Electrical equipment intelligent control method based on deep learning

Abstract

The invention discloses an electrical equipment intelligent control method based on deep learning, and relates to the technical field of electrical equipment intelligent control, and the method comprises the steps: carrying out the operation simulation according to preset operation parameters, and obtaining simulation vibration characteristic data, according to the correlation influence trend characteristics between the historical vibration characteristic data of the current electrical equipment and the historical vibration energy consumption value, predicting a vibration energy consumption estimated value, and counting the same space area enclosed by the opposite surfaces of the adjacent electrical equipment; predicting the dust distribution area and diffusion area of the same space area according to the correlation influence trend characteristics among the historical diffusion distribution area, the historical dust distribution average area and the historical vibration characteristic data in the historical monitoring record data to which the current electrical equipment belongs, and predicting the dust distribution area and diffusion area of the same space area according to the vibration energy consumption pre-estimated value and the pre-processed diffusion distribution area; and obtaining an actual energy consumption predicted value, and performing energy-saving regulation and control on the preset operation parameters. According to the invention, the efficiency of electrical equipment energy-saving operation parameter management and control work can be improved.

Description

Technical Field

[0001] This application relates to the field of intelligent control technology for electrical equipment, and in particular to an intelligent control method for electrical equipment based on deep learning. Background Technology

[0002] Deep learning can dynamically adjust the motor speed based on load changes. Traditional motor control methods may not be able to adapt to small fluctuations in load in a timely manner, causing the motor to always be in a relatively high energy consumption state. However, intelligent control systems can respond quickly, enabling the motor to operate in the most energy-efficient way in different production stages, reducing power loss.

[0003] In traditional technologies, the adjustment of operating parameters for energy-saving management of electrical equipment often relies on manual inspections and data experience. This significantly increases material resources, reduces the efficiency of management, and is highly subjective, leading to substantial errors in data adjustment and unsatisfactory energy-saving effects. Furthermore, the interconnected synchronous or asynchronous vibrations of electrical equipment in factory workshops, which affect temperature rise, and dust accumulation which affects heat dissipation efficiency, resulting in additional energy consumption, are not further analyzed, further reducing the efficiency of energy-saving management and parameter adjustment for electrical equipment. Summary of the Invention

[0004] To overcome the shortcomings of the prior art, this application provides a deep learning-based intelligent control method for electrical equipment.

[0005] This application provides a deep learning-based intelligent control method for electrical equipment, the method comprising: Step S1: Obtain the preset operating parameters of the current electrical equipment in the current period, test the energy consumption prediction model established based on the historical operating parameters and historical energy consumption data of the current electrical equipment, obtain the predicted value of operating energy consumption, perform operation simulation based on the preset operating parameters, obtain simulated vibration characteristic data, and predict the vibration energy consumption estimate based on the correlation influence trend characteristics between the historical vibration characteristic data and historical vibration energy consumption values ​​of the current electrical equipment. Step S2: Based on the correlation trend between the historical temperature rise amplitude and historical vibration characteristic data of the current electrical equipment, predict the preprocessed temperature rise amplitude generated by arbitrarily selected main vibration characteristic data and auxiliary vibration characteristic data in the simulated vibration characteristic data. Statistically analyze the spatial area enclosed by the adjacent opposing surfaces of the main measuring equipment to which the main measuring equipment belongs and the auxiliary measuring equipment to which the auxiliary measuring equipment belongs. Based on the correlation trend between the historical diffusion distribution area, the historical average dust distribution area, and the historical vibration characteristic data in the historical monitoring record data of the current electrical equipment, predict the dust distribution area and diffusion area of ​​the spatial area to obtain the preprocessed diffusion distribution area. Step S3: Based on the pre-processed temperature rise amplitude, pre-processed diffusion distribution area, and vibration energy consumption estimate, obtain the actual energy consumption prediction value. Based on the operating energy consumption prediction value and the actual energy consumption prediction value, perform energy-saving control on the preset operating parameters and output the electrical equipment management results.

[0006] Preferably, the historical operating parameters of the current electrical equipment in a historical period and the historical energy consumption data generated by the historical operating parameters are obtained, and an energy consumption prediction model is established based on the historical operating parameters and historical energy consumption data; Obtain the preset operating parameters of the current electrical equipment in the current period, input the preset operating parameters into the energy consumption prediction model for testing, and obtain the first energy consumption prediction value; Based on the preset operating parameters, the current electrical equipment is simulated to obtain simulated vibration characteristic data; Obtain historical monitoring record data of the current electrical equipment, and extract historical vibration characteristic data and historical vibration energy consumption value caused by the influence of historical vibration characteristic data on temperature rise from the historical monitoring record data based on the simulated vibration characteristic data; Based on the correlation trend characteristics between the historical vibration characteristic data and the historical vibration energy consumption value, the vibration energy consumption correlation coefficient between the historical vibration characteristic data and the historical additional energy consumption value is extracted. Based on the vibration loss correlation coefficient and simulated vibration characteristic data, the estimated vibration energy consumption of the current electrical equipment is obtained.

[0007] Preferably, the historical temperature rise amplitude value affected by historical vibration characteristic data is extracted from historical monitoring record data, and the temperature rise correlation coefficient between the historical temperature rise amplitude value and historical vibration characteristic data is extracted based on the correlation influence trend characteristics between the historical temperature rise amplitude value and historical vibration characteristic data. Randomly select primary vibration characteristic data and secondary vibration characteristic data from the simulated vibration characteristic data, wherein the primary electrical equipment to which the primary vibration characteristic data belongs and the secondary electrical equipment to which the secondary vibration characteristic data belongs are in an adjacent position relationship; Based on the main vibration characteristic data and the temperature rise correlation coefficient, the temperature rise amplitude value 1 of the main measuring equipment is obtained. Based on the auxiliary vibration characteristic data and the temperature rise correlation coefficient, the temperature rise amplitude value 2 of the auxiliary measuring equipment is obtained. The average value of the temperature rise amplitude value 1 and the temperature rise amplitude value 2 is calculated to obtain the preprocessed temperature rise amplitude value.

[0008] Preferably, the common spatial area enclosed by the adjacent opposing surfaces of the main measuring device and the auxiliary measuring device is statistically analyzed, and the spatial area in the same location is output. Establish a reference coordinate system by taking the edge perpendicular to the ground in the same spatial region as the y-axis and the lower edge parallel to the ground in the same spatial region as the x-axis; Based on the reference coordinate system, coordinate marking is performed on the areas where dust density is concentrated in the same spatial region to obtain the initial dust concentration coordinate characteristics.

[0009] Preferably, the historical dust concentration coordinate features of the same spatial area and the historical post-vibration concentrated distribution coordinates after the change of dust density concentration location coordinates due to vibration are extracted from historical monitoring record data. Based on the historical vibration feature data, the historical dust concentration coordinate features and the historical post-vibration concentrated distribution coordinates are correlated and aggregated to obtain the aggregated path features. Based on simulated vibration characteristic data, initial dust concentration coordinate characteristics, and convergence path characteristics, the preprocessed dust concentration location after the dust density concentration location coordinates of the area in the same spatial region change due to vibration is obtained. The location of the heat-prone components of the current electrical equipment is obtained, and the distance between the pre-treated dust concentration location and the location of the heat-prone components is converted into a percentage and statistically analyzed to obtain the differential weighting factor.

[0010] Preferably, the average coverage area of ​​the dust concentration coordinate region in the initial dust concentration coordinate feature is statistically analyzed to obtain the average area of ​​the dust distribution to be measured. Based on the average area of ​​the dust distribution to be measured, the historical average area of ​​dust distribution and the historical diffusion distribution area after the historical average area of ​​dust distribution is affected by vibration are extracted from the historical monitoring record data. Based on the correlation trend characteristics between the historical diffusion distribution area, the historical average dust distribution area, and the historical vibration characteristic data, the area diffusion influence factor is obtained; The preprocessed diffusion distribution area is obtained based on the average area of ​​the dust distribution to be tested, the area diffusion influence factor, and the simulated vibration characteristic data.

[0011] Preferably, the historical dust-related energy consumption value is extracted from historical monitoring records, which is affected by the heat dissipation obstruction of the historical diffusion distribution area. Based on the correlation trend characteristics between historical temperature rise amplitude, historical diffusion distribution area and historical dust-related energy consumption value, the dust energy consumption influence coefficient is obtained; Based on the dust energy consumption influence coefficient, pretreatment temperature rise amplitude, pretreatment diffusion distribution area, and differential weighting factor, the dust energy consumption estimate is obtained. The predicted vibration energy consumption, the predicted dust energy consumption, and the predicted operating energy consumption are summed to obtain the predicted actual energy consumption. Based on the predicted actual energy consumption, energy-saving control is applied to the preset operating parameters, and the electrical equipment management results are output.

[0012] Compared with the prior art, the present invention has the following characteristics and beneficial effects: By statistically analyzing the preset operating parameters of current electrical equipment and the simulated vibration characteristic data under the preset operating parameter conditions, it is possible to predict the performance consumption values ​​for these two different situations. In traditional technology, the energy consumption detection of electrical equipment is often based on empirical data statistics of corresponding energy consumption data according to the operating parameters of the electrical equipment. The additional energy consumption generated by the electrical equipment during operation is not further analyzed and processed, which can easily lead to the final energy consumption prediction result being underestimated. Subsequent adjustment of corresponding operating parameters will have a large error, thus increasing the waste of resources. By making differentiated predictions for the additional energy consumption value generated by vibration and the additional energy consumption value caused by the change of the original dust concentration distribution area and the diffusion of the dust distribution area caused by vibration, it is possible to fully consider the influence factors of multiple energy consumption and conduct correlation analysis. Through the above-mentioned analysis of individual performance consumption prediction for different situations, the overall data processing process is more systematic, the interference between multiple types of data is reduced, and the efficiency of the corresponding adjustment of operating parameters for energy-saving management of electrical equipment is improved. Attached Figure Description

[0013] Figure 1 This is a flowchart illustrating the steps of a deep learning-based intelligent control method for electrical equipment, which is the main feature of this embodiment. Detailed Implementation

[0014] The present invention will be further described in detail below with reference to the following embodiments.

[0015] Reference Figure 1 A deep learning-based intelligent control method for electrical equipment includes the following steps: Step S1: Obtain the preset operating parameters of the current electrical equipment in the current period; test the energy consumption prediction model established based on the historical operating parameters and historical energy consumption data of the current electrical equipment to obtain the predicted operating energy consumption value; perform operation simulation based on the preset operating parameters to obtain simulated vibration characteristic data; and predict the vibration energy consumption value based on the correlation and influence trend characteristics between the historical vibration characteristic data and historical vibration energy consumption value of the current electrical equipment.

[0016] Step S2: Based on the correlation trend between the historical temperature rise amplitude and historical vibration characteristic data of the current electrical equipment, predict the pre-processed temperature rise amplitude generated by arbitrarily selected main vibration characteristic data and auxiliary vibration characteristic data in the simulated vibration characteristic data. Statistically analyze the spatial area enclosed by the adjacent opposing surfaces of the main measuring equipment to which the main measuring equipment belongs and the auxiliary measuring equipment to which the auxiliary measuring equipment belongs. Based on the correlation trend between the historical diffusion distribution area, the historical average dust distribution area and the historical vibration characteristic data in the historical monitoring record data of the current electrical equipment, predict the dust distribution area and diffusion area of ​​the spatial area to obtain the pre-processed diffusion distribution area.

[0017] Step S3: Based on the pre-treatment temperature rise amplitude, pre-treatment diffusion distribution area, and vibration energy consumption estimate, obtain the actual energy consumption prediction value. Based on the operating energy consumption prediction value and the actual energy consumption prediction value, perform energy-saving control on the preset operating parameters and output the electrical equipment management results.

[0018] Specifically, by statistically analyzing the preset operating parameters of current electrical equipment and the simulated vibration characteristic data under the preset operating parameter conditions, it is possible to predict the performance consumption values ​​for these two different situations. In traditional technology, the energy consumption detection of electrical equipment is often based on empirical data statistics of corresponding energy consumption data according to the operating parameters of the electrical equipment. The additional energy consumption generated by the electrical equipment during operation is not further analyzed and processed, which can easily lead to the final energy consumption prediction result being underestimated. Subsequent adjustment of the corresponding operating parameters will have a large error, thus increasing the waste of resources. Differentiating the additional energy consumption value generated by vibration and the additional energy consumption value caused by the change of the original dust concentration distribution area and the diffusion of the dust distribution area caused by vibration, it is possible to fully consider the influence factors of multiple energy consumption and conduct correlation analysis. Through the above-mentioned individual performance consumption prediction analysis that distinguishes multiple situations, the whole data processing process is more systematic, the interference between multiple types of data is reduced, and the efficiency of the corresponding adjustment of operating parameters for energy-saving management of electrical equipment is improved.

[0019] The specific step S1 includes the following sub-steps: Obtain the historical operating parameters of the current electrical equipment in historical periods and the corresponding historical energy consumption data. Based on the historical operating parameters and historical energy consumption data, establish an energy consumption prediction model.

[0020] Obtain the preset operating parameters of the current electrical equipment in the current period, input the preset operating parameters into the energy consumption prediction model for testing, and obtain the first energy consumption prediction value.

[0021] Based on preset operating parameters, the current electrical equipment is simulated to obtain simulated vibration characteristic data.

[0022] Obtain historical monitoring records of current electrical equipment, and extract historical vibration characteristic data and historical vibration energy consumption values ​​caused by the influence of historical vibration characteristic data on temperature rise from the historical monitoring records based on simulated vibration characteristic data.

[0023] Based on the correlation trend between historical vibration characteristic data and historical vibration energy consumption values, the vibration energy consumption correlation coefficient between historical vibration characteristic data and historical additional energy consumption values ​​is extracted.

[0024] Based on the vibration loss correlation coefficient and simulated vibration characteristic data, the estimated vibration energy consumption of the current electrical equipment is obtained.

[0025] Specifically, in scenarios such as indoor and outdoor environments with no wind, an energy consumption prediction model is established (e.g., creating a database of corresponding operating parameters and energy consumption tables by matching historical operating parameters and historical energy consumption data). The energy consumption prediction value is then calculated (by matching the preset operating parameters with the energy consumption prediction model (operating parameters-energy consumption table database) to estimate the energy consumption generated by the current electrical equipment). Simulated vibration characteristic data (including vibration frequency and amplitude; here, vibration amplitude is used as the primary example, such as through a hardware-in-the-loop simulation system: combining the controller of the actual electrical equipment with a virtual simulation model, simulating the equipment's operating environment through a real-time computing platform, and generating dynamic response data (such as vibration signals). Applicable scenarios include motors, transformers, and other equipment with controllers, capable of reproducing vibration characteristics under complex operating conditions). Historical monitoring records (including data from electrical equipment) are also included. The data includes: temperature rise data caused by equipment operation vibration, additional energy consumption data after temperature rise, data on the location of dust concentration in the spatial area, additional energy consumption data caused by the diffusion of dust concentration area, historical vibration data, etc.); vibration energy consumption correlation coefficient (e.g., using historical additional energy consumption value as the value on the y-axis and historical vibration characteristic data as the value on the x-axis to establish a correlation trend characteristic graph between the two, i.e., curve T1); vibration energy consumption correlation coefficient (e.g., performing slope statistics (average value) on T1, which is the vibration energy consumption correlation coefficient K1); and vibration energy consumption prediction value (e.g., multiplying the vibration energy consumption correlation coefficient and simulated vibration characteristic data, and then summing the vibration error compensation value, if it is b1 (based on the average difference between the vibration energy consumption prediction data and the actual vibration energy consumption data under the same operating parameters in different historical periods), which is the vibration energy consumption prediction value, if it is g1).

[0026] The specific step S2 includes the following sub-steps: Historical temperature rise amplitude values ​​affected by historical vibration characteristic data are extracted from historical monitoring records. Based on the correlation trend characteristics between historical temperature rise amplitude values ​​and historical vibration characteristic data, the temperature rise correlation coefficient between historical temperature rise amplitude values ​​and historical vibration characteristic data is extracted.

[0027] Arbitrarily select primary and secondary vibration characteristic data from the simulated vibration characteristic data. The primary electrical equipment to which the primary vibration characteristic data belongs and the secondary electrical equipment to which the secondary vibration characteristic data belongs are in adjacent positions.

[0028] Based on the main vibration characteristic data and the temperature rise correlation coefficient, the temperature rise amplitude value 1 of the main measuring equipment is obtained. Based on the auxiliary vibration characteristic data and the temperature rise correlation coefficient, the temperature rise amplitude value 2 of the auxiliary measuring equipment is obtained. The average value of the temperature rise amplitude value 1 and the temperature rise amplitude value 2 is calculated to obtain the preprocessed temperature rise amplitude value.

[0029] The common spatial area enclosed by the adjacent opposing surfaces of the main measuring equipment and the auxiliary measuring equipment is statistically analyzed, and the spatial area in the same location is output.

[0030] Establish a reference coordinate system by taking the edge perpendicular to the ground in the same spatial region as the y-axis and the lower edge parallel to the ground in the same spatial region as the x-axis.

[0031] Based on the reference coordinate system, the coordinates of the concentrated dust density distribution areas within the same spatial region are marked to obtain the initial dust concentration coordinate characteristics.

[0032] The historical dust concentration coordinates of the same spatial area and the historical post-vibration concentrated distribution coordinates after the change of dust density concentration location coordinates due to vibration are extracted from historical monitoring records. Based on historical vibration characteristic data, the historical dust concentration coordinates and historical post-vibration concentrated distribution coordinates are correlated and aggregated to obtain aggregated path characteristics.

[0033] Based on simulated vibration characteristic data, initial dust concentration coordinate characteristics, and convergence path characteristics, the preprocessed dust concentration location after the dust density concentration location coordinates change due to vibration in the same spatial area are obtained.

[0034] The location of the heat-prone components of the current electrical equipment is obtained, and the distance between the pre-treated dust concentration location and the location of the heat-prone components is converted into a percentage and statistically analyzed to obtain the differential weighting factor.

[0035] The average coverage area of ​​the dust concentration coordinate region in the initial dust concentration coordinate feature is statistically analyzed to obtain the average area of ​​the dust distribution to be measured.

[0036] Based on the average area of ​​the dust distribution to be measured, the historical average area of ​​dust distribution and the historical diffusion distribution area after the historical average area of ​​dust distribution is affected by vibration are extracted from the historical monitoring record data.

[0037] Based on the correlation and influence trend characteristics between historical diffusion distribution area, historical average dust distribution area, and historical vibration characteristic data, the area diffusion influence factor is obtained.

[0038] The preprocessed diffusion distribution area is obtained based on the average area of ​​the dust distribution to be measured, the area diffusion influence factor, and the simulated vibration characteristic data.

[0039] Specifically, this includes the temperature rise correlation coefficient (e.g., using historical temperature rise amplitude as the y-axis value and historical vibration characteristic data as the x-axis value, establishing a correlation trend characteristic graph between the two, i.e., curve T2, and calculating the slope (average) of T2, which is the temperature rise correlation coefficient K2), the main measured vibration characteristic data and the auxiliary measured vibration characteristic data (since it targets the current electrical equipment (there are multiple devices), it is necessary to consider the correlation influence between adjacent electrical equipment, such as the temperature rise caused by vibration in the space area enclosed by adjacent facades, which leads to additional energy consumption), so arbitrarily select the main measured equipment and the auxiliary measured equipment from the current electrical equipment, and their positional relationship is adjacent. The main measured vibration characteristic data is the vibration data of the main measured equipment, and the auxiliary measured vibration characteristic data is the vibration data of the auxiliary measured equipment), temperature rise amplitude one, and temperature rise amplitude two (e.g., multiplying the main measured vibration characteristic data and the temperature rise correlation coefficient yields temperature rise amplitude one). Multiplying the amplitude vibration characteristic data and the temperature rise correlation coefficient yields the temperature rise amplitude (II). The spatial region (e.g., one measuring surface of the main measuring device (generator 1) and one measuring surface of the auxiliary measuring device (generator 2) are adjacent and opposite; the spatial region enclosed by these opposite surfaces is the common spatial region (if it is a cuboid spatial region)). A reference coordinate system (e.g., using the height of the common spatial region (the edge perpendicular to the ground in the same spatial region) as the y-axis and the length of the common spatial region (the lower edge parallel to the ground in the same spatial region) as the x-axis, thus constructing a reference coordinate system (a two-dimensional coordinate system, with the adjacent opposite surfaces of the main measuring device and the auxiliary measuring device as the coordinate markers)). Initial dust concentration coordinate characteristics (e.g., using a laser dust sensor: measuring the concentration of particulate matter in the air through laser scattering or light scattering methods; combined with multi-point deployment, the dust accumulation area can be located).Mark the spatial location of the concentrated dust distribution area (by extracting a regular region) (center point). If w1, w2, and wn exist, these represent the initial dust concentration coordinates. Historical post-vibration concentration coordinates (if the historical dust concentration coordinates are z1, z2, and zn, these refer to the dust concentration locations after z1, z2, and zn have changed due to historical vibration data; if z1, z2, and zn are Z, then the historical post-vibration concentration coordinates refer to the dust concentration locations after z1, z2, and zn have changed due to historical vibration data; if z1, z2, and zn are Z, then the historical post-vibration concentration coordinates refer to Z). Collective path characteristics (by comprehensively statistically analyzing the paths from z1, z2, and zn to Z, we obtain the comprehensive collection trend path characteristics; for example, z1, z2, and zn form a polygon, and Z is a polygon). The center location), for different historical vibration characteristic data, perform comprehensive statistics on the collection path characteristics), preprocess the dust concentration location (for example, match the simulated vibration characteristic data and historical vibration characteristic data to match the corresponding collection path characteristics, and then estimate the preprocessed dust concentration location based on the initial dust concentration coordinate characteristics (if it is the collection path characteristic of the center collection, then obtain the center location based on the figure enclosed by connecting w1, w2, and wn, that is, obtain the preprocessed dust concentration location Wx), the location of easily heated components (e.g., stator winding, obtain the location Wz of the stator winding), and the differential weighting factor. (Statistical analysis of the distance data between Wx and Wz; if it is B, then convert the numerical percentage to B%, the differential weighting factor is 1-B%. The smaller B is, the larger the differential weighting factor is, and the greater the impact of subsequent dust distribution on the additional energy consumption of the current electrical equipment). The average area of ​​the dust distribution to be measured (e.g., if the distribution concentration areas corresponding to w1, w2, and wn are q1, q2, and qn respectively, then calculate the average area of ​​q1, q2, and qn; if it is (m1+m2+mn) / n=S, then it is the average area of ​​the dust distribution to be measured). Historical diffusion distribution area (i.e., if the dust concentration area to which Z belongs is M1). Area diffusion influence factor. The first step is to use the historical diffusion distribution area as the value on the y-axis, compare the historical average area of ​​dust distribution with the amplitude value of historical vibration characteristic data to obtain the first ratio value, use this first ratio value as the value on the x-axis, construct the correlation influence trend characteristic diagram between the two, i.e., curve T3, and perform slope statistics (average value) on T3, which is the area diffusion influence factor K3. The second step is to use the first ratio value as the value on the x-axis, construct the correlation influence trend characteristic diagram between the two, i.e., curve T3, and perform slope statistics (average value) on T3, which is the area diffusion influence factor K3. The third step is to use the preprocessed diffusion distribution area (e.g., compare the average area of ​​dust distribution to be measured with the amplitude value of simulated vibration characteristic data to obtain the second ratio value, and multiply the second ratio value with the area diffusion influence factor K3 to obtain the preprocessed diffusion distribution area).

[0040] The specific step S3 includes the following sub-steps: Extract historical dust-related energy consumption values ​​from historical monitoring records, which are influenced by the historical temperature rise amplitude and the heat dissipation obstruction caused by the historical diffusion distribution area.

[0041] Based on the correlation trend characteristics between historical temperature rise amplitude, historical diffusion distribution area and historical dust-related energy consumption value, the dust energy consumption influence coefficient is obtained.

[0042] The dust energy consumption estimate is obtained based on the dust energy consumption influence coefficient, the pretreatment temperature rise amplitude, the pretreatment diffusion distribution area, and the differential weighting factor.

[0043] The predicted energy consumption of vibration, dust, and operation is summed to obtain the predicted actual energy consumption.

[0044] Based on the actual energy consumption forecast, the preset operating parameters are adjusted for energy conservation, and the electrical equipment management results are output.

[0045] Specifically, this includes the dust energy consumption influence coefficient (e.g., Y=(h+j)X+C, where Y refers to the historical dust-related energy consumption value, h refers to the historical temperature rise amplitude, j refers to the historical diffusion distribution area, X refers to the dust energy consumption influence coefficient, and C refers to the compensation error value. By substituting the historical temperature rise amplitude, historical diffusion distribution area, and historical dust-related energy consumption value under the same vibration characteristic data conditions in different historical periods, X and C can be obtained), the dust energy consumption prediction value (by substituting the dust energy consumption influence coefficient, pretreatment temperature rise amplitude, and pretreatment diffusion distribution area into Y=(h+j)X+C to calculate the energy consumption prediction value, and multiplying the energy consumption prediction value with the differential weighting factor, the dust energy consumption prediction value is obtained), and the electrical equipment control results (e.g., data matching based on the actual energy consumption prediction value and the energy-saving operation parameter correspondence table (an experience table derived from historical operation data statistics / clustering / optimization) (by looking up the table or interpolation to obtain the optimal / relatively good combination of control parameters), for example: "When the expected energy consumption is NkWh / h, it is recommended to set the motor frequency to PHz and the temperature setpoint to Z℃").

[0046] The above are all preferred embodiments of this application, and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. A method for intelligent control of electrical equipment based on deep learning, characterized in that, Includes the following steps: Step S1: Obtain the preset operating parameters of the current electrical equipment in the current period, test the energy consumption prediction model established based on the historical operating parameters and historical energy consumption data of the current electrical equipment, obtain the predicted value of operating energy consumption, perform operation simulation based on the preset operating parameters, obtain simulated vibration characteristic data, and predict the vibration energy consumption estimate based on the correlation influence trend characteristics between the historical vibration characteristic data and historical vibration energy consumption values ​​of the current electrical equipment. Step S2: Based on the correlation trend between the historical temperature rise amplitude and historical vibration characteristic data of the current electrical equipment, predict the preprocessed temperature rise amplitude generated by arbitrarily selected main vibration characteristic data and auxiliary vibration characteristic data in the simulated vibration characteristic data. Statistically analyze the spatial area enclosed by the adjacent opposing surfaces of the main measuring equipment to which the main measuring equipment belongs and the auxiliary measuring equipment to which the auxiliary measuring equipment belongs. Based on the correlation trend between the historical diffusion distribution area, the historical average dust distribution area, and the historical vibration characteristic data in the historical monitoring record data of the current electrical equipment, predict the dust distribution area and diffusion area of ​​the spatial area to obtain the preprocessed diffusion distribution area. Step S3: Based on the pre-processed temperature rise amplitude, pre-processed diffusion distribution area, and vibration energy consumption estimate, obtain the actual energy consumption prediction value. Based on the operating energy consumption prediction value and the actual energy consumption prediction value, perform energy-saving control on the preset operating parameters and output the electrical equipment management results.

2. The intelligent control method for electrical equipment based on deep learning according to claim 1, characterized in that, Step S1 includes: Obtain the historical operating parameters of the current electrical equipment in a historical period and the historical energy consumption data generated by the historical operating parameters. Based on the historical operating parameters and historical energy consumption data, establish an energy consumption prediction model. Obtain the preset operating parameters of the current electrical equipment in the current period, input the preset operating parameters into the energy consumption prediction model for testing, and obtain the first energy consumption prediction value; Based on the preset operating parameters, the current electrical equipment is simulated to obtain simulated vibration characteristic data; Obtain historical monitoring record data of the current electrical equipment, and extract historical vibration characteristic data and historical vibration energy consumption value caused by the influence of historical vibration characteristic data on temperature rise from the historical monitoring record data based on the simulated vibration characteristic data; Based on the correlation trend characteristics between the historical vibration characteristic data and the historical vibration energy consumption value, the vibration energy consumption correlation coefficient between the historical vibration characteristic data and the historical additional energy consumption value is extracted. Based on the vibration loss correlation coefficient and simulated vibration characteristic data, the estimated vibration energy consumption of the current electrical equipment is obtained.

3. The intelligent control method for electrical equipment based on deep learning according to claim 2, characterized in that, Step S1 also includes: The historical temperature rise amplitude caused by the influence of historical vibration characteristic data is extracted from the historical monitoring record data. Based on the correlation trend characteristics between the historical temperature rise amplitude and the historical vibration characteristic data, the temperature rise correlation coefficient between the historical temperature rise amplitude and the historical vibration characteristic data is extracted. Randomly select primary vibration characteristic data and secondary vibration characteristic data from the simulated vibration characteristic data, wherein the primary electrical equipment to which the primary vibration characteristic data belongs and the secondary electrical equipment to which the secondary vibration characteristic data belongs are in an adjacent position relationship; Based on the main vibration characteristic data and the temperature rise correlation coefficient, the temperature rise amplitude value 1 of the main measuring equipment is obtained. Based on the auxiliary vibration characteristic data and the temperature rise correlation coefficient, the temperature rise amplitude value 2 of the auxiliary measuring equipment is obtained. The average value of the temperature rise amplitude value 1 and the temperature rise amplitude value 2 is calculated to obtain the preprocessed temperature rise amplitude value.

4. The intelligent control method for electrical equipment based on deep learning according to claim 3, characterized in that, Step S2 includes: The common spatial area enclosed by the adjacent opposing surfaces of the main measuring equipment and the auxiliary measuring equipment is statistically analyzed, and the spatial area in the same location is output. Establish a reference coordinate system by taking the edge perpendicular to the ground in the same spatial region as the y-axis and the lower edge parallel to the ground in the same spatial region as the x-axis; Based on the reference coordinate system, coordinate marking is performed on the areas where dust density is concentrated in the same spatial region to obtain the initial dust concentration coordinate characteristics.

5. The intelligent control method for electrical equipment based on deep learning according to claim 4, characterized in that, Step S2 also includes: Extract the historical dust concentration coordinate features of the same spatial area from historical monitoring records, as well as the historical post-vibration concentrated distribution coordinates after the dust density concentration location coordinates changed due to vibration. Based on historical vibration feature data, statistical analysis of the correlation and aggregation path change features of historical dust concentration coordinate features and historical post-vibration concentrated distribution coordinates is performed to obtain aggregation path features. Based on simulated vibration characteristic data, initial dust concentration coordinate characteristics, and convergence path characteristics, the preprocessed dust concentration location after the dust density concentration location coordinates of the area in the same spatial region change due to vibration is obtained. The location of the heat-prone components of the current electrical equipment is obtained, and the distance between the pre-treated dust concentration location and the location of the heat-prone components is converted into a percentage and statistically analyzed to obtain the differential weighting factor.

6. The intelligent control method for electrical equipment based on deep learning according to claim 5, characterized in that, Step S2 also includes: The average coverage area of ​​the dust concentration coordinate region in the initial dust concentration coordinate feature is statistically analyzed to obtain the average area of ​​the dust distribution to be measured. Based on the average area of ​​the dust distribution to be measured, the historical average area of ​​dust distribution and the historical diffusion distribution area after the historical average area of ​​dust distribution is affected by vibration are extracted from the historical monitoring record data. Based on the correlation trend characteristics between the historical diffusion distribution area, the historical average dust distribution area, and the historical vibration characteristic data, the area diffusion influence factor is obtained; The preprocessed diffusion distribution area is obtained based on the average area of ​​the dust distribution to be tested, the area diffusion influence factor, and the simulated vibration characteristic data.

7. The intelligent control method for electrical equipment based on deep learning according to claim 6, characterized in that, Step S3 includes: Extract historical dust-related energy consumption values ​​from historical monitoring records, which are influenced by the obstruction of heat dissipation due to the historical diffusion distribution area. Based on the correlation trend characteristics between historical temperature rise amplitude, historical diffusion distribution area and historical dust-related energy consumption value, the dust energy consumption influence coefficient is obtained; Based on the dust energy consumption influence coefficient, pretreatment temperature rise amplitude, pretreatment diffusion distribution area, and differential weighting factor, the dust energy consumption estimate is obtained. The predicted vibration energy consumption, the predicted dust energy consumption, and the predicted operating energy consumption are summed to obtain the predicted actual energy consumption. Based on the predicted actual energy consumption, energy-saving control is applied to the preset operating parameters, and the electrical equipment management results are output.

Images