A method for fault diagnosis of photovoltaic array cables

Abstract

This invention discloses a method for fault diagnosis of photovoltaic array cables, comprising the following steps: acquiring photovoltaic-related parameters; cleaning the acquired data to obtain valid data; calculating the cleaned valid data to obtain the theoretical current data I0 of the photovoltaic string; and combining the theoretical current data I0 of the photovoltaic string with the actual current data I... m This invention involves capturing abnormal data series; performing preliminary diagnosis on the abnormal data series to obtain preliminary fault status; establishing a fault diagnosis model using meteorological data, string current anomaly data, and the preliminary fault status as input sources; and determining photovoltaic array cable faults based on the abnormal data series and the fault diagnosis model. In this invention, fault diagnosis using power plant operation data easily discovers hidden cable faults and accurately locates them. By filtering out environmental factors and unexpected fault effects, it achieves spatial and temporal dimension analysis. The diagnostic method is robust, providing guidance for power plant operation and maintenance and assisting in improving power plant power generation.

Description

Technical Field

[0001] This invention belongs to the field of photovoltaic power generation technology, and specifically relates to a method for diagnosing faults in photovoltaic array cables. Background Technology

[0002] Photovoltaic power generation is an effective technology for converting solar energy into electrical energy, playing a crucial role in addressing global climate change and providing sustainable energy. In China, photovoltaic power generation technology has experienced rapid and large-scale development. A photovoltaic power generation system mainly consists of photovoltaic modules, cables, combiner boxes, inverters, transformers, and other major equipment. Cables, in particular, are crucial bridges connecting power generation equipment, especially photovoltaic array cables. These cables are numerous and complex, playing a vital role in photovoltaic power plants, but they are also frequently prone to failure. Common faults include open circuits, short circuits, insulation failures, and poor contact. These faults are not always present, mainly affected by factors such as weather, which obviously increases the concealment of faults and the difficulty of fault diagnosis.

[0003] Currently, the main method for diagnosing cable faults in photovoltaic power plants relies on manual judgment combined with background alarm information. However, highly concealed faults require comprehensive consideration of meteorological data, historical data, and other indicators for identification. This process is not only time-consuming but also heavily dependent on the experience of the staff, making it prone to missed or false diagnoses. Consequently, fault diagnosis is inefficient and costly, ultimately leading to lost power generation revenue. Furthermore, for large-scale photovoltaic power plants, the land area is enormous, and the number of maintenance personnel is very limited. Relying on maintenance personnel for a complete inspection is extremely labor-intensive and difficult to complete in a short time. Moreover, the fault identification rate is low, and prolonged fault persistence can cause unnecessary power generation losses. Therefore, the ability to quickly and accurately identify and locate photovoltaic array faults is of great significance for the operation and maintenance of photovoltaic power plants and for improving their efficiency. Summary of the Invention

[0004] To address the technical problems existing in the prior art, the present invention aims to provide a method for diagnosing faults in photovoltaic array cables.

[0005] To achieve the above objectives and technical effects, the technical solution adopted by this invention is as follows:

[0006] A method for diagnosing faults in photovoltaic array cables includes the following steps:

[0007] S01: Obtain photovoltaic-related parameters, including the historical current I of the photovoltaic string. m Voltage data U m String temperature T model Component nominal operating current I mp Component nominal open-circuit voltage V ocComponent current temperature coefficient α, irradiance I r Wind speed W s Rainfall (R) and ambient temperature (T) emp ;

[0008] S02: Clean the acquired data to obtain valid data;

[0009] S03: Calculate the effective data after cleaning to obtain the theoretical current data I0 of the photovoltaic string;

[0010] S04: Combining the theoretical current data I0 and actual current data I of the photovoltaic string m Capture abnormal data series;

[0011] S05: Perform a preliminary diagnosis on the abnormal data series to obtain a preliminary fault status;

[0012] S06: Establish a fault diagnosis model using meteorological data, string flow anomaly data, and preliminary fault status as input sources;

[0013] S07: Determine the fault of the photovoltaic array cable based on the abnormal data series and fault diagnosis model.

[0014] Furthermore, in step S02, the steps of cleaning the acquired data to obtain valid data include:

[0015] S02-01: Irradiance data I r >100W / m 2 Given the first and last time points, for I m U m T model I r W s R, T emp Extract the data and denote it as Data;

[0016] S02-02: Irradiance data in Data r Perform cleaning;

[0017] S02-03: Historical current I of photovoltaic strings in Data m Perform cleaning;

[0018] S02-04: Regarding the voltage data U in Data... m Perform cleaning;

[0019] S02-05: Regarding the ambient temperature data T in Data... emp Perform cleaning;

[0020] S02-06: Regarding the wind speed data W in Data... sPerform cleaning;

[0021] S02-07: Component temperature data T in Data model Clean it.

[0022] Furthermore, in step S02-02, the irradiance data I in Data is... r The cleaning steps include:

[0023] S02-02-01: Calculate the theoretical irradiance I according to the following formula. r0 ;

[0024]

[0025] Among them, Ir stc The standard irradiance is 1000 W / m². 2 ;

[0026] S02-02-02: Calculate the deviation δIr(t) between real-time irradiance and theoretical irradiance using the following formula. i ):

[0027] δIr(t i )=|Ir(t i )-Ir0(t i )|

[0028] S02-02-03: Calculate δIr(t) i The median and standard deviation of () are denoted as M. Ir S Ir ;

[0029] S02-02-04: If δIr(t) i )>M Ir +α′*S Ir If α′ is an adjustable factor, then determine Ir(t) i () represents an outlier, and the value at time t is recorded. i ;

[0030] S02-02-05: Calculate time t respectively i-1 and t i+1 With t i I r and I r0 The slopes are denoted as ΔI. r and ΔI r0 ;

[0031] S02-02-06: Comparison of ΔI r and ΔI r0 The absolute value of |ΔI r |>|ΔI r0|, then Ir(t) i )=Ir0(t i );

[0032] S02-02-07: Repeat steps S02-02-02 to S02-02-06 until there are no outliers.

[0033] Furthermore, in steps S02-03, the historical current I of the photovoltaic string in Data is... m The cleaning steps include:

[0034] S02-03-01: Calculate the theoretical current I according to the following formula. m0 :

[0035]

[0036] S02-03-02: Calculate the deviation δIm(t) between the real-time current value and the theoretical current value using the following formula. i ):

[0037] δIm(t i )=|Im(t i )-Im0(t i )|

[0038] S02-03-03: Calculate δIm(t) i The median and standard deviation of () are denoted as M. Imm S Im ;

[0039] S02-03-04: If δIm(t) i )>M Im +α″*S Im If α″ is an adjustable factor, then determine Im(t) i If t is an outlier, then Im(t) i )=Im0(t i );

[0040] S02-03-05: Repeat steps S02-03-02 to S02-02-04 until there are no outliers.

[0041] Furthermore, in steps S02-04, the voltage data U in Data is... m The cleaning steps include:

[0042] S02-04-01: At 10-minute intervals, for U m Divide into segments, denoted as U mi ;

[0043] S02-04-02: Calculate U mjThe median and standard deviation are denoted as Muj and Suj, respectively.

[0044] S02-04-03: If |U mj (t i )-Muj|>β u Suj, β u If U is an adjustable factor, then mj (t i ) represents an outlier;

[0045] S02-04-04: Calculate and replace U using interpolation. mj (t i ).

[0046] Furthermore, in steps S02-05, the ambient temperature data T in Data is processed. emp The cleaning steps include:

[0047] S02-05-01: At 10-minute intervals, for T emp Divide into segments, denoted as T. empj ;

[0048] S02-05-02: Calculate T empj The median and standard deviation are denoted as Mtj and Stj, respectively;

[0049] S02-05-03: If |T empj (t i )-Mtj|>β Te Stj, β Te If T is an adjustable factor, then empj (t i ) represents an outlier;

[0050] S02-05-04: Calculate and replace T using interpolation. empj (t i );

[0051] In steps S02-07, the component temperature data T in Data is processed. model The cleaning steps include:

[0052] S02-07-01: At 10-minute intervals, for T model Divide into segments, denoted as T. modelj ;

[0053] S02-07-02: Calculate T modelj The median and standard deviation are denoted as Mmj and Smj, respectively;

[0054] S02-07-03: If |T modelj (t i)-Mmj|>β Tm Smj, β Tm If T is an adjustable factor, then modelj (t i ) represents an outlier;

[0055] S02-07-04: Calculate and replace T using interpolation. modelj (t i ).

[0056] Furthermore, in steps S02-06, the wind speed data W in Data is processed. s The cleaning steps include:

[0057] S02-06-01: Obtain the maximum wind speed at the photovoltaic power station project site over the past 10 years, denoted as W. m ;

[0058] S02-06-01: At 10-minute intervals, W... s Divide into segments, denoted as W. sj ;

[0059] S02-06-02: If W sj (t i )>W m or W sj (t i If W < 0, then use interpolation to calculate and replace W. sj (t i );

[0060] Furthermore, in step S04, the steps for capturing the series of abnormal data include:

[0061] S04-01: Based on the obtained valid string current data and theoretical current data, calculate the string current deviation rate using the following formula:

[0062]

[0063] S04-02: For δ(t) i Make a judgment if δ(t) i If ) < 0.2, then record that moment in set U. f In the middle, denoted as U f ={t1, t2, ..., t N}, N∈N * N * It is a natural number; if δ(t) i If the value is greater than 0.7, then that moment is recorded in set U. p In the middle, denoted as U p ={t1, t2, ..., t N}, N∈N * .

[0064] Furthermore, in step S05, the step of obtaining the preliminary fault status includes:

[0065] S05-01: Set the fault status word, denoted as F. s ;

[0066] S05-02: For U f and U p Make a judgment if U f and U p If all are non-empty sets, then F s =1; if U f Let U be a non-empty set, and U p If F is an empty set, then F s =2; if U f If it is empty, then F s =0;

[0067] S05-03: Regarding U f U p Perform a transformation, following the rules below: calculate t sequentially according to the data series order. i+1 -t i If t i+1 -t i If t > τ, then t i =(t i , t i If t i+1 -t i If t = τ, then continue calculating the next time step until t. i+l+1 -t i+l >τ, at this time t i =(t i , t i+l ), t i+1 To t i+1 Delete, transformed U f U p Let it be U f1 U p1 ;

[0068] S05-04: Regarding U f U p Perform a second transformation, with the following transformation rules: For U f1 U p1 Each element in the array increases by t i Maximum current I at time t m Maximum irradiance Ir, maximum wind speed Ws, and highest ambient temperature T emp The transformed set is denoted as U. f2 U p2 .

[0069] Furthermore, in step S06, the steps for establishing the fault diagnosis model include:

[0070] S06-01: Establish a data structure for string cable fault records. The data structure includes string number, date, set of fault times, rainfall condition R, fault status, and number of faults, denoted as D. s (SN, Date, U) f2 U p2 R, F s F t );

[0071] S06-02: Calculate the number of daily failures F t The calculation formula is as follows:

[0072] F t =Length(U f2 )

[0073] Where: Length is the computation set U f2 Number of elements;

[0074] S06-03: Calculate the correlation factors for each input source. The specific steps are as follows:

[0075] S06-03-1: Calculate the correlation factor between fault and wind speed, following these steps:

[0076] S06-03-1-1: Calculate U p2 The maximum wind speed Ws0 in each element range;

[0077] S06-03-1-2: Obtain U f2 Maximum wind speed Ws in each element range s The wind speed abrupt change factor is calculated using the following formula to obtain the abrupt change factor dataset, denoted as: ε={ε1, ε2, ε3, …, ε N}, N = 1, 2, 3…;

[0078]

[0079] S06-03-1-3: Calculate ε in the set ε within one week i The probability greater than 1.5 is defined as the wind speed correlation factor, denoted as P. w ;

[0080] S06-03-2: Calculate the correlation factor between faults and rainfall conditions, following these steps:

[0081] S06-03-2-1: Set the weight values ​​for rainfall in weather forecasts as follows:

[0082] No rain weight: 1%, light rain weight: 3%, moderate rain weight: 6%, heavy rain weight: 12%, rainstorm weight: 26%, heavy rainstorm weight: 52%;

[0083] S06-03-2-2: Calculate F within the most recent week s The average rainfall weight value with a value of 0 is denoted as β0.

[0084] S06-03-2-3: Calculate F within the most recent week s The average rainfall weight value with a value of 1 is denoted as β1.

[0085] S06-03-2-4: The correlation factor between faults and rainfall is defined as the ratio of β1 to β0, denoted as C. R ,Right now:

[0086]

[0087] S06-03-3: Calculate the correlation factor between the fault and ambient temperature, following these steps:

[0088] S06-03-3-1: Calculate F within the most recent week s The average ambient temperature with a value of 0 is denoted as T0.

[0089] S06-03-3-2: Calculate F within the most recent week s The average ambient temperature with a value of 1 is denoted as T1.

[0090] S06-03-3-3: Calculate U p2 The average ambient temperature for each element range is denoted as T. p ;

[0091] S06-03-3-4: Calculate U f2 The average ambient temperature for each element range is denoted as T. f ;

[0092] S06-03-3-5: The correlation factor between fault and ambient temperature is defined as T0, T1, T p T f The function, denoted as M T The calculation formula is as follows:

[0093]

[0094] S06-03-4: Calculate U for each day of the past week. f2 The repetition rate Rep for each element is determined by the following steps:

[0095] S06-03-4-1: Based on the sunrise and sunset times T within a week rise T setUsing τ as the step size, determine the time t at a certain moment. i Does +τ fall within all U values ​​within a week? f2 If it falls into the set U, then add that moment to the set U. com In the middle, denoted as U com ={t1, t2, ..., t N}, and record U for each day f2 The number of elements that fall into the middle is denoted as N. i ;

[0096] S06-03-4-2: If there exists a certain day N i ≤Ft i If the value is positive, then Rep = 1; otherwise, Rep = 0.

[0097] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0098] This invention discloses a method for fault diagnosis of photovoltaic array cables, comprising the following steps: acquiring photovoltaic-related parameters; cleaning the acquired data to obtain valid data; calculating the cleaned valid data to obtain the theoretical current data I0 of the photovoltaic string; and combining the theoretical current data I0 of the photovoltaic string with the actual current data I... m This invention involves capturing abnormal data series; performing preliminary diagnosis on the abnormal data series to obtain preliminary fault status; establishing a fault diagnosis model using meteorological data, string current anomaly data, and preliminary fault status as input sources; and determining photovoltaic array cable faults based on the abnormal data series and the fault diagnosis model. In this invention, fault diagnosis utilizes power plant operation data, making it highly practical, easier to discover hidden cable faults, and more accurately locates them. By filtering environmental factors and the impact of unexpected faults, it achieves spatial and temporal dimension analysis. The diagnostic method is robust, and the cable fault diagnosis model can accurately identify the causes of cable faults. It has complete influencing factors and comprehensive fault identification functions, providing guidance for power plant operation and maintenance and assisting in improving power plant power generation. Attached Figure Description

[0099] Figure 1 This is a flowchart of the present invention;

[0100] Figure 2 This is a flowchart of the data cleaning process of the present invention;

[0101] Figure 3 This is a flowchart of the abnormal data capture process of the present invention;

[0102] Figure 4 This is a flowchart illustrating the process of obtaining a preliminary fault state according to the present invention;

[0103] Figure 5 This is a flowchart illustrating the establishment of a fault diagnosis model for this invention.

[0104] Figure 6 This is a diagram of the fault diagnosis model of the present invention;

[0105] Figure 7 This is the photovoltaic string current curve of Embodiment 1 of the present invention;

[0106] Figure 8 This is the irradiance data curve for Embodiment 1 of the present invention;

[0107] Figure 9 This is the ambient temperature data curve for Embodiment 1 of the present invention;

[0108] Figure 10 This is the string temperature data curve of Embodiment 1 of the present invention;

[0109] Figure 11 This is the ambient temperature data curve for Embodiment 1 of the present invention;

[0110] Figure 12 This is the current curve after data cleaning in Embodiment 1 of the present invention;

[0111] Figure 13 This is the irradiance data curve after data cleaning in Embodiment 1 of the present invention;

[0112] Figure 14 This is the ambient temperature data curve after data cleaning in Embodiment 1 of the present invention;

[0113] Figure 15 This is the string temperature data curve after data cleaning in Embodiment 1 of the present invention;

[0114] Figure 16 The wind speed curve after data cleaning in Embodiment 1 of the present invention;

[0115] Figure 17 The set U of Embodiment 1 of the present invention f and U p ;

[0116] Figure 18 This is a fault diagnosis model diagram of Embodiment 1 of the present invention. Detailed Implementation

[0117] The present invention will now be described in detail so that its advantages and features can be more easily understood by those skilled in the art, thereby providing a clearer and more explicit definition of the scope of protection of the present invention.

[0118] The following provides a brief overview of one or more aspects to offer a basic understanding of them. This overview is not an exhaustive summary of all conceived aspects, nor is it intended to identify key or decisive elements of all aspects, nor to define the scope of any or all aspects. Its sole purpose is to present some concepts of one or more aspects in a simplified form to prepare for the more detailed descriptions that follow.

[0119] In the description of this invention, it should be understood that the terms "upper", "lower", "front", "rear", "left", "right", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0120] like Figure 1-18 As shown, a method for diagnosing faults in photovoltaic array cables includes the following steps:

[0121] S01: Obtain historical current I of the photovoltaic string m Voltage data U m String temperature T model Component nominal operating current I mp Component nominal open-circuit voltage V oc Component current temperature coefficient α, irradiance I r Wind speed W s Rainfall (R) and ambient temperature (T) emp ;

[0122] Among them, the historical current I of the photovoltaic string m Voltage data U m String temperature T model Component nominal operating current I mp Component nominal open-circuit voltage V oc Component current temperature coefficient α, irradiance I r Wind speed W s Rainfall (R) and ambient temperature (T) emp For a complete day's data, the data granularity is determined based on the actual data collection and the workload of computer technology, and is denoted as τ.

[0123] S02: Perform data cleaning on the acquired data to obtain valid data, specifically including the following steps:

[0124] S02-01: Irradiance data I r >100W / m 2 Given the first and last time points, for I m Um T model I r W s R, T emp Extract the data and denote it as Data;

[0125] S02-02: Irradiance data in Data r To perform the cleaning, follow these steps:

[0126] S02-02-01: Calculate the theoretical irradiance I according to the following formula. r0 ;

[0127]

[0128] Among them, Ir stc The standard irradiance is 1000 W / m². 2 ;

[0129] S02-02-02: Calculate the deviation δIr(t) between real-time irradiance and theoretical irradiance using the following formula. i ):

[0130] δIr(t i )=|Ir(t i )-Ir0(t i )|

[0131] S02-02-03: Calculate δIr(t) i The median and standard deviation of () are denoted as M. Ir S Ir ;

[0132] S02-02-04: If δIr(t) i )>M Ir +α″*S Ir If α′ is an adjustable factor, then determine Ir(t) i () represents an outlier, and the value at time t is recorded. i ;

[0133] S02-02-05: Calculate time t respectively i-1 and t i+1 With t i I r and I r0 The slopes are denoted as ΔI. r and ΔI r0 ;

[0134] S02-02-06: Comparison of ΔI r and ΔI r0 The absolute value of |ΔI r |>|ΔIr0 |, then Ir(t) i )=Ir0(t i );

[0135] S02-02-07: Repeat steps S02-02-02 to S02-02-06 until there are no outliers;

[0136] S02-03: Historical current I of photovoltaic strings in Data m To perform the cleaning, follow these steps:

[0137] S02-03-01: Calculate the theoretical current Im0(t) according to the following formula. i ):

[0138]

[0139] S02-03-02: Calculate the deviation δIm(t) between the real-time current value and the theoretical current value using the following formula. i ):

[0140] δIm(t i )=|Im(t i )-Im0(t i )|

[0141] S02-03-03: Calculate δIm(t) i The median and standard deviation of () are denoted as M. Im S Im ;

[0142] S02-03-04: If δIm(t) i )>M Im +α″*S Im If α″ is an adjustable factor, then determine Im(t) i If t is an outlier, then Im(t) i )=Im0(t i );

[0143] S02-03-05: Repeat steps S02-03-02 to S02-02-04 until there are no outliers;

[0144] S02-04: Regarding the voltage data U in Data... m To perform the cleaning, follow these steps:

[0145] S02-04-01: At 10-minute intervals, for U m Divide into segments, denoted as U mi ;

[0146] S02-04-02: Calculate U mjThe median and standard deviation are denoted as Muj and uj, respectively;

[0147] S02-04-03: If |U mj (t i )-Muj|>β u Suj, β u If U is an adjustable factor, then mj (t i ) represents an outlier;

[0148] S02-04-04: Calculate and replace U using interpolation. mj (t i );

[0149] S02-05: Regarding the ambient temperature data T in Data... emp To perform the cleaning, follow these steps:

[0150] S02-05-01: At 10-minute intervals, for T emp Divide into segments, denoted as T. empj ;

[0151] S02-05-02: Calculate T empj The median and standard deviation are denoted as Mtj and Stj, respectively;

[0152] S02-05-03: If |T empj (t i )-Mtj|>β Te Stj, β Te If T is an adjustable factor, then empj (t i ) represents an outlier;

[0153] S02-05-04: Calculate and replace T using interpolation. empj (t i );

[0154] S02-06: Regarding the wind speed data W in Data... s To perform the cleaning, follow these steps:

[0155] S02-06-01: Obtain the maximum wind speed at the photovoltaic power station project site over the past 10 years, denoted as W. m ;

[0156] S02-06-01: At 10-minute intervals, W... s Divide into segments, denoted as W. sj ;

[0157] S02-06-02: If W sj (t i )>Wm or W sj (t i If W < 0, then use interpolation to calculate and replace W. sj (t i ).

[0158] S02-07: Component temperature data T in Data model To perform the cleaning, follow these steps:

[0159] S02-07-01: At 10-minute intervals, for T model Divide into segments, denoted as T. modelj ;

[0160] S02-07-02: Calculate T modelj The median and standard deviation are denoted as Mmj and Smj, respectively;

[0161] S02-07-03: If |T modelj (t i )-Mmj|>β Tm Smj, β Tm If T is an adjustable factor, then modelj (t i ) represents an outlier;

[0162] S02-07-04: Calculate and replace T using interpolation. modelj (t i ).

[0163] S03: Calculate the effective data after cleaning to obtain the theoretical current data I0 of the photovoltaic string;

[0164] S04: Combining the theoretical current data I0 and actual current data I of the photovoltaic string m Capture abnormal data series;

[0165] S05: Use statistical analysis methods to perform a preliminary diagnosis of abnormal data series and obtain a preliminary fault status;

[0166] S06: Establish a fault diagnosis model using meteorological data, string flow anomaly data, and preliminary fault status as input sources;

[0167] S07: Determine the fault of the photovoltaic array cable based on the abnormal data series and fault diagnosis model.

[0168] Step S04, the steps for capturing the series of abnormal data include:

[0169] S04-01: Based on the obtained valid string current data and theoretical current data, calculate the string current deviation rate using the following formula:

[0170]

[0171] S04-02: For δ(t) i Make a judgment if δ(t) i If ) < 0.2, then record that moment in set U. f In the middle, denoted as U f ={t1, t2, ..., t N}, N∈N * N* is a natural number; if δ(t) i If the value is greater than 0.7, then that moment is recorded in set U. p In the middle, denoted as U p ={t1, t2, ..., t N}, N∈N * .

[0172] Step S05, the steps for obtaining the preliminary fault status include:

[0173] S05-01: Set the fault status word, denoted as F. s ;

[0174] S05-02: For U f and U p Make a judgment if U f and U p If all are non-empty sets, then F s =1; if U f Let U be a non-empty set, and U p If F is an empty set, then F s =2; if U f If it is empty, then F s =0;

[0175] S05-03: Regarding U f U p Perform a transformation, following the rules below: calculate t sequentially according to the data series order. i+1 -t i If t i+l -t i If t > τ, then t i =(t i , t i If t i+1 -t i If t = τ, then continue calculating the next time step until t. i+l+l -t i+l >τ, at this time t i =(t i , t i+l ), t i+1 To t i+l Delete, transformed U f Up Let it be U f1 U p1 ;

[0176] S05-04: Regarding U f U p Perform a second transformation, with the following transformation rules: For U f1 U p1 Each element in the array increases by t i Maximum current I at time t m Maximum irradiance Ir, maximum wind speed Ws, and highest ambient temperature T emp The transformed set is denoted as U. f2 U p2 .

[0177] Step S06, the steps for establishing the fault diagnosis model include:

[0178] S06-01: Establish a data structure for string cable fault records. The data structure includes string number, date, set of fault times, rainfall condition R, fault status, and number of faults, denoted as D. s (SN, Date, U) f2 U p2 R, F s F t );

[0179] S06-02: Calculate the number of daily failures F t The calculation formula is as follows:

[0180] F t =Length(U f2 )

[0181] Where: Length is the computation set U f2 Number of elements;

[0182] S06-03: Calculate the correlation factors for each input source. The specific steps are as follows:

[0183] S06-03-1: Calculate the correlation factor between fault and wind speed, following these steps:

[0184] S06-03-1-1: Calculate U p2 The maximum wind speed Ws0 in each element range;

[0185] S06-03-1-2: Obtain U f2 Maximum wind speed Ws in each element range s The wind speed abrupt change factor is calculated using the following formula to obtain the abrupt change factor dataset, denoted as: ε={ε1, ε2, ε3, ..., ε N}, N = 1, 2, 3...;

[0186]

[0187] S06-03-1-3: Calculate ε in the set ε within one week i The probability greater than 1.5 is defined as the wind speed correlation factor, denoted as P. w ;

[0188] S06-03-2: Calculate the correlation factor between faults and rainfall conditions, following these steps:

[0189] S06-03-2-1: Set the weight values ​​for rainfall in weather forecasts as follows:

[0190] No rain weight: 1%, light rain weight: 3%, moderate rain weight: 6%, heavy rain weight: 12%, rainstorm weight: 26%, heavy rainstorm weight: 52%;

[0191] S06-03-2-2: Calculate F within the most recent week s The average rainfall weight value with a value of 0 is denoted as β0.

[0192] S06-03-2-3: Calculate F within the most recent week s The average rainfall weight value with a value of 1 is denoted as β1.

[0193] S06-03-2-4: The correlation factor between faults and rainfall is defined as the ratio of β1 to β0, denoted as C. R ,Right now:

[0194]

[0195] S06-03-3: Calculate the correlation factor between the fault and ambient temperature, following these steps:

[0196] S06-03-3-1: Calculate F within the most recent week s The average ambient temperature with a value of 0 is denoted as T0.

[0197] S06-03-3-2: Calculate F within the most recent week s The average ambient temperature with a value of 1 is denoted as T1.

[0198] S06-03-3-3: Calculate U p2 The average ambient temperature for each element range is denoted as T. p ;

[0199] S06-03-3-4: Calculate U f2 The average ambient temperature for each element range is denoted as T. f ;

[0200] S06-03-3-5: The correlation factor between fault and ambient temperature is defined as T0, T1, T p T f The function, denoted as M T The calculation formula is as follows:

[0201]

[0202] S06-03-4: Calculate U for each day of the past week. f2 The repetition rate Rep for each element is determined by the following steps:

[0203] S06-03-4-1: Based on the sunrise and sunset times T within a week rise T set Using τ as the step size, determine the time t at a certain moment. i Does +τ fall within all U values ​​within a week? f2 If it falls into the set U, then add that moment to the set U. com In the middle, denoted as U com ={t1, t2, ..., t N}, and record U for each day f2 The number of elements that fall into the middle is denoted as N. i ;

[0204] S06-03-4-2: If there exists a certain day N i ≤Ft i If the value is positive, then Rep = 1; otherwise, Rep = 0.

[0205] Example 1

[0206] like Figure 1-18 As shown, a method for diagnosing faults in photovoltaic array cables includes the following steps:

[0207] S01: Obtain historical current I of the photovoltaic string m Voltage data U m String temperature T model Component nominal operating current I mp Component nominal open-circuit voltage V oc Component current temperature coefficient α, irradiance I r Wind speed W s Rainfall (R) and ambient temperature (T) emp ; Figure 7 This is the photovoltaic string current curve of Embodiment 1 of the present invention. Figure 8 This is the irradiance data curve for Embodiment 1 of the present invention. Figure 9 This is the ambient temperature data curve for Embodiment 1 of the present invention. Figure 10 This is the string temperature data curve of Embodiment 1 of the present invention. Figure 11The ambient temperature data curve for Embodiment 1 of the present invention and the rainfall data are shown in Table 1.

[0208] Table 1

[0209] date 10 / 24 10 / 25 10 / 26 10 / 27 10 / 28 10 / 29 10 / 30 R Light rain partly cloudy clear clear partly cloudy clear clear

[0210] S02: Clean the acquired data to obtain valid data;

[0211] S03: Calculate the effective data after cleaning to obtain the theoretical current data I0 of the photovoltaic string;

[0212] S04: Combining the theoretical current data I0 and actual current data I of the photovoltaic string m Extract abnormal data series and obtain set U f and U p ;

[0213] S05: Use statistical analysis methods to perform a preliminary diagnosis of abnormal data series and obtain a preliminary fault status;

[0214] S06: Establish a fault diagnosis model using meteorological data, string flow anomaly data, and preliminary fault status as input sources;

[0215] S07: Based on the abnormal data series and fault diagnosis model, it is determined that there is a loose connection problem in the photovoltaic array cable.

[0216] In step S01, the historical current I of the photovoltaic string m Voltage data U m String temperature T model Component nominal operating current I mp Component nominal open-circuit voltage V oc Component current temperature coefficient α, irradiance I r Wind speed W s Rainfall (R) and ambient temperature (T) emp For a complete day's data, the data granularity is determined based on the actual data collection and the workload of computer technology, denoted as τ, where τ = 1 minute.

[0217] In step S02, the data cleaning steps are as follows: Figure 2 As shown, the specific steps include:

[0218] S02-01: Irradiance data I r >100W / m 2 Given the first and last time points, for I m U m T model I r W s R, T empExtract the data and denote it as Data;

[0219] S02-02: Irradiance data in Data r Perform cleaning;

[0220] S02-03: Historical current I of photovoltaic strings in Data m Perform cleaning;

[0221] S02-04: Regarding the voltage data U in Data... m Perform cleaning;

[0222] S02-05: Regarding the ambient temperature data T in Data... emp Perform cleaning;

[0223] S02-06: Regarding the wind speed data W in Data... s Perform cleaning;

[0224] S02-07: Component temperature data T in Data model Clean it.

[0225] In step S02-02, the irradiance data I in Data is processed. r The cleaning steps include:

[0226] S02-02-01: Calculate the theoretical irradiance I according to the following formula. r0 ;

[0227]

[0228] Among them, Ir stc The standard irradiance is 1000 W / m². 2 ;

[0229] S02-02-02: Calculate the deviation δIr(t) between real-time irradiance and theoretical irradiance using the following formula. i ):

[0230] δIr(t i )=|Ir(t i )-Ir0(t i )|

[0231] S02-02-03: Calculate δIr(t) i The median and standard deviation of () are denoted as M. Ir S Ir ;

[0232] S02-02-04: If δIr(t) i )>M Ir +α′*S IrIf α′ is an adjustable factor, then determine Ir(t) i () represents an outlier, and the value at time t is recorded. i ;

[0233] S02-02-05: Calculate time t respectively i-1 and t i+1 With t i I r and I r0 The slopes are denoted as ΔI. r and ΔI r0 ;

[0234] S02-02-06: Comparison of ΔI r and ΔI r0 The absolute value of |ΔI r |>|ΔI r0 |, then Ir(t) i )=Ir0(t i );

[0235] S02-02-07: Repeat steps S02-02-02 to S02-02-06 until there are no outliers.

[0236] In steps S02-03, the historical current I of the photovoltaic string in Data is... m The cleaning steps include:

[0237] S02-03-01: Calculate the theoretical current I according to the following formula. m0 :

[0238]

[0239] S02-03-02: Calculate the deviation δIm(t) between the real-time current value and the theoretical current value using the following formula. i ):

[0240] δIm(t i )=|Im(t i )-Im0(t i )|

[0241] S02-03-03: Calculate δIm(t) i The median and standard deviation of () are denoted as M. Im S Im ;

[0242] S02-03-04: If δIm(t) i )>M Im +α″*S Im If α″ is an adjustable factor, then determine Im(t) i If t is an outlier, then Im(t)i )=Im0(t i );

[0243] S02-03-05: Repeat steps S02-03-02 to S02-02-04 until there are no outliers.

[0244] In steps S02-04, the voltage data U in Data is processed. m The cleaning steps include:

[0245] S02-04-01: At 10-minute intervals, for U m Divide into segments, denoted as U mi ;

[0246] S02-04-02: Calculate U mj The median and standard deviation are denoted as Muj and Suj, respectively.

[0247] S02-04-03: If |U mj (t i )-Muj|>β u Suj, β u If U is an adjustable factor, then mj (t i ) represents an outlier;

[0248] S02-04-04: Calculate and replace U using interpolation. mj (t i ).

[0249] In steps S02-05, the ambient temperature data T in Data is processed. emp The cleaning steps include:

[0250] S02-05-01: At 10-minute intervals, for T emp Divide into segments, denoted as T. empj ;

[0251] S02-05-02: Calculate T empj The median and standard deviation are denoted as Mtj and Stj, respectively;

[0252] S02-05-03: If |T empj (t i )-Mtj|>β Te Stj, β Te If T is an adjustable factor, then empj (t i ) represents an outlier;

[0253] S02-05-04: Calculate and replace T using interpolation. empj (t i).

[0254] In steps S02-06, the wind speed data W in Data is processed. s The cleaning steps include:

[0255] S02-06-01: Obtain the maximum wind speed at the photovoltaic power station project site over the past 10 years, denoted as W. m ;

[0256] S02-06-01: At 10-minute intervals, W... s Divide into segments, denoted as W. sj ;

[0257] S02-06-02: If W sj (t i )>W m or W sj (t i If W < 0, then use interpolation to calculate and replace W. sj (t i ).

[0258] In steps S02-07, the component temperature data T in Data is processed. model The cleaning steps include:

[0259] S02-07-01: At 10-minute intervals, for T model Divide into segments, denoted as T. modelj；

[0260] S02-07-02: Calculate T modelj The median and standard deviation are denoted as Mmj and Smj, respectively;

[0261] S02-07-03: If |T modelj (t i )-Mmj|>β Tm Smj, β Tm If T is an adjustable factor, then modelj (t i ) represents an outlier;

[0262] S02-07-04: Calculate and replace T using interpolation. modelj (t i ).

[0263] The current curve, irradiance curve, ambient temperature curve, string temperature curve, and wind speed curve after data cleaning are shown below. Figure 12-16 As shown.

[0264] In step S03, the theoretical current data series I0 of the photovoltaic string is calculated according to step S02-03-01.

[0265] In step S04, abnormal data series are captured to obtain set U. f and U p The steps include:

[0266] S04-01: Based on the obtained valid string current data and theoretical current data, calculate the string current deviation rate using the following formula:

[0267]

[0268] S04-02: For δ(t) i Make a judgment if δ(t) i If ) < 0.2, then record that moment in set U. f In the middle, denoted as U f ={t1, t2, ..., t N}, N∈N * N* is a natural number; if δ(t) i If the value is greater than 0.7, then that moment is recorded in set U. p In the middle, denoted as U p ={t1, t2, ..., t N}, N∈N * .

[0269] Step S05, the steps for obtaining the preliminary fault status include:

[0270] S05-01: Set the fault status word, denoted as F. s ;

[0271] S05-02: For U f and U p Make a judgment if U f and U p If all are non-empty sets, then F s =1; if U f Let U be a non-empty set, and U p If F is an empty set, then F s =2; if U f If it is empty, then F s =0;

[0272] S05-03: Regarding U f U p Perform a transformation, following the rules below: calculate t sequentially according to the data series order. i+1 -t i If t i+1 -t i If t > τ, then t i =(t i , t i If t i+1 -ti If t = τ, then continue calculating the next time step until t. i+l+1 -t i+l >τ, at this time t i =(t i , t i+l ), t i+1 To t i+l Delete, transformed U f U p Let it be U f1 U p1 ;

[0273] S05-04: Regarding U f U p Perform a second transformation, with the following transformation rules: For U f1 U p1 Each element in the array increases by t i Maximum current I at time t m Maximum irradiance Ir, maximum wind speed Ws, and highest ambient temperature T emp The transformed set is denoted as U. f2 U p2 As shown in Table 2.

[0274] Table 2

[0275]

[0276] Step S06, the steps for establishing the fault diagnosis model include:

[0277] S06-01: Establish a data structure for string cable fault records. The data structure includes string number, date, set of fault times, rainfall condition R, fault status, and number of faults, denoted as D. s (SN, Date, U) f2 U p2 R, F s F t );

[0278] S06-02: Calculate the number of daily failures F t The calculation formula is as follows:

[0279] F t =Length(U f2 )

[0280] Where: Length is the computation set U f2 Number of elements;

[0281] S06-03: Calculate the correlation factors for each input source. The specific steps are as follows:

[0282] S06-03-1: Calculate the correlation factor between fault and wind speed, following these steps:

[0283] S06-03-1-1: Calculate U p2 The maximum wind speed Ws0 in each element range;

[0284] S06-03-1-2: Obtain U f2 Maximum wind speed Ws in each element range s The wind speed abrupt change factor is calculated using the following formula to obtain the abrupt change factor dataset, denoted as: ε={ε1, ε2, ε3, ..., ε N}, N = 1, 2, 3...;

[0285]

[0286] S06-03-1-3: Calculate ε in the set ε within one week i The probability greater than 1.5 is defined as the wind speed correlation factor, denoted as P. w ;

[0287] S06-03-2: Calculate the correlation factor between faults and rainfall conditions, following these steps:

[0288] S06-03-2-1: Set the weight values ​​for rainfall in weather forecasts as follows:

[0289] No rain weight: 1%, light rain weight: 3%, moderate rain weight: 6%, heavy rain weight: 12%, rainstorm weight: 26%, heavy rainstorm weight: 52%;

[0290] S06-03-2-2: Calculate F within the most recent week s The average rainfall weight value with a value of 0 is denoted as β0.

[0291] S06-03-2-3: Calculate F within the most recent week s The average rainfall weight value with a value of 1 is denoted as β1.

[0292] S06-03-2-4: The correlation factor between faults and rainfall is defined as the ratio of β1 to β0, denoted as C. R ,Right now:

[0293]

[0294] S06-03-3: Calculate the correlation factor between the fault and ambient temperature, following these steps:

[0295] S06-03-3-1: Calculate F within the most recent week s The average ambient temperature with a value of 0 is denoted as T0.

[0296] S06-03-3-2: Calculate F within the most recent week s The average ambient temperature with a value of 1 is denoted as T1.

[0297] S06-03-3-3: Calculate U p2 The average ambient temperature for each element range is denoted as T. p ;

[0298] S06-03-3-4: Calculate U f2 The average ambient temperature for each element range is denoted as T. f ;

[0299] S06-03-3-5: The correlation factor between fault and ambient temperature is defined as T0, T1, T p T f The function, denoted as M T The calculation formula is as follows:

[0300]

[0301] S06-03-4: Calculate U for each day of the past week. f2 The repetition rate Rep for each element is determined by the following steps:

[0302] S06-03-4-1: Based on the sunrise and sunset times T within a week rise T set Using τ as the step size, determine the time t at a certain moment. i Does +τ fall within all U values ​​within a week? f2 If it falls into the set U, then add that moment to the set U. com In the middle, denoted as U com ={t1, t2, ..., t N}, and record U for each day f2 The number of elements that fall into the middle is denoted as N. i ;

[0303] S06-03-4-2: If there exists a certain day N i ≤Ft i If the value is positive, then Rep = 1; otherwise, Rep = 0, as shown in Table 3.

[0304] Table 3

[0305]

[0306] Any parts or structures not specifically described in this invention can be made using existing technologies or products, and will not be elaborated upon here.

[0307] The above description is merely an embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention specification, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for diagnosing faults in photovoltaic array cables, characterized in that, Includes the following steps: S01: Obtain photovoltaic-related parameters, including the historical current I of the photovoltaic string. m Voltage data U m String temperature T model Component nominal operating current I mp Component nominal open-circuit voltage V oc Component current temperature coefficient α, irradiance I r Wind speed W s Rainfall (R) and ambient temperature (T) emp ; S02: Clean the acquired data to obtain valid data; S03: Calculate the effective data after cleaning to obtain the theoretical current data I0 of the photovoltaic string; S04: Combining the theoretical current data I0 and actual current data I of the photovoltaic string m Capture abnormal data series; S05: Perform a preliminary diagnosis on the abnormal data series to obtain a preliminary fault status; S06: Establish a fault diagnosis model using meteorological data, abnormal string current data, and preliminary fault status as input sources; S07: Determine photovoltaic array cable faults based on abnormal data series and fault diagnosis models; Step S06, the steps for establishing the fault diagnosis model include: S06-01: Establish a data structure for string cable fault records. The data structure includes string number, date, set of fault times, rainfall condition R, fault status, and number of faults, denoted as D. s (SN, Date, U) f2 U p2 R, F s F t ); S06-02: Calculate the number of daily failures F t The calculation formula is as follows: F t =Length(U f2 ) Where: Length is the computation set U f2 Number of elements; S06-03: Calculate the correlation factors for each input source. The specific steps are as follows: S06-03-1: Calculate the correlation factor between fault and wind speed, following these steps: S06-03-1-1: Calculate U p2 The maximum wind speed Ws0 in each element range; S06-03-1-2: Obtain U f2 Maximum wind speed Ws in each element range s The wind speed abrupt change factor is calculated using the following formula to obtain the abrupt change factor dataset, denoted as: ε={ε1, ε2, ε3, …, ε N }, N = 1, 2, 3…; S06-03-1-3: Calculate ε in the set ε within one week i The probability greater than 1.5 is defined as the wind speed correlation factor, denoted as P. w ; S06-03-2: Calculate the correlation factor between faults and rainfall conditions, following these steps: S06-03-2-1: Set the weight values ​​for rainfall in weather forecasts as follows: No rain weight: 1%, light rain weight: 3%, moderate rain weight: 6%, heavy rain weight: 12%, rainstorm weight: 26%, heavy rainstorm weight: 52%; S06-03-2-2: Calculate F within the most recent week s The average rainfall weight value with a value of 0 is denoted as β0. S06-03-2-3: Calculate F within the most recent week s The average rainfall weight value with a value of 1 is denoted as β1. S06-03-2-4: The correlation factor between faults and rainfall is defined as the ratio of β1 to β0, denoted as C. R ,Right now: S06-03-3: Calculate the correlation factor between the fault and ambient temperature, following these steps: S06-03-3-1: Calculate F within the most recent week s The average ambient temperature with a value of 0 is denoted as T0. S06-03-3-2: Calculate F within the most recent week s The average ambient temperature with a value of 1 is denoted as T1. S06-03-3-3: Calculate U p2 The average ambient temperature for each element range is denoted as T. p ; S06-03-3-4: Calculate U f2 The average ambient temperature for each element range is denoted as T. f ; S06-03-3-5: The correlation factor between fault and ambient temperature is defined as T0, T1, T p T f The function, denoted as M T The calculation formula is as follows: S06-03-4: Calculate U for each day of the past week. f2 The repetition rate Rep for each element is determined by the following steps: S06-03-4-1: Based on the sunrise and sunset times T within a week rise T set Using τ as the step size, determine the time t at a certain moment. i Does +τ fall within all U values ​​within a week? f2 If it falls into the set U, then add that moment to the set U. com In the middle, denoted as U com ={t1, t2, ..., t N }, and record U for each day f2 The number of elements that fall into the middle is denoted as N. i ; S06-03-4-2: If there exists a certain day N i ≤Ft i If the value is positive, then Rep = 1; otherwise, Rep = 0.

2. The method for fault diagnosis of photovoltaic array cables according to claim 1, characterized in that, Step S02, which involves cleaning the acquired data to obtain valid data, includes the following steps: S02-01: Irradiance data I r >100W / m 2 Given the first and last time points, for I m U m T model I r W s R, T emp Extract the data and denote it as Data; S02-02: Irradiance data in Data r Perform cleaning; S02-03: Historical current I of photovoltaic strings in Data m Perform cleaning; S02-04: Regarding the voltage data U in Data... m Perform cleaning; S02-05: Regarding the ambient temperature data T in Data... emp Perform cleaning; S02-06: Regarding the wind speed data W in Data... s Perform cleaning; S02-07: Component temperature data T in Data model Clean it.

3. The method for fault diagnosis of photovoltaic array cables according to claim 2, characterized in that, In step S02-02, the irradiance data I in Data is processed. r The cleaning steps include: S02-02-01: Calculate the theoretical irradiance I according to the following formula. r0 ; Among them, Ir stc The standard irradiance is 1000 W / m². 2 ; S02-02-02: Calculate the deviation δIr(t) between real-time irradiance and theoretical irradiance using the following formula. i ): δIr(t i )=|Ir(t i ))-Ir0(t i )| S02-02-03: Calculate δIr(t) i The median and standard deviation of () are denoted as M. Ir S Ir ; S02-02-04: If δIr(t) i )>M Ir +a′*S Ir If α′ is an adjustable factor, then determine Ir(t) i () represents an outlier, and the value at time t is recorded. i ; S02-02-05: Calculate time t respectively i-1 and t i+1 With t i I r and I r0 The slopes are denoted as ΔI. r and △I r0 ; S02-02-06: Comparison with △I r and △I r0 The absolute value of |△I r |>|△I r0 |, then Ir(t) i )=Ir0(t i ); S02-02-07: Repeat steps S02-02-02 to S02-02-06 until there are no outliers.

4. The method for fault diagnosis of photovoltaic array cables according to claim 2, characterized in that, In steps S02-03, the historical current I of the photovoltaic string in Data is... m The cleaning steps include: S02-03-01: Calculate the theoretical current I according to the following formula. m0 : S02-03-02: Calculate the deviation δIm(t) between the real-time current value and the theoretical current value using the following formula. i ): δIm(t i )=|Im(t i )-Im0(t i )| S02-03-03: Calculate δIm(t) i The median and standard deviation of () are denoted as M. Im S Im ; S02-03-04: If δIm(t) i )>M Im +α″*S Im If a″ is an adjustable factor, then determine Im(t) i If t is an outlier, then Im(t) i )=Im0(t i ); S02-03-05: Repeat steps S02-03-02 to S02-02-04 until there are no outliers.

5. A method for diagnosing photovoltaic array cable faults according to claim 2, characterized in that, Step S In 02-04, regarding the voltage data U in Data... m The cleaning steps include: S02-04-01: At 10-minute intervals, for U m Divide into segments, denoted as U mj ; S02-04-02: Calculate U mj The median and standard deviation are denoted as Muj and Suj, respectively. S02-04-03: If |U mj (t i )-Muj|>β u Suj, β u If U is an adjustable factor, then mj (t i ) represents an outlier; S02-04-04: Calculate and replace U using interpolation. mj (t i ).

6. A method for diagnosing photovoltaic array cable faults according to claim 2, characterized in that, In steps S02-05, the ambient temperature data T in Data is processed. emp The cleaning steps include: S02-05-01: At 10-minute intervals, for T emp Divide into segments, denoted as T. empj ; S02-05-02: Calculate T empj The median and standard deviation are denoted as Mtj and Stj, respectively; S02-05-03: If |T empj (t i )-Mtj|>β Te Stj, β Te If T is an adjustable factor, then empj (t i ) represents an outlier; S02-05-04: Calculate and replace T using interpolation. empj (t i ); In steps S02-07, the component temperature data T in Data is processed. model The cleaning steps include: S02-07-01: At 10-minute intervals, for T model Divide into segments, denoted as T. modelj ; S02-07-02: Calculate T modelj The median and standard deviation are denoted as Mmj and Smj, respectively; S02-07-03: If |T modelj (t i )-Mmj|>β Tm Smj, β Tm If T is an adjustable factor, then modelj (t i ) represents an outlier; S02-07-04: Calculate and replace T using interpolation. modelj (t i ).

7. A method for diagnosing photovoltaic array cable faults according to claim 2, characterized in that, In steps S02-06, the wind speed data W in Data is processed. s The cleaning steps include: S02-06-01: Obtain the maximum wind speed at the photovoltaic power station project site over the past 10 years, denoted as W. m ; S02-06-02: At 10-minute intervals, W... s Divide into segments, denoted as W. sj ; S02-06-03: If W sj (t i W m or W sj (t i If W < 0, then use interpolation to calculate and replace W. sj (t i ).

8. The method for fault diagnosis of photovoltaic array cables according to claim 1, characterized in that, Step S04, the steps for capturing the series of abnormal data include: S04-01: Based on the obtained valid string current data and theoretical current data, calculate the string current deviation rate using the following formula: S04-02: For δ(t) i Make a judgment if δ(t) i If ) < 0.2, then record that moment in set U. f In the middle, denoted as U f ={t1, t2, ..., t N }, N∈N * N* is a natural number; if δ(t) i If the value is greater than 0.7, then that moment is recorded in set U. p In the middle, denoted as U p ={t1, t2, ..., t N }, N∈N * .

Images