Photovoltaic array line fault diagnosis method based on shadow power tracking

Abstract

The invention provides a photovoltaic array line fault diagnosis method based on shadow power tracking. The method comprises the steps of obtaining a sampling point array current and an array voltage of a target photovoltaic array system; calculating the sampling point power of the target photovoltaic array system; determining whether the target photovoltaic array system has a fault or not; determining the fault type of the target photovoltaic array system; line fault detection of the target photovoltaic array system is realized; and multi-peak maximum power point tracking control of the target photovoltaic array system is realized.

Description

Technical Field

[0001] The invention relates to renewable energy, artificial intelligence and power electronic conversion technology, and belongs to the field of electrical engineering. Background Art

[0002] Photovoltaic power generation is the main way to use solar energy, and its installed capacity is increasing globally. However, due to the limitation of available land resources, the installed capacity of photovoltaic power generation is gradually approaching saturation. Therefore, how to improve the operating efficiency of photovoltaic power generation under limited resource conditions has become a major problem.

[0003] The light intensity and temperature of the external environment are the main factors affecting photovoltaic power generation, and the volatility of the two determines the randomness and volatility of photovoltaic power generation. Conventional maximum power point tracking control technologies such as the conductance increment method and the perturbation observation method can well achieve maximum power point tracking under uniform illumination, but changes in weather and environment have caused the photovoltaic array to be in a multi-peak shadow state for a long time. Conventional tracking control technologies are difficult to achieve global maximum power point tracking, which reduces the operating efficiency of the photovoltaic system.

[0004] Photovoltaic arrays are exposed to the outdoors for a long time to obtain light energy. Factors such as animal invasion and rain corrosion make photovoltaic arrays very prone to line faults. The existence of these faults will also cause great interference to the operating efficiency and stability of the system, and even cause safety hazards such as fire in serious cases. At present, the research methods of line fault detection mainly include those based on signals, statistical probability and neural networks, but it is difficult to locate the fault and the real-time performance cannot be guaranteed.

[0005] Therefore, to address this problem, it is necessary to realize photovoltaic array line fault diagnosis under complex conditions and reduce the impact of shadows and line faults on the power generation efficiency of the photovoltaic array system to meet actual needs. Summary of the invention

[0006] In order to address the deficiencies of the prior art, the present invention proposes a photovoltaic array line fault diagnosis method based on shadow power tracking, which realizes photovoltaic array line fault diagnosis under complex conditions and reduces the impact of shadows and line faults on the power generation efficiency of the photovoltaic array system.

[0007] A photovoltaic array line fault diagnosis method based on shadow power tracking is characterized in that: the target photovoltaic array system is initially diagnosed according to the following steps S1 to S4 in real time to determine whether there is a fault and the type of fault:

[0008] Step S1, based on a preset sampling period, sampling the target photovoltaic array system, obtaining the array current and array voltage corresponding to the current sampling moment and the previous adjacent sampling moment at each sampling point, and then executing step S2;

[0009] Step S2, calculating and obtaining the power value of each sampling point corresponding to the current acquisition moment and the adjacent previous acquisition moment, and then executing step S3;

[0010] Step S3, based on the power values ​​of each sampling point corresponding to the current collection time and the previous adjacent collection time, respectively, obtain the power difference value and the power change percentage value of each sampling point at the current collection time and the previous adjacent collection time, and combine the preset range of the power difference value and the preset range of the power change percentage to determine whether the target photovoltaic array system has a fault, and if so, execute step S4;

[0011] Step S4, based on the power change percentage values ​​at the current collection moment and the previous adjacent collection moment, combined with the preset range of abnormality detection, determine the fault type of the target photovoltaic array system.

[0012] Furthermore, in step S3, the preset range of the power difference is ΔP<0, and the preset range of the power change percentage is Where ΔP is the power difference, Δ is the power change percentage, and fs is the sampling frequency.

[0013] Furthermore, in step S4, the abnormality detection preset range is Where Δ is the power change percentage, N s is the target PV array system row value, N p is the target photovoltaic array system column value, I m is the maximum power point current of the component, U m is the maximum power point voltage of the component.

[0014] Further, based on step S1 to step S4, it is determined that there is a line fault in the target photovoltaic array, and step S5 is also included to implement line fault detection of the target photovoltaic array system according to the fault characteristics:

[0015] Step S51, constructing a circuit model of the photovoltaic module according to the connection mode of the photovoltaic module in the photovoltaic array system, obtaining the working characteristics of the photovoltaic array, and then executing step S52;

[0016] Step S52, based on the circuit model of the photovoltaic module, construct a fault model, realize the collection of fault characteristics, and form a fault characteristic table, and then execute step S53;

[0017] Step S53, based on the preset test points of each group of strings in the photovoltaic array system, select the test point voltages, and determine the test point voltage sequence in order from small to large, and then execute step S54;

[0018] Step S54, according to the test point voltage sorting, sampling and testing the preset test points of each string in turn, obtaining the positive and negative current values ​​of each string, and then executing step S55;

[0019] Step S55, according to the test point voltage and the fault feature table, combined with the positive and negative current values ​​of each string, determine the line fault location and line fault type, and judge the number of faulty components.

[0020] Furthermore, when the test point voltage of each preset test point of each string is selected in step S53, the following formula is used:

[0021] U cs (i) = i*U oc +ε

[0022] Where: U cs is the test point voltage, U oc is the open circuit voltage of a single component, and ε is a decimal greater than 0.

[0023] Further, based on step S3, it is judged that there is no fault in the target photovoltaic array system, or based on steps S1 to S4, it is judged that there is no line fault in the target photovoltaic array, or based on step S5, line fault detection of the target photovoltaic array system is achieved, and step S6 is also included, according to the hybrid algorithm, to achieve multi-peak maximum power point tracking control of the target photovoltaic array system:

[0024] Step S61, using Circle chaotic mapping to generate an initial population of size N, each individual in the population represents a duty cycle particle, and then executing step S62;

[0025] Step S62, according to the generation order of the duty cycle particles in the initial population, the fitness of each duty cycle particle is calculated in turn to obtain the output power of each duty cycle particle, and the best duty cycle particle is determined according to the power optimization principle, and then step S63 is executed;

[0026] Step S63, based on the optimal duty cycle particle, according to the chaotic mapping principle, update each duty cycle particle, and then execute step S64;

[0027] Step S64, according to the generation order of each duty cycle particle after the update, the output power of each duty cycle particle is obtained in sequence, and the best duty cycle particle after the update is determined in combination with the best duty cycle particle at the target historical iteration moment, and then step S65 is executed;

[0028] Step S65, if the termination condition is met, the iteration is terminated to achieve stable output of maximum power; otherwise, steps S63 to S64 are iteratively repeated until the termination condition is met.

[0029] Furthermore, the step S61 generates duty cycle particles according to the following formula:

[0030] x i+1 =αxi (1-x i )0<α≤4

[0031] X i+1 =x i+1 *(u b -l b )+l b

[0032] In the formula, x i is the duty cycle particle generated according to the chaotic mapping formula, α is the mapping parameter, X i Based on the actual range (l b ,u b ) to generate the duty cycle particles.

[0033] Furthermore, the step S63 includes:

[0034] Step S631, based on each duty cycle particle, according to the chaotic mapping principle, generate a pseudo-random number p corresponding to each duty cycle particle and in the range of (0,1), and then execute step S632;

[0035] Step S632, based on the pseudo-random number corresponding to each duty cycle particle and according to the preset pseudo-random number range, determine whether to update each duty cycle particle according to the spiral predation principle of the Blue Whale algorithm. If yes, update according to the following formula; otherwise, execute step S633:

[0036] X i (t+1)=w*e bl cos(2πl)+X best (t)

[0037]

[0038] D′=|X best (t)-X i (t)|

[0039] l∈(-1,1)

[0040] b=1

[0041] Where i is the individual number in the population, i∈{1, 2, ..., N}; N is the population size of the WOA algorithm; b is a constant related to the spiral shape; l is a random number between (-1, 1); D' is the distance vector; t is the current iteration number, t max is the maximum number of iterations; w is the step inertia factor; X best is the particle with the best duty cycle;

[0042] Step S633, based on the number of iterations and the sine function, generate the parameter α and the vector parameter A using the following formula, and then execute step S634:

[0043]

[0044] A=2α*r1-α

[0045] Where α is a control parameter related to the number of iterations; t is the current number of iterations; t max is the maximum number of iterations; r1 is a random number (0,1);

[0046] Step S634, based on the value of the vector parameter A and the range of the preset vector parameter A, it is determined whether to update each duty cycle particle according to the shrinking and surrounding position principle of the Blue Whale algorithm. If yes, update each duty cycle particle according to the following formula, otherwise, execute step S635:

[0047] X i (t+1)=X best (t)-A*D*w

[0048] D=|C*X best (t)-X i (t)|

[0049] C=2*r2

[0050] Where, X i Based on the actual range (l b ,u b ) generated duty cycle particles; X best is the particle with the best duty cycle; C is a vector parameter, where |A|∈(0,1) in the encirclement stage; r2 is a random number of (0,1).

[0051] Step S635, based on the value of the vector parameter A and the Levy flight strategy, update each duty cycle particle according to the following formula:

[0052]

[0053]

[0054] α=α0*(X i (t)-X best (t))

[0055] Where, X i Based on the actual range (l b ,u b ) generated duty cycle particles; X bestis the particle with the best duty cycle; Levy(β) is the Levy flight strategy, β is 1.5; α is the distance vector, α0 is 0.01; μ and v are normal distributions between (0,1); Γ() is the Gamma distribution function.

[0056] Furthermore, in step S632, the pseudo-random number range is preset to p≥0.5.

[0057] Furthermore, in step S634, the range of the vector parameter A is preset to |A|<1. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 The figure is a schematic diagram of the implementation process of a photovoltaic array line fault diagnosis method based on shadow power tracking, where (a) is a line fault anomaly detection diagram, and (b) is a fault location detection diagram.

[0059] Figure 2 This is a schematic diagram of the fault classification process.

[0060] Figure 3 A 5*6 photovoltaic array system under shadow.

[0061] Figure 4 It is a manifestation of common line faults in photovoltaic array systems.

[0062] Figure 5 The normalized IV characteristic curves of line faults under series blocking diodes, where (a) is the IV characteristic curve under shadow conditions, (b) is the IV characteristic curve under line short circuit conditions, (c) is the IV characteristic curve under grounding conditions, and (d) is the IV characteristic curve under line short circuit conditions.

[0063] Figure 6 The normalized IV characteristic curves of line faults without series blocking diodes, where (a) is the IV characteristic curve under shadow conditions, (b) is the IV characteristic curve under line short circuit conditions, (c) is the IV characteristic curve under grounding conditions, and (d) is the IV characteristic curve under line short circuit conditions.

[0064] Figure 7 This is a curve diagram of the system parameters changes of the three components of the grounding of the photovoltaic array system without series blocking diodes.

[0065] Figure 8 The curve diagram of the changes in various system parameters of a component in the photovoltaic array system with no series blocking diodes and an internal short circuit.

[0066] Fig. 9 This is a curve diagram of the changes in system parameters of two components with short circuits between lines in a photovoltaic array system without series blocking diodes.

[0067] Fig.10 This is a curve diagram of the changes in system parameters of the photovoltaic array system without series blocking diodes.

[0068] Fig.11 This is the control flow chart of multi-peak maximum power point tracking.

[0069] Fig.12 This is the normalized array power-voltage curve of the photovoltaic array system under shadow conditions.

[0070] Fig.13 To improve the steady-state power tracking comparison of the Blue Whale algorithm with the Blue Whale algorithm, Grey Wolf algorithm, and Particle Swarm Optimizer, (a) is the steady-state power tracking comparison under a single peak condition, (b) is the steady-state power tracking comparison under a double peak condition, (c) is the steady-state power tracking comparison under a three-peak condition, and (d) is the steady-state power tracking comparison under a four-peak condition. Fig.14 To improve the number of multi-peak shadow tracking iterations of the Blue Whale algorithm, the Blue Whale algorithm, the Grey Wolf algorithm, and the Particle Swarm Optimizer are compared, where (a) is the comparison of the number of multi-peak shadow tracking iterations under unimodal conditions, (b) is the comparison of the number of multi-peak shadow tracking iterations under bimodal conditions, (c) is the comparison of the number of multi-peak shadow tracking iterations under trimodal conditions, and (d) is the comparison of the number of multi-peak shadow tracking iterations under quadrumodal conditions. DETAILED DESCRIPTION

[0071] refer to Figure 1 , (a) is the line fault anomaly detection diagram, (b) is the fault location detection diagram and Figure 2 Fault classification process diagram, perform fault diagnosis on the target photovoltaic array system, determine whether there is a fault and the fault type, and implement multi-peak maximum power point tracking control of the target photovoltaic array system:

[0072] Step S1, based on a preset sampling period, sampling the target photovoltaic array system, obtaining the array current and array voltage corresponding to the current sampling moment and the previous adjacent sampling moment at each sampling point, and then executing step S2;

[0073] Step S2, calculating the power value of each sampling point corresponding to the current acquisition moment and the adjacent previous acquisition moment by multiplying the array current and the array voltage, and then executing step S3;

[0074] Step S3, based on the power values ​​of each sampling point corresponding to the current acquisition moment and the adjacent previous acquisition moment, obtain the power difference and power change percentage values ​​of each sampling point at the current acquisition moment and the adjacent previous acquisition moment, combined with the power difference preset range ΔP<0 and the power change percentage preset range Wherein ΔP is the power difference, Δ is the power change percentage, and fs is the sampling frequency. It is determined whether the target photovoltaic array system has a fault. If so, step S4 is executed;

[0075] Step S4: Based on the power change percentage value of the current acquisition time and the previous acquisition time, combined with the abnormal detection preset range Where Δ is the power change percentage, N s is the target PV array system row value, N p is the target photovoltaic array system column value, I m is the maximum power point current of the component, U m is the maximum power point voltage of the component, and determines the type of fault existing in the target photovoltaic array system;

[0076] Based on steps S1 to S4, it is determined that there is a line fault in the target photovoltaic array, and further includes step S5, which implements line fault detection of the target photovoltaic array system according to the fault characteristics:

[0077] Step S51, according to Figure 3 The 5*6 photovoltaic array system under the shadow shown is constructed, the circuit model of the component is built, the circuit model of the photovoltaic component is constructed, the working characteristics of the photovoltaic array are obtained, and the relevant parameters of a single photovoltaic component are shown in Table 1, and then step S52 is executed;

[0078] Table 1

[0079] parameter Numeric <![CDATA[Open-circuit voltage U OC / V]]> 44.4 <![CDATA[Short-circuit current I SC / A]]> 5.4 <![CDATA[Maximum power point voltage U m / V]]> 35.4 <![CDATA[Maximum power point current I m / A]]> 4.95 <![CDATA[Maximum power P m / W]]> 175.23

[0080] Step S52: construct a fault model based on the circuit model of the photovoltaic module, that is, randomly perform fault processing on the photovoltaic module based on the circuit model of the photovoltaic module, such as ground fault. For details, see Figure 4 , and on this basis, the IV characteristic of the ground fault is simulated to obtain the normalized IV characteristic curve. For details, see Figure 5 and Figure 6 , realize the collection of fault characteristics and form a fault characteristic table, specifically N s Row N p The fault characteristics of the photovoltaic array system are shown in Table 2. pv is the array voltage, U C is the theoretical open circuit voltage of the faulty string, following the formula: U C =(N s -N F )*U OC , where N F is the number of faulty components in the string, U OC is the open circuit voltage of a single component, I T is the positive current of the string, I B is the negative electrode current of the string, k is the sequence of the string, and then step S53 is executed;

[0081] Table 2

[0082]

[0083]

[0084] Step S53: Based on the preset test points of each string in the photovoltaic array system, according to U cs (i) = i * U oc + ε to select the test point voltage, where U cs is the test point voltage, U oc is the open-circuit voltage of a single component, ε is a decimal greater than 0, and for the case of a series blocking diode, a test point voltage U CS (0) = U OC - ε can be added to distinguish open-circuit faults and in-line short circuits. The selection table of test voltages and corresponding duty cycles for a photovoltaic array system without a series blocking diode is shown in Table 3. Then, determine the test point voltage sorting in ascending order of the test point voltage and execute Step S54;

[0085] Table 3

[0086] <![CDATA[U CS (1)]]> <![CDATA[U CS (2)]]> <![CDATA[U CS (3)]]> <![CDATA[U CS (4)]]> Voltage / V 44.4+4.44 88.8+4.44 133.2+4.44 177.6+4.44 Duty cycle D 0.88 0.76 0.65 0.54

[0087] Step S54: According to the test point voltage sorting, sequentially perform point-by-point sampling tests on the preset test points of each string to obtain the positive and negative current values of each string, and then execute Step S55;

[0088] Step S55: Based on the test point voltage and the fault characteristic table, combined with the positive and negative current values of each string, determine the line fault location and line fault type, and judge the number of faulty components. Specifically, refer to Figures 7 to 10 , and the specific process is as follows, where k is the string sequence, I T is the positive current of the string, I B is the negative current of the string, N F is the number of faulty components in the string, N s is the row value of the target photovoltaic array system:

[0089] ① Any I Tk > 0, no line fault;

[0090] In the case of a series blocking diode:

[0091] ② At U CS (0), if I Tk = 0, then the k-sequence string has an open-circuit fault;

[0092] ③ At the remaining test voltages, if there is I Tk = I BK = 0, then the k-sequence string has an in-line short circuit fault, and the number of faulty components NF =N S -i;

[0093] ④ If there is I Tk >0,I BK <0,I Tm =0,I Bm >0, then the m-sequence and k-sequence strings have line-to-line short circuit faults, and the number of faulty components is N F =i;

[0094] ⑤Other cases are ground faults, and the number of faulty components is N F =N S -i.

[0095] ⑥After the inspection is completed, wait for manual replacement of components;

[0096] Without series blocking diode:

[0097] ②Existence I Tk =0, then the k-sequence string has a circuit breaker fault;

[0098] ③I Tk <0, if there is I Tk =I Bk , then a short circuit occurs in the k-sequence string, and the number of faulty components is N F =N S -i;

[0099] ④Existence I Tk >0,I BK <0,I Bm >0, then the m-sequence and k-sequence strings have line-to-line short circuit faults, and the number of faulty components is N F =i;

[0100] ⑤Other cases are ground faults, and the number of faulty components is N F =N S -i;

[0101] ⑥After the inspection is completed, wait for manual replacement of components;

[0102] Based on the fact that the line fault detection of the target photovoltaic array system has been achieved in step S5, the method further includes step S6. Table 4 is a table showing the distribution of the light intensity of the strings in the photovoltaic array system under shadow conditions. Fig.11 Multi-peak maximum power point tracking control flow chart, based on the hybrid algorithm, to achieve the multi-peak maximum power point tracking control of the target photovoltaic array system:

[0103] Table 4

[0104] Peak conditions <![CDATA[1000W / m 2 ]]> <![CDATA[600W / m 2 ]]> <![CDATA[400W / m 2 ]]> <![CDATA[200W / m 2 ]]> Single Peak 5 0 0 0 Twin Peaks 3 2 0 0 Three Peaks 2 2 1 0 Four Peaks 1 2 1 1

[0105] In step S61, Circle chaotic mapping is used to generate an initial population of size N according to the following formula, where each individual in the population represents a duty cycle particle, and then step S62 is executed:

[0106] x i+1 =αx i (1-x i )0<α≤4

[0107] X i+1 =x i+1 *(u b -l b )+l b

[0108] In the formula, x i is the duty cycle particle generated according to the chaotic mapping formula, α is the mapping parameter, X i Based on the actual range (l b ,u b ) generated duty cycle particles;

[0109] Step S62, according to the generation order of the duty cycle particles in the initial population, the fitness of each duty cycle particle is calculated in turn to obtain the output power of each duty cycle particle, and the best duty cycle particle is determined according to the power optimization principle, and then step S63 is executed;

[0110] Step S63, based on the optimal duty cycle particle and according to the chaotic mapping principle, updating each duty cycle particle, specifically includes:

[0111] Step S631, based on each duty cycle particle, according to the chaotic mapping principle, generate a pseudo-random number p corresponding to each duty cycle particle and in the range of (0,1), and then execute step S632;

[0112] Step S632, based on the pseudo-random number corresponding to each duty cycle particle, according to p ≥ 0.5, determine whether to update each duty cycle particle according to the spiral predation principle of the Blue Whale algorithm, if yes, update according to the following formula, otherwise execute step S633:

[0113] X i (t+1)=w*e bl cos(2πl)+X best (t)

[0114]

[0115] D′=|X best (t)-X i (t)|

[0116] l∈(-1,1)

[0117] v=1

[0118] Where i is the individual number in the population, i∈{1, 2, ..., N}; N is the population size of the WOA algorithm; b is a constant related to the spiral shape; l is a random number between (-1, 1); D' is the distance vector; t is the current iteration number, t max is the maximum number of iterations; w is the step inertia factor; X best is the particle with the best duty cycle;

[0119] Step S633, based on the number of iterations and the sine function, generate the parameter α and the vector parameter A using the following formula, and then execute step S634:

[0120]

[0121] A=2α*r1-α

[0122] Where α is a control parameter related to the number of iterations; t is the current number of iterations; t max is the maximum number of iterations; r1 is a random number (0,1);

[0123] Step S634, based on the value of the vector parameter A, according to the condition |A|<1, determine whether to update each duty cycle particle according to the shrinking and surrounding position principle of the Blue Whale algorithm. If yes, update each duty cycle particle according to the following formula, otherwise execute step S635:

[0124] X i (t+1)=X best (t)-A*D*w

[0125] D=|C*X best (t)-X i (t)|

[0126] C=2*r2

[0127] Where, X i Based on the actual range (l b ,u b ) generated duty cycle particles; X best is the particle with the best duty cycle; C is a vector parameter, where |A|∈(0,1) in the encirclement stage; r2 is a random number of (0,1).

[0128] Step S635, based on the value of the vector parameter A and the Levy flight strategy, update each duty cycle particle according to the following formula:

[0129]

[0130] α=α0*(Xi(t)-Xbest(t))

[0131] Where, X i Based on the actual range (l b ,u b ) is the duty cycle particle generated; Xbest is the best duty cycle particle; Levy(β) is the Levy flight strategy, β is 1.5; α is the distance vector, α0 is 0.01; μ and ν are normal distributions between (0,1); Γ() is the Gamma distribution function.

[0132] Step S64, according to the generation order of each duty cycle particle after the update, the output power of each duty cycle particle is obtained in turn, and the best duty cycle particle after the update is determined in combination with the best duty cycle particle at the target historical iteration time, and the next iteration value of each duty cycle particle is determined according to the pseudo-random number p and the vector parameter A, and each particle is run after the position is updated, and each output power is sampled to obtain the optimal particle, which is compared with the optimal particle at the previous iteration time. If the output power increases, the optimal particle is replaced, otherwise it remains unchanged, and then step S65 is executed;

[0133] Step S65: If the termination condition is met, that is, the maximum number of iterations tmax is reached, the iteration is terminated to achieve stable output of maximum power; otherwise, steps S63 to S64 are iteratively repeated until the termination condition is met.

[0134] The results show that the improved Blue Whale algorithm based on chaotic mapping and Levy flight strategy is used in this paper to achieve multi-peak shadow tracking of photovoltaic array system, avoiding the disadvantage of inaccurate maximum power point tracking of traditional MPPT algorithm under shadow conditions. Fig.12 ;

[0135] In addition, through the comparison of steady-state power tracking, the comparison of the number of multi-peak shadow tracking iterations, and the comparison of the average results of 15 multi-peak shadow tracking, it can be seen that the improved Blue Whale algorithm is better than the Blue Whale algorithm (WOA), the Grey Wolf Algorithm (GWO), and the Particle Swarm Optimization (PSO) in achieving multi-peak maximum power point tracking. For the steady-state power comparison, see Fig.13 , for comparison of the number of iterations, see Fig.14 ,The comparison of the average results of 15 multi-peak shadow tracking is shown in Table 5;

[0136] Table 5

[0137]

[0138]

[0139] At the same time, chaotic mapping is used to solve the unevenness of random updates of initialized particles, reducing the possibility of the algorithm falling into local optimality; the Levy flight strategy is used to replace the random search update formula of the Blue Whale algorithm, avoiding a large number of invalid iterations and improving the convergence rate of the algorithm;

[0140] In addition, the fault characteristics of common photovoltaic array line faults were obtained through circuit simulation, and the faults were detected, classified and located using the test method of special working voltage points, realizing real-time diagnosis of photovoltaic array line faults. The fault characteristics obtained based on IV characteristics ensure the accuracy of diagnosis.

[0141] In summary, the proposed invention effectively solves the two main problems causing power loss in photovoltaic power generation systems.

[0142] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any way. Any simple modification, change and equivalent structural change made to the above embodiment based on the technical essence of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A photovoltaic array line fault diagnosis method based on shadow power tracking, characterized in that: Perform initial fault diagnosis on the target photovoltaic array system in real time according to the following steps S1 to S4 to determine whether there is a fault and the type of fault: Step S1, based on a preset sampling period, sampling the target photovoltaic array system, obtaining the array current and array voltage corresponding to the current sampling moment and the previous adjacent sampling moment at each sampling point, and then executing step S2; Step S2, calculating and obtaining the power value of each sampling point corresponding to the current acquisition moment and the adjacent previous acquisition moment, and then executing step S3; Step S3, based on the power values ​​of each sampling point corresponding to the current collection time and the previous adjacent collection time, respectively, obtain the power difference value and the power change percentage value of each sampling point at the current collection time and the previous adjacent collection time, and combine the preset range of the power difference value and the preset range of the power change percentage to determine whether the target photovoltaic array system has a fault, and if so, execute step S4; Step S4, based on the power change percentage values ​​at the current collection moment and the previous adjacent collection moment, combined with the preset range of abnormality detection, determine the fault type of the target photovoltaic array system.

2. A photovoltaic array line fault diagnosis method based on shadow power tracking according to claim 1, characterized in that: In step S3, the preset range of the power difference is ΔP<0, and the preset range of the power change percentage is Where ΔP is the power difference, Δ is the power change percentage, and fs is the sampling frequency.

3. A photovoltaic array line fault diagnosis method based on shadow power tracking according to claim 1, characterized in that: In step S4, the preset range of abnormal detection is Where Δ is the power change percentage, N s is the target PV array system row value, N p is the target photovoltaic array system column value, I m is the maximum power point current of the component, U m is the maximum power point voltage of the component.

4. A photovoltaic array line fault diagnosis method based on shadow power tracking according to claim 1, characterized in that: Based on steps S1 to S4, it is determined that there is a line fault in the target photovoltaic array, and further includes step S5, which implements line fault detection of the target photovoltaic array system according to the fault characteristics: Step S51, constructing a circuit model of the photovoltaic module according to the connection mode of the photovoltaic module in the photovoltaic array system, obtaining the working characteristics of the photovoltaic array, and then executing step S52; Step S52, based on the circuit model of the photovoltaic module, construct a fault model, realize the collection of fault characteristics, and form a fault characteristic table, and then execute step S53; Step S53, based on the preset test points of each group of strings in the photovoltaic array system, select the test point voltages, and determine the test point voltage sequence in order from small to large, and then execute step S54; Step S54, according to the test point voltage sorting, sampling and testing the preset test points of each string in turn, obtaining the positive and negative current values ​​of each string, and then executing step S55; Step S55, according to the test point voltage and the fault feature table, combined with the positive and negative current values ​​of each string, determine the line fault location and line fault type, and judge the number of faulty components.

5. A photovoltaic array line fault diagnosis method based on shadow power tracking according to claim 4, characterized in that: The step S53 selects the test point voltage of each preset test point of each string according to the following formula: IN cs (i)=i*U oc +ε Where: U cs is the test point voltage, U oc is the open circuit voltage of a single component, and ε is a decimal greater than 0.

6. A photovoltaic array line fault diagnosis method based on shadow power tracking according to claim 1 or claim 4, characterized in that: Based on step S3, it is judged that there is no fault in the target photovoltaic array system, or based on steps S1 to S4, it is judged that there is no line fault in the target photovoltaic array, or based on step S5, line fault detection of the target photovoltaic array system is achieved, and step S6 is also included, according to the hybrid algorithm, to achieve multi-peak maximum power point tracking control of the target photovoltaic array system: Step S61, using Circle chaotic mapping to generate an initial population of size N, each individual in the population represents a duty cycle particle, and then executing step S62; Step S62, according to the generation order of the duty cycle particles in the initial population, the fitness of each duty cycle particle is calculated in turn to obtain the output power of each duty cycle particle, and the best duty cycle particle is determined according to the power optimization principle, and then step S63 is executed; Step S63, based on the optimal duty cycle particle, according to the chaotic mapping principle, update each duty cycle particle, and then execute step S64; Step S64, according to the generation order of each duty cycle particle after the update, the output power of each duty cycle particle is obtained in sequence, and the best duty cycle particle after the update is determined in combination with the best duty cycle particle at the target historical iteration moment, and then step S65 is executed; Step S65, if the termination condition is met, the iteration is terminated to achieve stable output of maximum power; otherwise, steps S63 to S64 are iteratively repeated until the termination condition is met.

7. A photovoltaic array line fault diagnosis method based on shadow power tracking according to claim 6, characterized in that: The step S61 generates duty cycle particles according to the following formula: x i+1 =αx i (1-x i )0<α≤4 X i+1 =x i+1 *(u b -l b )+l b In the formula, x i is the duty cycle particle generated according to the chaotic mapping formula, α is the mapping parameter, X i Based on the actual range (l b ,u b ) to generate the duty cycle particles.

8. A photovoltaic array line fault diagnosis method based on shadow power tracking according to claim 6, characterized in that: The step S63 comprises: Step S631, based on each duty cycle particle, according to the chaotic mapping principle, generate a pseudo-random number p corresponding to each duty cycle particle and in the range of (0,1), and then execute step S632; Step S632, based on the pseudo-random number corresponding to each duty cycle particle and according to the preset pseudo-random number range, determine whether to update each duty cycle particle according to the spiral predation principle of the Blue Whale algorithm. If yes, update according to the following formula; otherwise, execute step S633: X i (t+1)=w*e bl cos(2πl)+X best (t) D′=|X best (t)-X i (t)| l∈(-1,1) b＝1 Where i is the individual number in the population, i∈{1, 2, ..., N}; N is the population size of the WOA algorithm; b is a constant related to the spiral shape; l is a random number between (-1, 1); D' is the distance vector; t is the current iteration number, t max is the maximum number of iterations; w is the step inertia factor; X best is the particle with the best duty cycle; Step S633, based on the number of iterations and the sine function, generate the parameter α and the vector parameter A using the following formula, and then execute step S634: A＝2α*r1-α Where α is a control parameter related to the number of iterations; t is the current number of iterations; t max is the maximum number of iterations; r1 is a random number (0,1); Step S634, based on the value of the vector parameter A and the range of the preset vector parameter A, it is determined whether to update each duty cycle particle according to the shrinking and surrounding position principle of the Blue Whale algorithm. If yes, update each duty cycle particle according to the following formula, otherwise, execute step S635: X i (t+1)=X best (t)-A*D*w D=|C*X best (t)-X i (t)| C＝2*r2 Where, X i Based on the actual range (l b ,u b ) generated duty cycle particles; X best is the particle with the best duty cycle; C is a vector parameter, where |A|∈(0,1) in the encirclement stage; r2 is a random number of (0,1). Step S635, based on the value of the vector parameter A and the Levy flight strategy, update each duty cycle particle according to the following formula: α=α0*((X i (t)-X best (t)) Where, X i Based on the actual range (l b ,u b ) generated duty cycle particles; X best is the particle with the best duty cycle; Levy(β) is the Levy flight strategy, β is 1.5; α is the distance vector, α0 is 0.01; μ and v are normal distributions between (0,1); Γ() is the Gamma distribution function.

9. A photovoltaic array line fault diagnosis method based on shadow power tracking according to claim 8, characterized in that: The pseudo-random number range in step S632 is preset to p≥0.

5.

10. A photovoltaic array line fault diagnosis method based on shadow power tracking according to claim 8, characterized in that: The range of the vector parameter A in step S634 is preset to |A|<1.

Images