Photovoltaic power station short-term power intelligent prediction method based on TXP-MTGNN model

Abstract

The invention relates to the field of intelligent prediction, in particular to a photovoltaic power station short-term power intelligent prediction method based on a TXP-MTGNN model, and the method comprises the steps: carrying out the data collection of current data in a photovoltaic power station through arranging a plurality of current sensors, carrying out the data analysis of an adjustment group, and judging whether the current data reaches a set value or not; the infrared sensor performs electric signal conversion on an optical signal corresponding to the current data, the collector collects and preprocesses the electric signal, the preprocessed data is learned according to a TXP-MTGNN model, a corresponding power generation power distribution diagram is generated, and a damage distribution result of the ground cable in the photovoltaic power station is analyzed and judged. The specific damaged position of the ground cable is judged and repaired through machine learning, the accuracy of the short-term power prediction result of the photovoltaic power station is improved, the situation that the model cannot recognize the specific position due to the fact that the ground cable is damaged due to special conditions is avoided, the running safety of the photovoltaic power station is enhanced, and the service life of the photovoltaic power station is prolonged. And real-time intelligent prediction, regulation and control of the photovoltaic power station are realized.

Description

Technical Field

[0001] The present invention relates to the field of intelligent prediction, and in particular to a method for intelligent short-term power prediction of a photovoltaic power station based on a TXP-MTGNN model. Background Art

[0002] Photovoltaic power generation is widely used in power generation systems as a power generation mode with low pollution, wide sources and large reserves. In my country's new energy development strategy, solar energy has been developed and utilized on a large scale, and photovoltaic power generation systems have developed rapidly. According to the latest data provided by the National Energy Administration, my country's new photovoltaic power generation grid-connected capacity in 2022 was 87.408GW, a year-on-year increase of 60.3%, including 36.3GW of centralized photovoltaics and 51.11GW of distributed photovoltaics. By the end of 2022, my country's photovoltaic grid-connected installed capacity reached 392.61GW, a year-on-year increase of 28.1%. Large-scale photovoltaic power generation grid-connected is an effective way to achieve emission reduction. Due to the characteristics of photovoltaic power generation with large output volatility and randomness, the grid-connected photovoltaic system will have an impact on the reliability and stability of the original power grid, affecting the real-time dispatch and safe operation of the power grid. Therefore, providing accurate and effective photovoltaic power station power forecast is very important for the economic and safe operation of photovoltaic power stations.

[0003] China Patent Application Publication No.: CN116247651A discloses a distributed photovoltaic power intelligent prediction method and device, the method comprising: training a preset prediction model to be trained according to the target training data of all weather types determined in advance for training, obtaining a trained prediction model, and judging whether the trained prediction model converges, if so, determining the trained prediction model as the target prediction model; the target prediction model is used to predict the future photovoltaic power parameters of the photovoltaic power station. It can be seen that the implementation of the present invention can train a prediction model for predicting future photovoltaic power parameters, so as to intelligently predict the future photovoltaic power parameters according to the current weather type of the photovoltaic power station, which is conducive to reducing the impact of weather conditions on photovoltaic power parameters, and then to improving the predicted future photovoltaic power parameters, so as to facilitate timely dispatching of the power grid according to the future photovoltaic power parameters, so that the power grid can operate normally and ensure normal electricity consumption for users.

[0004] China Patent Authorization Announcement No.: CN112512110B discloses a wireless transmission power prediction control method constrained by reliability demand of intelligent substation, and its steps include: firstly, by analyzing the relationship among reliability, signal-to-noise ratio and transmission power, using the loss model of wireless signal propagation and Kalman filter, constructing the state space equation of wireless signal propagation with state observation; secondly, compensating the expected signal-to-noise ratio by estimating the confidence interval of signal-to-noise ratio fluctuation in real time; finally, defining a quadratic cost function with the transmission power and the difference between the actual signal-to-noise ratio and the expected signal-to-noise ratio as the target, and realizing the optimal control of the transmission power through a dynamic programming algorithm. The present invention can solve the coordination problem between the transmission power of wireless sensor networks and the reliability of wireless communication in substation environment, thereby reducing the transmission power of nodes as much as possible while ensuring the reliability of wireless communication.

[0005] However, the above method has the following problems: the prediction model has a simple structure, resulting in insufficient accuracy of the short-term power prediction results of the photovoltaic power station. Summary of the invention

[0006] To this end, the present invention provides a photovoltaic power station short-term power intelligent prediction method based on the TXP-MTGNN model to overcome the problem that the prediction model structure of the prior art is simple, resulting in insufficient accuracy of the photovoltaic power station short-term power prediction results.

[0007] To achieve the above object, the present invention provides a photovoltaic power station short-term power intelligent prediction method based on the TXP-MTGNN model, comprising:

[0008] Several current sensors are set up to collect current data of each photovoltaic device in the photovoltaic power station;

[0009] Using the adjustment group to perform data analysis on the current data to determine whether the current data reaches a set value;

[0010] Using an infrared sensor to convert an optical signal corresponding to the current data into an electrical signal; the optical signal is converted into an electrical signal;

[0011] Collecting the electrical signal, and preprocessing the electrical signal to form corresponding preprocessing data;

[0012] The pre-processed data is learned according to the TXP-MTGNN model to generate a corresponding power generation distribution diagram;

[0013] The power generation distribution diagram is analyzed to determine the distribution result of the ground cable damage in the photovoltaic power station.

[0014] Furthermore, the photovoltaic power station is provided with a plurality of current sensors, each of which is provided with a corresponding position number.

[0015] Wherein, the current sensor adopts a parallel transmission mode to receive the current data at the corresponding acquisition line position.

[0016] Furthermore, the current data collected by the current sensor is position-marked according to the position number, and the current data that has been position-marked is collected and packaged.

[0017] Furthermore, the step of the adjustment group performing data analysis on the current data includes:

[0018] When the current data does not reach the set value, keeping the current switch in an open state;

[0019] The current sensor continues to collect the current data;

[0020] When the current data reaches the set value, the current switch is turned off.

[0021] Furthermore, the step of converting the optical signal into an electrical signal comprises:

[0022] After the current switch is turned off, the infrared sensor amplifies the optical signal corresponding to the current data through an operational amplifier;

[0023] Performing data analysis on the optical signal after signal amplification;

[0024] The optical signal is converted into a corresponding electrical signal.

[0025] Further, the step of constructing a data set of the TXP-MTGNN model according to the electrical signal includes:

[0026] Set the standard number of single training samples for the TXP-MTGNN model;

[0027] Determine the initialization parameters of the TXP-MTGNN model according to the number of standard single training samples;

[0028] Construct the TXP-MTGNN model according to the initialization parameters;

[0029] Input the data set into the TXP-MTGNN model to obtain a prediction result, calculate the accuracy of the prediction result and verify it;

[0030] Among them, if the accuracy rate is less than the accuracy rate threshold, the training of the TXP-MTGNN model continues, and if the accuracy rate is greater than the accuracy rate threshold, the verified TXP-MTGNN model is obtained.

[0031] Furthermore, the step of collecting and preprocessing the electrical signal includes:

[0032] basing the electrical signal on data acquisition features of the data set;

[0033] Perform data cleaning to obtain corresponding cleaned data;

[0034] Performing standardization processing on the cleaned data;

[0035] The standardization process is to segment the cleaned data according to a standard sampling rate to obtain corresponding pre-processed data.

[0036] Furthermore, the pre-processed data is transmitted to the TXP-MTGNN model for learning to obtain a corresponding power generation distribution diagram.

[0037] Wherein, a power threshold is provided in the power generation distribution diagram;

[0038] The power threshold is the lowest power value corresponding to the current data.

[0039] Furthermore, the power generation distribution diagram is analyzed, and when the power generation distribution diagram is greater than the power threshold, no adjustment is made to the aboveground cable.

[0040] Furthermore, when the power generation distribution diagram is less than the power threshold, the damaged position of the aboveground cable is determined according to the position number, and the damaged aboveground cable is repaired.

[0041] Compared with the prior art, the present invention collects current data in a photovoltaic power station by setting up a number of current sensors, the adjustment group performs data analysis to determine whether the current data reaches a set value, the infrared sensor converts the optical signal corresponding to the current data into an electrical signal, the collector collects the electrical signal and performs preprocessing, and learns the preprocessed data according to the TXP-MTGNN model to generate a corresponding power generation distribution map, and analyzes it to determine the distribution results of ground cable damage in the photovoltaic power station, and determines the specific damaged location of the ground cable through machine learning and repairs it, thereby improving the accuracy of the short-term power prediction results of the photovoltaic power station, avoiding the situation where the model cannot identify the specific location due to the damage of the ground cable due to special circumstances, enhancing the safety of the operation of the photovoltaic power station, and realizing real-time intelligent prediction and regulation of the photovoltaic power station.

[0042] Furthermore, by setting a position number and performing position marking processing on the current data, the specific position corresponding to the current data can be accurately identified, which facilitates locating the damaged position of the cable and repairing it.

[0043] Furthermore, by analyzing the current data, it is possible to accurately determine whether the current data reaches the set value, thereby improving the accuracy of the short-term power prediction results of the photovoltaic power station.

[0044] Furthermore, by converting optical signals into electrical signals, conversion between optical signals and electrical signals can be achieved, which facilitates monitoring of the power corresponding to the current data, further improves the accuracy of short-term power prediction results of photovoltaic power stations, and enhances the safety of photovoltaic power station operation.

[0045] Furthermore, by constructing the TXP-MTGNN model, it is convenient to predict the short-term power of photovoltaic power stations in real time, reducing the difficulty of intelligent prediction.

[0046] Furthermore, the accuracy of intelligent prediction of the TXP-MTGNN model is improved by collecting and preprocessing electrical signals, avoiding the interference of erroneous data on the prediction results during the prediction process.

[0047] Furthermore, by analyzing the power generation distribution diagram and adjusting the damaged portion of the ground cable, the safety of the photovoltaic power station operation is enhanced, and real-time intelligent prediction and regulation of the photovoltaic power station is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 It is a flow chart of the photovoltaic power station short-term power intelligent prediction method based on the TXP-MTGNN model of the present invention;

[0049] Figure 2 A flow chart of data analysis of current data by the adjustment group of the present invention;

[0050] Figure 3 A flow chart of converting an optical signal into an electrical signal according to the present invention;

[0051] Figure 4 A flowchart of a data set for constructing a TXP-MTGNN model based on electrical signals in the present invention;

[0052] Figure 5 The present invention is a flow chart of collecting and preprocessing electrical signals. DETAILED DESCRIPTION

[0053] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0054] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0055] It should be noted that, in the description of the present invention, terms such as "up", "down", "left", "right", "inside" and "outside" indicating directions or positional relationships are based on the directions or positional relationships shown in the drawings. This is merely for the convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it cannot be understood as a limitation on the present invention.

[0056] In addition, it should be noted that in the description of the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0057] See also Figure 1 As shown, it is a flow chart of the photovoltaic power station short-term power intelligent prediction method based on the TXP-MTGNN model of the present invention, including:

[0058] Step S1, setting a number of current sensors to collect current data of each photovoltaic device in the photovoltaic power station;

[0059] Step S2, using the adjustment group to analyze the current data to determine whether the current data reaches the set value;

[0060] Step S3, using an infrared sensor to convert the optical signal corresponding to the current data into an electrical signal;

[0061] Step S4, collecting electrical signals and preprocessing the electrical signals to form corresponding preprocessing data;

[0062] Step S5, learning the preprocessed data according to the TXP-MTGNN model to generate a corresponding power generation distribution map;

[0063] Step S6, analyzing the power generation distribution diagram to determine the distribution result of ground cable damage in the photovoltaic power station.

[0064] Several current sensors are set up to collect current data in the photovoltaic power station. The adjustment group performs data analysis to determine whether the current data reaches the set value. The infrared sensor converts the optical signal corresponding to the current data into an electrical signal. The collector collects the electrical signal and performs preprocessing. The preprocessed data is learned according to the TXP-MTGNN model to generate the corresponding power distribution map. The map is analyzed to determine the distribution of ground cable damage in the photovoltaic power station. The specific damaged location of the ground cable is determined and repaired through machine learning. This improves the accuracy of the short-term power prediction results of the photovoltaic power station, avoids the situation where the model cannot identify the specific location due to the damage of the ground cable due to special circumstances, enhances the safety of the operation of the photovoltaic power station, and realizes real-time intelligent prediction and regulation of the photovoltaic power station.

[0065] Specifically, a photovoltaic power station is provided with a number of current sensors, each of which has a corresponding position number.

[0066] Among them, the current sensor adopts a parallel transmission mode to receive the current data of the corresponding acquisition line position.

[0067] Specifically, the current data collected by the current sensor is position-marked according to the position number, and the current data that has been position-marked is collected and packaged.

[0068] By setting the position number and performing position marking processing on the current data, the specific position corresponding to the current data can be accurately identified, making it easier to locate the damaged cable and repair it.

[0069] See also Figure 2 As shown, it is a flow chart of the adjustment group of the present invention performing data analysis on current data, including:

[0070] Step St1, when the current data does not reach the set value, keep the current switch in the open state;

[0071] Step St2, the current sensor continues to collect current data;

[0072] Step St3, until the current data reaches the set value, the current switch is turned off.

[0073] Embodiment 1:

[0074] The set value is 100A, and the A current data is 80A. The set value is not reached, so the current switch remains open and the current sensor continues to collect A current data.

[0075] Embodiment 2:

[0076] The set value is 100A, the A current data is 180A, and when the set value is reached, the current switch is turned off.

[0077] By analyzing the current data, it is possible to accurately determine whether the current data reaches the set value, thereby improving the accuracy of the short-term power prediction results of the photovoltaic power station.

[0078] See also Figure 3 As shown, it is a flow chart of converting an optical signal into an electrical signal according to the present invention, including:

[0079] Step Sp1, after the current switch is turned off, the infrared sensor amplifies the optical signal corresponding to the current data through an operational amplifier;

[0080] Step Sp2, performing data analysis on the optical signal after signal amplification;

[0081] Step Sp3, converting the optical signal into a corresponding electrical signal.

[0082] By converting optical signals into electrical signals, the conversion between optical signals and electrical signals can be realized, which facilitates the monitoring of the power corresponding to the current data, further improves the accuracy of the short-term power prediction results of the photovoltaic power station, and enhances the safety of the operation of the photovoltaic power station.

[0083] See also Figure 4 As shown, it is a flow chart of constructing a data set of a TXP-MTGNN model according to an electrical signal in the present invention, including:

[0084] Step Stp1, setting the standard number of single training samples for the TXP-MTGNN model;

[0085] Step Stp2, determining the initialization parameters of the TXP-MTGNN model according to the number of standard single training samples;

[0086] Step Stp3, construct the TXP-MTGNN model according to the initialization parameters;

[0087] Step Stp4, input the data set into the TXP-MTGNN model to obtain the prediction results, calculate the accuracy of the prediction results and verify them;

[0088] Among them, if the accuracy rate is less than the accuracy rate threshold, the training of the TXP-MTGNN model continues, and if the accuracy rate is greater than the accuracy rate threshold, the verified TXP-MTGNN model is obtained.

[0089] By constructing the TXP-MTGNN model, it is convenient to predict the short-term power of photovoltaic power stations in real time, reducing the difficulty of intelligent prediction.

[0090] Specifically, the steps of collecting and preprocessing the electrical signals include:

[0091] Step Ss1, collecting data features of the electrical signal based on the data set;

[0092] Step Ss2, performing data cleaning to obtain corresponding cleaned data;

[0093] Step Ss3, standardizing the cleaned data;

[0094] Among them, the standardization processing is to divide the cleaned data according to the standard sampling rate to obtain the corresponding pre-processed data.

[0095] In the specific implementation, when the standard sampling rate is 10 / second, the TXP-MTGNN model has the best analysis effect on the preprocessed data and the most accurate prediction result for the short-term power of the photovoltaic power station.

[0096] By collecting and preprocessing electrical signals, the accuracy of the TXP-MTGNN model's intelligent prediction is improved, and the interference of erroneous data on the prediction results during the prediction process is avoided.

[0097] Specifically, the preprocessed data is transferred to the TXP-MTGNN model for learning to obtain the corresponding power generation distribution map.

[0098] Among them, a power threshold is set in the power generation distribution diagram;

[0099] The power threshold is the minimum power value corresponding to the current data.

[0100] Specifically, the power generation distribution diagram is analyzed, and when the power generation distribution diagram is greater than a power threshold, no adjustment is made to the aboveground cable.

[0101] Embodiment 3:

[0102] The power threshold is 100Kw / h. When the values ​​of the power distribution diagram are all greater than 100Kw / h, no adjustment is made to the ground cables.

[0103] Specifically, when the power distribution diagram is less than the power threshold, the damaged position of the aboveground cable is determined according to the position number, and the damaged aboveground cable is repaired.

[0104] Embodiment 4:

[0105] The power threshold is 100Kw / h. When the value of the power generation distribution diagram appears to be 80Kw / h, 80Kw / h<100Kw / h. It is found that the position number corresponding to the electrical signal corresponding to the power generation is 12. It is determined that the damaged position of the ground cable is 12, and the ground cable at the damaged position is repaired.

[0106] In specific implementations, if the aboveground cable is damaged due to external forces (such as rats biting the wires), causing the equipment to go offline, the TXP-MTGNN model can still identify the damaged location of the aboveground cable and repair it.

[0107] By analyzing the power generation distribution diagram and adjusting the damaged part of the ground cable, the safety of the operation of the photovoltaic power station is enhanced, and real-time intelligent prediction and regulation of the photovoltaic power station is achieved.

[0108] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

[0109] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A photovoltaic power station short-term power intelligent prediction method based on the TXP-MTGNN model is characterized by: include: Several current sensors are set up to collect current data of each photovoltaic device in the photovoltaic power station; Using the adjustment group to perform data analysis on the current data to determine whether the current data reaches a set value; Using an infrared sensor to convert an optical signal corresponding to the current data into an electrical signal; the optical signal is converted into an electrical signal; Collecting the electrical signal, and preprocessing the electrical signal to form corresponding preprocessing data; The pre-processed data is learned according to the TXP-MTGNN model to generate a corresponding power generation distribution diagram; The power generation distribution diagram is analyzed to determine the distribution result of the ground cable damage in the photovoltaic power station.

2. The photovoltaic power station short-term power intelligent prediction method based on the TXP-MTGNN model according to claim 1 is characterized in that: The photovoltaic power station is provided with a plurality of current sensors, each of which is provided with a corresponding position number. Wherein, the current sensor adopts a parallel transmission mode to receive the current data at the corresponding acquisition line position.

3. The photovoltaic power station short-term power intelligent prediction method based on the TXP-MTGNN model according to claim 2 is characterized in that: The current data collected by the current sensor is subjected to position marking processing according to the position number, and the current data subjected to the position marking processing is collected and packaged.

4. The photovoltaic power station short-term power intelligent prediction method based on the TXP-MTGNN model according to claim 3 is characterized in that: The step of the adjustment group performing data analysis on the current data comprises: When the current data does not reach the set value, keeping the current switch in an open state; The current sensor continues to collect the current data; When the current data reaches the set value, the current switch is turned off.

5. The photovoltaic power station short-term power intelligent prediction method based on the TXP-MTGNN model according to claim 4 is characterized in that: The step of converting the optical signal into an electrical signal comprises: After the current switch is turned off, the infrared sensor amplifies the optical signal corresponding to the current data through an operational amplifier; Performing data analysis on the optical signal after signal amplification; The optical signal is converted into a corresponding electrical signal.

6. The photovoltaic power station short-term power intelligent prediction method based on the TXP-MTGNN model according to claim 5 is characterized in that: The step of constructing a data set of the TXP-MTGNN model according to the electrical signal comprises: Set the standard number of single training samples for the TXP-MTGNN model; Determine the initialization parameters of the TXP-MTGNN model according to the number of standard single training samples; Construct the TXP-MTGNN model according to the initialization parameters; Input the data set into the TXP-MTGNN model to obtain a prediction result, calculate the accuracy of the prediction result and verify it; Among them, if the accuracy rate is less than the accuracy rate threshold, the training of the TXP-MTGNN model continues, and if the accuracy rate is greater than the accuracy rate threshold, the verified TXP-MTGNN model is obtained.

7. The photovoltaic power station short-term power intelligent prediction method based on the TXP-MTGNN model according to claim 6 is characterized in that: The steps of collecting and preprocessing the electrical signal include: basing the electrical signal on data acquisition features of the data set; Perform data cleaning to obtain corresponding cleaned data; Performing standardization processing on the cleaned data; The standardization process is to segment the cleaned data according to a standard sampling rate to obtain corresponding pre-processed data.

8. The photovoltaic power station short-term power intelligent prediction method based on the TXP-MTGNN model according to claim 7 is characterized in that: The pre-processed data is transmitted to the TXP-MTGNN model for learning to obtain the corresponding power generation distribution diagram. Wherein, a power threshold is provided in the power generation distribution diagram; The power threshold is the lowest power value corresponding to the current data.

9. The photovoltaic power station short-term power intelligent prediction method based on the TXP-MTGNN model according to claim 8 is characterized in that: The power generation distribution diagram is analyzed, and when the power generation distribution diagram is greater than the power threshold, no adjustment is made to the aboveground cable.

10. The photovoltaic power station short-term power intelligent prediction method based on the TXP-MTGNN model according to claim 9 is characterized in that: When the power generation distribution diagram is less than the power threshold, the damaged position of the aboveground cable is determined according to the position number, and the damaged aboveground cable is repaired.

Images