable parameters in each stage of the drying process are set according to the obtained optimal values, and the drying process is adaptively controlled according to the set adjustable parameters.

[0016] The transformer drying process control system based on artificial intelligence technology comprises:

[0017] A data acquisition module is configured to divide the drying process of the transformer device to be controlled into n stages, and obtain the jth adjustable parameter x ij and real-time data in the drying process;

[0018] A data processing module is configured to extract index parameters for evaluating drying efficiency and quality for weighted processing, and determine a comprehensive evaluation factor Ce h of the hth drying process.

[0019] A data analysis module is configured to perform correlation analysis on the adjustable parameters x ij and the comprehensive evaluation factor Ce h , calculate a correlation degree r ij between the adjustable parameters and the comprehensive evaluation factor, set a correlation degree threshold r min , and screen the adjustable parameters with r ij ≥ r min as strong correlation parameters of the comprehensive evaluation factor.

[0020] A model construction module is configured to construct an adjustable parameter optimization model by using historical data of the obtained comprehensive evaluation factor and strong correlation parameters, and obtain optimal values of each of the strong correlation parameters under different initial conditions by using the optimization model.

[0021] A parameter setting module is configured to set the adjustable parameters in each stage of the drying process according to the obtained optimal values, and adaptively control the drying process according to the set adjustable parameters.

[0022] The system further comprises a memory, a processor, and an electronic program stored in the memory and capable of running on the processor.

[0023] Further, the calculation formula of the comprehensive evaluation factor Ce h is as follows:

[0024]

[0025] In the formula, d represents the distance value between all index parameters in the hth drying process and the optimal state thereof. Distance value between the kth index parameter and the worst state in the hth drying process;

[0026] wherein, is the weight of the kth index parameter, and z hk is the normalized value of the kth index parameter in the hth drying process; is the maximum value in the normalized value of the kth index parameter;

[0027] is the minimum value in the normalized value of the kth index parameter.

[0028] Further, the drying process is divided into a closed-loop hot air stage, a preheating stage, a step-by-step vacuum stage, an intermittent vacuum stage, a high vacuum stage, and an end point judgment stage.

[0029] The index parameters include power consumption, drying time, humidity

[0030] The adjustable parameters of the closed-loop hot air stage include a hot air temperature set value and a hot air end temperature set value.

[0031] The adjustable parameters of the preheating stage include a preheating vacuum set value, a target temperature set value, a vacuum waiting time set value, and a preheating end temperature set value.

[0032] The adjustable parameters of the step-by-step vacuum stage include four-stage vacuum set values and four-stage vacuum intermittent waiting time set values.

[0033] The adjustable parameters of the intermittent vacuum stage include a vacuum time set value, a Roots vacuum set value, a Roots temperature set value, and an intermittent end vacuum set value.

[0034] The adjustable parameter of the high vacuum stage includes a high vacuum time set value.

[0035] The adjustable parameter of the end point judgment stage includes an end point judgment number set value.

[0036] Further, the real-time data includes core initial state data, tank initial state data, and process monitoring data.

[0037] The core initial state data includes the initial humidity and temperature of the core.

[0038] The tank initial state data includes the initial temperature and air pressure in the tank.

[0039] Further, the correlation degree r ij The calculation process includes the following steps:

[0040] Step S21, using the comprehensive evaluation factor Ce of h drying processes h Construct data sequence E={Ce h}, with h adjustable parameters of the drying process Construct a data sequence F = {F1, F2, ..., F ij};F q ={x qh};

[0041] Step S22, using formula Standardize each data point in data series E and F;

[0042] Step S23, using formula Calculate the correlation coefficient between each data point in data sequence F and data sequence E Where ρ is the resolution coefficient, ranging from (0, 1], and ρ = 0.5; Δmin and Δmax are the minimum and maximum differences between the two levels of the data sequence, respectively;

[0043] Step S24, using formula Calculate the correlation between the adjustable parameter and the comprehensive evaluation factor r ij ,Q is the type of adjustable parameter.

[0044] The calculation formulas for the two-level minimum difference and the two-level maximum difference of the data sequence are:

[0045]

[0046]

[0047] in, Indicates first-take Calculate the minimum value in the result, and then take the minimum value of all minimum values;

[0048] Indicates first-take Calculate the maximum value among the results, and then take the maximum value of all the maximum values.

[0049] The present invention has the following beneficial effects:

[0050] Compared with the existing technology, the drying process of the transformer equipment to be controlled is divided into n stages, and the j-th adjustable parameter x in the i-th stage is obtained. ij , collect real-time data during the drying process, extract index parameters used to evaluate drying efficiency and quality, perform weighted processing, and determine the comprehensive evaluation factor Ce of the h-th drying process h , for the adjustable parameter x ij and the comprehensive evaluation factor Ce hCarrying out correlation analysis, calculating the correlation degree r between the adjustable parameters and the comprehensive evaluation factor ij , setting the correlation degree threshold r min , screening the adjustable parameters with r ij ≥r min as the strong correlation parameters of the comprehensive evaluation factor, using the historical data of the obtained comprehensive evaluation factor and strong correlation parameters to construct an adjustable parameter optimization model, using the optimization model to obtain the optimal values of each of the strong correlation parameters under different initial conditions, setting the adjustable parameters obtained in the drying process of each stage, and performing adaptive control on the drying process according to the set adjustable parameters, improving the efficiency and precision of the transformer in the closed cycle hot air-vacuum drying process, and reducing unnecessary energy consumption in the production process. BRIEF DESCRIPTION OF DRAWINGS

[0051] Figure 1 is a flowchart of the transformer drying process method based on artificial intelligence technology of the present application;

[0052] Figure 2 is a structural schematic diagram of the transformer drying process control system based on artificial intelligence technology of the present application. DETAILED DESCRIPTION

[0053] The present application will be further described below in conjunction with specific embodiments, wherein the drawings are only used for exemplary description, and the representations are only schematic diagrams, not physical drawings, and cannot be understood as limiting the present application. In order to better illustrate the specific embodiments of the present application, some components of the drawings will be omitted, enlarged or reduced, and do not represent the size of the actual product.

[0054] The specific implementation process of the technical scheme of the present application includes the following steps:

[0055] Step 1: The drying process of the transformer equipment to be controlled is divided into a closed cycle hot air stage, a preheating stage, a step-by-step vacuum stage, an intermittent vacuum stage, a high vacuum stage, and an end point judgment stage.

[0056] Specifically, the closed cycle hot air stage includes two sub-stages, which are used to improve the temperature of the iron core and condense the humidity. The adjustable parameters of this stage include hot air temperature setting and hot air end temperature setting.

[0057] The preheating stage is divided into two sub-stages, which are used to adjust and improve the vacuum degree by circulating air extraction and air charging, and to make the temperature of the iron core reach the set value. The adjustable parameters of this stage include preheating vacuum setting, target temperature setting, vacuum waiting time setting, and preheating end temperature setting.

[0058] Step-by-step vacuum stage: The vacuum level is gradually adjusted and increased until the core temperature and vacuum level meet the requirements of the next stage. The adjustable parameters in this stage include 4-level vacuum settings and the interval waiting time settings for the 4-level vacuuming.

[0059] Intermittent vacuum stage: Utilizing intermittent and relay vacuum pumping with a conventional vacuum pump and a Roots pump, the vacuum level inside the tank is further reduced. Adjustable parameters during this stage include pumping time, Roots vacuum start setting, Roots vacuum start temperature setting, and intermittent vacuum end setting.

[0060] High vacuum stage: Continuous vacuuming to achieve the high vacuum conditions required for the drying endpoint. The adjustable parameter in this stage is the high vacuum time setting.

[0061] End point determination stage: Intermittent vacuuming is performed, and the drying process is determined by the increase in vacuum level. The adjustable parameter in this stage is the number of endpoint determination settings.

[0062] Step 2: Get the j-th adjustable parameter x of the i-th stage ij , collect real-time data during the drying process; among them, the real-time data includes the initial state data of the iron core, the initial state data in the tank, and process monitoring data; the initial state data of the iron core includes the initial humidity and temperature of the iron core; the initial state data in the tank includes the initial temperature and air pressure in the tank.

[0063] Step 3: Extract the index parameters used to evaluate the drying efficiency and quality, including power consumption, drying time, and humidity, and perform weighted processing on the obtained index parameters to determine the comprehensive evaluation factor Ce of the h-th drying process h The calculation formula is:

[0064]

[0065] Where, It is expressed as the distance between all index parameters and their optimal state during the h-th drying process; It is expressed as the distance between all index parameters and their worst state in the h-th drying process;

[0066] in, is represented as the weight of the k-th indicator parameter, and z hk It is expressed as the normalized value of the kth index parameter in the hth drying process; It is expressed as the maximum value among the normalized values ​​of the k-th indicator parameter;

[0067] It is expressed as the minimum value among the normalized values ​​of the k-th indicator parameter.

[0068] Step 4: Adjust the parameter xij correlation analysis is performed to calculate the correlation degree r between the adjustable parameters and the comprehensive evaluation factor Ce h correlation analysis is performed to calculate the correlation degree r between the adjustable parameters and the comprehensive evaluation factor Ce ij The calculation process includes the following steps:

[0069] Step S41, the comprehensive evaluation factor Ce of the hth drying process is calculated according to the following formula: h A data sequence E = {Ce1, Ce2,..., Ce h} is constructed, in which Ce A data sequence F = {F1, F2,..., F ij} is constructed, in which F q = {x qh};

[0070] Step S42, the data points in the data sequences E and F are standardized according to the following formula:

[0071] Step S43, the correlation coefficient of each data point in the data sequence F and the data sequence E is calculated according to the following formula: In the formula, p is a resolution coefficient, and the value range is (0, 1], and p = 0.5 is taken; Δmin and Δmax are the two-level minimum difference and the two-level maximum difference of the data sequence respectively; the calculation formula is as follows:

[0072]

[0073]

[0074] In the formula, min (x) represents the minimum value of x, and max (x) represents the maximum value of x. min (x) represents the minimum value of x, and max (x) represents the maximum value of x. min (x) represents the minimum value of x, and max (x) represents the maximum value of x.

[0075] min (x) represents the minimum value of x, and max (x) represents the maximum value of x. min (x) represents the minimum value of x, and max (x) represents the maximum value of x.

[0076] Step S44, the correlation degree r between the adjustable parameters and the comprehensive evaluation factor is calculated according to the following formula: ij Q is the type of the adjustable parameter.

[0077] Step 5: set the correlation degree threshold r min , and screen the adjustable parameters with r ij ≥ r min as the strong correlation parameters of the comprehensive evaluation factor;

[0078] ​​​Step 6: using the obtained comprehensive evaluation factor and the historical data of the strong correlation parameter, a parameter-adjustable optimization model is constructed, wherein the expression of the optimization model is:

[0079]

[0080]

[0081] In the formula, A represents the adjustable space of the strong correlation parameter; represents the mapping relationship between the qth strong correlation parameter and the comprehensive evaluation factor when the initial state is s μ x q represents the qth strong correlation parameter; qmin represents the adjustable lower limit of the qth strong correlation parameter; qmax represents the adjustable upper limit of the qth strong correlation parameter; and S represents the space set of the initial state.

[0082] Step 7: using the optimization model to obtain the optimal value of each strong correlation parameter under different initial conditions, setting the adjustable parameters of each stage in the drying process according to the obtained optimal value, and performing self-adaptive control on the drying process according to the set adjustable parameters.

[0083] The basic principles and main features of the present application and the advantages of the present application are shown and described above. It should be understood by those skilled in the art that the present application is not limited by the above examples, and the above examples and descriptions in the specification are only to illustrate the principles of the present application. Without departing from the spirit and scope of the present application, various changes and improvements can be made to the present application, and these changes and improvements all fall within the scope of the present application. The scope of protection of the present application is defined by the appended claims and their equivalents.

Claims

1. A transformer drying process control method based on artificial intelligence technology, characterized in that: include: The drying process of the transformer equipment to be controlled is divided into n stages, and the jth adjustable parameter x in the i-th stage is obtained. ij , collect real-time data during the drying process, extract index parameters used to evaluate drying efficiency and quality, perform weighted processing, and determine the comprehensive evaluation factor Ce of the h-th drying process h ; For the adjustable parameter x ij and the comprehensive evaluation factor Ce h Perform correlation analysis to calculate the correlation degree r between the adjustable parameters and the comprehensive evaluation factors ij , set the correlation threshold r min , filter r ij ≥r min The adjustable parameter is used as a strong correlation parameter of the comprehensive evaluation factor; Using the acquired comprehensive evaluation factors and historical data of strongly correlated parameters, an adjustable parameter optimization model is constructed, wherein the expression of the optimization model is: Where A represents the adjustable space of strong correlation parameters; Indicated as the initial state s μ The mapping relationship between the strong correlation parameter of the q term and the comprehensive evaluation factor; q Expressed as the qth strong correlation parameter; x qmin It is expressed as the adjustable lower limit of the qth strong correlation parameter; x qmax It is represented as the adjustable upper limit of the qth strong correlation parameter; S is the spatial set of initial states; The optimization model is used to obtain the optimal values ​​of the strongly correlated parameters under different initial conditions, and the adjustable parameters of each stage in the drying process are set based on the obtained optimal values, and the drying process is adaptively controlled according to the set adjustable parameters.

2. The transformer drying process control method based on artificial intelligence technology according to claim 1 is characterized in that: Comprehensive evaluation factor Ce h The calculation formula is: Where, It is expressed as the distance between all index parameters and their optimal state during the h-th drying process; It is expressed as the distance between all index parameters and their worst state in the h-th drying process; in, is represented as the weight of the k-th indicator parameter, and z hk It is expressed as the normalized value of the kth index parameter in the hth drying process; It is expressed as the maximum value among the normalized values ​​of the k-th indicator parameter; It is expressed as the minimum value among the normalized values ​​of the k-th indicator parameter.

3. The transformer drying process control method based on artificial intelligence technology according to claim 1 is characterized in that: The drying process is divided into a closed-circulation hot air stage, a preheating stage, a step-by-step vacuum stage, an intermittent vacuum stage, a high vacuum stage, and an endpoint judgment stage; The index parameters include power consumption, drying time, humidity The adjustable parameters of the closed-cycle hot air stage include a hot air temperature setting value and a hot air end temperature setting value; The adjustable parameters of the preheating stage include a preheating vacuum setting value, a target temperature setting value, a vacuum waiting time setting value, and a preheating end temperature setting value; The adjustable parameters of the step-by-step vacuum stage include 4-level vacuum setting values ​​and 4-level vacuum intermittent waiting time setting values; The adjustable parameters of the intermittent vacuum stage include the evacuation time setting value, the Roots vacuum setting value, the Roots temperature setting value, and the intermittent vacuum end setting value; The adjustable parameters of the high vacuum stage include a high vacuum time setting value; The adjustable parameters of the endpoint judgment stage include a set value for the number of endpoint judgments.

4. The transformer drying process control method based on artificial intelligence technology according to claim 1 is characterized in that: The real-time data includes the initial state data of the core, the initial state data in the tank, and the process monitoring data; The iron core initial state data includes the initial humidity and temperature of the iron core; The initial state data in the tank include the initial temperature and pressure in the tank.

5. The transformer drying process control method based on artificial intelligence technology according to claim 1 is characterized in that: Correlation r ij The calculation process includes the following steps: Step S21, using the comprehensive evaluation factor Ce of h drying processes h Construct data sequence E={Ce h }, with h adjustable parameters of the drying process Construct a data sequence F = {F1, F2, ..., F ij };F q ={x qh }; Step S22, using formula Standardize each data point in data series E and F; Step S23, using formula Calculate the correlation coefficient between each data point in data sequence F and data sequence E Where ρ is the resolution coefficient, ranging from (0, 1], and ρ = 0.5; Δmin and Δmax are the minimum and maximum differences between the two levels of the data sequence, respectively; Step S24, using formula Calculate and obtain the correlation degree r between the adjustable parameter and the comprehensive evaluation factor ij ,Q is the type of adjustable parameter.

6. The transformer drying process control method based on artificial intelligence technology according to claim 5 is characterized in that: The calculation formulas for the two-level minimum difference and the two-level maximum difference of the data sequence are: in, Indicates first-take Calculate the minimum value in the result, and then take the minimum value of all minimum values; Indicates first-take Calculate the maximum value among the results, and then take the maximum value of all the maximum values.

7. The transformer drying process control system based on artificial intelligence technology is characterized by: The system is used to implement the steps of the transformer drying process control method based on artificial intelligence technology according to any one of claims 1 to 6, including: The data acquisition module is used to divide the drying process of the transformer equipment to be controlled into n stages and obtain the jth adjustable parameter x in the i-th stage. ij and real-time data during the drying process; The data processing module is used to extract the index parameters used to evaluate the drying efficiency and quality, perform weighted processing, and determine the comprehensive evaluation factor Ce of the h-th drying process. h ; A data analysis module is used to analyze the adjustable parameter x ij and the comprehensive evaluation factor Ce h Perform correlation analysis to calculate the correlation degree r between the adjustable parameters and the comprehensive evaluation factors ij , set the correlation threshold r min , filter r ij ≥r min The adjustable parameter is used as a strong correlation parameter of the comprehensive evaluation factor; A model building module is used to use the obtained comprehensive evaluation factors and historical data of strongly correlated parameters to build an adjustable parameter optimization model, and use the optimization model to obtain the optimal value of each of the strongly correlated parameters under different initial conditions; The parameter setting module is used to set the adjustable parameters at each stage of the drying process according to the obtained optimal values, and to perform adaptive control on the drying process according to the set adjustable parameters.

8. The transformer drying process control system based on artificial intelligence technology according to claim 7 is characterized in that: The system also includes a memory, a processor, and an electronic program stored in the memory and capable of running on the processor, wherein the processor can implement the steps of the transformer drying process control method based on artificial intelligence technology according to any one of claims 1 to 6 when running the electronic program.

Images