A method and system for green energy power data transmission based on 5G slicing

Abstract

This invention relates to the field of data transmission technology, and provides a method and system for green energy power data transmission based on 5G slicing. The method includes: analyzing environmental data streams to determine multiple segmentation periods; segmenting the motor data stream according to the multiple segmentation periods to obtain multiple motor data segments; inputting the motor data segments into a classification model to obtain waveform categories of the motor data segments; inputting the motor data segments into a waveform extraction model for the corresponding waveform categories to obtain key parameters of the motor data segments; and sending the start time, end time, waveform category, and key parameters of each motor data segment as real-time data to an analysis and control center so that the analysis and control center can use the real-time data to simulate and obtain the real-time observation waveform of the corresponding motor data stream. This invention extracts real-time data and uses uRLLC slicing for real-time data transmission, thereby enabling real-time observation of data waveforms.

Description

Technical Field

[0001] This invention relates to the field of data transmission technology, and in particular to a method and system for transmitting green energy power data based on 5G slicing. Background Technology

[0002] Wind and solar energy have become the most widely developed and utilized new energy sources in my country. Wind power generation technology boasts significant economic value, short construction periods, high independence, and environmental friendliness. Compared to previous power generation technologies, it can also be applied promptly and effectively to provide regional power supply. Photovoltaic power generation offers advantages such as readily available solar resources, localized power supply, high theoretical power generation efficiency, and environmental friendliness. With the development of digital twin technology, the need for detailed monitoring of various indicators of wind and solar power generation has arisen to ensure the normal operation of wind turbines and photovoltaic panels. However, due to the relatively large volume of monitoring data generated by wind and solar power generation, it is difficult to transmit this data to a remote analysis and control center in real time. Consequently, the data monitored at the analysis and control center is often not timely, making it difficult for users to observe data changes in real time.

[0003] Therefore, overcoming the shortcomings of the existing technology is an urgent problem to be solved in this technical field. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a method and system for transmitting green energy power data based on 5G slicing.

[0005] The present invention adopts the following technical solution:

[0006] In a first aspect, the present invention provides a green energy power data transmission method based on 5G slicing, which sets up a coordination data server for a motor array and establishes a local area network between the motor array, environmental sensors, and the coordination data server. The method includes:

[0007] The coordination data server requests the establishment of a first network slice and a second network slice from the 5G core network in advance; wherein, the first network slice is a uRLLC slice and the second network slice is an mMTC slice;

[0008] The data server coordinates the use of the local area network to receive data streams from each motor in the motor array and environmental data streams from each environmental sensor.

[0009] The environmental data stream is analyzed to determine multiple segmentation periods; each segmentation period corresponds to a stable power generation process.

[0010] The motor data stream is segmented according to the multiple segmentation cycles to obtain multiple motor data segments. The motor data segments are input into a classification model to obtain the waveform category of the motor data segments. The motor data segments are then input into the waveform extraction model of the corresponding waveform category to extract the key parameters of the motor data segments.

[0011] The start time, end time, waveform type, and key parameters of each motor data segment are used as real-time data. The real-time data is sent to the analysis and control center using the first network slice so that the analysis and control center can use the real-time data to simulate and obtain the real-time observation waveform of the corresponding motor data stream.

[0012] The coordination data server also uses a second network slice to send the data streams of each motor to the analysis and control center so that the analysis and control center can trace the data later.

[0013] Wherein, when the motor array is a wind turbine array, the environmental data stream includes at least a wind speed data stream; when the motor array is a photovoltaic array, the environmental data stream includes at least a light intensity data stream.

[0014] Preferably, when the motor array is a fan array, the analysis of the environmental data stream to determine multiple segmentation periods specifically includes:

[0015] The wind speed data stream is segmented once to divide the wind speed data stream into a first wind speed data segment less than or equal to a first preset value, a second wind speed data segment greater than the first preset value and less than or equal to a second preset value, a third wind speed data segment greater than the second preset value and less than a third preset value, and a fourth wind speed data segment greater than or equal to a third preset value.

[0016] Differential identification is performed on the second wind speed data segment to identify the troughs in the second wind speed data segment. Using the troughs as the dividing boundary, the second wind speed data segment is further segmented to obtain multiple fifth wind speed data segments.

[0017] Multiple segmentation periods are determined based on the time periods of the first, third, fourth, and fifth wind speed data segments.

[0018] Preferably, when the motor array is a photovoltaic array, the analysis of the environmental data stream to determine multiple segmentation periods specifically includes:

[0019] The light intensity data stream is segmented once to divide the light intensity data stream into a first light intensity data segment that is less than a fourth preset value and a second light intensity data segment that is greater than or equal to the fourth preset value;

[0020] Multiple segmentation periods are determined based on the time periods of the first and second light intensity data segments.

[0021] Preferably, the analysis and control center uses the real-time data simulation to obtain the real-time observation waveform of the corresponding motor data stream, specifically including:

[0022] Input the start time, end time, and key parameters into the waveform fitting model of the corresponding waveform category to fit the real-time observed waveform segment corresponding to the real-time data.

[0023] The real-time observation waveform is obtained by splicing together multiple real-time observation waveform segments in chronological order.

[0024] Preferably, the waveform fitting model and waveform extraction model for the corresponding waveform category are trained together, specifically including:

[0025] The training dataset was obtained by annotating historical motor data segments by experts; the annotations included key parameters.

[0026] The training dataset is divided into a first dataset and a second dataset according to a preset ratio;

[0027] The first dataset is input into the waveform extraction initial model for the first stage of training to obtain the waveform extraction intermediate model; wherein, the first stage of training uses the first loss function.

[0028] The first dataset is input into the initial waveform fitting model for the second stage of training to obtain an intermediate waveform fitting model; wherein, the second stage of training uses the second loss function.

[0029] The second dataset is input into the waveform extraction intermediate model, and the key parameters output by the waveform extraction intermediate model are input into the waveform fitting intermediate model to perform a third-stage training on the waveform extraction intermediate model and the waveform fitting intermediate model to obtain the waveform extraction model and the waveform fitting model. The third-stage training uses a third loss function.

[0030] The first loss function is: The second loss function is: The third loss function is: ; For the predicted i-th key parameter, For the i-th key parameter obtained from the annotation, To predict the j-th value in the obtained instantaneous observed waveform segment, For the j-th value in the predicted historical motor data segment, and Preset weights.

[0031] Preferably, the coordination data server requests the establishment of a first network slice and a second network slice from the 5G core network in advance, specifically including:

[0032] After establishing a local area network with the coordination data server, each motor array and each environmental sensor sends a heartbeat packet to the coordination data server, which carries the device identifier of each motor in the motor array and the device identifier of the environmental sensor.

[0033] The coordination data server determines the first slice size of the required first network slice and the second slice size of the required second network slice based on the number of heartbeat packets and / or the device identifiers in the heartbeat packets;

[0034] The Coordinating Data Server sends a network slice registration request to the AMF node; wherein the network slice registration request carries the signed NSSAI, the first slice size, and the second slice size, so that the AMF node can establish a first network slice that meets the first slice size and a second network slice that meets the second slice size for the Coordinating Data Server; wherein the signed NSSAI includes the S-NSSAI of uRLLC slices and the S-NSSAI of mMTC slices.

[0035] Preferably, the method further includes: the analysis control center, at preset intervals, finding data points in the real-time observed waveform that correspond to the latest timestamp of the motor data stream according to the latest received motor data stream, and recording the data points as interval points;

[0036] When the next refresh cycle arrives, the portion of the real-time observed waveform before the interval point is drawn using the first color, and the portion of the real-time observed waveform after the interval point is drawn using the second color, so that the user can distinguish the portion of the motor data stream that has been received from the interval point from the portion of the motor data stream that has not been received based on the first color and the second color.

[0037] The analysis and control center also responds to the user's click action on the real-time observed waveform, obtains the click coordinates based on the click action, obtains the corresponding data point in the real-time observed waveform based on the click coordinates, and if the data point is located before the interval point, obtains the first motor data segment within a preset interval length centered on the interval point, and displays the first motor data segment in the details window.

[0038] Preferably, the waveform category includes one or more of the following: constant category, triangular wave type, sine wave type, composite wave type, and data point fitting type;

[0039] The key parameters corresponding to the constant category include constant values;

[0040] The key parameters corresponding to the triangular wave type include amplitude, period, phase shift, and vertical shift;

[0041] The key parameters corresponding to the sine wave type include amplitude, frequency, phase shift, and vertical shift;

[0042] The key parameters corresponding to the composite wave type include multiple amplitudes, multiple frequencies, multiple phase shifts, and vertical shifts;

[0043] The key parameters corresponding to the data point fitting type include multiple key data points.

[0044] Secondly, the present invention also provides a green energy power data transmission device based on 5G slicing, used to implement the green energy power data transmission method based on 5G slicing described in the first aspect, the device comprising:

[0045] At least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor for performing the 5G slicing-based green energy power data transmission method described in the first aspect.

[0046] Thirdly, the present invention also provides a non-volatile computer storage medium storing computer-executable instructions that are executed by one or more processors to perform the method described in the first aspect.

[0047] Fourthly, a chip is provided, comprising: a processor and an interface for calling and running a computer program stored in a memory, performing the method as described in any of the first aspects.

[0048] Fifthly, a computer program product comprising instructions is provided, which, when executed on a computer or processor, cause the computer or processor to perform the method as described in any of the first aspects.

[0049] In a sixth aspect, a green energy power data transmission system based on 5G slicing is provided, including a motor array, environmental sensors, a coordination data server, and an analysis and control center; the coordination data server uses the green energy power data transmission method based on 5G slicing described in the first aspect to report data.

[0050] This invention analyzes environmental data streams to divide them into multiple stable power generation cycles. Then, it performs waveform classification and extracts key parameters from the motor data segments within each stable power generation cycle to obtain real-time data. This reduces the amount of data that needs to be transmitted in real time. Furthermore, it uses low-latency uRLLC slices to transmit real-time data, making real-time observation of data waveforms possible. Attached Figure Description

[0051] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments of the present invention will be briefly described below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0052] Figure 1 This is a flowchart illustrating a green energy power data transmission method based on 5G slicing provided in an embodiment of the present invention.

[0053] Figure 2 This is a schematic diagram of a green energy power data transmission method based on 5G slicing provided in an embodiment of the present invention;

[0054] Figure 3 This is a schematic diagram of a green energy power data transmission method based on 5G slicing provided in an embodiment of the present invention;

[0055] Figure 4 This is a schematic diagram of a green energy power data transmission method based on 5G slicing provided in an embodiment of the present invention;

[0056] Figure 5 This is a flowchart illustrating a green energy power data transmission method based on 5G slicing provided in an embodiment of the present invention.

[0057] Figure 6 This is a schematic diagram of the segmentation period in a green energy power data transmission method based on 5G slicing provided in an embodiment of the present invention;

[0058] Figure 7 This is a flowchart illustrating a green energy power data transmission method based on 5G slicing provided in an embodiment of the present invention.

[0059] Figure 8 This is a flowchart illustrating a green energy power data transmission method based on 5G slicing provided in an embodiment of the present invention.

[0060] Figure 9 This is a flowchart illustrating a green energy power data transmission method based on 5G slicing provided in an embodiment of the present invention.

[0061] Figure 10 This is a schematic diagram of a green energy power data transmission method based on 5G slicing provided in an embodiment of the present invention;

[0062] Figure 11 This is a flowchart illustrating a green energy power data transmission method based on 5G slicing provided in an embodiment of the present invention.

[0063] Figure 12 This is a flowchart illustrating a green energy power data transmission method based on 5G slicing provided in an embodiment of the present invention.

[0064] Figure 13 This is a schematic diagram of a green energy power data transmission method based on 5G slicing provided in an embodiment of the present invention;

[0065] Figure 14 This is a schematic diagram of the architecture of a green energy power data transmission device based on 5G slicing provided in an embodiment of the present invention. Detailed Implementation

[0066] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0067] Unless the context otherwise requires, throughout the specification and claims, the term "comprising" is interpreted as openly inclusive, meaning "including, but not limited to." In the description of the specification, terms such as "one embodiment," "some embodiments," "exemplary embodiment," "example," "specific example," or "some examples" are intended to indicate that a particular feature, structure, material, or characteristic associated with that embodiment or example is included in at least one embodiment or example of this disclosure. The illustrative representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics mentioned may be included in any suitable manner in any one or more embodiments or examples; that is, although they may be incorporated into embodiments or examples using the above terms for reasons such as order and position, it does not limit them to be incorporated in combination by a single embodiment or example.

[0068] In the description of this invention, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined with "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of embodiments of this disclosure, unless otherwise stated, "a plurality of" means two or more. Furthermore, for example, the description may use the prefix "A" or "B" to describe the same type of nouns as two independent entities. In this case, the corresponding features defined with "A" and "B" are used only to distinguish between similar entities and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features.

[0069] In the description of this invention, the expression “A and / or B” (where A and B are used to formally represent specific features) will be used. The corresponding expression includes the following three combinations: only A, only B, and a combination of A and B.

[0070] As used in this invention, “about,” “approximately,” or “approximately” includes the stated value and the average value within an acceptable range of deviation from a particular value, wherein the acceptable range of deviation is determined by a person skilled in the art taking into account the measurement under discussion and the error associated with the measurement of the particular quantity (i.e., the limitations of the measurement system).

[0071] Furthermore, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0072] Example 1:

[0073] Considering that in practical use, users often only need detailed monitoring data for later data tracing or fault analysis, and only the fluctuations in the monitoring data need to be observed in real time, based on this consideration, Embodiment 1 of this invention provides a green energy power data transmission method based on 5G slicing. A coordination data server is set up for the motor array, and a local area network is established between the motor array, environmental sensors, and the coordination data server, such as... Figure 1 and Figure 2 As shown, the method includes:

[0074] In step 201, the coordination data server requests the establishment of a first network slice and a second network slice from the 5G core network in advance. The first network slice is an ultra-reliable and low-latency communication (uRLLC) slice, and the second network slice is a massive machine-type communication (mMTC) slice. uRLLC slices have high reliability and relatively low latency, while mMTC slices have relatively high latency but are more suitable for large-scale data transmission.

[0075] In step 202, the coordination data server uses the local area network to receive data streams from each motor in the motor array and environmental data streams from each environmental sensor.

[0076] In step 203, the environmental data stream is analyzed to determine multiple segmentation periods; each segmentation period corresponds to a stable power generation process; for example, for a wind turbine array, the duration of a continuous gust of wind can be used as a segmentation period, and for a photovoltaic array, the duration of continuous sunlight can be used as a segmentation period.

[0077] In step 204, the motor data stream is segmented according to the multiple segmentation periods to obtain multiple motor data segments. These motor data segments are then input into a classification model to obtain their waveform categories. Finally, the motor data segments are input into a waveform extraction model corresponding to each waveform category to extract key parameters. These key parameters can be understood as parameters or data that, combined with the waveform category, can reconstruct the general trend of the fitted data. Both the classification model and the waveform extraction model are pre-trained by those skilled in the art. Specifically, historical motor data streams and historical environmental data streams are pre-collected. The segmentation period is determined based on the historical environmental data stream. The historical motor data stream is segmented according to the segmentation period. Experts annotate the multiple historical motor data segments obtained from the segmentation. The annotation results include the waveform category and key parameters of the historical motor data segments.

[0078] Multiple historical motor data segments and corresponding annotation results are used to generate a training dataset. The training dataset is then used to train the corresponding deep learning model to obtain a classification model and a waveform extraction model.

[0079] In one alternative implementation, the deep learning models used by the classification model and the waveform extraction model can be convolutional neural networks (CNNs).

[0080] In step 205, the start time, end time, waveform type, and key parameters of each motor data segment are used as real-time data. The real-time data is sent to the analysis and control center using the first network slice so that the analysis and control center can use the real-time data to simulate and obtain the real-time observation waveform of the corresponding motor data stream. The real-time observation waveform can be understood as a waveform similar to the trend of the motor data stream.

[0081] In step 206, the coordination data server also uses a second network slice to send the data streams of each motor to the analysis and control center so that the analysis and control center can trace the data later.

[0082] Steps 203-205 can be understood as converting a large amount of data into smaller, real-time data through waveform extraction, and then transmitting it to the analysis and control center using a uRLLC slice with low transmission latency. This allows the analysis and control center to quickly observe and display the waveform in real time. The complete motor data stream, on the other hand, is transmitted using an mMTC slice with higher transmission latency. Because the network slices are isolated from each other, the transmission of the motor data stream does not affect the transmission of real-time data. Even if the motor data stream is transmitted at a slower speed, it does not affect the analysis and control center's real-time monitoring of the motor array's operating status using real-time waveform observation. For the environmental data stream, waveform extraction can be performed based on the same concept as steps 203-205 to obtain real-time data, which is then sent to the analysis and control center using the first network slice. Alternatively, it can be transmitted directly as in step 206. The specific transmission method for the environmental data stream is determined by those skilled in the art based on the real-time monitoring requirements of the environmental data.

[0083] The motor array can be a wind turbine array or a photovoltaic array, etc. When the motor array is a wind turbine array, the environmental data stream includes at least a wind speed data stream (i.e., wind speed). In practical application scenarios, the environmental data stream may also include one or more of the following data streams: wind direction, temperature, and air pressure. The motor data stream includes one or more of the following data streams: wind turbine rotation speed, wind turbine torque, wind turbine power generation, and other equipment data. When the motor array is a photovoltaic array, the environmental data stream includes at least a light intensity data stream. In practical application scenarios, the environmental data stream may also include one or more of the following data streams: temperature, humidity, air pressure, wind speed, and wind direction. The motor data stream includes one or more of the following data streams: power generation of the photovoltaic panels in the photovoltaic array, and other equipment data.

[0084] This embodiment analyzes the environmental data stream to divide it into multiple stable power generation cycles. Then, it performs waveform classification and extracts key parameters from the motor data segments within each stable power generation cycle to obtain real-time data, thereby reducing the amount of data that needs to be transmitted in real time. Furthermore, it uses low-latency uRLLC slices to transmit real-time data, making real-time observation of data waveforms possible.

[0085] In some practical application scenarios, the first and second network slices can be implemented by the operator's 5G network functions, or they can be provided by a dedicated 5G network for the power grid. Figure 3 As shown, Figure 3 Three transmission methods (a), (b), and (c) are demonstrated. Method (a) directly utilizes a network slice provided by the dedicated power grid network for data transmission. In this method, the dedicated power grid network and the operator's 5G network reuse the same base station, such as... Figure 4 As shown, public network service data is forwarded by the base station to the operator's network for transmission, while power grid service data is forwarded by the base station to the dedicated power grid network for transmission. In method (b), the operator's 5G network provides network slicing, and in method (c), the operator provides network slicing, which transmits data to the dedicated power grid network, and then reaches the monitoring master station through the dedicated power grid network.

[0086] In a practical application scenario, when the motor array is a wind turbine array, the environmental data stream is analyzed to determine multiple segmentation periods, such as... Figure 5 As shown, it specifically includes:

[0087] In step 301, the wind speed data stream is segmented once to divide the wind speed data stream into a first wind speed data segment less than or equal to a first preset value, a second wind speed data segment greater than the first preset value and less than or equal to a second preset value, a third wind speed data segment greater than the second preset value and less than a third preset value, and a fourth wind speed data segment greater than or equal to a third preset value.

[0088] In step 302, differential identification is performed on the second wind speed data segment to identify the troughs in the second wind speed data segment. Using the troughs as the dividing boundary, the second wind speed data segment is further divided to obtain multiple fifth wind speed data segments.

[0089] In step 303, multiple segmentation periods are determined according to the time periods of the first wind speed data segment, the third wind speed data segment, the fourth wind speed data segment, and the fifth wind speed data segment.

[0090] The first and second preset values ​​are obtained by those skilled in the art based on the working characteristics of the wind turbine array. In actual use, the first preset value may be the wind turbine's cut-in speed, the second preset value may be the wind turbine's rated speed, and the third preset value may be the wind turbine's cut-out speed.

[0091] Since the most important monitoring data for wind turbines is their power generation, and the relationship between power generation and wind speed is as follows:

[0092]

[0093] in, This represents the current wind speed of the fan. The cut-in wind speed of the fan. The rated wind speed of the fan. This refers to the cut-out velocity of the fan.

[0094] Therefore, it is possible , and As a dividing line, four types of data segments are obtained (i.e., the first wind speed data segment, the second wind speed data segment, the third wind speed data segment, and the fourth wind speed data segment). Among them, the power generation of the wind turbines corresponding to the first, third, and fourth wind speed data segments is usually a constant value, making it easy to analyze the characteristics of their power generation. In the second wind speed data segment, by analyzing the changes in wind speed, the second wind speed data segment can be divided into multiple segments (i.e., the fifth wind speed data segment). Each segment represents a continuous wind, so that the power generation within a segment shows a relatively regular variation pattern, which facilitates the analysis of the characteristics of the power generation.

[0095] by Figure 6 For example, let's start with... As a dividing line, two first wind speed data segments, cut_1 and cut_5, are obtained, along with two second wind speed data segments composed of cut_3, cut_4, and cut_5, and two third wind speed data segments composed of cut_6 and cut_7. Differential identification is then performed on the second wind speed data segments to obtain peaks and troughs. Based on the trough positions, five fifth wind speed data segments, cut_3, cut_4, cut_5, cut_6, and cut_7, are obtained. Observing the power generation of each wind speed data segment reveals that the power generation in each segment is either a convex wave or a constant value of 0. Thus, a complex long waveform is divided into multiple simple waveforms through segmentation, facilitating subsequent waveform extraction and fitting.

[0096] In another practical application scenario, since the key observation data for photovoltaic panels is primarily their power generation, which depends mainly on sunlight intensity, when the motor array is a photovoltaic array, the environmental data stream is analyzed to determine multiple segmentation periods, such as... Figure 7 As shown, it specifically includes:

[0097] In step 401, the light intensity data stream is segmented once to divide the light intensity data stream into a first light intensity data segment that is less than a fourth preset value and a second light intensity data segment that is greater than or equal to the fourth preset value; the fourth preset value is obtained by those skilled in the art based on the working characteristics of the photovoltaic array, and in actual use, the fourth preset value is the minimum light intensity at which the photovoltaic motor can generate photovoltaic power.

[0098] In step 402, multiple segmentation periods are determined according to the time period of the first light intensity data segment and the time period of the second light intensity data segment.

[0099] Since the change in light intensity is relatively regular, that is, the light intensity gradually increases after sunrise and gradually decreases near sunset, the fourth preset value can be used directly for periodic division.

[0100] In a preferred embodiment, the analysis and control center uses the real-time data simulation to obtain the real-time observation waveform of the corresponding motor data stream, such as... Figure 8 As shown, it specifically includes:

[0101] In step 501, the start time (i.e., the start time of the corresponding segmentation period), the end time (i.e., the end time of the corresponding segmentation period), and key parameters are input into the waveform fitting model of the corresponding waveform category to fit and obtain the real-time observed waveform segment corresponding to the real-time data.

[0102] In step 502, the real-time observation waveform is obtained by splicing multiple real-time observation waveform segments in chronological order.

[0103] In this process, the waveform fitting model and the waveform extraction model for the corresponding waveform category are trained together, such as... Figure 9 and Figure 10 As shown, it specifically includes:

[0104] In step 601, experts annotate historical motor data segments to obtain a training dataset. The annotation content includes key parameters. The historical motor data segments are obtained by analyzing the segmentation period from historically acquired motor data streams and environmental data streams, and then segmenting the historically acquired motor data streams according to the segmentation period. In practical use, the annotation content also includes waveform categories. Based on different waveform categories, the corresponding annotated historical motor data segments are assigned to the corresponding training datasets.

[0105] In step 602, the training dataset is divided into a first dataset and a second dataset according to a preset ratio; the preset ratio is obtained by those skilled in the art based on experience.

[0106] In step 603, the first dataset is input into the waveform extraction initial model for the first stage of training to obtain the waveform extraction intermediate model; wherein, the first stage of training uses the first loss function.

[0107] In step 604, the first dataset is input into the waveform fitting initial model for the second stage of training to obtain the waveform fitting intermediate model; wherein the second stage of training uses the second loss function.

[0108] In step 605, the second dataset is input into the waveform extraction intermediate model, and the key parameters output by the waveform extraction intermediate model are input into the waveform fitting intermediate model to perform a third-stage training on the waveform extraction intermediate model and the waveform fitting intermediate model, resulting in a waveform extraction model and a waveform fitting model. The third-stage training uses a third loss function. Furthermore, in the third-stage training, the parameters of the waveform fitting intermediate model are adjusted first. If, after M adjustments, the convergence of the third loss function is still not achieved, then the parameters of the waveform extraction intermediate model are adjusted. Figure 10 In the diagram, CNN_1 represents the initial and intermediate waveform extraction models, while CNN_2 represents the initial and intermediate waveform fitting models.

[0109] The initial waveform extraction model and the initial waveform fitting model can be CNN models. The first and second training stages are both for training the intermediate waveform extraction model and the intermediate waveform fitting model to a usable level. In actual use, since waveform fitting is an approximate fitting, most waveform fittings cannot obtain a completely accurate fitting result. Similarly, for the same data segment, the key parameters labeled by different experts may also be different. In actual use, this embodiment needs to ensure the accuracy of the waveforms fitted in the group, not the accuracy of the key parameters. Therefore, this embodiment adds a third training stage after the first and second training stages. This training stage uses the intermediate waveform extraction model and the intermediate waveform fitting model to be trained together, thereby ensuring that the finally trained waveform extraction model and waveform fitting model can be successfully integrated, thus ensuring the accuracy of waveform fitting.

[0110] The first loss function is: The second loss function is: The third loss function is: ; For the predicted i-th key parameter, For the i-th key parameter obtained from the annotation, To predict the j-th value in the obtained instantaneous observed waveform segment, For the j-th value in the predicted historical motor data segment, and The preset weights were obtained by those skilled in the art based on experience.

[0111] The waveform categories include one or more of the following: constant category, triangular wave type, sine wave type, composite wave type, and data point fitting type; the key parameters corresponding to the constant category include constant values; the waveform of the constant category can be represented as follows: , It is a constant value.

[0112] The key parameters corresponding to the triangular wave type include amplitude, period, phase shift, and vertical shift; the waveform of the triangular wave type can be represented as follows: ,in, For amplitude, For a period of time, For phase shift, This is a vertical offset.

[0113] The key parameters corresponding to the aforementioned sine wave type include amplitude, frequency, phase shift, and vertical shift. The waveform of the sine wave type can be represented as follows: ; For amplitude, For frequency, For phase shift, This is a vertical offset.

[0114] The key parameters corresponding to the composite wave type include multiple amplitudes, multiple frequencies, multiple phase shifts, and vertical shifts. For example, if it includes three sets of amplitudes, frequencies, phase shifts, and vertical shifts, the waveform of the composite wave type can be represented as follows: .in, , and For amplitude, , and For frequency, , and For phase shift, This is a vertical offset.

[0115] The key parameters corresponding to the data point fitting type include multiple key data points, each of which includes a time and a value. The corresponding waveform is obtained by fitting multiple key data points. When a motor data segment does not conform to the constant category, triangular wave type, sine wave type, or composite wave type, it is identified as a data point fitting type.

[0116] When a composite wave is obtained by superimposing two sine waves... , , The value can be 0, and so on.

[0117] In some application scenarios, the number of motors in motor arrays in different regions may vary. In such cases, the required data size may also differ significantly. To address this issue, this embodiment also provides a preferred implementation method, whereby the coordination data server pre-requests the establishment of a first network slice and a second network slice from the 5G core network. Figure 11 As shown, it specifically includes:

[0118] In step 701, after establishing a local area network with the coordination data server, each motor array and each environmental sensor sends a heartbeat packet to the coordination data server. The heartbeat packet carries the device identifier of each motor in the motor array and the device identifier of the environmental sensor. The device identifier can be a Subscription Concealed Identifier (SUCI).

[0119] In step 702, the coordination data server determines the first slice size of the required first network slice and the second slice size of the required second network slice based on the number of heartbeat packets and / or the device identifiers in the heartbeat packets. In practical use, two fixed resource sizes S1 and S2 can be preset for each motor, where S1 represents the resource size required for the motor data stream and S2 represents the resource size required for real-time data. A fixed resource size Ss can be preset for each environmental sensor. Based on the device identifiers carried in the heartbeat packets, the number of motors Ne and the number of environmental sensors Ns are identified, thereby calculating the first slice size as Ne×S2+Sr1 and the second slice size as Ne×S2+Ns×Ss+Sr2, where Sr1 is the reserved resource size of the first network slice and Sr2 is the reserved resource size of the first network slice.

[0120] In step 703, the Coordination Data Server sends a network slice registration request to the Access and Mobility Management Function (AMF) node. The network slice registration request carries the subscribed NSSAI (Network Slice Selection Assistance Information), the first slice size, and the second slice size, so that the AMF node can create a first network slice that meets the first slice size and a second network slice that meets the second slice size for the Coordination Data Server. The subscribed NSSAI includes the S-NSSAI for uRLLC slices and the S-NSSAI for mMTC slices. Within the 5G core network, after the AMF node sends the first and second slice sizes to the Network Slice Selection Function (NSSF) node, the NSSF node selects the corresponding network slice (i.e., the first network slice) based on the first slice size and the corresponding network slice (i.e., the second network slice) based on the second slice size, and returns the S-NSSAI of the selected network slice to the AMF node, which then returns it to the Coordination Data Server for use.

[0121] In practical use, such as Figure 12 As shown, the method described in this embodiment further includes:

[0122] In step 801, the analysis and control center finds the data point in the real-time observed waveform that corresponds to the latest timestamp of the motor data stream according to the latest received motor data stream at preset intervals, and records the data point as the interval point.

[0123] In step 802, and when the next refresh cycle arrives, the portion of the instantaneous observation waveform before the interval point is drawn using a first color, and the portion of the instantaneous observation waveform after the interval point is drawn using a second color, so that the user can distinguish the portion of the motor data stream that has been received from the interval point from the portion of the motor data stream that has not been received based on the first color and the second color.

[0124] In step 803, the analysis and control center also responds to the user's click action on the real-time observed waveform, obtains the click coordinates based on the click action, obtains the corresponding data point in the real-time observed waveform based on the click coordinates, and if the data point is located before the interval point, then obtains the preset interval length centered on the interval point (e.g., Figure 13 The first motor data segment (shown) is displayed in the details window for user observation. The preset interval length is obtained by those skilled in the art based on requirements analysis. The interval before and after the interval point are measured in terms of time dimension, that is, data points with times earlier than the interval point are data points before the interval point, and data points with times later than the interval point are data points after the interval point.

[0125] Example 2:

[0126] like Figure 14 The diagram shown is an architectural schematic of a green energy power data transmission device based on 5G slicing according to an embodiment of the present invention. This embodiment of the green energy power data transmission device based on 5G slicing includes one or more processors 21 and a memory 22. Wherein, Figure 14 Take a processor 21 as an example.

[0127] Processor 21 and memory 22 can be connected via a bus or other means. Figure 14 Taking the example of a connection between China and Israel via a bus.

[0128] The memory 22, as a non-volatile computer-readable storage medium, can be used to store non-volatile software programs and non-volatile computer-executable programs, such as the green energy power data transmission method based on 5G slicing in Embodiment 1. The processor 21 executes the green energy power data transmission method based on 5G slicing by running the non-volatile software programs and instructions stored in the memory 22.

[0129] Memory 22 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some embodiments, memory 22 may optionally include memory remotely located relative to processor 21, which can be connected to processor 21 via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0130] The program instructions / modules are stored in the memory 22. When executed by one or more processors 21, they execute the green energy power data transmission method based on 5G slicing in Embodiment 1.

[0131] This embodiment also provides a green energy power data transmission system based on 5G slicing, including a motor array, environmental sensors, a coordination data server, and an analysis and control center; the coordination data server uses the green energy power data transmission method based on 5G slicing described in Embodiment 1 to report data.

[0132] It is worth noting that the information interaction and execution process between the modules and units in the above-mentioned device and system are based on the same concept as the processing method embodiment of the present invention. For details, please refer to the description in the method embodiment of the present invention, and will not be repeated here.

[0133] Those skilled in the art will understand that all or part of the steps in the various methods of the embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, which may include: read-only memory (ROM), random access memory (RAM), magnetic disk or optical disk, etc.

[0134] 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, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for transmitting green energy power data based on 5G slicing, characterized in that, To set up a coordination data server for the motor array, a local area network is established between the motor array, environmental sensors, and the coordination data server. Methods include: The coordination data server requests the establishment of a first network slice and a second network slice from the 5G core network in advance; wherein, the first network slice is a uRLLC slice and the second network slice is an mMTC slice; The data server coordinates the use of the local area network to receive data streams from each motor in the motor array and environmental data streams from each environmental sensor. Analyze the environmental data stream to determine multiple segmentation periods; The motor data stream is segmented according to the multiple segmentation cycles to obtain multiple motor data segments. The motor data segments are input into a classification model to obtain the waveform category of the motor data segments. The motor data segments are then input into the waveform extraction model of the corresponding waveform category to extract the key parameters of the motor data segments. The start time, end time, waveform type, and key parameters of each motor data segment are used as real-time data. The real-time data is sent to the analysis and control center using the first network slice so that the analysis and control center can use the real-time data to simulate and obtain the real-time observation waveform of the corresponding motor data stream. The coordination data server also uses a second network slice to send the data streams of each motor to the analysis and control center so that the analysis and control center can trace the data later. Wherein, when the motor array is a wind turbine array, the environmental data stream includes at least a wind speed data stream; when the motor array is a photovoltaic array, the environmental data stream includes at least a light intensity data stream.

2. The green energy power data transmission method based on 5G slicing according to claim 1, characterized in that, When the motor array is a wind turbine array, the analysis of the environmental data stream to determine multiple segmentation periods specifically includes: The wind speed data stream is segmented once to divide the wind speed data stream into a first wind speed data segment less than or equal to a first preset value, a second wind speed data segment greater than the first preset value and less than or equal to a second preset value, a third wind speed data segment greater than the second preset value and less than a third preset value, and a fourth wind speed data segment greater than or equal to a third preset value. Differential identification is performed on the second wind speed data segment to identify the troughs in the second wind speed data segment. Using the troughs as the dividing boundary, the second wind speed data segment is further segmented to obtain multiple fifth wind speed data segments. Multiple segmentation periods are determined based on the time periods of the first, third, fourth, and fifth wind speed data segments.

3. The green energy power data transmission method based on 5G slicing according to claim 1, characterized in that, When the motor array is a photovoltaic array, the analysis of the environmental data stream to determine multiple segmentation periods specifically includes: The light intensity data stream is segmented once to divide the light intensity data stream into a first light intensity data segment that is less than a fourth preset value and a second light intensity data segment that is greater than or equal to the fourth preset value; Multiple segmentation periods are determined based on the time periods of the first and second light intensity data segments.

4. The green energy power data transmission method based on 5G slicing according to claim 1, characterized in that, The analysis and control center uses the real-time data simulation to obtain the real-time observation waveform of the corresponding motor data stream, specifically including: Input the start time, end time, and key parameters into the waveform fitting model of the corresponding waveform category to fit the real-time observed waveform segment corresponding to the real-time data. The real-time observation waveform is obtained by splicing together multiple real-time observation waveform segments in chronological order.

5. The method for green energy power data transmission based on 5G slicing according to claim 4, characterized in that, The waveform fitting model and waveform extraction model for the corresponding waveform category are trained together, specifically including: The training dataset was obtained by annotating historical motor data segments by experts; the annotations included key parameters. The training dataset is divided into a first dataset and a second dataset according to a preset ratio; The first dataset is input into the waveform extraction initial model for the first stage of training to obtain the waveform extraction intermediate model; wherein, the first stage of training uses the first loss function. The first dataset is input into the initial waveform fitting model for the second stage of training to obtain an intermediate waveform fitting model; wherein, the second stage of training uses the second loss function. The second dataset is input into the waveform extraction intermediate model, and the key parameters output by the waveform extraction intermediate model are input into the waveform fitting intermediate model to perform a third-stage training on the waveform extraction intermediate model and the waveform fitting intermediate model to obtain the waveform extraction model and the waveform fitting model. The third-stage training uses a third loss function. The first loss function is: The second loss function is: The third loss function is: ; For the predicted i-th key parameter, For the i-th key parameter obtained from the annotation, To predict the j-th value in the obtained instantaneous observed waveform segment, For the j-th value in the predicted historical motor data segment, and Preset weights.

6. The green energy power data transmission method based on 5G slicing according to claim 1, characterized in that, The coordination data server requests the establishment of a first network slice and a second network slice from the 5G core network in advance, specifically including: After establishing a local area network with the coordination data server, each motor array and each environmental sensor sends a heartbeat packet to the coordination data server, which carries the device identifier of each motor in the motor array and the device identifier of the environmental sensor. The coordination data server determines the first slice size of the required first network slice and the second slice size of the required second network slice based on the number of heartbeat packets and / or the device identifiers in the heartbeat packets; The Coordinating Data Server sends a network slice registration request to the AMF node; wherein the network slice registration request carries the signed NSSAI, the first slice size, and the second slice size, so that the AMF node can establish a first network slice that meets the first slice size and a second network slice that meets the second slice size for the Coordinating Data Server; wherein the signed NSSAI includes the S-NSSAI of uRLLC slices and the S-NSSAI of mMTC slices.

7. The green energy power data transmission method based on 5G slicing according to claim 1, characterized in that, The method also includes: analyzing the control center at preset intervals, finding the data point in the real-time observed waveform that corresponds to the latest timestamp of the motor data stream according to the latest received motor data stream, and recording the data point as the interval point; And when the next refresh cycle arrives, the portion of the real-time observed waveform before the interval point is drawn using the first color, and the portion of the real-time observed waveform after the interval point is drawn using the second color, so that the user can distinguish the portion of the motor data stream that has been received from the motor data stream and the portion of the motor data stream that has not been received from the motor data stream based on the first color and the second color. The analysis and control center also responds to the user's click action on the real-time observed waveform, obtains the click coordinates based on the click action, obtains the corresponding data point in the real-time observed waveform based on the click coordinates, and if the data point is located before the interval point, obtains the first motor data segment within a preset interval length centered on the interval point, and displays the first motor data segment in the details window.

8. The green energy power data transmission method based on 5G slicing according to claim 1, characterized in that, The waveform categories include one or more of the following: constant category, triangular wave type, sine wave type, composite wave type, and data point fitting type; The key parameters corresponding to the constant category include constant values; The key parameters corresponding to the triangular wave type include amplitude, period, phase shift, and vertical shift; The key parameters corresponding to the sine wave type include amplitude, frequency, phase shift, and vertical shift; The key parameters corresponding to the composite wave type include multiple amplitudes, multiple frequencies, multiple phase shifts, and vertical shifts; The key parameters corresponding to the data point fitting type include multiple key data points.

9. A non-volatile computer storage medium, characterized in that, The computer storage medium stores computer-executable instructions, which are executed by one or more processors to perform the green energy power data transmission method based on 5G slicing as described in any one of claims 1-8.

10. A green energy power data transmission system based on 5G slicing, characterized in that, This includes motor arrays, environmental sensors, a coordinated data server, and an analysis and control center; The coordinated data server uses the green energy power data transmission method based on 5G slicing as described in any one of claims 1-8 to report data.

Images