A distributed photovoltaic fire early warning method

Abstract

The application discloses a kind of distributed photovoltaic fire early warning method, it is related to fire early warning technical field, and its technical solution points are: the current, voltage and infrared temperature matrix of photovoltaic module are collected by detection module;The data collected by detection module is analyzed based on space-time correlation analysis module, and the dynamic weight coefficient of current is output in real time, and dynamically fused to generate weighted feature vector;The position P of photovoltaic module is obtained by positioning component, and the space-time correlation matrix between photovoltaic modules is calculated by space-time correlation analysis module;Weighted feature vector and space-time correlation matrix are analyzed based on LSTM network, and the risk score R (t) of photovoltaic module is output;According to risk score, trigger hierarchical analysis response.The application realizes efficient, accurate photovoltaic fire early warning by multi-source detection, dynamic analysis, hierarchical response.

Description

Technical Field

[0001] This invention relates to the field of fire early warning technology, and more specifically, to a method for early warning of distributed photovoltaic fires. Background Technology

[0002] With the widespread application of renewable energy, solar photovoltaic power generation systems, as a common way of utilizing solar energy resources, have developed rapidly in recent years and are widely used in the field of clean energy supply. However, photovoltaic power generation systems may face various safety hazards during long-term operation. Among them, photovoltaic power generation systems generate heat while generating electricity, which may cause fires. Fires will cause serious damage to the power generation system and even threaten personal safety. Therefore, an efficient and reliable photovoltaic fire early warning method is very important.

[0003] However, existing photovoltaic fire early warning methods still have shortcomings. The existing monitoring methods are very simple, only detecting a single parameter, lacking accuracy and comprehensiveness, which leads to inaccuracy in detection. Moreover, existing photovoltaic fire early warning generally uses traditional temperature sensors, which have low sampling rates and rely on cloud processing for data upload, resulting in low spatiotemporal correlation between photovoltaic modules. This leads to long fire early warning delays, usually only issuing warnings some time after the fire has started, thus increasing the damage and losses caused by the fire.

[0004] Therefore, a new solution is needed to address this problem. Utility Model Content

[0005] The purpose of this invention is to provide a distributed photovoltaic fire early warning method to solve the above-mentioned problems.

[0006] The above-mentioned technical objectives of the embodiments of the present invention are achieved through the following technical solutions:

[0007] In a first aspect, a method for early warning of distributed photovoltaic fires is provided, characterized in that the fire early warning method includes the following steps:

[0008] The current, voltage, and infrared temperature matrix of the photovoltaic module are collected through the detection module.

[0009] The spatiotemporal correlation analysis module analyzes the data collected by the detection module and outputs the dynamic weighting coefficients α(t) of the current, β(t) of the voltage, and γ(t) of the infrared temperature matrix in real time, and dynamically fuses them to generate a weighted feature vector.

[0010] The photovoltaic modules are located by positioning components to obtain the position P of the photovoltaic modules, and the spatiotemporal correlation matrix between the photovoltaic modules is calculated by the spatiotemporal correlation analysis module.

[0011] The spatio-temporal correlation analysis module calculates the spatio-temporal correlation matrix between photovoltaic modules, including:

[0012] The positions of photovoltaic modules A and B obtained by positioning are respectively and , is the weighted eigenvector of photovoltaic module A, is the weighted eigenvector of photovoltaic module B. Then the calculation formula of the spatio-temporal correlation matrix is:

[0013] ;

[0014] Among them, represents the spatio-temporal correlation matrix between photovoltaic modules A and B, d0 is the thermal diffusion characteristic distance, and the calculation formula of the said d0 is:

[0015] d0 = k·υ(t)

[0016] Among them, k is the material thermal conductivity of the photovoltaic module, and υ(t) is the real-time wind speed;

[0017] Based on the LSTM network, analyze the weighted eigenvector and the spatio-temporal correlation matrix, and output the risk score R(t) of the photovoltaic module;

[0018] Trigger a hierarchical analysis response according to the risk score. When 0.6 < R(t) ≤ 0.8, trigger a secondary alarm, reduce the load of the photovoltaic module power, and push the alarm information at the same time; when R(t) > 0.8, trigger a primary alarm, cut off the current, and start the fire extinguishing device at the same time.

[0019] The present invention is further set as: the spatio-temporal correlation analysis module analyzes the data collected by the detection module, including normalizing the collected current and voltage;

[0020] ;

[0021] Among them, represents the data after current normalization processing, represents the data after voltage normalization processing;

[0022] Compress the infrared temperature matrix to obtain .

[0023] The present invention is further set as: the calculation methods of the dynamic weight coefficient α(t) of the output current, the dynamic weight coefficient β(t) of the voltage, and the dynamic weight coefficient of the infrared temperature matrix are as follows:

[0024] ;

[0025] The numerator represents the rate of change of the current, voltage, and infrared temperature matrix, while the denominator is the sum of the rates of change of the current, voltage, and infrared temperature matrix.

[0026] The present invention is further configured such that the calculation method for the dynamically fused weighted feature vector is as follows:

[0027] ;

[0028] Among them, the This represents the weighted eigenvector.

[0029] The present invention is further configured such that: the LSTM network includes a fully connected output layer; the LSTM network captures the evolution law of weighted feature vectors and spatiotemporal correlation matrix by introducing an attention mechanism; the fully connected output layer generates a risk score for photovoltaic modules; and the LSTM network aggregates the risk scores of all photovoltaic modules to generate an overall risk assessment.

[0030] In a second aspect, a computer-readable storage medium is provided, the computer-readable storage medium storing a computer program that, when executed by a computer or processor, implements the method described in any one of the preceding claims.

[0031] Thirdly, a computer program product comprising a computer program that, when executed by a computer or processor, causes the computer or processor to perform the method described in any one of the preceding claims.

[0032] In summary, the present invention has the following beneficial effects:

[0033] 1. The detection module simultaneously collects current, voltage, and infrared temperature data from the photovoltaic module. The collected data is then dynamically fused to generate a weighted feature vector. This weighted feature vector transforms the current, voltage, and temperature parameters collected by different sensors into a unified mathematical expression, facilitating the analysis and comparison of the collected data. By inputting the weighted feature vector into an LSTM network, the state of the photovoltaic module can be analyzed. By collecting and analyzing multiple data sources, the detection of the photovoltaic module becomes more accurate and comprehensive, enabling a more precise determination of its state.

[0034] 2. By locating each photovoltaic module using a positioning component, the position P of each photovoltaic module is obtained. Then, a spatiotemporal correlation matrix between each photovoltaic module is generated through a spatiotemporal correlation analysis module, which enhances the spatiotemporal correlation between the photovoltaic modules. This allows for the prediction of fault propagation paths based on the thermal diffusion and electrical coupling relationships between photovoltaic modules. When local anomalies have not yet exceeded the standard, an alarm is triggered in advance based on the correlation, thereby reminding staff in advance and reducing the occurrence of fires and the harm and losses caused by fires. This invention achieves efficient and accurate photovoltaic fire early warning through multi-source detection, dynamic analysis, and hierarchical response. Attached Figure Description

[0035] Figure 1 This is a schematic diagram of the structure of a distributed photovoltaic fire early warning method according to the present invention; Detailed Implementation

[0036] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0037] See the column for a feasible implementation. Figure 1 As shown, a distributed photovoltaic fire early warning method includes the following steps:

[0038] Step 101: Collect the current, voltage, and infrared temperature matrix of the photovoltaic module through the detection module;

[0039] Step 102: Analyze the data collected by the detection module based on the spatiotemporal correlation analysis module, and output the dynamic weighting coefficients α(t) of the current, β(t) of the voltage, and γ(t) of the infrared temperature matrix in real time, and dynamically fuse them to generate a weighted feature vector;

[0040] Step 103: The photovoltaic module is located by the positioning component to obtain the position P of the photovoltaic module, and the spatiotemporal correlation matrix between the photovoltaic modules is calculated by the spatiotemporal correlation analysis module;

[0041] Step 104: Analyze the weighted feature vector and spatiotemporal correlation matrix based on the LSTM network, and output the risk score R(t) of the photovoltaic module.

[0042] Step 105: Trigger hierarchical analysis response according to the risk score. When 0.6 < R(t) ≤ 0.8, trigger a secondary alarm, reduce the power load of the photovoltaic module, and push the alarm information at the same time. When R(t) > 0.8, trigger a primary alarm, cut off the current, and start the fire extinguishing device at the same time.

[0043] Specifically, the detection module performs multi-source acquisition on the current, voltage, and infrared temperature matrix of the photovoltaic module, and then analyzes the collected data through the spatio-temporal correlation analysis module to obtain the dynamic weight coefficients α(t), β(t), and γ(t) of the current, voltage, and infrared temperature matrix respectively. Then, the spatio-temporal correlation analysis module dynamically fuses to generate a weighted feature vector, and the weighted feature vector will be updated in real time as time changes. By collecting and analyzing various data, the detection of the photovoltaic module is more accurate and comprehensive, and the state of the photovoltaic module can be detected more accurately.

[0044] Further, please refer to Figure 1 As shown, the photovoltaic module is positioned by the positioning component to obtain the position P of each photovoltaic module, and then the spatio-temporal correlation matrix between each photovoltaic module is calculated through the spatio-temporal correlation analysis model, so as to associate each photovoltaic module, enhancing the spatio-temporal correlation between each photovoltaic module. The weighted feature vector and the spatio-temporal correlation matrix are analyzed through the LSTM network, and then the fault spread path can be predicted based on the thermal diffusion and electrical coupling relationship between the photovoltaic modules. When the local anomaly has not exceeded the standard, an alarm is triggered in advance based on the correlation, which can remind the staff in advance and reduce the occurrence of fires and the hazards and losses caused by fires.

[0045] Further, after analyzing the weighted feature vector and the spatio-temporal correlation matrix, the LSTM network will output the risk score R(t) of the photovoltaic module. Different values of R(t) will trigger different reactions. The present invention realizes efficient and accurate photovoltaic fire warning through multi-source detection, dynamic analysis, and hierarchical response.

[0046] Specifically, please refer to Figure 1 As shown, in step 101, the detection module includes Hall sensors for collecting current and voltage, such as Allegro ACS712. Allegro ACS712 has the advantages of high precision, low noise, and fast response. The sampling rate of Allegro ACS712 ≥ 1 kHz, and the infrared temperature matrix is generated by an infrared thermal imager, such as FLIR A315. FLIR A315 can identify those high-temperature problems that are easily overlooked and has built-in logic devices. The resolution of FLIR A315 is 0.1 degree Celsius.

[0047] Further, please refer to Figure 1As shown, in step 102, the spatiotemporal correlation analysis module analyzes the data collected by the detection module, including normalizing the collected current and voltage.

[0048] ;

[0049] in, This represents the data after current normalization. This represents the data after voltage normalization.

[0050] By normalizing current and voltage to the [0,1] interval, the dimensional difference between current and voltage is eliminated, which can accelerate the convergence speed of optimization algorithms such as gradient descent, enabling the spatiotemporal correlation analysis module to find the optimal solution more quickly, thereby speeding up the training process.

[0051] Specifically, the infrared temperature matrix is ​​compressed to obtain... The infrared temperature matrix is ​​dimensionality reduced, the main components are extracted, and the computational load is reduced.

[0052] The dynamic weighting coefficient α(t) of the output current, the dynamic weighting coefficient β(t) of the voltage, and the dynamic weighting coefficient of the infrared temperature matrix. The calculation method is as follows:

[0053] ;

[0054] The numerator represents the rate of change of the current, voltage, and infrared temperature matrices, while the denominator is the sum of the rates of change of the current, voltage, and infrared temperature matrices. The weighting coefficients change dynamically over time, thus improving accuracy.

[0055] The calculation method for dynamically fused weighted feature vectors is as follows:

[0056] ;

[0057] in, This represents a weighted feature vector. The generated weighted feature vector is uploaded to the cloud, triggering spatiotemporal correlation analysis. By dynamically fusing the generated weighted feature vector, the sensitivity of detecting anomalies in photovoltaic modules can be improved. The weights are adjusted according to the rate of change of parameters, so that sudden anomalies such as electric arcs and local overheating can be quickly captured. The dynamic fusion of weighted feature vectors can be optimized through real-time weight adjustment, multi-data complementarity and edge computing, which solves the shortcomings of traditional methods in terms of sensitivity, false alarm rate and adaptability. It can accurately capture early faults and discover and solve problems as early as possible.

[0058] Furthermore, the spatiotemporal correlation analysis module calculates the spatiotemporal correlation matrix between photovoltaic modules, including:

[0059] The positions of photovoltaic modules A and B obtained by positioning are respectively and , then the calculation formula of the spatio-temporal correlation matrix is:

[0060] ;

[0061] Among them, represents the spatio-temporal correlation matrix between photovoltaic modules A and B, is the weighted feature vector of photovoltaic module A, is the weighted feature vector of photovoltaic module B, d0 is the thermal diffusion characteristic distance, and the calculation formula of d0 is:

[0062] d0 = k·υ(t)

[0063] Among them, k is the material thermal conductivity of the photovoltaic module, υ(t) is the real-time wind speed. The spatio-temporal correlation matrix can capture the correlation between photovoltaic modules, avoid isolated misjudgment, and can accurately locate the fault source through the spatio-temporal correlation matrix and predict the propagation path in advance for early response.

[0064] Specifically, the position of the photovoltaic module is located by the positioning component, which can be located by GPS. d0 is the thermal diffusion characteristic distance, which represents the effective range of heat transfer on the surface of the photovoltaic module. By calculating d0, the thermal diffusion distance under different wind speeds can be obtained for timely adjustment and warning.

[0065] Furthermore, the LSTM network includes a fully connected output layer. The LSTM network captures the evolution rules of the weighted feature vector and the spatio-temporal correlation matrix by introducing a spatio-temporal attention mechanism, generates a risk score for the photovoltaic module through the fully connected output layer, aggregates the risk scores of all photovoltaic modules by the LSTM network to generate an overall risk assessment, and triggers a hierarchical analysis response according to the risk score at the same time. When 0.6 < R(t) ≤ 0.8, a secondary alarm is triggered, the power of the photovoltaic module is reduced, and alarm information is pushed at the same time; when R(t) > 0.8, a primary alarm is triggered, the current is cut off, and the fire extinguishing device is started at the same time. The LSTM network is an improved recurrent neural network, which has a powerful long-distance dependence modeling ability, wide application adaptability, and powerful prediction ability. The core advantage of the LSTM network lies in its gated structure and cell state design, making it one of the preferred models for processing sequence data, especially outstanding in scenarios that require long-term memory.

[0066] Specifically, LSTM combines the weighted feature vector with the spatio-temporal correlation rectangle to generate an enhanced feature

[0067] , and its calculation formula is:

[0068] ;

[0069] N is the total number of components, and the normalized score is then output through the activation function:

[0070] ;

[0071] Where σ is the Sigmoid function, is the trainable weight parameter, which can be -1, 0, or 1, and b is the bias term.

[0072] Specifically, for example, if the temperature of a faulty photovoltaic module A rises from 40 degrees Celsius to 120 degrees Celsius in 60 seconds, and the temperature of a nearby photovoltaic module B rises from 40 degrees Celsius to 65 degrees Celsius due to heat diffusion, then when the time is 0 seconds, the calculation shows that... When RA(t) is 0.2, RB(t) is 0.3, and RB(t) is 0.2, the system has no response. After 20 seconds, the result is calculated. With RA(t) = 0.6, RB(t) = 0.7, and RB(t) = 0.4, a level-two warning is activated, and the photovoltaic module A undergoes load reduction. After 40 seconds, the calculation shows that... The values ​​are 0.9, RA(t) is 0.95, and RB(t) is 0.7. At this time, the secondary warning of the nearby photovoltaic module B is activated, reducing the load on the faulty photovoltaic module B. The faulty photovoltaic module A activates the primary alarm and cuts off the circuit of photovoltaic module A. From this analysis, it can be seen that the traditional method does not have a spatiotemporal correlation matrix. Therefore, after 40 seconds, only photovoltaic module A issues an alarm, and the risk of photovoltaic module B is not predicted. However, this invention, after 20 seconds, uses the spatiotemporal correlation matrix to... Early warning can buy 20 crucial seconds for emergency response, thus effectively reducing the losses caused by fire.

[0073] This application embodiment also provides a computer-readable storage medium, which can be any available medium that a computing device can store or a data storage device such as a data center containing one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state drive). The computer-readable storage medium contains instructions that instruct the computing device to perform the aforementioned time synchronization method.

[0074] This application also provides a computer program product containing instructions. The computer program product may be a software or program product containing instructions that can run on a computing device or be stored in any available medium. When the computer program product runs on a computer device, it causes the computing device to perform the aforementioned time synchronization method.

[0075] In the description of this specification, references to terms such as "an embodiment," "example," and "specific example" indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0076] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for early warning of distributed photovoltaic fires, characterized in that, The fire warning method includes the following steps: Collect the current, voltage, and infrared temperature matrix of the photovoltaic module through a detection module; Based on a spatio-temporal correlation analysis module, analyze the data collected by the detection module, and output the dynamic weight coefficient α(t) of the current, the dynamic weight coefficient β(t) of the voltage, and the dynamic weight coefficient γ(t) of the infrared temperature matrix in real time, and dynamically fuse to generate a weighted feature vector; Locate the photovoltaic module through a positioning component to obtain the position P of the photovoltaic module, and calculate the spatio-temporal correlation matrix between photovoltaic modules through the spatio-temporal correlation analysis module; The spatio-temporal correlation analysis module calculates the spatio-temporal correlation matrix between photovoltaic modules, including: The locations of photovoltaic modules A and B are obtained as follows: as well as Then the formula for calculating the spatiotemporal correlation matrix is: ； in, This represents the spatiotemporal correlation matrix between photovoltaic modules A and B. Let be the weighted feature vector of photovoltaic module A. Let be the weighted feature vector of photovoltaic module B, and d0 be the thermal diffusion feature distance. The formula for calculating d0 is: d0 = k·υ(t) where k is the material thermal conductivity of the photovoltaic module, and υ(t) is the real-time wind speed; Analyze the weighted feature vector and the spatio-temporal correlation matrix based on an LSTM network, and output the risk score R(t) of the photovoltaic module; Trigger a hierarchical analysis response according to the risk score. When 0.6 < R(t) ≤ 0.8, trigger a secondary alarm, reduce the load of the photovoltaic module power, and push alarm information at the same time. When R(t) > 0.8, trigger a primary alarm, cut off the current, and start the fire extinguishing device at the same time.

2. The distributed photovoltaic fire early warning method according to claim 1, characterized in that: The spatio-temporal correlation analysis module analyzes the data collected by the detection module, including normalizing the collected current and voltage; ； in, This represents the data after current normalization. This represents the data after voltage normalization. The infrared temperature matrix is ​​compressed to obtain... .

3. The distributed photovoltaic fire early warning method according to claim 2, characterized in that: The dynamic weighting coefficients α(t) of the output current, β(t) of the voltage, and the dynamic weighting coefficients of the infrared temperature matrix are... The calculation method is as follows: ； where the numerator represents the change rate of the current, voltage, and infrared temperature matrix, and the denominator is the sum of the change rates of the current, voltage, and infrared temperature matrix.

4. A distributed photovoltaic fire early warning method according to claim 3, characterized in that: The calculation method of dynamically fusing to generate a weighted feature vector is as follows: ； Among them, the This represents the weighted eigenvector.

5. A distributed photovoltaic fire early warning method according to claim 1, characterized in that: The LSTM network includes a fully connected output layer. The LSTM network captures the evolution law of the weighted feature vector and the spatio-temporal correlation matrix by introducing a spatio-temporal attention mechanism, generates a risk score for the photovoltaic module through the fully connected output layer, and aggregates the risk scores of all photovoltaic modules by the LSTM network to generate an overall risk assessment.

6. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a computer or a processor, it implements the method described in any one of claims 1-5 above.

7. A computer program product, characterized in that: The computer program product includes a computer program, and when the computer program is executed by a computer or a processor, it causes the computer or the processor to execute the method described in any one of claims 1-5.

Images