An AI-based photovoltaic power station fire hazard identification method

Abstract

The application relates to the field of fire hazard identification, in particular to a photovoltaic power station fire hazard identification method based on AI; various parameters of a cable during work of a photovoltaic power station are acquired to form first data, the average temperature of an insulation layer is calculated, the aging degree of the cable is estimated according to the average temperature, the aging degree is used as one of factors influencing fire of the cable in the photovoltaic power station, then various factors influencing fire of the cable during work are obtained, an influence degree matrix is further formed, fire hazards are calculated according to actual parameters of the various factors, whether the cable has fire hazards is judged according to a risk threshold, information whether the cable has fire hazards is obtained in a more forward-looking manner, and operation and maintenance personnel can have more sufficient time to take measures on the cable with fire hazards; the method solves the problem that existing modes have relatively poor forward-looking performance when identifying fire hazards, and operation and maintenance personnel have difficulty in having sufficient time to deal with fire hazards.

Description

Technical Field

[0001] This invention relates to the field of fire hazard identification, specifically to an AI-based method for identifying fire hazards in photovoltaic power plants. Background Technology

[0002] Currently, fire early warning systems for photovoltaic power plants typically rely on a combination of multiple sensors and image analysis. For example, they can detect the gas that appears in the early stages of a fire due to the high temperature melting the internal materials of the cables, and identify fire hazards by detecting the composition of the gas. They can also identify fire hazards by recognizing smoke that appears before a cable fire, and monitor cable temperature using infrared thermography, issuing a notification of a fire hazard when the cable temperature is abnormal.

[0003] However, for current unmanned or sparsely staffed photovoltaic power stations, when using this method to identify fire hazards, it may take several hours from the time a fire hazard is identified to the time when maintenance personnel handle it. During this time, the cables may have already burned, which will have a significant impact on the normal operation of the photovoltaic power station. Therefore, the above-mentioned existing methods have relatively poor foresight in identifying fire hazards, and maintenance personnel do not have enough time to deal with fire hazards. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides an AI-based method for identifying fire hazards in photovoltaic power plants. This method solves the problem that existing methods have relatively poor foresight in identifying fire hazards, and that maintenance personnel do not have enough time to deal with fire hazards.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] An AI-based method for identifying fire hazards in photovoltaic power plants, comprising the following steps:

[0007] S1. Obtain first data including cable specifications, infrared image, ambient humidity, ambient temperature, number of failures, current carrying capacity, laying method, and usage time;

[0008] S2. Calculate the average temperature of the cable insulation layer based on the first data;

[0009] S3. Calculate the total aging of the cable based on the average temperature of the insulation layer at any given time and the first data, set several aging levels corresponding to different total aging amounts, and determine the aging level of the cable.

[0010] S4. Construct an influence matrix of each influencing factor on the degree of influence of cable fire hazards based on the first data and aging level;

[0011] S5. Calculate the risk coefficient of the cable based on the influence matrix;

[0012] S6. Based on whether the risk coefficient is greater than the risk threshold, output a prompt message indicating whether there is a fire hazard in the cable;

[0013] If so, a warning message will be displayed indicating a fire hazard in the output cable;

[0014] If not, then the process ends.

[0015] Preferably, step S21 specifically includes the following steps:

[0016] S21. Calculate the first heat generated at any cross-section of the cable fiber core; the formula for calculating the first heat is:

[0017]

[0018] In the above formula, Q 0 Let I represent the first heat, R represent the current in the cable, and L represent the resistance of the cable. The length of the cable corresponding to this resistance is L.

[0019] S22. Calculate the cross-sectional temperature at any location on any cross-section of the cable based on the first heat source.

[0020] S23. Calculate the average temperature of the insulation layer on any cross-section of the cable based on the cross-sectional temperature at any location on any cross-section of the cable.

[0021] Preferably, step S23 specifically includes the following steps:

[0022] S231. Calculate the coordinates of the center of any cross-section of the cable based on the set of points on that cross-section; the expression for the coordinates of the center of the cross-section is:

[0023]

[0024] In the above formula, (x0, y0, z0) are the x, y, and z coordinates of the center of the circle, respectively. i and y i Let x and y represent the coordinates of a point on any cross-section within the cable, respectively, and let n be the number of points obtained on that cross-section.

[0025] S232. Obtain the radius range of the cable insulation layer;

[0026] S233. Construct a first circle to obtain the average temperature of the insulation layer based on the radius range and the center coordinates of the cross-section of the cable insulation layer; the expression for the first circle is:

[0027]

[0028] In the above formula, r min and r maxThese are the minimum and maximum values ​​of the radius range of the cable insulation layer, respectively, and α is the coefficient for taking the value;

[0029] S234. Obtain the temperature of any point on the first circle based on the cross-sectional temperature at any position on any cross-section of the cable, and use it as the average temperature of the insulation layer on any cross-section of the cable.

[0030] Preferably, step S3 specifically includes the following steps:

[0031] S31. Calculate the aging rate of the cable per unit time based on the average temperature of the insulation layer;

[0032] S32. Calculate the aging amount of the cable over a certain period of time based on the aging rate of the cable per unit time. The formula for calculating the aging amount is:

[0033]

[0034] In the above formula, L represents the aging amount, t represents the service life of the cable, and V(t) represents the relationship between aging rate and time.

[0035] S33. Obtain several factors that affect the aging amount of the cable, and calculate the total aging amount of the cable based on the aging amount.

[0036] S34. Set several aging total ranges and corresponding aging levels for each range, and determine the aging level of the cable based on the aging total range corresponding to the cable's total aging.

[0037] Preferably, step S4 specifically includes the following steps:

[0038] S41. Set up several factors that affect the occurrence of cable fires to obtain a set of factors, select any one factor from the set of factors to set up an influence degree matrix, set the first weight of the factors in the influence degree matrix, and mark the factors in the influence degree matrix as the first factors.

[0039] S42. Select any one of the influencing factors from the set of influencing factors and mark it as the second influencing factor. Set the second weight corresponding to the second influencing factor. Then mark the second influencing factor as the first influencing factor and add it to the influence matrix.

[0040] S43. Compare the influence of the second influencing factor and the first influencing factor in the influence matrix on cable fires, so that the second weight is less than the first weight of the first influencing factor which has a greater influence, and the second weight is greater than the first weight of the first influencing factor which has a smaller influence.

[0041] S44. Determine whether the set of influencing factors is an empty set;

[0042] If so, proceed to the next step;

[0043] If not, return to step S42;

[0044] S45. Normalize the first weight in the influence matrix to obtain the influence matrix; the formulas for normalizing the first and second weights are as follows:

[0045]

[0046] In the above formula, Z' i Z represents the first weight of the i-th first influencing factor in the normalized influence matrix. i This represents the first weight of the i-th first influencing factor in the influence matrix before normalization. There are n first weights in the influence matrix.

[0047] Preferably, step S43 specifically includes the following steps:

[0048] S431. Set several levels of influence and corresponding influence ratios;

[0049] S432. Determine whether there exists a first influencing factor in the influence matrix whose first weight is the same as the second weight of the second influencing factor.

[0050] If so, then set the second weight of the second influencing factor to the first weight of the first influencing factor;

[0051] If not, proceed to step S433;

[0052] S433. Determine whether there is a first influencing factor in the influence matrix that has a smaller influence than the second influencing factor.

[0053] If it exists, obtain the first influencing factor and its corresponding first weight that are closest to the influence degree ratio of the second influencing factor, and mark the first influencing factor as the third influencing factor;

[0054] If not, proceed to step S437;

[0055] S434. Calculate the second weight of the second influencing factor based on the ratio of the third influencing factor to its degree of influence; the formula for calculating the second weight is:

[0056] Z i =τ i，i-1 ·Z i-1

[0057] In the above formula, Z i Z represents the second weight of the second influencing factor. i-1 As the first weight of the third influencing factor, τ i，i-1This represents the ratio of the influence of the second influencing factor to that of the third influencing factor.

[0058] S435. Determine whether there is a first influencing factor in the influence matrix that has a greater influence than the second influencing factor.

[0059] If it exists, obtain the first influencing factor and its corresponding first weight, and mark the first influencing factor as the fourth influencing factor;

[0060] If not, proceed to step S44;

[0061] S436. Scale or expand the first weight of the fourth influencing factor according to the degree of influence ratio; the formula for calculating the scaled or expanded first weight of the fourth influencing factor is:

[0062]

[0063] In the above formula, Z' j Z represents the first weight of the fourth influencing factor after scaling or enlarging. i Z represents the second weight of the second influencing factor. i Z represents the first weight of the fourth influencing factor before scaling or enlarging. i+1 τ represents the first influencing factor in the influence matrix, which has a greater influence than the second influencing factor, and whose influence ratio with the second influencing factor is the smallest. i，i-1 This represents the ratio of the influence of the second influencing factor to that of the third influencing factor.

[0064] S437. Determine whether there is a first influencing factor in the influence matrix that has a greater influence than the second influencing factor.

[0065] If it exists, obtain the first influencing factor and its corresponding first weight that are closest to the influence degree ratio of the second influencing factor, and mark the first influencing factor as the fifth influencing factor;

[0066] If not, proceed to step S44;

[0067] S438. Based on the first weight of the fifth influencing factor and the ratio of the influence of the fifth influencing factor to that of the second influencing factor, calculate the second weight of the second influencing factor; the formula for calculating the first weight is:

[0068]

[0069] In the above formula, Z i Z represents the second weight of the second influencing factor. i+1 τ represents the first weight of the fifth influencing factor. i+1，i This represents the ratio of the influence of the fifth influencing factor to that of the second influencing factor.

[0070] Preferably, step S5 specifically includes the following steps:

[0071] S51. Obtain the parameters of the influencing factors corresponding to the influence matrix;

[0072] S52. Calculate the risk coefficient of cable fire based on the parameters and influence matrix of the influencing factors; the formula for calculating the risk coefficient is:

[0073]

[0074] In the above formula, F represents the risk factor for cable fires, and Z' i δ represents the first weight of the i-th first influencing factor in the normalized influence matrix. i This represents the parameter corresponding to the i-th influencing factor. There are a total of n influencing factors, and each influencing factor corresponds one-to-one with the first influencing factor in the influence degree matrix.

[0075] Preferably, in step S22, the formula for calculating the cross-sectional temperature is:

[0076]

[0077] In the above formula, Q 0 Let x represent the first heat, y represent the coordinates of any cross-section inside the cable, where the temperature is T, A is the thermal conductivity of the material inside the cable, and m is the temperature at that point. 0 The value represents the mass per unit area of ​​the material at any cross-section within the cable, c represents the specific heat capacity of the material at that location, and t represents the service life of the cable.

[0078] Preferably, in step S31, the formula for calculating the aging rate is:

[0079]

[0080] In the above formula, V represents the aging rate of the cable per unit time, a and b are the first and second constants, respectively, which are related to the activation energy of the insulation material, and T 0 T represents the average temperature of the insulation layer, T' represents the first temperature, and B represents the aging baseline coefficient, which is taken as 1.6089.

[0081] Preferably, in step S34, the formula for calculating the total aging amount is:

[0082] LZ=K1·K2·K3·K4·K5·L·β

[0083] in,

[0084] K1=γ n , 1 < γ < 2

[0085]

[0086] In the above formula, K1 represents the failure frequency influencing factor, γ represents the first coefficient, n represents the failure frequency, K2 represents the laying method influencing factor, the corresponding value varies depending on the laying method, K3 represents the environmental humidity influencing factor, different values ​​correspond to different environmental humidity ranges, K4 represents the service life influencing factor, A1 is the second coefficient, A2 is the third coefficient, t0 is the service life of the cable, K5 is the load influencing factor, its value corresponds to the range of cable load rate, and β represents other influencing factors.

[0087] Compared with existing technologies, this invention provides an AI-based method for identifying fire hazards in photovoltaic power plants, which has the following beneficial effects:

[0088] 1. This invention calculates fire hazards based on the actual parameters of various influencing factors collected, and then judges whether there are fire hazards in cables based on risk thresholds. This allows for more proactive information on whether cables have fire hazards, giving maintenance personnel more time to take measures for cables with fire hazards.

[0089] 2. This invention calculates the temperature at different locations on the cross-section of the cable based on the material data and fiber core load, according to the laws of heat transfer. This allows for the calculation of the temperature at various points on the cable cross-section, and further calculation of the average temperature of the insulation layer. Based on this, the total aging amount can be estimated according to the subsequent aging rate and the average temperature of the insulation layer.

[0090] 3. This invention calculates which location on the cable insulation layer has the same average temperature as the average temperature of the cable during the experiment, thereby directly obtaining the temperature at that location on the cable as the average temperature of the insulation layer, thus reducing the computational workload of calculating the average temperature.

[0091] 4. This invention obtains various influencing factors affecting cable fire hazards in photovoltaic power plants, and quantifies the degree of influence of each influencing factor on fire hazards through experiments or experience to obtain an influence matrix. Based on the parameters of various influencing factors obtained in real time and combined with the influence matrix, the risk coefficient can be calculated, and then compared with the risk threshold to determine whether there is a fire hazard in the cable. Attached Figure Description

[0092] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0093] Figure 1 This is a flowchart of the photovoltaic power plant fire hazard identification method of the present invention. Detailed Implementation

[0094] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. This will allow for a full understanding of how the present application uses technical means to solve technical problems and achieve technical effects, and to facilitate its implementation.

[0095] Those skilled in the art will understand that all or part of the steps in the methods of the following embodiments can be implemented by a program instructing related hardware. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0096] To address the relatively poor foresight of existing methods in identifying fire hazards, which leaves maintenance personnel with insufficient time to address potential fire risks, this invention provides an AI-based method for identifying fire hazards in photovoltaic power plants. By acquiring the cable's own parameters and environmental parameters during operation, the probability of a cable fire is estimated, thereby estimating the magnitude of the fire hazard. This method provides a more proactive understanding of the likelihood of cable fires, allowing maintenance personnel in unmanned or sparsely staffed photovoltaic power plants more time to maintain cables and eliminate significant fire hazards. The method includes the following steps:

[0097] S1. Obtain first data including cable specifications, infrared images, ambient humidity, ambient temperature, number of failures, current carrying capacity, laying method, and usage time. Cable specifications include the cable's maximum current carrying capacity, wire diameter, insulation layer diameter range, insulation layer specific heat capacity, etc. Infrared images can be obtained through robot inspection using infrared cameras, or directly using an infrared camera fixed at a certain location to monitor a certain section of cable. Ambient humidity and ambient temperature can be obtained through temperature and humidity sensors.

[0098] S2. Calculate the average temperature of the cable insulation layer based on the first data. The temperature of the cable insulation layer is mainly affected by the heat generated by the cable core during operation. The temperature inside the cable gradually decreases from the core to the outside. Therefore, the heat generated by the core can be calculated by measuring the current and cable resistance. The temperature of the insulation layer can be estimated based on the specific heat capacity of the insulation layer and the materials inside the cable. Specifically, step S21 includes the following steps:

[0099] S21. Calculate the first heat generated at any cross-section of the cable fiber core; the formula for calculating the first heat is:

[0100]

[0101] In the above formula, Q 0 Let I represent the first heat, R represent the current in the cable, and L represent the resistance of the cable. The length of the cable corresponding to this resistance is L.

[0102] S22. Calculate the cross-sectional temperature at any location on any cross-section within the cable based on the first heat source; the formula for calculating the cross-sectional temperature is:

[0103]

[0104] In the above formula, Q 0 Let x represent the first heat, y represent the coordinates of any cross-section inside the cable, where the temperature is T, A is the thermal conductivity of the material inside the cable, and m is the temperature at that point. 0 The mass per unit area of ​​the material at any cross-section within the cable is represented by c, where c represents the specific heat capacity of the material at that point, and t represents the service life of the cable.

[0105] S23. Based on the cross-sectional temperature at any location on any cross-section of the cable, calculate the average temperature of the insulation layer on any cross-section of the cable. The insulation layer is a ring centered on the cable center. Therefore, which part of the ring to use as the basis for calculating the average temperature needs further consideration. Here, to save on calculations, the temperature at a certain location on the insulation layer is directly used as the average temperature. This can be further clarified. Step S23 specifically includes the following steps:

[0106] S231. Calculate the coordinates of the center of any cross-section of the cable based on the set of points on that cross-section; the expression for the coordinates of the center of the cross-section is:

[0107]

[0108] In the above formula, (x0, y0, z0) are the x, y, and z coordinates of the center of the circle, respectively. i and y i Let x and y represent the coordinates of a point on any cross-section within the cable, respectively, and let n be the number of points obtained on that cross-section.

[0109] S232. Obtain the radius range of the cable insulation layer;

[0110] S233. Construct a first circle to obtain the average temperature of the insulation layer based on the radius range and the center coordinates of the cross-section of the cable insulation layer; the expression for the first circle is:

[0111]

[0112] In the above formula, r min and r max These are the minimum and maximum values ​​of the radius range of the cable insulation layer, respectively, and α is the coefficient for taking the value;

[0113] S234. Obtain the temperature of any point on the first circle based on the cross-sectional temperature at any position on any cross-section of the cable, and use it as the average temperature of the insulation layer on any cross-section of the cable. The cross-sectional temperature at any position on any cross-section can be freely set. In actual operation, the temperature at each point on the cross-section of the cable can be obtained first, and the average temperature can be calculated. Then, the temperature at which diameter on the insulation layer is equal to the average temperature can be determined by calculation. In subsequent calculations, the temperature at that point can be directly obtained, thus saving a lot of calculations and making it more convenient.

[0114] S3. Calculate the total aging of the cable based on the average temperature of the cable's insulation layer at any given time and the first data. Set several aging levels corresponding to different total aging amounts and determine the aging level of the cable. During use, the cable is affected by external environmental factors, its current carrying capacity, and cable materials. Therefore, by experimentally analyzing the impact of different influencing factors on the cable's aging rate, the aging rate of the cable is simulated to facilitate the calculation and monitoring of the cable's aging degree. This serves as a factor in determining whether the cable poses a fire hazard. Specifically, step S3 includes the following steps:

[0115] S31. Calculate the aging rate of the cable per unit time based on the average temperature of the insulation layer; the formula for calculating the aging rate is:

[0116]

[0117] In the above formula, V represents the aging rate of the cable per unit time, a and b are the first and second constants, respectively, which are related to the activation energy of the insulation material, and T 0 T represents the average temperature of the insulation layer, T' represents the first temperature, and B represents the aging baseline coefficient, which is generally taken as 1.6089.

[0118] S32. Calculate the aging amount of the cable over a certain period of time based on the aging rate of the cable per unit time. The formula for calculating the aging amount is:

[0119]

[0120] In the above formula, L represents the aging amount, t represents the service life of the cable, and V(t) represents the relationship between aging rate and time.

[0121] S33. Obtain several factors affecting the aging amount of cables, and calculate the total aging amount of the cables based on the aging amount; the formula for calculating the total aging amount is:

[0122] LZ=K1·K2·K3·K4·K5·L·β

[0123] in,

[0124] K1=γ n , 1 < γ < 2

[0125]

[0126] In the above formula, K1 represents the failure frequency impact factor, γ represents the first coefficient, n represents the failure frequency, K2 represents the laying method impact factor, the corresponding value varies depending on the laying method, for example, direct burial is set to 1.2, and pipe laying is set to 1.3, etc., K3 represents the environmental humidity impact factor, different values ​​correspond to different environmental humidity ranges, for example, the value is 1 when the humidity is 0-40%, the value is 1.1 when the humidity is 40%-60%, the value is 1.2 when the humidity is 60%-80%, and the value is 1.3 when the humidity is 80%-100%, K4 represents the service life impact factor, A1 is the second coefficient, generally set to 0.9601. A2 is the third coefficient, generally taken as 0.01899. t0 is the service life of the cable. K5 is the load impact factor, the value of which corresponds to the range of cable load rate. For example, when the load rate is less than 50%, its value is 1; when the load rate is between 50% and 70%, it is set to 1.2; and when the load rate is greater than 70%, its value is set to 1.4. β is other influencing factors. Currently, in the actual use of photovoltaic power stations, the factors affecting cable aging are mainly divided into five types: the number of cable failures, the laying method, the ambient humidity, the service life, and the cable load. Of course, there may be other influencing factors, so adjustments are made through other influencing factors.

[0127] S34. Set several aging total ranges and corresponding aging levels for each range, and determine the aging level of the cable based on the aging total range corresponding to the total aging of the cable. Currently, the aging levels of the cable are set as good, light, moderate and heavy, and their corresponding aging total ranges are LZ<30%, 30%≤LZ<50%, 50%≤LZ<70%, and 70%≤LZ.

[0128] S4. Construct an influence matrix of each influencing factor on the fire hazard of cables based on the first data and aging level. For the fire hazard of photovoltaic power station cables, the influencing factors include not only the aging degree of the cables, but also dust, ambient temperature, ambient humidity, cable parameters, and cable load. Therefore, it is necessary to estimate the degree of influence on the fire hazard through experiments, so as to facilitate the calculation of the risk coefficient reflecting the degree of fire hazard by statistically analyzing the parameters of each influencing factor in subsequent actual use. Therefore, step S4 specifically includes the following steps:

[0129] S41. Set up several factors that affect the occurrence of cable fires to obtain a set of factors, select any one factor from the set of factors to set up an influence degree matrix, set the first weight of the factors in the influence degree matrix, and mark the factors in the influence degree matrix as the first factors.

[0130] S42. Select any one of the influencing factors from the set of influencing factors and mark it as the second influencing factor. Set the second weight corresponding to the second influencing factor. Then mark the second influencing factor as the first influencing factor and add it to the influence matrix.

[0131] S43. Compare the impact of the second influencing factor with that of the first influencing factor in the influence matrix on cable fires. Make the second weight less than the first weight of the first influencing factor, which has a greater influence, and greater than the first weight of the first influencing factor, which has a smaller influence. The magnitude of the first weight of each first influencing factor in the influence matrix needs to be estimated based on experimental data. The following is a specific method for estimating the magnitude of the first weight of each first influencing factor. Step S43 specifically includes the following steps:

[0132] S431. Set several levels of influence and corresponding influence ratios; for example, when comparing the influence of the i-th and j-th factors on cable fires, if the i-th factor is equally important as the j-th factor, the difference in their influence is 1; if the i-th factor is slightly more important than the j-th factor, the influence ratio is 3; if the i-th factor is significantly more important than the j-th factor, the influence ratio is 5; if the i-th factor is much more important than the j-th factor, the influence ratio is 7; if the i-th factor is extremely important compared to the j-th factor, the influence ratio is 9.

[0133] S432. Determine whether there exists a first influencing factor in the influence matrix whose first weight is the same as the second weight of the second influencing factor.

[0134] If so, then set the second weight of the second influencing factor to the first weight of the first influencing factor;

[0135] If not, proceed to step S433;

[0136] S433. Determine whether there is a first influencing factor in the influence matrix that has a smaller influence than the second influencing factor.

[0137] If it exists, obtain the first influencing factor and its corresponding first weight that are closest to the influence degree ratio of the second influencing factor, and mark the first influencing factor as the third influencing factor;

[0138] If not, proceed to step S437;

[0139] S434. Calculate the second weight of the second influencing factor based on the ratio of the third influencing factor to its degree of influence; the formula for calculating the second weight is:

[0140] Z i =τ i，i-1 ·Z i-1

[0141] In the above formula, Z i Z represents the second weight of the second influencing factor. i-1 As the first weight of the third influencing factor, τ i，i-1 This represents the ratio of the influence of the second influencing factor to that of the third influencing factor.

[0142] S435. Determine whether there is a first influencing factor in the influence matrix that has a greater influence than the second influencing factor.

[0143] If it exists, obtain the first influencing factor and its corresponding first weight, and mark the first influencing factor as the fourth influencing factor;

[0144] If not, proceed to step S44;

[0145] S436. Scale or expand the first weight of the fourth influencing factor according to the degree of influence ratio; the formula for calculating the scaled or expanded first weight of the fourth influencing factor is:

[0146]

[0147] In the above formula, Z' j Z represents the first weight of the fourth influencing factor after scaling or enlarging. i Z represents the second weight of the second influencing factor. i Z represents the first weight of the fourth influencing factor before scaling or enlarging. i+1 τ represents the first influencing factor in the influence matrix, which has a greater influence than the second influencing factor, and whose influence ratio with the second influencing factor is the smallest. i，i-1This represents the ratio of the influence of the second influencing factor to that of the third influencing factor.

[0148] S437. Determine whether there is a first influencing factor in the influence matrix that has a greater influence than the second influencing factor.

[0149] If it exists, obtain the first influencing factor and its corresponding first weight that are closest to the influence degree ratio of the second influencing factor, and mark the first influencing factor as the fifth influencing factor;

[0150] If not, proceed to step S44;

[0151] S438. Based on the first weight of the fifth influencing factor and the ratio of the influence of the fifth influencing factor to that of the second influencing factor, calculate the second weight of the second influencing factor; the formula for calculating the first weight is:

[0152]

[0153] In the above formula, Z i Z represents the second weight of the second influencing factor. i+1 τ represents the first weight of the fifth influencing factor. i+1，i This represents the ratio of the influence of the fifth influencing factor to that of the second influencing factor.

[0154] S44. Determine whether the set of influencing factors is an empty set;

[0155] If so, proceed to the next step;

[0156] If not, return to step S42;

[0157] S45. Normalize the first weight in the influence matrix to obtain the influence matrix; the formulas for normalizing the first and second weights are as follows:

[0158]

[0159] In the above formula, Z' i Z represents the first weight of the i-th first influencing factor in the normalized influence matrix. i This represents the first weight of the i-th first influencing factor in the influence matrix before normalization. There are n first weights in the influence matrix.

[0160] S5. Calculate the risk coefficient of the cable based on the influence matrix. After quantitatively obtaining the first weight of each first influencing factor in the influence matrix, the actual parameters corresponding to the first influencing factors are obtained to facilitate the calculation of the risk coefficient. Specifically, step S5 includes the following steps:

[0161] S51. Obtain the parameters of the influencing factors corresponding to the influence matrix;

[0162] S52. Calculate the risk coefficient of cable fire based on the parameters and influence matrix of the influencing factors; the formula for calculating the risk coefficient is:

[0163]

[0164] In the above formula, F represents the risk factor for cable fires, and Z' i δ represents the first weight of the i-th first influencing factor in the normalized influence matrix. i This represents the parameter corresponding to the i-th influencing factor. There are a total of n influencing factors, and each influencing factor corresponds one-to-one with the first influencing factor in the influence degree matrix.

[0165] S6. Based on whether the risk coefficient is greater than the risk threshold, output a prompt message indicating whether there is a fire hazard in the cable. The risk threshold is generally set between 40 and 50, which can ensure the service life of the cable, reduce cost waste, and also make the identification of fire hazards more accurate. When the risk coefficient is less than the risk threshold, there is no fire.

[0166] If so, the system will output a warning message indicating a fire hazard in the cable. The warning message includes the cable number and location, allowing maintenance personnel to identify which cable segment is at risk of fire.

[0167] If not, then the process ends.

[0168] This invention acquires various parameters of cables operating in a photovoltaic power station to form first data. Then, it calculates the average temperature of the cable insulation layer based on this first data. Next, it estimates the cable's aging degree based on the first parameters and the average insulation temperature, considering this as one of the influencing factors for cable fires in photovoltaic power stations. Furthermore, based on the cable's operating environment in the photovoltaic power station, it collects various influencing factors that may cause fires and obtains the magnitude of each factor's impact on fire through experiments. It then forms an influence degree matrix using the method mentioned in step S4. Based on the actual parameters of each collected influencing factor and the influence degree matrix, it calculates the fire hazard, i.e., the risk coefficient of cable fires. Based on the risk threshold, it determines whether the cable has a fire hazard, thus enabling more proactive information acquisition about potential fire hazards. This allows maintenance personnel to identify potential fire hazards earlier and take more time to address them.

[0169] The above embodiments provide a detailed description of the present invention. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for identifying fire hazards in photovoltaic power plants based on AI, characterized in that, The method includes the following steps: S1. Obtain first data including cable specifications, infrared image, ambient humidity, ambient temperature, number of failures, current carrying capacity, laying method, and usage time; S2. Calculate the average temperature of the cable insulation layer based on the first data; S3. Calculate the total aging of the cable based on the average temperature of the insulation layer at any given time and the first data, set several aging levels corresponding to different total aging amounts, and determine the aging level of the cable. S4. Construct an influence matrix of each influencing factor on the degree of impact of cable fire hazards based on the first data and aging level; Step S4 specifically includes the following steps: S41. Set several factors that affect the occurrence of cable fires to obtain a set of factors, select any one factor from the set of factors to set an influence degree matrix, set the first weight of the factors in the influence degree matrix, and mark the factors in the influence degree matrix as the first factors. S42. Select any one of the influencing factors from the set of influencing factors and mark it as the second influencing factor. Set the second weight corresponding to the second influencing factor. Then mark the second influencing factor as the first influencing factor and add it to the influence matrix. S43. Compare the influence of the second influencing factor and the first influencing factor in the influence matrix on cable fires, so that the second weight is less than the first weight of the first influencing factor which has a greater influence, and the second weight is greater than the first weight of the first influencing factor which has a smaller influence. S44. Determine whether the set of influencing factors is an empty set; If so, proceed to the next step; If not, return to step S42; S45. Normalize the first weight in the influence matrix to obtain the influence matrix; the formulas for normalizing the first and second weights are as follows: In the above formula, This represents the first weight of the i-th first influencing factor in the influence matrix after normalization. This represents the first weight of the i-th first influencing factor in the influence matrix before normalization. The number of first weights in the influence matrix is ​​n1. S5. Calculate the risk coefficient of the cable based on the influence matrix; S6. Based on whether the risk coefficient is greater than the risk threshold, output a prompt message indicating whether there is a fire hazard in the cable; If so, a warning message will be displayed indicating a fire hazard in the output cable; If not, then the process ends.

2. The identification method according to claim 1, characterized in that, Step S2 specifically includes the following steps: S21. Calculate the first heat generated at any cross-section of the cable fiber core; the formula for calculating the first heat is: In the above formula, Indicates the first calorie. Indicates the current in the cable. This indicates the resistance of the cable, and the corresponding cable length is... ; S22. Calculate the cross-sectional temperature at any location on any cross-section of the cable based on the first heat source. S23. Calculate the average temperature of the insulation layer on any cross-section of the cable based on the cross-sectional temperature at any location on any cross-section of the cable.

3. The identification method according to claim 2, characterized in that, Step S23 specifically includes the following steps: S231. Calculate the coordinates of the center of any cross-section of the cable based on the set of points on that cross-section; the expression for the coordinates of the center of the cross-section is: In the above formula, Let x, y, and z be the coordinates of the center of the circle. and These represent the x-axis and y-axis coordinates of a point on any cross-section within the cable, respectively, and the number of points obtained on this cross-section is n4; S232. Obtain the radius range of the cable insulation layer; S233. Construct a first circle to obtain the average temperature of the insulation layer based on the radius range and the center coordinates of the cross-section of the cable insulation layer; the expression for the first circle is: In the above formula, and These represent the minimum and maximum values ​​of the radius range of the cable insulation layer, respectively. The coefficient is used to select values. S234. Obtain the temperature of any point on the first circle based on the cross-sectional temperature at any position on any cross-section of the cable, and use it as the average temperature of the insulation layer on any cross-section of the cable.

4. The identification method according to claim 1, characterized in that, Step S3 specifically includes the following steps: S31. Calculate the aging rate of the cable per unit time based on the average temperature of the insulation layer; S32. Calculate the aging amount of the cable over a certain period of time based on the aging rate of the cable per unit time. The formula for calculating the aging amount is: In the above formula, Indicates the amount of aging. Indicates the duration of cable use. An expression representing the relationship between aging rate and time; S33. Obtain several factors that affect the aging amount of the cable, and calculate the total aging amount of the cable based on the aging amount. S34. Set several aging total ranges and corresponding aging levels for each range, and determine the aging level of the cable based on the aging total range corresponding to the cable's total aging.

5. The identification method according to claim 1, characterized in that, Step S43 specifically includes the following steps: S431. Set several levels of influence and corresponding influence ratios; S432. Determine whether there exists a first influencing factor in the influence matrix whose first weight is the same as the second weight of the second influencing factor. If so, then set the second weight of the second influencing factor to the first weight of the first influencing factor; If not, proceed to step S433; S433. Determine whether there is a first influencing factor in the influence matrix that has a smaller influence than the second influencing factor. If it exists, obtain the first influencing factor and its corresponding first weight that are closest to the influence degree ratio of the second influencing factor, and mark the first influencing factor as the third influencing factor; If not, proceed to step S437; S434. Calculate the second weight of the second influencing factor based on the ratio of the third influencing factor to its degree of influence; the formula for calculating the second weight is: In the above formula, This indicates the second weight of the second influencing factor. As the third influencing factor with the highest weight, This represents the ratio of the influence of the second influencing factor to that of the third influencing factor. S435. Determine whether there is a first influencing factor in the influence matrix that has a greater influence than the second influencing factor. If it exists, obtain the first influencing factor and its corresponding first weight, and mark the first influencing factor as the fourth influencing factor; If not, proceed to step S44; S436. Scale or expand the first weight of the fourth influencing factor according to the degree of influence ratio; the formula for calculating the scaled or expanded first weight of the fourth influencing factor is: In the above formula, This indicates the first weight of the fourth influencing factor after scaling or enlarging. This indicates the second weight of the second influencing factor. This indicates the first weight of the fourth influencing factor before scaling or enlarging. This indicates the first influencing factor in the influence matrix that has a greater influence than the second influencing factor, and the ratio of the influence of the first influencing factor to the influence of the second influencing factor is the smallest. This represents the ratio of the influence of the second influencing factor to that of the third influencing factor. S437. Determine whether there is a first influencing factor in the influence matrix that has a greater influence than the second influencing factor. If it exists, obtain the first influencing factor and its corresponding first weight that are closest to the influence degree ratio of the second influencing factor, and mark the first influencing factor as the fifth influencing factor; If not, proceed to step S44; S438. Calculate the second weight of the second influencing factor based on the first weight of the fifth influencing factor and the ratio of the influence of the fifth influencing factor to that of the second influencing factor. The formula for calculating the second weight is: In the above formula, This indicates the second weight of the second influencing factor. This indicates the first weight of the fifth influencing factor. This represents the ratio of the influence of the fifth influencing factor to that of the second influencing factor.

6. The identification method according to claim 1, characterized in that, Step S5 specifically includes the following steps: S51. Obtain the parameters of the influencing factors corresponding to the influence matrix; S52. Calculate the risk coefficient of cable fire based on the parameters and influence matrix of the influencing factors; the formula for calculating the risk coefficient is: In the above formula, This indicates the risk factor for cable fires. This represents the first weight of the i-th first influencing factor in the influence matrix after normalization. This represents the parameter corresponding to the i-th influencing factor. There are a total of n2 influencing factors, and each influencing factor corresponds one-to-one with the first influencing factor in the influence degree matrix.

7. The identification method according to claim 2, characterized in that, In step S22, the formula for calculating the cross-sectional temperature is: In the above formula, Indicates the first calorie. and Let T and A represent the coordinates of any cross-section within the cable, where the temperature is T, and A is the thermal conductivity of the material within the cable. This represents the mass per unit area of ​​material at any cross-section within the cable. This indicates the specific heat capacity of the material at that location, and t is the service life of the cable.

8. The identification method according to claim 4, characterized in that, In step S31, the formula for calculating the aging rate is: In the above formula, This represents the aging rate of the cable per unit time, where a and b are the first and second constants, respectively, and are related to the activation energy of the insulation material. This indicates the average temperature of the insulation layer. Indicates the first temperature. This represents the aging baseline coefficient, which is taken as 1.6089.

9. The identification method according to claim 4, characterized in that, In step S34, the formula for calculating the total aging amount is: in, In the above formula, Indicates the factor affecting the number of failures. Indicates the first coefficient. 3 indicates the number of failures. This indicates the influence factor of the laying method; the corresponding value varies depending on the laying method. This represents the environmental humidity influencing factor, with different values ​​corresponding to different environmental humidity ranges. Indicates the impact factor of the number of years of use. As the second coefficient, The third coefficient, The service life of the cable. This is the load impact factor, and its value corresponds to the range of cable load rates. Other influencing factors.

Images